Rice

Science
and
Technology
 FOOD SCIENCE AND TECHNOLOGY

A Series of Monographs, Textbooks, and Reference Books

EDITORIAL BOARD

Owen R. Fennema University of Wisconsin-Madison

Marcus Karel Massachusetts Institute of Technology
Gary W. Sanderson Universal Foods Corporation
Steven R. Tannenbaum Massachusetts Institute of Technology
Pieter Walstra Wageningen Agricultural University
John R. Whitaker University of California-Davis

1. Flavor Research: Principles and Techniques, R. Teranishi, I. Hornstein,

P. lssenberg, and E. L. Wick
2. Principles of Enzymology for the Food Sciences, John R. Whitaker
3. Low-Temperature Preservation of Foods and Living Matter, Owen R.
Fennema, William D. Powrie, and Elmer H. Marth
4. Principles of Food Science
Part 1: Food Chemistry, edited by Owen R. Fennema
Part II: Physical Methods of Food Preservation, Marcus Karel, Owen R.
Fennema, and Daryl B. Lund
5. Food Emulsions, edited by Stig E. Friberg
6. Nutritional and Safety Aspects of Food Processing, edited by Steven R.
Tannenbaum
7. Flavor Research: Recent Advances, edited by R. Teranishi, Robert A.
Flath, and Hiroshi Sugisawa
8. Computer-Aided Techniques in Food Technology, edited by Israel Saguy
9. Handbook of Tropical Foods, edited by Harvey T. Chan
10. Antimicrobials in Foods, edited by Alfred Larry Branen and P. Michael
Davidson
11. Food Constituents and Food Residues: Their Chromatographic Determi-
nation, edited by James F. Lawrence
12. Aspartame: Physiology and Biochemistry, edited by Lewis D. Stegink
and L. J. Filer, Jr.
13. Handbook of Vitamins: Nutritional, Biochemical, and Clinical Aspects,
edited by Lawrence J. Mach/in
14. Starch Conversion Technology, edited by G. M.A. van Beynum and J.
A. Roels
15. Food Chemistry: Second Edition, Revised and Expanded, edited by
Owen R. Fennema
16. Sensory Evaluation of Food: Statistical Methods and Procedures, Mi-
chael O'Mahony
17. Alternative Sweetners, edited by Lyn 0 'Brien Nabors and Robert C.
Gelardi
18. Citrus Fruits and Their Products: Analysis and Technology, S. V. Ting
and Russell L. Rouseff
19. Engineering Properties of Foods, edited by M.A. Rao and S. S. H. Rizvi
20. Umami: A Basic Taste, edited by Yojiro Kawamura and Morley R. Kare
21 . Food Biotechnology, edited by Dietrich Knorr
22. Food Texture: Instrumental and Sensory Measurement, edited by How-
ard R. Moskowitz
23. Seafoods and Fish Oils in Human Health and Disease, John E. Kinsella
24. Postharvest Physiology of Vegetables, edited by J. Weichmann
25. Handbook of Dietary Fiber: An Applied Approach, Mark L. Dreher
26. Food Toxicology, Parts A and B, Jose M. Concan
27. Modern Carbohydrate Chemistry, Roger W Binkley
28. Trace Minerals in Foods, edited by Kenneth T. Smith
29. Protein Quality and the Effects of Processing, edited by R. Dixon Phillips
and John W Finley
30. Adulteration of Fruit Juice Beverages, edited by Steven Nagy, John A.
Attaway, and Martha E. Rhodes
31. Foodborne Bacterial Pathogens, edited by Michael P. Doyle
32. Legumes: Chemistry, Technology, and Human Nutrition, edited by Ruth
H. Matthews
33. Industrialization of Indigenous Fermented Foods, edited by Keith H.
Steinkraus
34. International Food Regulation Handbook: Policy ·Science· Law, edited
by Roger D. Middlekauff and Philippe Shubik
35. Food Additives, edited by A. Larry Branen, P. Michael Davidson, and
Seppo Salminen
36. Safety of Irradiated Foods, J. F. Diehl
37. Omega-3 Fatty Acids in Health and Disease, edited by Robert S. Lees
and Marcus Karel
38. Food Emulsions: Second Edition, Revised and Expanded, edited by K~re
Larsson and Stig E. Friberg
39. Seafood: Effects of Technology on Nutrition, George M. Pigott and
Barbee W Tucker
40. Handbook of Vitamins: Second Edition, Revised and Expanded, edited
by Lawrence J. Mach/in
41. Handbook of Cereal Science and Technology, Klaus J. Lorenz and Karel
Kulp
42. Food Processing Operations and Scale-Up, Kenneth J. Valentas, Leon
Levine, and J. Peter Clark
43. Fish Quality Control by Computer Vision, edited by L. F. Pau and R.
0/afsson
44. Volatile Compounds in Foods and Beverages, edited by Henk Maarse
45. Instrumental Methods for Quality Assurance in Foods, edited by Daniel
Y. C. Fung and Richard F. Matthews
46. Listeria, Listeriosis, and Food Safety, Elliot T. Ryser and Elmer H. Marth
4 7. Acesulfame-K, edited by D. G. Mayer and F. H. Kemper
48. Alternative Sweeteners: Second Edition, Revised and Expanded, edited
by Lyn O'Brien Nabors and Robert C. Gelardi
49. Food Extrusion Science and Technology, edited byJozef L. Kokini, Chi-
Tang Ho, and Mukund V. Karwe
50. Surimi Technology, edited by Tyre C. Lanier and Chong M. Lee
51. Handbook of Food Engineering, edited by Dennis R. Heldman and Daryl
B. Lund
52. Food Analysis by HPLC, edited by Leo M. L. No/let
53. Fatty Acids in Foods and Their Health Implications, edited by Ching
Kuang Chow
54. Clostridium botulinum: Ecology and Control in Foods, edited by Andreas
H. W. Hauschild and Karen L. Dodds
55. Cereals in Breadmaking: A Molecular Colloidal Approach, Anne-Char-
lotte Eliasson and Ktlre Larsson
56. Low-Calorie Foods Handbook, edited by Aaron M. Altschul
57. Antimicrobials in Foods: Second Edition, Revised and Expanded, edited
by P. Michael Davidson and Alfred Larry Branen
58. Lactic Acid Bacteria, edited by Seppo Salminen and Atte von Wright
59. Rice Science and Technology, edited by Wayne E. Marshall and James
I. Wadsworth
60. Food Biosensor Analysis, edited by Gabriele Wagner and George G.
Guilbault

Additional Volumes in Preparation

Principles of Food Enzymology for the Food Sciences: Second Edition,

John R. Whitaker
 Rice
Science
and
Technology
edited by
Wayne E. Marshall
James I. Wadsworth
Agricultural Research Service
U. S. Department of Agriculture
New Orleans, Louisiana

Marcel Dekker, Inc. New York• Basel• Hong Kong

Library of Congress Cataloging-in-Publication Data

Rice science and technology I edited by Wayne E. Marshall, James

I. Wadsworth.
p. em. -- (Food science and technology ; 59)
Includes bibliographical references and index.
ISBN 0-8247-8887-7 (acid-free paper)
1. Rice--Processing. 2. Rice--Storage. 3. Rice--Quality.
4. Food--Composition. 5. Rice--Utilization. I. Marshall, Wayne
E. II. Wadsworth, James I. III. Series: Food
science and technology (Marcel Dekker, Inc.) ; 59.
TS2159.R5R523 1993
664'.725--dc20 93-28919
CIP

The publisher offers discounts on this book when ordered in bulk quantities.
For more information, write to Special Sales/Professional Marketing at the
address below.

This book is printed on acid-free paper.

Copyright @ 1994 by MARCEL DEKKER, INC. All Rights Reserved.

Neither this book nor any part may be reproduced or transmitted in any form
or by any means, electronic or mechanical, including photocopying, micro-
filming, and recording, or by any information storage and retrieval system,
without permission in writing from the publisher.

MARCEL DEKKER, INC.

270 Madison Avenue, New York, New York 10016

Current printing (last digit):

10 9 8 7 6 5 4 3 2 1

PRINTED IN THE UNITED STATES OF AMERICA

 Preface

Rice research and development in the United States is a multidisciplinary

effort involving scientists from academia, government, and industry. This
book brings together researchers from all three areas to focus on particu-
larly active problems and opportunities in rice research and development
where the next major advances in rice storage, quality, processing, and
utilization may be made.
In preparation for their contributions, each author was asked to include
(a) an introduction to the topic with a short historical perspective, (b) a
description of his/her research and its importance to an end user (grower,
processor, consumer), with emphasis on current research underway in his/
her laboratory, and (c) a conclusion that identifies promising new areas for
investigation based on an extension of the author's current research. With
this kind of information, we hope to stimulate the reader's interest in the
broad field of rice science and technology and show the reader where some
of the major problems and solutions lie. For veteran rice researchers, we
hope to supply information of value that they can bring to bear on their
particular area of investigation.
The book is intended for a wide audience of rice scientists seeking
current information on selected topics in rice storage, quality, processing,
and utilization. The book may also be of interest to investigators working

iii
iv PREFACE

with other cereal crops, since many of the approaches and methods de-
scribed for rice can apply to wheat, corn, barley, oats, etc.
Although the pursuit of knowledge concerning rice is an international
endeavor, the editors have selected authors working in the United States on
problems of interest to the U.S. rice industry. We have chosen this path
because we wanted to support and document the strong current activity in
U.S. rice research.
We extend our grateful appreciation to the contributors of this book.
Authorities in their field, they contributed their time and expert knowledge
to making this book a reality.
Sincere thanks also go to the editorial staff at Marcel Dekker, Inc.,
who eased the editors' burden, wherever possible, during the publication
process.

Wayne E. Marshall
James I. Wadsworth
 Contents

Preface iii
Contributors ix

1. Introduction 1
Wayne E. Marshall and James I. Wadsworth

Part 1: Storage

2. Brown Rice Stabilization 17

Elaine T. Champagne

3. Storage Stability of Extrusion Stabilized and

Parboiled Rice Bran 37
Don R. McCaskill and Frank T. Orthoefer

4. Effect of Storage on the Physicochemical Properties and

Quality Factors of Rice 49
Joseph Chrastil

v
vi CONTENTS

Part II: Quality

5. Breeding for Rice Quality 83

Kent S. McKenzie
6. New Methods for the Evaluation of Rice Quality and
Related Terminology 113
David E. Kohlwey
7. Degree of Milling 139
James I. Wadsworth
8. The Influence of Rice Protein on Rice Quality 177
Bruce R. Hamaker
9. Enrichment of Rice 195
Diane W. Hoffpauer and Salmen L. Wright III
10. Starch Gelatinization in Brown and Milled Rice: A Study
Using Differential Scanning Calorimetry 205
Wayne E. Marshall

Part III: Processing

11. New Methods and Equipment for Processing Rice 229

RobertS. Satake
12. Parboiling Rice with Microwave Energy 263
Lakshman Velupillai
13. New Methods for On-the-Farm Rice Drying:
Solar and Biomass 275
Lalit R. Verma
14. Microwave-Vacuum Drying 299
James I. Wadsworth
15. Role of Moisture Content in Affecting Head Rice Yield 341
Terry J. Siebenmorgen

Part IV: Utilization

16. Processing and Utilization of Rice Bran in the United States 381
Keith L. Hargrove, Jr.
CONTENTS vii

17. Food Applications for Modified Rice Starches 405

Roy N. Sharp and Carolyn Q. Sharp
18. Rice Bran Oil and Its Heath Benefits 421
Robert J. Nicolosi, Eugene J Rogers, Lynne M. Ausman and
Frank T. Orthoefer
19. Changing Market Demands for Rice and Rice Products 439
Richard A. Meyers

Index 465
 Contributors

Lynne M. Ausman Human Nutrition Research Center on Aging and

School of Nutrition, Tufts University, Boston, Massachusetts

Elaine T. Champagne Southern Regional Research Center, Agricultural

Research Service, U.S. Department of Agriculture, New Orleans, Louisiana

Joseph Chrastil Southern Regional Research Center, Agricultural Re-

search Service, U.S. Department of Agriculture, New Orleans, Louisiana

Bruce R. Hamaker Department of Food Science, Purdue University,

West Lafayette, Indiana

Keith L. Hargrove, Jr. Farmers' Rice Cooperative, Sacramento, Califor-

nia

Diane W. Hoffpauer Gourmet Technologies, Inc., Crowley, Louisiana

David E. Kohlwey Riviana Foods, Inc., Houston, Texas

Wayne E. Marshall Southern Regional Research Center, Agricultural Re-

search Service, U.S. Department of Agriculture, New Orleans, Louisiana

ix
x CONTRIBUTORS

Don R. McCaskill Research & Development, Riceland Foods, Inc.,

Stuttgart, Arkansas

Kent S. McKenzie Rice Experiment Station, California Cooperative Rice

Research Foundation, Biggs, California

Richard A. Meyers U.S.A. Rice Council, Houston, Texas

Robert J. Nicolosi Department of Clinical Sciences, University of

Massachusetts-Lowell, Lowell, Massachusetts

Frank T. Orthoefer Research & Development, Riceland Foods, Inc.,

Stuttgart, Arkansas

Eugene J. Rogers Department of Clinical Sciences, University of

Massachusetts-Lowell, Lowell, Massachusetts

Robert S. Satake Satake Corporation, Hiroshima, Japan

Carolyn Q. Sharp Department of Food Science, University of Arkansas,

Fayetteville, Arkansas

Roy N. Sharpt Department of Food Science, University of Arkansas, Fay-

etteville, Arkansas

Terry J. Siebenmorgen Department of Biological & Agricultural Engi-

neering, University of Arkansas, Fayetteville, Arkansas

Lakshman Velupillai International Programs, Louisiana State University

Agricultural Center, Baton Rouge, Louisiana

Lalit R. Verma Department of Biological and Agricultural Engineering,

Louisiana State University, Agricultural Center, Baton Rouge, Louisiana

James I. Wadsworth Southern Regional Research Center, Agricultural

Research Service, U.S. Department of Agriculture, New Orleans, Louisiana

Salmen L. Wright ill Wright Enrichment, Inc., Crowley, Louisiana

tDeceased.
 Rice
Science
and
Technology
 1
Introduction
Wayne E. Marshall and James I. Wadsworth
Agricultural Research Service, U.S. Department of Agriculture,
New Orleans, Louisiana

I. HISTORICAL PERSPECTIVE
Rice is a semiaquatic, annual grass which can be grown under a broad
range of climatic conditions. Cultivated rice is designated as either Oryza
sativa L. or Oryza glaberrima Steud. 0. sativa is the predominant species;
0. glaberrima is grown only in Africa on a limited scale.
The date and geographical location of the first cultivated rice (0. sativa
L.) is obscure. Historical records indicate that rice was simultaneously and
independently cultivated between 2000 and 1500 B.c. in an area extending
roughly from central India, through northern Burma, northern Thailand,
Laos, Vietnam, and into southeastern China (1). From this broad belt, the
cultivation of rice spread to Indonesia, the Philippines, and northern Aus-
tralia. Later, traders carried the grain throughout Asia, the Middle East,
and Europe.
Rice was introduced into the New World (North and South America)
comparatively recently. The first documentation of rice as a commercial
crop in the United States was in 1686 in Charleston, South Carolina, al-
though trial plantings occurred in Virginia in 1609 (2). Rice cultivation
moved further south and west in the ensuing 200 years and became an
established crop in Louisiana and Texas in 1888, Arkansas in 1904, Califor-
nia in 1912, and Mississippi in 1942 (3). More recently, Missouri and Flor-
ida have initiated commercial production.

1
2 MARSHALL AND WADSWORTH

II. STRUCTURE AND COMPOSITION

The gross structure of the mature rough rice grain is shown in Figure 1. The
principal parts of the grain are the hull, pericarp, seed coat, nucellus,
embryo, aleurone layer, and endosperm. This section will give a brief
description of these major parts and the chemical composition of the vari-
ous milling fractions of which these parts play a role. Recent comprehen-
sive descriptions of rough rice structure and composition exist in the litera-
ture (4,5) for the interested reader.

Seed coat
Nucellus
Hull leurone layer
-Subaleurone}
layer
..___..,.__ _ Starchy Endosperm
endosperm

~f~~e~\~m}
Embryo
;;;;;;,.,:rr---- Radicle
t~~!fil&;iii~J;-7+----Epiblast
Sterile
lemmae

Figure 1 Structure of the mature rice grain. (From Ref. 4.)

 INTRODUCTION 3

A. Hull
The hull is the outer covering for the caryopsis (brown rice). It comprises
18-20% by weight of the rough rice (4). The hull serves a protective
function against insect infestation and against rapid changes in moisture
content of the grain due to large humidity fluctuations in the external
environment. It consists of the lemma, which covers the dorsal part of the
grain, and the palea, which covers the ventral portion (see Fig. 1). The
lemma and palea are joined together longitudinally. Hulls are low in pro-
tein, fat, and starch but high in crude fiber, crude ash (mostly silica), and
dietary fiber (Table 1). The hull is the only part of the grain that contains
appreciable silica (silicon dioxide). Hulls also contain small quantities of
calcium, sodium, magnesium, potassium, manganese, aluminum, iron, cop-
per, and zinc (4). Their gross composition makes them nutritionally insig-
nificant (except for their fiber content) and therefore unattractive as a
human food.

B. Pericarp, Seed Coat, Nucellus, and Aleurone

Removal of the hull from rough rice by dehulling exposes the rice
caryopsis. The outer four morphologically distinct layers of the caryopsis
are the pericarp, seed coat (tegmen), nucellus, and aleurone (see Fig. 1).
Along with much of the embryo (germ), these layers comprise the bran
portion of the rice grain. Although the aleurone layer is botanically part of
the endosperm, it is removed as part of the bran fraction during milling.
The bran portion accounts for 5-8% of the brown rice weight (4).
The bran is the most nutritious part of the caryopsis. The cells of the
aleurone layers, in particular, consist of many inclusions called protein

Table 1 Range of Proximate Composition of Rough Rice, Brown Rice, Milled Rice,
Rice Hulls, Rice Bran, Rice Embryo, and Rice Polish(% dry basis)
Constituent Rough Brown Milled Hulls Bran Embryo Polish
Protein (N 6.7-8.3 8.3-9.6 7.3-8.3 2.3-3.2 13.2-17.3 17.7-23.9 13.0-14.4
X 5.95)
Crude fat 2.1-2.7 2.1-3.3 0.4-0.6 0.4-0.7 17.0-22.9 19.3-23.8 11.7-14.4
Crude fiber 8.4-12.1 0.7-1.2 0.3-0.6 40.1-53.4 9.5-13.2 2.8-4.1 2.7-3.7
Crude ash 3.4-6.0 1.2-1.8 0.4-0.9 15.3-24.4 9.2-11.5 6.8-10.1 6.1-8.5
Starch 62.1 77.2 90.2 1.8 16.1 2.4 48.3-55.4
Dietary 19.1 4.5 2.7 77.3 27.6-33.3
fiber
Source: Ref. 11.
4 MARSHALL AND WADSWORTH

bodies and lipid bodies. Most of the bran protein and oil are stored in these
structures. Bran contains a good balance of protein, fat (rice oil), carbohy-
drate (starch), and dietary fiber (see Table 1). In addition, the bran con-
tains many vitamins and minerals in amounts beneficial to humans (4).

C. Subaleurone and Starchy Endosperm

Further milling of the rice caryopsis removes the subaleurone layer and a
small part of the starchy endosperm (see Fig. 1). This milling fraction is
referred to as "polish." The end result of bran and polish removal is milled
(white) rice. The polish comprises 3-4% by weight of brown rice (4). The
subaleurone layer and the starchy endosperm differ in the type of inclusion
bodies found in the cells. The subaleurone layer is rich in protein bodies, it
has fewer lipid bodies than the aleurone layer, but contains only a small
number of starch granules. The starchy endosperm is rich in starch gran-
ules, contain some protein bodies, especially in the outer endosperm lay-
ers, and almost no lipid bodies.
Polish contains only slightly less protein and lipid but considerably more
starch than bran (see Table 1). The polish fraction is only slightly less nutri-
tious than the bran fraction, primarily because it contains lower levels of
minerals and vitamins than the bran (4). When brown rice is commercially
milled to white rice, the bran and polish fractions are not separated. This
entire milling fraction is called rice bran. Although nutritious for humans,
most of the rice bran is sold as animal feed. One of the challenges facing the
rice industry is to utilize the food value of rice bran more effectively.
The milling product of greatest commercial importance is white rice. It is
composed entirely of starchy endosperm. Starchy endosperm can be fur-
ther divided into an outer endosperm, just below the subaleurone layer,
and the inner endosperm or central core (5). The regions differ primarily in
the number of starch granules and protein bodies contained within the
cells, which reside in the two regions. The protein content is highest in the
outer layers, and starch content is highest in the central core. In the central
core, the starch granules are hexagonal in shape, highly compact, and take
up most of the intracellular space within the starch cells. Between the
center and the outside of the grain, the starch cells become elongated and
radiate outward. The radial wall of the starch cells forms potential cleavage
planes that may result in grain breakage caused by mechanical impact or
moisture stress (6).
The starchy endosperm contains more starch per unit weight than any
other milling fraction (hulls, bran, polish) and also has one of the lowest
lipid contents (see Table 1). Protein content is lower than in the bran and
polish fractions, but higher than in the hulls. White rice is considered a
INTRODUCTION 5

good source of protein and carbohydrate but not a significant source of

minerals and vitamins unless fortified (7).

ill. PRODUCTION AND CONSUMPTION

Rice is one of the leading food crops of the world and is second only to
wheat in terms of annual production for food use. It is the main staple food
for about 60% of the world's population. About 90% of the world's rice is
produced and consumed in Asia.

A. World Production and Consumption

Figure 2 shows world rice production and consumption from 1980 to 1993.
This period is marked by a steady increase in both production and consump-
tion until the 1985/1986 crop year, when a slowdown occurred. Since the
1985/1986 crop year, both parameters have shown only a gradual increase.
The overall increase in production since 1980 has been the result of higher
yields since acres planted worldwide have remained fairly constant. Con-
sumption has kept pace with production throughout the period. This signi-

Milled Rice (Mt)

400
-World Production ~World Consumption

360 ........................................................................................ .

300 ........................ .

260

200
1880-81 1882-83 1884-86 1888-87 1888-88 1880-81 1882-83

Crop Year
Figure 2 World rice production and consumption in million metric tons. Data for
1991-92 are forecast; data for 1992-93 are projected. (From Ref. 8.)
6 MARSHALL AND WADSWORTH

fies the value of rice as a food. Corn, for example, is produced mainly as
feed. Food consumption of corn is only a fraction of total production.

B. Export Market
Thailand and the United States are currently the two top rice-exporting
countries and have been since at least 1987 (Table 2). However, they only
contribute 4-5% and 1-2%, respectively, to the total world crop (8). China
and India are first and second, respectively, in world rice production (8), but
they are not considered major rice-exporting countries, since their large
populations consume about all of the rice they grow. The United States
exports about one half of its annual production and has contributed on the
average about 20% of the total rice on the export market (Fig. 3). There is
considerable fluctuation on a year-to-year basis that is highly dependent on

Table 2 International Export Trade, 1987-1993, of Major Rice-Exporting Coun-

tries (1000 metric tons of milled rice equivalent)

Calendar year

Country 1987 1988 1989 1990 1991 1992• 1993b

United States 2,444 2,241 2,967 2,420 2,197 2,200 2,300

Argentina 150 160 130 70 125 175 175
Australia 338 417 450 470 400 450 500
Burma 493 368 456 186 300 200 200
China 1,020 698 320 300 689 750 700
Taiwan 240 104 68 50 200 200 200
EC-12 981 920 963 990 1,090 1,040 900
Egypt 105 108 100 32 85 160 170
Guyana 69 56 26 30 53 55 75
India 350 200 450 420 500 400 450
Indonesia 100 0 104 50 0 0 0
North Korea 154 199 175 75 0 0 0
Pakistan 1,226 950 779 904 1,297 1,200 1,200
Thailand 4,355 4,791 6,036 3,937 3,993 4,600 4,400
Uruguay 190 244 251 250 262 300 350
Vietnam 153 97 1,400 1,500 1,000 1,200 1,200
Other 560 355 419 350 358 412 452
World total 12,928 11,908 15,094 12,034 12,549 13,342 13,272
•Forecast.
hProjected as of July 1992.
Source: Ref. 8.
INTRODUCTION 7

U.S. Share of World Trade (%)

30

28

22

18

14

10+-r.-r.-r.-r.-ro-r.-ro-ro-TO-ro-TO-ro-TOrTo-~
1982 1986 1988 1971 197 4 1977 1980 1983 1988 1989 1992
Year
Figure 3 Percentage of world trade in rice between 1962 and 1992 supplied by
U.S. rice exports. Data for 1992 are forecast. (From Ref. 8.)

world production/consumption. The United States is able to export a signifi-

cant percentage of its rice because of high production efficiency (high yields)
and low per capita domestic consumption compared to other countries.
Although U.S. rice exports have been fairly stable at about 20% of the
world market over the past decade, fundamental changes in the market have
altered the destination and type of U.S. exports (9). These changes are
shown in Tables 3 and 4. Over the past three decades, the Middle East,
Africa, and Western Europe replaced Asia as primary export markets. Since
the early 1970s, parboiled rice, demanded by growing markets in the Middle
East, has increased its share of U.S. exports. Brown rice, once demanded in
large quantities by South Korea, has declined. Most of the exported rice is
the long-grain type, which reached about 70% of the total exported in the
1980s. This was due to the expansion of the Middle Eastern markets and
declines in the Asian markets. Two world political events occurring since
1989 have had major impact on U.S. rice exports. Before the Persian Gulf
war, Iraq was the single largest buyer of U.S. rice, purchasing about 20% of
total exports. This market has disappeared, and the U.S. share of the world
rice trade has decreased. The other political event was the dissolution of the
U.S.S.R. The republics of the former Soviet Union (FSU-12) have begun
purchasing U.S. rice through export programs. In crop year 1991-92, the
 8 MARSHALL AND WADSWORTH

Table 3 Top 10 U.S. Rice Export Markets(% of exports• to country of destination)

FY 1986 FY 1987 FY 1988
Rank Country (%) Country (%) Country (%)
1 Iraq 22.2 Iraq 22.1 Iraq 21.4
2 Brazil 14.4 Saudi Arabia 13.1 Saudi Arabia 14.2
3 Saudi Arabia 12.8 Belgium 6.0 Belgium 6.0
4 Belgium 6.2 Haiti 4.7 Philippines 5.9
5 Canada 4.9 Canada 4.4 Canada 5.3
6 Liberia 3.2 South Africa 3.4 South Africa 4.5
7 South Africa 2.8 Guinea 2.7 Haiti 3.3
8 Switzerland 2.2 Netherlands 2.5 Switzerland 3.0
9 Jamaica 2.0 Liberia 2.4 Jamaica 2.9
10 Dominican Rep 1.9 Turkey 2.4 Bangladesh 2.7
Subtotal 72.5 63.7 69.3
•Percent calculated as proportion of total value of U.S. rice exports.
Source: Refs. 8, 10.

FSU-12 bought 41.9 thousand metric tons of U.S. rice to rank 12th on the list
of largest buyers (10). Rice market analysts anticipate that this market will
continue to grow. U.S. government-sponsored export programs will play a
major role in promotion of rice exports to FSU-12.

C. U.S. Production
U.S. rice production and consumption (including exports) since 1980 is
shown in Figure 4. Both production and consumption showed a decline in
the early/mid-1980s, but have exhibited only minor year to year variations
since the 1988-89 crop year. Measurable rice production is confined to
Arkansas, California, Louisiana, Mississippi, Missouri, and Texas. Produc-
tion by each of these states for the period 1980-1991 is given in Table 5.
Arkansas is the leading rice-producing state with 43% of the national total
in 1991, followed by California and Louisiana at 16%, Texas 13%, Missis-
sippi 8%, and Missouri 3%.
Traditionally, rice varieties are classified as long-, medium- and short-
grain types. The majority of rice grown in the United States in 1991 was the
long-grain type, which was grown mainly in the mid-south (Arkansas, Loui-
siana, Texas, Mississippi, Missouri). This type comprised 71% of the total
rice grown. Medium-grain rice, grown predominantly in California, Louisi-
ana, and Arkansas, accounted for 28% of the total. Short-grain rice was
grown almost exclusively in California and represented less than 1% of the
total grown.
 INTRODUCTION 9

FY 1989 FY 1990 FY 1991 FY 1992

Country (%) Country (%) Country (%) Country (%)
Iraq 18.8 Iraq 12.1 Saudi Arabia 11.1 Brazil 10.6
Saudi Arabia 8.7 Saudi Arabia 9.5 Brazil 8.2 Turkey 8.7
Belgium 5.1 Mexico 7.5 Canada 6.8 Saudi Arabia 8.3
Turkey 4.4 Peru 6.3 Haiti 6.1 Netherlands 8.2
Spain 4.3 Canada 5.4 Turkey 5.7 Mexico 5.8
Mexico 3.8 Turkey 5.3 South Africa 4.9 South Africa 5.8
Canada 3.5 Haiti 3.1 Switzerland 4.1 Haiti 5.5
Switzerland 3.2 South Africa 4.1 Liberia 3.9 Canada 5.0
Haiti 3.1 Belgium 4.1 Netherlands 3.5 Ivory Coast 3.2
South Africa 3.1 Jordan 3.7 Mexico 3.5 Jamaica 3.2
58.1 62.4 57.8 64.4

Table 4 U.S. Rice Exports by Type

Crop Milled Brown Parboiled Rough Brokens

year (%) (%) (%) (%) (%)
1977/78 63.9 10.6 21.7 2.0 1.9
1978/79 58.3 11.4 25.8 3.7 0.9
1979/80 56.8 17.6 22.1 2.0 1.5
1980/81 33.4 39.7 25.8 0.4 0.6
1981182 36.4 18.7 37.3 7.0 0.4
1982/83 44.8 16.0 38.1 0.8 0.2
1983/84 42.8 14.7 36.2 4.6 1.7
1984/85 51.5 8.6 32.2 5.2 2.4
1985/86 50.6 14.5 27.8 2.8 4.3
1986/87 56.8 9.0 24.3 9.7 0.2
1987/88 55.9 7.8 28.0 1.6 6.7
1988/89 51.0 11.5 30.0 4.6 2.9
1989/90 45.8 12.2 37.3 2.0 2.6

Source: Ref. 8.

Breeding programs in the United States have produced varieties for each
grain type that are associated with specific cooking, eating, and processing
qualities. Long-grain varieties are most desired for direct table use because
they cook dry and fluffy and the cooked grains tend to separate rather than
stick together in a clump. High-quality medium- and short-grain types are
most moist, chewy, and sticky, and for home use they appeal to the U.S.
10 MARSHALL AND WADSWORTH

Table 5 State and U.S. Rice Production by Class (1000 cwt), 1980-1991

State 1980 1981 1982 1983 1984

Long-grain rice
Arkansas 42,480 57,280 47,608 33,012 46,320
California 826 1,309 4,288
Louisiana 8,875 10,554 10,962 7,622 13,899
Mississippi 9,086 14,432 10,094 6,440 8,265
Missouri 2,100 2,747 3,160 2,460 3,358
Texas 24,310 25,413 20,774 13,475 19,899
United States 86,851 110,426 93,424 64,318 96,029
Medium-grain rice
Arkansas 9,073 11,094 8,400 5,784 6,400
California 29,606 31,373 27,202 14,129 20,520
Louisiana 11,893 16,524 13,900 7,071 8,033
Mississippi 140 360
Missouri 191 320 400 74 90
Texas 504 1,826 1,440 330 261
United States 51,407 61,497 51,342 27,388 35,304
Short-grain rice
Arkansas 1,062 1,236 1,029 363 180
California 6,780 9,551 7,820 7,651 7,252
Missouri 50 32 22 45
United States 7,892 10,819 8,871 8,014 7,477
Total grains
Arkansas 52,615 69,610 57,037 39,159 52,900
California 36,386 40,924 35,848 23,089 32,060
Louisiana 20,768 27,078 24,862 14,693 21,932
Mississippi 9,226 14,792 10,094 6,440 8,265
Missouri 2,341 3,099 3,582 2,534 3,493
Texas 24,814 27,239 22,214 3,805 20,160
United States 146,150 182,742 153,637 99,720 138,810

•No grain estimates.

Source: Ref. 8.

immigrant population, especially from Asia. Because of the diverse cui-

tural makeup in the United States, there is a demand for all three grain
types. Also, rice processors require different quality characteristics for use
in various prepared and convenience food products.

D. U.S. Consumption
The United States mirrors the rest of the world in that rice is almost
exclusively used for food and rarely for animal feed. The United States is
INTRODUCTION 11

1985 1986 1987 1988 1989 1990 1991

50,712 49,462 45,259 57,447 57,458 53,034 58,328

3,834 1,520 2,592 4,200 2,250 1,314 1,080
14,418 14,061 12,079 17,538 13,128 14,805 12,500
10,058 10,692 10,098 13,275 13,395 14,250 12,320
3,415 3,335 3,420 4,080 4,056 3,713 4,641
17,930 17,703 15,547 22,824 18,874 20,690 20,180
100,367 96,773 88,995 119,364 109,161 107,806 109,049

3,809 4,544 7,656 7,236 6,322 6,912 8,392

18,628 21,917 22,496 22,050 26,315 28,215 23,510
5,383 5,319 7,031 6,542 8,360 11,664 12,235
a 505 a

48 99 144 102 52 47 51
141 360 324 456 392 490 400
28,464 32,239 37,651 36,891 41,441 47,358 44,589

76 54 110 52 60 54 60
6,006 4,290 2,847 3,590 3,825 900 760
a a a

6,082 4,344 2,957 3,642 3,885 954 820

54,597 54,060 53,025 64,735 63,840 60,000 66,780

28,468 27,727 27,935 29,840 32,390 30,429 25,350
20,256 19,380 19,110 24,080 21,488 26,469 24,735
10,058 10,692 10,098 13,780 13,395 14,250 12,320
3,463 3,434 3,564 4,182 4,108 3,760 4,692
18,071 18,063 15,871 23,280 19,266 21,180 20,580
134,913 133,356 129,603 159,847 154,487 156,088 154,457

not a conspicuous consumer of rice compared to many other countries.

However, per capita consumption in the United States has doubled since
1975 and is now approaching 10 kg (22lb) (Fig. 5). This growth has bene-
fited from several factors, including (a) the growing trend in U.S. diets
away from high-fat animal products and towards low-fat grain-based foods,
(b) the increase in the Asian and Hispanic segments of the U.S. population,
and (c) the aggressive marketing campaign by the U.S. Rice Council to
make rice more visible to a wider range of consumers.
Domestic rice has three different uses (a) direct food use, (b) processed
foods, and (c) brewing. Direct food use is the largest category and com-
prised about 58% of the total domestic market in 1991 (8). Direct food use
 Milled Rice (Mt)
10
- U. 8. Production ~ U. 8. Coneumptlon

8 .................................................................................................................. ..

4
1980-81 1882-83 1884-86 1888-87 1888-88 1980-81 1882-93
Crop Year
Figure 4 United States rice production and consumption in million metric tons.
Data for 1991-92 are forecast; data for 1992-93 are projected. (From Ref. 8.)

Per Capita Consumption (kg)

10

o~~~~~~~~~--~~--~~--~~--~~--~~--

nnnnn~~H~M~~~~6~~~

Year
Figure 5 United States per capita consumption of rice from 1975 to 1992 in
kilograms. Value for 1992 is forecast. (From Ref. 8.)

12
INTRODUCTION 13

U.S. Domestic Rice Consumption (Mt)

3
-Direct Food Uee ~ Proceeeed Food• IHH Beer
2.6 ···········································································································

2 ················································································

1.6

0.6

0
66 69 71 73 76 78 80 82 84 88 88 90 92
Marketing Years
Figure 6 United States domestic rice consumption in million metric tons between
1966 and 1992 broken down into three categories: direct food use, processed foods,
and brewing. Results for 1992 are forecast. (From Ref. 8.)

includes the conventional white rice plus specialty rice products (parboiled,
precooked, aromatic, brown, and prepackaged mixes). Specialty products
account for approximately 20% of direct food use. About 70% of the rice
destined for direct food use reaches the consumer thought retail outlets and
the other 30% through food service outlets.
Rice consumption through brewing and processed foods share equally
the remainder of the U.S. market, but use in the processed food market has
almost doubled over the last decade, while use for brewing has remained
fairly stable (8). Breakfast cereals lead the processed food category and
consist of mainly medium-grain rice. Other processed foods include pre-
packaged mixes and rice cakes (predominantly long-grain), canned soup
and frozen dinners (long-grain), candy (medium-grain and brokens), and
baby and pet food (brokens and rice flour). Rice (brokens) for brewing is
primarily used for the production of beer.
Total consumption of rice for domestic use and its consumption by cate-
gory for the period 1966 to 1992 is portrayed in Figure 6. The continued
increase since the 1982 marketing year has resulted from gains in the pro-
cessed foods and direct food use categories.
14 MARSHALL AND WADSWORTH

IV. CURRENT RESEARCH AND DEVELOPMENT

The steady increase in both total (see Fig. 6) and per capita consumption
(see Fig. 5) of rice in the United States has brought about increased rice
research and development activity from universities, government laborato-
ries, and industry. All of these groups have developed active rice research
programs. These programs use the most recent technology to provide basic
and applied information to ensure high-quality rice and rice products for
U.S. domestic and export markets. The chapters that follow address par-
ticularly active areas of rice research and development where the editors
believe the next major advances in rice storage, quality, processing, and
utilization may be made.
Consumer recognition of brown rice as a healthy food has shifted inter-
est toward a greater use of brown rice in the U.S. diet. However, a major
obstacle in increased marketing of brown rice is the short shelf life of the
product. After removal of the hull, brown rice quality deteriorates rapidly
while sitting on the supermarket shelf. Chapters 2 and 3 describe current
research designed to solve the deterioration problem and significantly ex-
tend the shelf life of brown rice. Storage of milled rice can also affect its
quality. Chapter 4 presents a review of current knowledge describing desir-
able and undesirable changes in milled rice during storage.
Rice quality, in a positive sense, is a term used to describe the specific traits
or characteristics of rice that make it appealing to the grower, processor, and
consumer. Chapters 5 through 10 cover rice quality from all of these three
perspectives. Chapter 5 describes current breeding strategies used to satisfy
grower demands for certain quality features. Chapter 6 discusses the latest
methods used to quantify rice quality from a rice processor's perspective.
Chapter 7 describes a critical determinant in achieving rice quality, namely,
degree of milling. It also evaluates the potential of a relatively new technique,
near infrared reflectance (NIR) spectroscopy, as a suitable on-line method for
the determination and control of degree of milling. Chapters 8 and 10 assess
the role played by protein and starch, the two largest components in rice, in
determining rice quality, particularly in the cooked grain. The basic research
presented in these chapters have application to both the processor and con-
sumer. Chapter 9 gives the reader up-to-date information on the enrichment
process that is widely used in the rice industry to improve the nutritional
quality of the grain with its direct, beneficial effects on human health.
The emphasis in Chapters 11 through 15 is on the newest technology
used for processing rice. The Japanese have long been leaders in the devel-
opment of new, innovative technology (methods and equipment) for rice
processing. A description of state-of-the-art methods and equipment sold
by the U.S. subsidiary of a Japanese company is presented in Chapter 11.
INTRODUCTION 15

Chapters 12, 13, and 14 discuss innovative approaches to rice parboiling

and drying, two processing techniques practiced for centuries without ma-
jor change. Emphasis is placed on the use of microwave energy, a relatively
recent technology, for greater control and efficiency of the parboiling and
drying processes. Chapter 15 evaluates recent data on the relationship
between rice moisture content and head yield. This information is critically
important to rice millers, who must continually monitor moisture content
to maximize yield of head rice in order to stay competitive.
Storage, quality, and processing of rice are important steps that lead to
the ultimate value of rice as a food grain. The ultimate value of rice resides
in its utilization by the processor or consumer as a food or an ingredient in
foods. The final chapters highlight a very active area of rice utilization,
namely, the increasing use bran and bran oil (Chapters 16 and 18) and
starch (Chapter 17) as nutritional and functional food ingredients. Chapter
19 concludes by presenting an excellent overview of current and future
prospects for rice utilization as seen from a marketing perspective.

REFERENCES
1. Chang, T. T. (1976). The origin, evolution, cultivation, dissemination, and
diversification of Asian and African rices. Euphytica, 25:425.
2. Dethloff, H. C. (1982). The colonial rice trade. Agric. Hist., 56:231.
3. Adair, C. R. (1973). Introduction, Rice in the United States: Varieties and Pro-
duction. ARS, USDA, Agricultural Handbook 289, Washington, D.C., p. 1.
4. Juliano, B. 0. and Bechtel, D. B. (1985). The rice grain and its gross composi-
tion, Rice: Chemistry and Technology (B.O. Juliano, ed.), American Associa-
tion of Cereal Chemists, Inc., St. Paul, MN, p. 17.
5. Lu, S., and Luh, B. S. (1991). Properties of the rice caryopsis, Rice: Produc-
tion, Vol. I (B.S. Luh, ed.), AVI Publishing Co., Westport, CT, p. 389.
6. Araullo, E. V., DePadua, D. B., and Graham, M. (1976). Rice Postharvest
Technology, International Development Research Center, IDRC-053e, Ot-
tawa, Canada.
7. de Lumen, B. 0., and Chow, H. (1991). Nutritional quality of rice endo-
sperm, Rice: Utilization, Vol. II (B.S. Luh, ed.), AVI Publishing Co., West-
port, CT p. 363.
8. USDA. (1992). Rice, Situation and Outlook Yearbook. Economic Research
Service, RS-64, July, Washington, D.C.
9. Childs, N. W. (1989). The changing role of the United States in the world rice
market, Rice. Situation and Outlook Yearbook. USDA, Economic Research
Service, RS-55, July, Washington, D.C.
10. USDA. (1992). Rice Market News, 73:8.
11. Pomeranz, Y., and Ory, R. L. (1982). Rice processing and utilization, CRC
Handbook of Processing and Utilization in Agriculture, Vol. II (I. A. Wolff,
ed.), CRC Press, West Palm Beach, FL.
 2
Brown Rice Stabilization
Elaine T. Champagne
Agricultural Research Service, U.S. Department of Agriculture,
New Orleans, Louisiana

I. INTRODUCTION
Brown rice is nutritionally superior to white rice. It has higher percentages
of all nutrients except carbohydrate (1). Thiamine and oil contents of
brown rice are approximately five times that of white rice, while fiber,
niacin, phosphorous, potassium, iron, sodium, and riboflavin contents are
approximately two to three times greater (1). The bran layers that are
milled from brown rice to obtain white rice account for the higher nutritive
content of brown rice. Besides being nutrient-rich, these bran layers also
have hypocholesterolemic properties (2-5). The cholesterol-lowering activ-
ity of the bran is believed to be associated with unsaponifiables found in
processed rice bran oil (6-8).
In spite of its nutritional value, consumption of brown rice worldwide
has been limited. Only 3% of the rice eaten in the United States is
brown rice (9). A major deterrent to greater use of brown rice is its short
shelf life of 3-6 months due to rancid off-flavors and off-odors being
imparted to the rice as its oil rapidly deteriorates. This susceptibility to
rancidity has limited the commercial production, marketing, and con-
sumption of not only brown rice kernels but also of its products-flour,
bran, and oil.

17
18 CHAMPAGNE

II. LIPOLYTIC HYDROLYSIS AND OXIDATION IN BROWN

RICE
The oil in brown rice is subject to hydrolytic and oxidative deterioration, as
depicted in Figure 1. Lipases, both endogenous to the bran and of microbial
origin, catalyze the hydrolysis of kernel oil. In the intact rough rice kernel,
lipases are dormant because the enzyme and its substrate (oil) are not in
contact. Lipases and oil are compartmentalized in the testa layer and in the
aleurone and germ (10), respectively, as depicted in Figure 2. Dehulling rice
disrupts these outer layers, oil diffuses to make contact with lipases, and the
hydrolysis of triglycerides to free fatty acids (FFA) readily proceeds. Lipases
produced by mold and bacteria located on kernel surfaces also have access to
bran oil following dehulling and promote its hydrolysis (11).
The rate of FFA formation in brown rice depends on the extent of
surface disruption, moisture and microflora contents, and temperature of
storage. Typically the level of FFA in brown rice after 6 months of storage
ranges from 6 to 25% (12). In brown rice flours and bran, oil and lipases
are mingled and the rate of FFA formation is very high; approximately 30%
of the oil can be converted to FFA within a week under high humidity and
temperature (13).
The contribution of FFA to the flavor of brown rice is limited to possibly
a soapy taste (14). The off-flavors and off-odors associated with rancidity
result from oxidative and not hydrolytic deterioration of bran oil. Oxida-

lipases
RICE BRAN OIL .... ,. FREE FATTY ACIDS

HYDROPEROXIDES
•ii
'
SECONDARY OXIDATION PRODUCTS
diperoxides, aldehydes, semi- aldehydes,
alcohols, ketones, acids, epoxides, cyclic
fatty acid monomers, dimers, polymers,
etc.

Figure 1 Routes of hydrolytic and oxidative deterioration of rice bran oil.

BROWN RICE STABILIZATION 19

Lemma (Hull)

Palea (Hull)

Embryo (Germ)

Figure 2 Cross-section of a rice kernel.

tive deterioration can be either enzyme-catalyzed or nonenzymatic (see

Fig. 1). Enzymatic oxidation in brown rice is primarily attributed to
lipoxygenase, an enzyme found in the germ (15). Lipoxygenase catalyzes
the oxidation of free unsaturated fatty acids to hydroperoxides, which react
to yield products (e.g., aldehydes, hydroxy acids) that impart off-flavors
and off-odors. Because lipoxygenase preferentially acts on free unsaturated
fatty acids, the extent of oxidation initiated by this enzyme depends on the
amount of substrate made available from hydrolytic degradation of oil (16).
Thus, lipolytic hydrolysis may predispose the oil in brown rice to subse-
quent lipoxygenase-initiated oxidation.
Nonenzymatic oxidation is inhibited by antioxidants present in the bran,
i.e., tocopherols, and is slow in undamaged rice grains (17,18). Metal ions
naturally present in the bran or introduced by contamination from shelling
equipment catalyze nonenzymatic oxidation. Light, high-energy radiation,
and heat also act as catalysts. Nonenzymatic oxidation can occur by free
radical (autoxidation) or photo-oxidation routes. The free radical route
depends on the production of free radicals from lipid molecules via their
interaction with oxygen in the presence of a catalyst. Hydroperoxides are
the initial reaction products. The rate at which lipids undergo autoxidation
20 CHAMPAGNE

increases with the degree of unsaturation in the acyl chain (14). The fatty
acids can be in the unesterified form (free fatty acid) or as esters. In the
photo-oxidation route, a photosensitive molecule (e.g., riboflavin, heavy
metal ions) is converted to an excited state by absorbing light. The excited
sensitizer reacts directly or indirectly with oxygen to form singlet oxygen,
which reacts with fatty acids forming hydroperoxides. As for enzymatic
oxidation, the hydroperoxides formed during nonenzymatic oxidation yield
products in subsequent reactions that cause objectionable flavors and odors
in brown rice.
The susceptibility of brown rice to nonenzymatic oxidation may be in-
creased by treatments designed to stabilize rice with respect to lipolytic
hydrolysis. Processes involving heat, moisture, or solvents lead to redistri-
bution of oil, destruction of natural antioxidants, and increased surface
area exposed to oxygen (16). These processes also denature the hemo-
proteins catalase and peroxidase found in brown rice kernels, allowing
exposed heme iron to initiate oxidation (19).

III. STABILIZING BROWN RICE TO LIPOLYTIC

HYDROLYSIS
Processes developed for stabilizing brown rice to lipolytic hydrolysis follow
three approaches: (a) heat denaturation and inactivation of kernellipases,
(b) extraction with an organic solvent to remove "freed" oil that serves as a
substrate for lipase, and (c) ethanolic denaturation and inactivation of bran
lipases and lipase-producing bacteria and mold. Table 1 lists processes
using these three approaches and characteristics of their products. An over-
view of these processes and their products follows.

A. Processes Using Heat

Technology developed during the past 30 years for producing quick-
cooking white rice products has been expanded to apply to quick-cooking
brown rice products. These processes not only reduce the cooking time of
brown rice but also stabilize it to lipolytic hydrolysis.
Precooking processes used for producing stable, quick-cooking brown
rice products are variations of the Ozai-Durrani "soak-boil-steam-dry" pro-
cess (20), which was used by General Foods Corporation in the late 1940s
to make the first quick-cooking white rice product, Minute Rice. In these
patents brown rice is soaked, cooked by steaming (21,22), boiling (23), or
baking (24), and dried. Hydration and gelatinization is accomplished by
two or more successive steps. Precooking processes fully gelatinize kernel
starch and give kernels porous structures allowing rapid rehydration. Ker-
 t::=
Table 1 Processes for Stabilizing Brown Rice to Lipolytic Hydrolysis and Characteristics of Their Products ~
~
Partial loss of z

Process
Gelatinized
starch Oil
B
Vitamins Minerals
Kernel appearance
(brown rice)
Quick-
cooking
n=
t!j
rJl
Heat
Soaked-Cooked-Dried Yes No Yes Yes porous, fissured, possibly Yes ~
~

enlarged
Dry heat No No No No endosperm opaque, chalky; Yes ~
(dextrinized) (low kernel fissured, at high ~
~

temp.) temp. puffed 0

Parboiling Yes No Yes Yes endosperm glassy and trans- No
z
lucent, light yellow or am-
ber; kernels enlarged and
darkened
Blanching rough rice with heated, No No No No not described No
moist gas
Organic solvent extraction of freed
oil
Hexane extractions No Yes Yes No slightly abraded No
Ethanolic denaturation of lipases
24°C liquid ethanol No No Yes No slightly abraded No
(<10%)
Fresh 70°C liquid ethanol No Yes Yes No slightly abraded No
Recycled 70°C liquid ethanol No No Yes No slightly abraded No
Ethanol vapors No No No No same as untreated rice No
to
~
22 CHAMPAGNE

nel lipases are inactivated by the hydrothermal treatments. Losses of

water-soluble vitamins and minerals by dissolution occur during soaking
and cooking steps. Roberts et al. (23) report losses of thiamin, riboflavin,
potassium, and iron in their "soaked-boiled-dried" brown rice product of
approximately 31, 67, 22, and 24%, respectively. Protein, fat, fiber, and
niacin levels in this product were comparable to those of raw brown rice.
Dry heat treatments subject brown rice to a hot air stream (90-315°C) fol-
lowed by a cool air stream (25-27). Dry heat processes fissure brown rice ker-
nel surfaces, which facilitates cooking by allowing rapid penetration of water.
Lipases are heat denatured and inactivated. The endosperm is left opaque
and chalky (26). Kernel starch is probably dextrinized to some extent but is
not gelatinized (28). Nutrients are not lost and, if the temperature is not too
high, vitamins are not destroyed (26). At high processing temperatures, the
brown rice produced has a puffed appearance and a toasted flavor (26).
Brown rice products produced from rough rice subjected to parboiling or
blanching with heated, moist gas (air, nitrogen, carbon dioxide) are also
stable to lipolytic hydrolysis (29-32). Cooking time is not reduced by these
processes. Parboiling is a process consisting of steeping rough rice, steaming
it to gelatinize the starch, and then slowly drying the kernels. Kernellipases
are partially or totally inactivated by parboiling during the steaming step
and original FFA content is lowered (29,30,33), as illustrated in Figure 3.

80 0

70 0

~ 60
(3
<50
>-
1- 40
!;;:
u.:
w 30
w
a:
u. 20
c/!.

0~~~~~-=--~-.L~-~-~--=~=-~-~-~--
0 10 20 30 40 50

DAYS

Figure 3 Free fatty acid (FFA) formation in bran from raw (0), soaked (70°C, 3
hr; D), steamed (100°C, 5 min; x ), and parboiled (6) rice during storage at 37°C.
FFA is expressed as percent of kernel oil. (From Ref. 30.)
BROWN RICE STABILIZATION 23

During the parboiling process, water-soluble vitamins and minerals diffuse

from the bran layers into the endosperm, while oil from disrupted bodies in
the aleuorone and germ migrates outward as a band (34-36). Total B vita-
min contents in brown rice are reduced by parboiling due to dissolution of
the vitamins during soaking and loss to heat destruction (34,37 ,38). Little or
no loss of mineral contents occurs (39). Oil and protein contents are unaf-
fected by parboiling (35,40,41). However, protein bodies are ruptured dur-
ing parboiling, and protein solubility and extractability decreases ( 40).
Van Atta et al. (31) blanched rough rice with steam at temperatures
ranging from 85 to 100°C for 0.5-15 minutes. Steam deactivated lipases
without gelatinizing kernel starch. A short steam treatment of 5 minutes at
100°C destroyed most of the lipase activity in the kernels; FFA in the oil
increased by approximately 1.7% during 140 days storage compared with
10% for the unsteamed control (32). Figure 3 depicts the development of
FFA in bran prepared from paddy steamed for 5 minutes at 100°C. In
contrast to parboiling, nutrients did not migrate in the kernel during this
process (31).
The value of parboiling, precooking, or dry or moist heat treatments for
stabilizing brown rice to lipolytic hydrolysis is offset by loss of resistance to
oxidation (32,42). Increased susceptibility to oxidation arises from destruc-
tion of heat-labile antioxidants, i.e., tocopherols, heat denaturation of
hemoproteins, and redistribution of oil in the kernels during processing.
Researchers have observed that the tendency of kernel oil to disperse and
move outward increases with temperature (43) and duration (44) of soak-
ing as well as with increase in pressure of steaming (34). This outward
migration of oil leaves a greater proportion of oil on or near kernel surfaces
and exposed to oxygen.

B. Organic Solvent Extraction of "Freed" Kernel Oil

Kester (45) stabilized brown rice to lipolytic hydrolysis by extracting ker-
nels with either room temperature petroleum ether or boiling hexane.
"Freed" oil was extracted from rice by these fat solvents. As discussed
earlier, dehulling rice disrupts the outer bran layers, freed oil diffuses
through the bran making contact with lipases, and lipolytic hydrolysis pro-
ceeds. Solvent extraction of freed oil prevents the substrate from interact-
ing with lipases, thus stabilizing the kernels to lipolytic hydrolysis.
Kester ( 45) found that room temperature petroleum ether extracted
40% of the total oil in brown rice. After a storage period of 80 days, FFA
levels in extracted and control kernels were 3.7 and 22.3% (expressed as%
of oil), respectively. Brown rice kernels extracted with hexane were re-
ported to have a FFA level of 3.8% following one month of storage.
24 CHAMPAGNE

An in-depth investigation of the effects of extracting brown rice kernels

with boiling (68°C) hexane for one hour on nutrient composition and stor-
age stabilities of the kernels and their flours was recently reported (46).
Extraction of brown rice with boiling hexane removed approximately 15%
of the kernel oil. No loss of protein, total dietary fiber, carbohydrate, or
minerals occurred during extraction with boiling hexane. Twenty percent of
the kernel thiamine was lost. Kernel starch was not gelatinized. Extraction
of freshly dehulled brown rice kernels with boiling hexane yielded a prod-
uct with a FFA level approximately 85% that of unextracted kernels. Ex-
traction of kernels with a FFA content of 27% led to a 57% reduction in
FFA.
As initially observed by Kester, boiling hexane was effective in stabiliz-
ing brown rice kernels to lipolytic hydrolysis; the FFA levels in kernels
extracted with boiling hexane increased from 3.3 to 4.3% during 6 months
storage at 36°C, as shown in Figure 4. In contrast, the FFA level of control
kernels increased from 4 to 10% during storage. Flours prepared from
kernels extracted with boiling hexane were only partially stable to lipolytic
hydrolysis. During 6-month storage at 36°C, FFA level in flour prepared
from kernels extracted with boiling hexane increased from 3.3 to 36%,
while that in control flour increased from 4 to 47% (see Fig. 4). When

50
45

(/) 40
c
C3 35
<
30
>-
1-
!;;: 25
u.:
w 20
w
a: 15
u.
"#. 10
5

0
0 2 345 6

MONTHS

Figure 4 Free fatty acid (FFA) formation in kernels extracted with hexane at
68°C (D), flour prepared from extracted kernels (~), control kernels (0), and
control flour ( x) during storage at 36°C. FFA is expressed as percent of kernel oil.
(From Ref. 46.)
BROWN RICE STABILIZATION 25

flour was produced from hexane-extracted kernels, residual oil was min-
gled with lipases allowing lipolytic hydrolysis to readily proceed.
Extracting brown rice kernels with boiling hexane did not affect the
susceptibility of the kernels to oxidation, as determined by comparing the
change in conjugated diene hydroperoxide contents (CDHP) in extracted
and control kernels during 6-month storage.

C. Processes Utilizing Ethanol (EtOH)

Processes for stabilizing brown rice to lipolytic hydrolysis by liquid EtOH
extraction and by treatment with EtOH vapors were recently reported (46-
50). The stabilizing action of EtOH was attributed to (1) ethanolic denatu-
ration of bran lipases with concomitant deactivation and (b) killing of
lipase-producing bacteria and mold located on kernel surfaces.
Liquid EtOH (95%, v/v), at temperatures ranging from 24°C (room
temperature) to 70°C, stabilized brown rice to lipolytic hydrolysis (46-48).
The effectiveness of EtOH in stabilizing kernels increased with higher
extraction temperatures and longer extraction times, as illustrated in Fig-
ure 5. Brown rice kernels extracted with 70°C EtOH for 60 minutes were
the most stable to lipolytic hydrolysis; the FFA level in these kernels in-
creased from 1.0 to 1.4% during 6 months of storage at 36°C. To produce

30

25
en 0
0
~ 20
>-
I-
15
~
u.: 0

w
w 10
a:
u.
# 5

0
0 2 3 4 5 6

MONTHS

Figure 5 Effect ofEtOH extraction time and temperature on free fatty acid (FFA)
formation in brown rice kernels stored at 36°C. Kernels were extracted at 24°C for 10
(.0,.) or 60 minutes (D) or at 70°C for 10 ( x) or 60 minutes (\7). Control (0) kernels
were untreated. FFA is expressed as percent of kernel oil. (From Ref. 48.)
26 CHAMPAGNE

stabilized kernels, liquid EtOH needs to make contact with lipases that
have been activated during dehulling of the kernel and with lipase-
producing bacteria and mold found on kernel surfaces. Since these interac-
tions are on or near the kernel surface, high extraction temperatures and/or
long extraction times providing "deep" penetration of EtOH into the ker-
nel are not required. To produce stabilized flours from extracted kernels,
EtOH must penetrate the testa layer and denature all or nearly all of the
lipase. This requires higher extraction temperatures and longer extraction
times. As indicated by increases in FFA levels during storage (Fig. 6), an
extraction temperature higher than 54°C was required to produce highly
stable flours from EtOH-extracted kernels. Flours prepared from kernels
extracted for 20 minutes or longer at 70°C had high stability to lipolytic
hydrolysis.
Vapors from boiling aqueous EtOH (95%, v/v) were also effective in
stabilizing brown rice kernels and their flours to lipolytic hydrolysis, as
depicted in Figure 7 (49,50). During 6 months of storage at 36°C, FFA
levels in brown rice kernels treated with EtOH vapors for 3 or 5 minutes
increased from 3.0 to 3.9 and 3.6%, respectively. FFA content did not
change in kernels treated with vapors for 10 minutes, while that of control
kernels increased from 3 to 28%. Flours produced from kernels treated

100

U) 80
0
0
~
>- 60-
.....
!cc
u.;
w 40
w
0::
u.
';/!. 20

0
0 2 3 4 5 6

MONTHS

Figure 6 Effect of EtOH extraction temperature on free fatty acid (FFA) forma-
tion in flours prepared from extracted kernels and stored at 36°C. Kernels were
extracted at 24°C (.6.), 46°C (D), 54°C ( x ), or 70°C (\7) for 60 minutes. Control
(0) was flour made from untreated kernels. FFA is expressed as percent of kernel
oil. (From Ref. 48.)
BROWN RICE STABILIZATION 27

30

5
en
c
0<( 0
0 2 3 4 5 6
>-
1-
~ 100
u:
w 90
w
a: 80
u.
?Ji! 70
60-
0
8

5
MONTHS

Figure 7 Free fatty acid formation (FFA) in (A) brown rice kernels treated with
EtOH vapors, heat-treated (6), or untreated (0) and in (B) flours prepared from
these kernels during storage at 36°C. Kernels were treated with EtOH vapors for 3
(D), 5 ( x ), or 10 (\7) minutes. Heat-treated kernels were heated at 78°C for 10
minutes without vapor treatment. FFA is expressed as percent of kernel oil. (From
Ref. 50.)
28 CHAMPAGNE

with EtOH vapors had low residual lipase activities. Following 5 months of
storage at 36°C, FFA levels in flours prepared from kernels treated with
vapors for 3, 5, and 10 minutes increased from 3 to 9, 7, and 6%, respec-
tively. In contrast, the FFA level in flour prepared from control kernels
increased from 3 to 80%.
Stabilization with EtOH, as a liquid or vapor, increased the susceptibil-
ity of the oil in brown rice to oxidative deterioration, as indicated by
increases in CDHP contents during storage (Fig. 8). When using liquid
EtOH, the higher the temperature of extraction, the more susceptible the
kernels were to oxidative deterioration during storage. Extraction time was
not a factor. Microscopic examinations (46,50) of kernel surfaces revealed
that liquid and vapor treatments with EtOH disturbed the caryopsis coat
and increased kernel porosity by causing increased waviness and narrow,
superficial transverse fissures, respectively. The higher the temperature of
liquid EtOH, the greater was its penetration into the kernel, and the more
disruptive it was to the caryopsis coat. By disrupting and increasing the
porosity of the caryopsis coat, EtOH treatment leaves kernel oil exposed to
oxygen and susceptible to oxidation. EtOH treatments also caused both
ethanolic and heat denaturation of the hemoproteins, catalase and peroxi-
dase. Unfolding of these enzymes causes greater exposure of the heme
groups to oil, allowing heme iron to initiate oxidation (19). Nonenzymatic

3
2.8
~
2.6
.ri 2.4
-c:i
Q)
2.2
.2
....
2
1.8
.....U)
Cl
1.6
Q)
1.4
0
E 1.2
~ 1
a.. 0.8
J: 0.6
c
(.)
<] 0.2
0
0 2 3 4 5 6

MONTHS

Figure 8 Change in conjugated diene hydroperoxide (6 CDHP) content in

kernels treated with liquid EtOH at 24°C for 10 (6) or 60 (0) minutes or at 70°C
for 10 ( x) or 60 (\7) minutes or with EtOH vapors for 5 ( 0) or 10 ( +) minutes.
Control (0) kernels were untreated. (From Refs. 48, 50.)
BROWN RICE STABILIZATION 29

oxidation initiated by heme groups of denatured hemoproteins has been

observed to proceed more rapidly than enzymatic oxidation initiated by
active hemoproteins and lipoxygenase (19,51).
Full fat ( <3% oil loss) products are produced by extracting brown rice
with EtOH at 24°C or at higher temperatures with recycled EtOH (oil
saturated) and by treating brown rice with vapors from boiling EtOH.
Extraction with fresh EtOH at temperatures higher than 24°C removes up
to 15% of the kernel oil. Little or no loss of protein, dietary fiber, carbohy-
drates, or minerals occurred during EtOH extraction or vapor treatments.
Thiamine was retained in kernels extracted at temperatures ::::54°C and in
kernels treated with EtOH vapors. Sixty-three percent of the thiamine in
kernels extracted at 70°C was lost. Bacterial and mold populations on
EtOH-extracted and vapor-treated brown rice were very low or not count-
able. Starch was not gelatinized during processing with EtOH. EtOH-
stabilized brown rice has the general appearance and cooking properties of
natural brown rice.

IV. FLOUR, BRAN, AND OIL FROM STABILIZED BROWN

RICE KERNELS
Brown rice, stabilized by processes that eliminate all or nearly all of the
kernel lipase activity, is a suitable starting material for stable flour or stable
bran. Brown rice flours produced from rice stabilized by methods that
render the starch gelatinized (e.g., precooking processes asnd parboiling)
have functional properties, unlike raw flours, and have limited use in bak-
ing (52). Stable, ungelatinized brown rice flours are produced by grinding
kernels that are extracted with EtOH at elevated temperatures or treated
with EtOH vapors (46-50). Processes subjecting brown rice kernels to dry
heat (25-27) or rough rice to moist heat at elevated temperatures (> 100°C)
(31) are also suitable for producing stable brown rice flours with ungela-
tinized starch. These processes may partially dextrinize the starch and im-
part a toasted flavor to the brown rice flour. Residual lipase activity is too
high in brown rice kernels stabilized by hexane extraction to allow stable
flour to be produced.
Food-quality, stabilized bran is produced by milling kernels that are ex-
tracted with EtOH at elevated temperatures or treated with EtOH vapors
(46-50). Protein, dietary fiber, carbohydrate, and mineral contents of bran
from EtOH-treated products are the same as those of bran from untreated
brown rice. Oil and thiamine are fully retained in bran from kernels treated
with EtOH vapors and are lower in bran produced from kernels extracted
with EtOH at elevated temperatures. Bran from parboiled rice contains
more oil than untreated bran and lower levels of minerals and B vitamins
30 CHAMPAGNE

(34). Starchy endosperm contamination is low in parboiled-rice bran, since

the endosperm of parboiled rice has a greater resistance to milling, allowing
the bran to effectively separate from it (34). Although parboiling inactivates
kernellipases, bran from parboiled rice is stable to lipolytic hydrolysis only if
properly dried and stored (53). If not dried, the high moisture content in
parboiled-rice bran causes the bran to be susceptible to insect and fungal
infestation, which leads to FFA formation during storage (54). Stabilized
bran is also obtained by milling brown rice produced from the VanAtta et al.
process (31) in which rough rice is treated with moist, heated gas. Brown rice
produced by precooking (quick-cooking) processes has altered kernel sur-
faces (fissured, porous, brittle, puffed) and is not suited for milling.
The rapid manner in which lipases hydrolyze oil in unstabilized rice bran
does not provide industry with the time needed to transport the bran and
extract oil low in FFA. Losses for potentially edible oil during refining are
two to three times the FFA content of the oil (13). Bran stabilized by the
brown rice treatments described above is suitable for extraction to produce
edible oil. Problems are encountered in extracting oil from the bran of
parboiled rice that are not present when untreated rice bran is used (55-
57). Parboiled bran does not produce hard pellets, as untreated bran does,
which leads to poor extraction and solvent loss (57). Oil obtained from
parboiled bran is difficult to bleach (56). Also, parboiled bran tends to stick
to the steam-jacketed dryer wall, which may cause self-ignition during
preheating (55). Oil from parboiled rice bran has a lower saponification
number and a lower iodine value than does oil from untreated rice (58).
Flour, bran, and oil produced from kernels stabilized by processes using
heat or ethanol are susceptible to oxidative deterioration.

V. STABILIZING BROWN RICE TO OXIDATION

Removal of oxygen prevents oxidative rancidity. However, both enzymatic
and nonenzymatic oxidation require only low levels of oxygen. Technically
and economically it is difficult to achieve and maintain oxygen levels below
1% in food packaging (16). Researchers (12,18,59-61) have demonstrated
that storing brown rice in the dark at low temperatures under modified
atmospheres, e.g., carbon dioxide, or under vacuum slows but does not
prevent oxidation. Adding an iron oxide absorbant to a sealed container
having a high oxygen barrier reduces oxygen levels to 100 ppm or less (62)
and could possibly be a suitable method for preventing oxidation in brown
rice products.
Enzymatic oxidation initiated by lipoxygenase is halted by heat treat-
ments designed to stabilize brown rice to lipolytic hydrolysis. Antioxidants
have little effect on lipoxygenase-initiated oxidation but are effective in
BROWN RICE STABILIZATION 31

inhibiting nonenzymatic oxidation. However, tocopherols and their deriva-

tives, which serve as naturally occurring antioxidants, may be destroyed by
lipoxygenase-catalyzed oxidation and by processing with heat (14). Adding
antioxidants to brown rice products provides protection against oxidative
rancidity. An extract from rosemary leaves at 300 ppm was recently shown
to be effective in inhibiting oxidation in brown rice (62). Metal chelators,
e.g., citric acid, phosphoric acid, ascorbic acid, EDTA, and phytic acid,
prevent iron and other metals from initiating oxidation. Studies are cur-
rently underway in our laboratory at the Southern Regional Research Cen-
ter to evaluate the effectiveness of adding iron chelators and/or antioxi-
dants to EtOH used in brown rice extractions.

VI. NEED FOR STABILIZED BROWN RICE

Two factors will determine the need for stabilized brown rice. The first is
consumer interest in brown rice as a table rice. Through efforts of the U.S.
Rice Council and others, consumers are being made aware of the nutri-
tional benefits of brown rice and their interest in adding this food to the
diet is increasing. However, consumer interest will wane if brown rice does
not retain a fresh, nutty taste and readily develops rancid off-flavors and
off-odors during storage. Without stabilization to hydrolytic and oxidative
deterioration, the nutritional advantage of brown rice will be lost, and its
future as a table rice will be bleak.
A second factor is market demand for products derived from stabilized
brown rice, such as flour, food-grade bran, and oil. Processes using ethanol
are well suited for production of these commodities since kernel compo-
nents, including the oil, show little or no modification from the process. In
the last 2-3 years, processors have shown considerable interest in food uses
of rice bran and oil. Interest has been spurred by the potential LDL
cholesterol-lowering effects of the oil when added in reasonable quantities
to the diet and by the health benefits of dietary fiber in bran. These factors
have also increased the interest in stabilized brown rice flour, particularly
that with ungelatinized starch. Processes that treat brown rice with ethanol
or dry heat or rough rice with moist heat are attractive for the production of
stabilized brown rice flours with ungelatinized starch. The market for
brown rice flour, bran, and oil has the potential to grow rapidly. However,
the growth of this market is dependent upon economically feasible pro-
cesses being available to stabilize brown rice and its products to hydrolytic
and oxidative deterioration, while retaining their nutritional value. Stabi-
lized brown rice and its products have the potential to play a significant role
in increasing the demand for rice in the United States and returning greater
profits to the farmer and rice-processing industry.
32 CHAMPAGNE

REFERENCES
1. Kennedy, B. M. (1980). Nutritional quality of rice endosperm, Rice: Produc-
tion & Utilization (B.S. Luh, ed.), AVI Publishing Co., Westport, CT, p. 439.
2. Kahlon, T. S., Saunders, R. M., Chow, F. 1., Chiu, M. C., and Betschart, A.
A. (1989) Effect of rice bran and oat bran on plasma cholesterol in hamsters,
Cereal Foods World, 34:768.
3. Gerhardt, A. L. (1989). Food Chemical News, Nov. 13.
4. Kestin, M., Moss, R., Clifton, P.M., and Nestel, P. J. (1990). Comparative
effects of three cereal brans on plasma lipids, blood pressure, and glucose
metabolism in mildly hypercholesterolemic men. Am. J. Clin. Nutr. 52:661.
5. Hegsted, M., Windhauser, M. M., Lester, F. B., and Morris, S. K. (1990).
Stabilized rice bran and oat bran lower cholesterol in humans. FASEB 4:368a.
6. Sharma, R. D., and Rukmini, C. (1986). Rice bran oil and hypocho-
lesterolemia in rats. Lipids, 21:715.
7. Nicolosi, R. J., Ausman, L. M., and Hegsted, D. M. (1989). Lipoprotein
levels in monkeys fed a diet containing rice oil. Circulation, 80:86A.
8. Applewhite, T. H. (ed.) (1990). Unsaponifiables in rice bran oil under study,
Inform, 1:831.
9. Childs, N. W. (1991). Changes in U.S. rice distribution patterns, 1988/1989.
Cereal Foods World, 36:719.
10. Shastry, B. S., and Raghavendra Rao, M. R. (1971). Studies on rice bran
lipase. Ind. J. Biochem. Biophys., 8:327.
11. DeLucca, A. J., Plating, S. J., and Ory, R. L. (1978). Isolation and identifica-
tion of lipolytic microorganisms found on rough rice from two growing areas.
J. Food Prot., 41:28.
12. Sharp, R. N., and Timme, L. K. (1986). Effects of storage time, storage
temperature, and packaging method on shelf life of brown rice. Cereal Chern.,
63:247.
13. Enochian, R. V., Saunders, R. M., Schultz, W. G., Beagle, E. C., and
Crowley, P. R. (1980). Stabilization of rice bran with extruder cookers and
recovery of edible oil: A preliminary analysis of operational and financial
feasibility, U.S. Dept. of Agriculture Marketing Research Report 1120.
14. Barnes, P., and Galliard, T. (1991). Rancidity in cereal products. Lipid
Techno/., 3:23.
15. Ohta, H., Ida, S., Mikami, B., and Morita, Y. (1986). Purification and charac-
terization of rice lipoxygenase component 3 from embryos. Agric. Bioi.
Chern., 50:3165.
16. Galliard, T. (1989). Rancidity in cereal products, Rancidity in Foods (J. C.
Allen and R. J. Hamilton, eds.), Elsevier Applied Science, New York, p. 141.
17. Gyorgy, P., and Tomarelli, R. (1943). Antioxidant activity in sources of the B
vitamins. J. Bioi. Chern., 147:515.
18. Sowbhagya, C. M., and Bhattacharya, K. R. (1976). Lipid autoxidation in
rice. J. Food Sci., 41:1018.
19. Eriksson, C. E., Olsson, P. A., and Svensson, S. G. (1969). Oxidation offatty
acids by heat treated hemoproteins, Lipids, 5:365.
BROWN RICE STABILIZATION 33

20. Ozai-Durrani, A. K. (1948). Quick-cooking rice and process for making

same. U.S. Patent 2,438,939, April6.
21. Ozai-Durrani, A. K. (1956). Quick-cooking rice and process therefor. U.S.
Patent 2,740,719, April3.
22. Miller, F. J. (1963). Process of preparing a quick-cooking brown rice. U.S.
Patent 3,086,867, April23.
23. Roberts, R. L., Carlson, R. A., and Farkas, D. F. (1980). Preparation of a
quick-cooking brown rice product using a centrifugal fluidizer bed drier. J.
Food Sci. 45:1080.
24. McCabe, D. (1976). Process for preparing quick-cooking brown rice and the
resulting product. U.S. Patent 3,959,515, May 25.
25. Alexander, W. P. (1954). Process of preparing quick-cooking rice. U.S. Patent
507,242, November 9.
26. Bardet, G. V., and Giesse, R. C. (1961). Processing of brown rice. U.S.
Patent 2,992,921, July 18.
27. Tadahiko Hirokawa, Y., Katsumoto Oki, M., Yoh Kumagai, Y., and Hajime
Sasaki, A. (1986). Method for heat treating brown rice, U.S. Patent 4,582,713,
April15.
28. Roberts, R. L. (1972). Quick-cooking rice, Rice Chemistry and Technology
(D. F. Houston, ed.), American Association of Cereal Chemists, St. Paul,
MN, p. 381.
29. Shaheen, A. B., El-Dash, A. A., and El-Shirbeeny, A. E. (1975). Effect of
parboiling of rice on the rate of lipid hydrolysis and deterioration of rice bran.
Cereal Chern., 52:1.
30. Viraktamath, C. S., and Desikachar, H. S. R. (1971). Inactivation of lipase in
rice bran in Indian rice mills. J. Food Sci. Techno[., 8:70.
31. VanAtta, G. R., Kester, E. B., and Olcott, H. S. (1952). Method of stabiliz-
ing rice products. U.S. Patent 2,585,978, February 19.
32. Houston, D. F., Hunter, I. R., and Kester, E. B. (1951). Effect of steaming
fresh paddy rice on the development of free fatty acids during storage of
brown rice. Cereal Chern., 28:394.
33. Anthoni Raj, S., and Singaravadivel, K. (1982). Reduction in free fatty acids
due to parboiling of paddy. J. Food Sci., 47:692.
34. Padua, A. B ., and Juliano, B. 0. (1974). Effect of parboiling on thiamine,
protein and fat of rice. J. Sci. Food Agric., 25:697.
35. Benedito de Barber, C., Martinez, J., and Barber, S. (1977). Effects of parboil-
ing processes on the chemical composition and nutritional characteristics of rice
bran, Rice By-Products Utilization, Vol. IV (S. Barber and E. Tortosa, eds.),
Int. Conf. Valencia, Spain, 1974. Inst. Agroquimica Tecnol. Alimentos, Valen-
cia, Spain, p. 121.
36. Mahadevappa, M., and Desikachar, H. S. R. (1968). Some observations on
histology of raw and parboiled rice. J. Food Sci. Techno[., 5:72.
37. Done, J. (1949). The effects of various treatments of paddy rice on the aneurin
and nicotinic acid content of milled grain. Br. J. Nutr., 3:335.
38. Subba Rao, P. V., and Bhattacharya, K. R. Effect of parboiling on thiamine
content of rice. J. Agric. Food Chern., 14:479.
34 CHAMPAGNE

39. Bhattacharya, K. R., and Ali, S. Z. (1985). Changes in rice during parboiling
and properties of parboiled rice, Advances in Cereal Science and Technology,
Vol. VII (Y. Pomeranz, ed.), American Association of Cereal Chemists, St.
Paul, MN, p. 105.
40. Raghavendra Rao, S. N., and Juliano, B. 0. (1970). Effect of parboiling on
some physiochemical properties of rice. J. Agric. Food Chern., 18:289.
41. Sondi, A. B., Reddy, I. M., and Bhattacharya, K. R. (1980). Effect of process-
ing conditions on the oil content of parboiled-rice bran. Food Chern., 5:277.
42. Houston, D. F., Hunter, I. R., McComb, E. A., and Kester, E. B. (1954).
Deteriorative changes in the oil fraction of stored parboiled rice. J. Agric.
Food Chern., 2:1185.
43. Chakravarty, H. B., and Ghose, T. K. (1966). Studies on the hydration of
Indian paddy. IliA: Efficiency of elevated temperature on composition of oil
in bran. Souvenir, Seminar on Modern Technology of Rice Milling, East India
Rice Mills Association, Calcutta, p. 21.
44. Vasan, B.S., Iengar, N. G. C., Subramanyam, T.V., Chandrasekaran, R., and
Subrahmanyan, V. (1971). Effect of processing on the movement of oil in the
rice kernel and its relation to the oil content of paddy and the yield of oil from
rice bran, Interregional Seminar on the Industrial Processing of Rice, Food and
Agriculture Organization, ECAFE, and Government of India, Madras.
45. Kester, E. B. (1951). Stabilization of brown rice. U.S. Patent 2,538,007,
January 16.
46. Champagne, E. T., Hron, Sr., R. J., and Abraham, G. (1991). Stabilizing
brown rice products by aqueous ethanol extraction. Cereal Chern., 68:267.
47. Champagne, E. T., Hron, Sr., R. J., and Abraham, G. (1990). Stabilizing
unmilled brown rice by ethanol extraction. U.S. Patent Application 07/
557,822, July 26.
48. Champagne, E. T., and Hron, Sr., R. J. (1992). Stability of ethanol-extracted
brown rice to hydrolytic and oxidative deterioration. J. Food Sci., 57: 433.
49. Champagne, E. T., Hron, Sr., R. J., and Abraham, G. (1991). Stabilizing
unmilled brown rice by ethanol vapors. U.S. Patent Application 07/694,964,
May2.
50. Champagne, E. T., and Hron, Sr., R. J. (1992). Stabilizing brown rice to
lipolytic hydrolysis by ethanol vapors. Cereal Chern., 69: 152
51. St. Angelo, A. J., Ory, R. L., and Brown, L. E. (1972). A comparison of
minor constituents in peanut butter as possible sources of fatty acid peroxi-
dation. J. Am. Peanut Res. Educ. Assoc., 4:186.
52. Bean, M. M., and Nishita, K. D. (1985). Rice flours for baking, Rice Chemis-
try and Technology (B. 0. Juliano, ed.), American Association of Cereal
Chemists, St. Paul, MN, p. 539.
53. Kumaravel, S., Singaravadivel, K., Vasan, B.S., and Anthoni Raj, S. Storage
of parboiled bran. J. Oil Techno/. Assoc. India, 12:49.
54. Bhattacharya, K. R. (1985). Parboiling of rice, Rice Chemistry and Technol-
ogy (B. 0. Juliano, ed.), American Association of Cereal Chemists, St. Paul,
MN, p. 289.
BROWN RICE STABILIZATION 35

55. Pe, M. (1977). Parboiled rice bran as a source of edible oil in Burma. I.
Storage of stabilized and parboiled rice bran. II. Processing problems, Pro-
ceedings Rice By-Products Utilization Int. Conference, Valencia, Spain, 1974,
Vol 3. (S. Barber and E. Tortosa, eds.), Inst. Agroquim. Tecnol. aliment.,
Valencia, p. 39.
56. Pe, M. (1977). Bran oil extraction, Working Papers, Action Oriented Field
Workshop for Prevention of Post-Harvest Rice Losses, Alor setar, Kedah,
Malaysia, 1977 (I. Shamsudin and F. Abdullah, eds.), Res. Planning Div.
Lembaga Padi dan Beras Negara, Kuala Lumpur, p. 452.
57. Sarda, P. S. (1977). Rice bran oil from parboiled bran (edible grade) and its
extraction, Proceedings Rice By-Products Utilization Int. Conference, Valen-
cia, Spain, 1974 (S. Barber and E. Tortosa, eds.), Inst. Agroquim. Tecnol.
Aliment., Valencia, p. 53.
58. Kumaresan, K., and Sree Ramulu, U.S. (1978). Factors influencing the quan-
tity and quality of rice bran oil. III. Parboiling. Riso, 27:341.
59. Ory, R. L., DeLucca, A. J., St. Angelo, A. J., and Dupuy, H. P. (1980).
Storage quality of brown rice as affected by packaging with and without car-
bon dioxide. I Food Prot., 43:929.
60. Mitsuda, H., Kawai, F., and Yamamoto, A. (1972). Underwater and under-
ground storage of cereal grains. Food Technol., 26:50.
61. Santoprete, G. (1980). Preservation of slightly milled rice, Riso, 29:61.
62. Loliger, J. (1989). Natural antioxidants, Rancidity in Foods (J. C. Allen, ed.),
Elsevier Applied Science, New York, p. 105.
 3
Storage Stability of Extrusion Stabilized and
Parboiled Rice Bran
Don R. McCaskill and Frank T. Orthoefer
Rice/and Foods, Inc., Stuttgart, Arkansas

I. INTRODUCTION
A. Quality of Rice Bran
Rice bran as a food is a rich source of nutrients. Several obstacles have
hindered its utilization. In rice milling, the historical emphasis has been on
the economical production and maximum output of whole grains. Bran
quality has typically received little attention. Some mills were set up in a
"single-break" design, which involves the simultaneous removal of hulls,
bran, and germ. Even in "double-break" systems, especially those involv-
ing friction-type bran removal, relatively high levels of hulls are often
allowed to enter the bran removal step. The presence of hulls with a high
silica and cellulose content, even at low levels in the bran, severely limits its
acceptability as human food. Carefully controlling the quality of rough rice
used in milling results in bran composition with limited contamination by
rice endosperm from immature, damaged, or otherwise weak grains. Cal-
cium carbonate or ground limestone is often used as a milling aid with
parboiled rice because of the increased difficulty in achieving adequate
bran removal. This also becomes a contaminant in the resultant bran.

B. Dietary Considerations
Recent studies have indicated that a diet supplemented with rice bran may
be an effective means of reducing serum cholesterol (1,2). As a result,

37
38 MCCASKILL AND ORTHOEFER

interest in the use of rice bran as a food ingredient has increased. Rice
bran, when compared to other food fiber sources that reduce cholesterol, is
a rich nutrient source. Rice bran, like most cereal brans, contains an-
tinutritional properties, including phytin content, presence of trypsin inhibi-
tor, potential antigenicity, and mineral binding (3). There are indications
that the heat treatment associated with stabilization may minimize or elimi-
nate at least some of these characteristics. In feeding trials, Sayre et al. (4)
observed significant improvement in weight gain of chicks fed diets contain-
ing stabilized versus raw rice bran.

C. Mechanism of Deterioration
The greatest restriction to the use of rice bran as a food ingredient, or even
as a source of edible oil, is its reputation for being highly unstable in
storage. At 18-22% lipid, it contains the necessary substrate for rancidity
development. Indeed, rice bran possesses all the components necessary for
rapid flavor deterioration.
There are two types of deterioration of lipids in the bran layer: hydro-
lytic and oxidative rancidity. Hydrolytic rancidity, catalyzed by lipase, is
primarily responsible for deterioration of bran.
Lipase activity increases early during seed development and reaches a
maximum 12 days after flowering and levels off to 50% of maximum activ-
ity (5). Lipase activity is not limited to rice. Oat lipase has higher activity
than rice lipase. Lipase is concentrated in the bran and germ layers. Rice
bran lipase has an optimum pH between 7.5 and 8.0 and an optimum
temperature of 37°C (6). It is inactivated by heating for 15 minutes at 60°C.
Bran lipase, similar to pancreatic lipase, preferentially hydrolyzes fatty acid
ester bonds at the 1,3 positions. Other lipolytic acylhydrolase activity oc-
curs with galactolipase and phospholipase and their isozymes (7).
Lipoxygenase activity, which usually results in oxidative rancidity, in-
creases to 20 days after flowering, then levels off. Lipoxygenase shows
lower activity in the endosperm than in the bran layer. It promotes oxida-
tion of unsaturated fatty acids, leading to the formation of shorter-chain
aldehydes and ketones associated with rancid odors and flavors. Isozymes
are present in both the germ and bran layers. Lipoxygenase produces
mainly the nine hydroperoxides from linoleic acid. The optimum pH for
the major lipoxygenase of rice bran is 8.5. Lipoxygenase activity is gener-
ally not associated with oil degradation in rice bran.
DeLucca et al. (8) reported that approximately 10% of the bacterial
population of rough rice samples grown in Arkansas and Louisiana were
lipolytic and that all molds isolated from the samples demonstrated at least
some lipolytic activity. Finally, the actions involved in rice milling bring the
STORAGE STABIUTY OF RICE BRAN 39

reactive components (lipids, enzymes, and oxygen) together in a homoge-

neous mixture. After milling, bran quality degradation will occur rapidly
unless it is stabilized.

IT. STABILIZATION RESEARCH

A. General
Sayre et al. (9) have reviewed the methods used for the stabilization of rice
bran. These include dry heat, wet heat, and extrusion methods. Most rice
bran stabilization research has sought to develop methods to prepare the
bran for economically feasible extraction of oil. Studies have repeatedly
shown that free fatty acids (FFA) develop rapidly in untreated rice bran or
loosely milled rice during the first few days or weeks after milling (10-14).
This change profoundly affects the value of bran for the extraction of oil.
As FFA content rises, oil-refining losses increase proportionately. While
oxidative changes also negatively impact oil quality, these changes are not
as rapid or obvious. Therefore, primary attention in stabilization efforts
has been paid to destruction or inhibition of lipase, the enzyme responsible
for FFA development.

B. Effects of Rice Processing

Parboiled rice has been shown to exhibit a reduced level of FFA as com-
pared to unprocessed rice (15) and to be more resistant to the development
of FFA during storage (10,12). It is generally accepted that lipase enzymes
are destroyed in the process, due to the treatment times and temperatures
typically involved in parboiling. Some preformed free fatty acids appar-
ently are leached out, oxidized, and/or complexed with amylose, which
accounts for the initial reduction. The oil in parboiled rice and in bran from
parboiled rice has been shown to be highly susceptible to oxidative deterio-
ration (13,16). This is generally attributed to destruction or removal of
natural antioxidants during parboiling.
A process for stabilizing brown rice by aqueous ethanol treatment was
recently described by Champagne et al. (17). Treatment at 78°C inactivated
lipase, but accelerated oxidative changes as measured by development of
conjugated diene hydroperoxides. This work is discussed in greater detail
in Chapter 2.

C. Enzyme Inhibition
Enzymes are typically inactivated or denatured through the application of
heat or by chemical treatment. Enzyme activity may be inhibited by modify-
40 MCCASKILL AND ORTHOEFER

ing the moisture, pH, or temperature of the sample. Drying to low mois-
ture levels was shown to limit lipase activity in bran (10) and rice polish
(18), but actually hastened oxidation (13,18). Adjustment of pH requires
the addition of undesirable acidulants, alters the flavor profile, necessitates
elaborate processing steps, and ultimately reduces the overall compatibility
of bran as a food ingredient. Storage of bran (10) or brown rice (19) at
refrigerated temperatures has been found to prolong storage life. How-
ever, the refrigerated shelf life is still too short to enable practical utiliza-
tion of rice bran as a food ingredient. In most cases, the cost of refrigerated
storage is prohibitive as well.

D. Requirements of Food Ingredients

The ability of rice bran supplements to reduce serum cholesterol triggered
considerable interest within the food industry and prompted several rice-
milling companies to enter the stabilized rice bran market. Continued in-
creases in use of rice bran in packaged consumer goods is likely. Since its
cholesterol-reducing properties were first reported, several major food
companies have either introduced or announced plans for products contain-
ing rice bran. Breakfast cereals and bread containing rice bran are being
produced and marketed. In addition, packaged rice bran is now available
in many supermarkets.
For stabilized rice bran, insufficient information is available to ade-
quately answer questions regarding optimum packaging systems, antici-
pated shelf life, and ingredient storage requirements so as to avoid unex-
pected problems in the marketplace. Additionally, both extrusion-stabilized
and parboiled rice bran are being marketed. Due to the inherent differences
in lipolytic and oxidative characteristics of these brans, their performance in
commercial distribution would be expected to differ. Storage trials were
initiated by a Riceland Foods research group to develop the data required to
resolve these questions.

E. Lipolysis and Oxidative Deterioration

Both lipolysis and oxidative deterioration of the lipids in bran occur during
storage. In the presence of active lipase enzymes, lipolysis occurs very
rapidly. The free fatty acid content of rice bran is commonly determined by
simple titration. Long-chain free fatty acids formed during lipolysis of rice
bran oil are not responsible for typical rancid odors but are believed to
impart an acidic or "soapy" taste.
Oxidative deterioration of fats generally occurs by a free radical mecha-
nism. In the initiation step, an active hydrogen, especially in the presence of
a metal catalyst such as copper, is removed from a triglyceride to yield a free
STORAGE STABILITY OF RICE BRAN 41

radical. The free radical can then combine with oxygen to form a peroxide-
free radical, which removes hydrogen from another unsaturated molecule to
yield a peroxide and a new free radical. This propagation stage becomes a
chain reaction and may continue until the free radicals react with each other
to form inactive products, leading to termination of the cycle.
Peroxides are the primary oxidation products. Peroxides are quite unsta-
ble and decompose into a range of secondary products including aldehydes,
alcohols, and ketones, which produce the typical rancid oil aroma. It is only
during the initial stages of oxidative deterioration that peroxide value may
be used to indicate oxidative deterioration. Peroxide value is a term used to
quantify the content, expressed in milliequivalents of peroxide per kilo-
gram of sample, of all substances that oxidize potassium iodide under
specified conditions. In fat and oil testing, these substances are assumed to
be peroxides or similar products of fat oxidation.
Storage and shelf life of fat-bearing foods have been studied by anum-
ber of techniques that rely on a combination of analytical and sensory
methods covering a series of storage times and conditions (20). Sharp and
Timme (19) coupled HPLC analysis for conjugated diene hydroperoxide
(CDHP) with a duo-trio sensory method for off-odor and off-flavor in
studying the shelf life of brown rice. Champagne et al. (17) monitored
changes in ethanol treated (stabilized) brown rice by analysis for free fatty
acids (FFA) and CDHP. Shin et al. (21) cited n-hexanal as the compound
most responsible for stale flavor in rice and demonstrated the correlation (r
= 0.99) between levels of oxidized linoleic acid and n-hexanal during stor-
age of brown rice. Kroll et al. (22) utilized gas chromatographic analysis of
headspace gas for lipid oxidation products, particularly hexanal, to study
storage changes in oatmeal cereals, tortilla chips, and soybean oil.

lll. RICELAND FOODS STORAGE STUDY

A. Methodology
At Riceland Foods, lipolysis and oxidative deterioration of rice bran lipids
were followed during storage. This study was designed to monitor changes
in free fatty acids, peroxide value (PV), and odor of two bran types stored
in two packaging materials and at two storage temperatures over a period
of one year. PV was chosen to monitor oxidative changes since an increase
in this value is the first indication of lipid oxidation. Variables are summa-
rized in Table 1.
The extrusion-stabilized bran used was processed through a Wenger TX-
80 twin screw cooker/extruder system. Extruder configuration and operat-
ing conditions that would meet the criteria identified as "optimal stabiliza-
42 MCCASKILL AND ORTHOEFER

Table 1 Independent Variables in Riceland Rice Bran Storage Study

Variable
Bran type Extrusion-stabilized
Parboiled
Packaging type Low barrier (2-mil polyethylene, heat-sealed)
High barrier (laminated 25-lb paper outside, 7.5-lb polyethyl-
ene, 0.00035-inch aluminum foil, 15-lb polyethylene in-
side, heat-sealed)
Storage temperature 22°C
38°C
Storage time None (Initial)
1 Month
2 Months
3 Months
4Months
6 Months
12 Months

tion conditions" reported by Randall et al. (14) were employed. Parboiled

bran was randomly collected from the normal bran stream. Each bran
sample was then split into multiple 100-g portions for packaging, storage,
and subsequent analysis.
Weighed portions were filled into one of two types of flexible packaging
material and the packages heat-sealed. A 2-mil polyethylene film similar to
that commonly used for packaging rice was used to represent a package
having relatively low oxygen and water vapor barrier properties. Pouches
formed with a laminate of 25-lb paper (outside), 7.5-lb polyethylene,
0.00035-inch aluminum foil, and 15-lb polyethylene (inside) were used to
represent packaging with high oxygen and vapor barrier characteristics.
One-half of each package type was stored under ambient temperature
(22°C) conditions. The other half of each package type was stored at 38°C.
No humidity control was applied, and all samples were stored in an un-
lighted environment.
Samples were analyzed at 0 (initial), 1, 2, 3, 4, 6, and 12 months.
Although no formal sensory panel was available, a team of five persons
familiar with the evaluation of lipid-bearing products for rancidity and odor
development was selected so that at least two of them could be used to
assess the odor of samples prior to analysis at the end of each storage
period. Analysis consisted of determination of FFA as a percent of oil and
PV as meq/kg of oil. Samples were prepared for analysis by extraction of
the oil from 100 g of bran using hexane at ambient temperature. The bran/
STORAGE STABILITY OF RICE BRAN 43

solvent mixture was agitated periodically for 5 minutes, then allowed to

stand until bran settling was essentially complete. The miscella was de-
canted and filtered to remove the remaining fines. The solvent was then
removed under aspirator vacuum at 70°C using a Brinkman Rotavapor.
The resulting oil was analyzed for FFA and PV by AOCS official methods
Ca5a-40 and CdS-53, respectively (23).

B. Results
Our results were consistent with earlier reports regarding lipolytic and
oxidative characteristics of parboiled rice bran. As shown in Table 2, initial
FFA levels were slightly lower in parboiled bran than in extrusion-
stabilized bran. Conversely, peroxide values were initially much higher in
parboiled bran and increased rapidly during the first 4 months of storage.
To confirm that the parboiled bran used in the storage test was typical,
additional samples of a commercially marketed parboiled bran were ana-
lyzed for FFA and PV with similar results. In general, both brans appeared
to be quite stable to lipolytic activity. Additionally, the extrusion-stabilized
bran showed good resistance to oxidation as measured by peroxide value.
Looking at the effects of packaging material type, data in Table 3 demon-
strate that free fatty acid development was not materially influenced by the
choice of packaging, although high-barrier packaging generally tended to
result in slightly higher free fatty acid levels in extrusion stabilized bran. As
shown in Table 4, peroxide values tended to rise in parboiled bran at
similar rates through the first 2 months in both packages. But beyond this
point, peroxide development leveled off in the high-barrier package, while

Table 2 Changes in Free Fatty Acid Content and Peroxide Value of Extrusion-
Stabilized and Parboiled Rice Bran During Storagea
FFA (%oil basis) PV (meq/kg oil basis)
Storage time
(months) ESB PBB ESB PBB

0 4.0 2.4 4.0 26.7

1 3.8 1.3 2.8 43.6
2 4.6 1.3 3.8 81.0
3 5.4 1.4 3.2 112.8
4 5.2 1.4 3.4 136.1
6 6.9 4.0 3.6 100.1
12 8.6 8.9 8.3 124.2
•Average of two storage temperatures and two packaging materials.
ESB, Extrusion-stabilized rice bran; PBB, parboiled rice bran.
44 MCCASKILL AND ORTHOEFER

Table 3 Influence of Packaging Type Upon Changes in Free

Fatty Acid Content of Extrusion-Stabilized and Parboiled Rice
Bran During Storage•
FFA (%oil basis)
ESB PBB
Storage time
(months) Poly Foil Poly Foil
0 4.0 4.0 2.4 2.4
1 3.6 4.0 1.2 1.3
2 4.0 5.2 1.3 1.2
3 4.4 6.3 1.3 1.4
4 4.0 6.4 1.3 1.4
6 5.0 8.8 6.5 1.6
12 5.8 11.4 7.9 10.0
•Average of two storage temperatures.

Table 4 Influence of Packaging Type Upon Changes in

Peroxide Value of Extrusion Stabilized and Parboiled Rice Bran
During Storage•
PV (meq/kg oil basis)
ESB PBB
Storage time
(months) Poly Foil Poly Foil
0 4.0 4.0 26.7 26.7
1 3.2 2.4 39.6 47.5
2 4.1 3.4 71.8 90.2
3 3.0 3.3 135.4 90.2
4 4.0 2.8 168.6 103.6
6 4.4 2.9 129.2 70.9
12 11.9 4.8 113.8 134.6
•Average of two storage temperatures.

continuing to rise in the low-barrier package. It is speculated that oxygen

availability became limiting in the high-barrier package. Peroxide value
remained relatively unchanged in extrusion-stabilized bran regardless of
packaging type through the first 6 months. However, a moderate increase
occurred between 6 and 12 months in the low-barrier package.
Data presented in Table 5 indicate that storage temperature produced
no clear effect on free fatty acid development in these enzyme-inactivated
STORAGE STABILITY OF RICE BRAN 45

Table 5 Effect of Storage Temperature Upon Changes in Free

Fatty Acid Content and Peroxide Value of Stabilized Rice Bran
During Storagea
FFA (%oil PV (meq/kg oil
basis) basis)
Storage time
(months) 22°C 38°C 22°C 38°C
0 3.2 3.2 15.4 15.4
1 2.3 2.8 16.8 29.6
2 2.6 3.2 35.3 49.5
3 2.9 3.8 48.0 67.9
4 2.8 3.7 54.7 84.8
6 3.5 7.4 61.6 42.2
12 4.6 13.0 110.0 22.5
•Average of two bran types and two packaging materials.

Table 6 Summary of Observations from Informal Sensory Evaluations of Rice

Bran Storage Samples
Storage
Bran Packaging temperature
type type ("C) Observations
ESB Poly 22 No off-odors noted through 12 months
ESB Poly 38 Rancid after 12 months
ESB Foil 22 Slight unidentified odor after 3 months; soapy
aroma after 4 months
ESB Foil 38 Slight unidentified odor after 3 months; soapy
aroma after 4 months
PBB Poly 22 No off-odors noted through 12 months
PBB Poly 38 Rancid after 4 months; second-strongest rancid
odor after 12 months
PBB Foil 22 Soapy aroma after 4 months; progressed to ran-
cid odor at 12 months
PBB Foil 38 Slight unidentified odor after 3 months; soapy
aroma after 4 months; strongest rancid odor
of all samples after 12 months

samples, but that storage at 38°C clearly accelerated the formation of perox-
ides. As noted earlier, increases in peroxide value were limited almost
totally to parboiled bran.
Interestingly, no readily apparent off-odors were noted in either bran type
during the 12 months of storage at 22°C in low-barrier poly bags (Table 6).
46 MCCASKILL AND ORTHOEFER

The data for peroxide value indicate that substantial oxidation occurred in
the parboiled bran stored under these conditions. It is unclear why odor was
not detected at any of the evaluation times. Since only secondary oxidative
reactions generate compounds responsible for off-odor, the permeable na-
ture of the poly packaging may have resulted in nondetection of odor. All
other samples developed some type of age-related odor, in some cases as
soon as 3 months after storage. "Soapy" was the word most frequently used
to describe aromas noted in the extrusion stabilized bran, while "sour" or
"rancid" were often used in describing aromas in parboiled bran samples.
While no specific pattern was apparent, samples stored at 38°C tended to
develop earlier and more intense off-odors. When individual sample data for
free fatty acid and peroxide values were compared, a loose relationship
appeared to exist. Due to the nature of oxidative changes in lipids involving
primary and secondary reactions and the fact that peroxides are not directly
responsible for off-odors, effort to establish a precise correlation was not
practical. Nevertheless, the monitoring of free fatty acid levels and peroxide
value is an effective means of following chemical changes in rice bran prod-
ucts during storage.

C. Discussion
Results of this study have generally confirmed previous reports (10,12-
14,16) and may provide practical guidance for the storage, packaging, and
utilization of stabilized rice bran as a food ingredient. As indicated by
relatively stable free fatty acid levels, lipase was inactive in both extrusion-
stabilized and parboiled rice bran. Extrusion-stabilized bran demonstrated
good oxidative stability. Parboiled bran, however, proved to be quite sus-
ceptible to oxidative deterioration.
High-barrier packaging does not appear to be necessary for storage of
extrusion-stabilized bran. In fact, our results suggest that low-barrier pack-
aging may be better for this type of bran so that volatile reaction products
may escape, rather than accumulating to detectable undesirable odor lev-
els. Conversely, barrier packaging may be more important for storage of
parboiled bran, although barrier packaging alone did not prevent oxida-
tion. Gas-impermeable packaging, coupled with a controlled atmosphere
(nitrogen flush) or the addition of antioxidants to replace those presumably
lost or destroyed during parboiling, could provide satisfactory results. Addi-
tional studies are required.
Storage temperature appeared to have little effect on the rate of free
fatty acid development. In parboiled bran, elevated storage temperatures
accelerated peroxide formation. In addition, storage at 38°C tended to
accelerate the development of undesirable odors.
STORAGE STABILITY OF RICE BRAN 47

IV. SUMMARY
It appears that properly processed extrusion-stabilized rice bran may be
safely stored for up to one year at :szzoc in gas-permeable packaging.
Parboiled rice bran would require additional protective measures to
achieve a similar storage life. The maximum safe storage life of parboiled
bran appears to be 3-4 months. Addition of antioxidants and/or the use of
controlled atmosphere packaging may effectively extend its storage life,
but further research is required before specific recommendations for stor-
age of parboiled rice bran are appropriate.

REFERENCES
1. Kahlon, T. S., Saunders, R. M., Chow, F. 1., Chiu, M. C., and Betschart, A.
A. (1989). Influence of rice, oat and wheat bran on plasma cholesterol in
hamsters. FASEB 1, 3:958A.
2. Hegsted, M., Windhauser, M. M., Lester, S. B., and Morris, S. K. (1990).
Rice bran and oat bran lower cholesterol in humans. FASEB J., 4:368A.
3. Juliano, B. 0. (1985). Rice bran, Rice: Chemistry and Technology (B. 0.
Juliano, ed.), AACC, St. Paul, MN, p. 654.
4. Sayre, R. N., Earl, L., Kratzer, F. H., and Saunders, R. M. (1987). Nutri-
tional qualities of stabilized and raw rice bran for chicks. Poultry Sci., 66:493.
5. Choudhury, N.H., and Juliano, B. 0. (1980). Lipids in developing and ma-
ture rice grain. Phytochemistry, 19:1063.
6. Aizono, Y., Funatsu, M., Sugano, M., Hayashi, K., and Fujiki, Y. (1973).
Enzymatic properties of rice bran lipase. Agric. Bioi. Chern., 37:2031.
7. Matsuda, H., and Hirayama, 0. (1975). Purification and characterization of
lipolytic acyl-hydrolases from rice endosperm. Nippon Nogei Kagaku Kaishi,
49:577.
8. DeLucca, A. J. II, Plating, S. J., and Ory, R. L. (1978). Isolation and identifi-
cation of lipolytic microorganisms found on rough rice from two growing
areas. J. Food. Prot., 41:28.
9. Sayre, R.N., Saunders, R. M., Enochian, R. V., Schultz, W. G., and Beagle,
E. C. (1982). Review of rice bran stabilization systems with emphasis on
extrusion cooking. Cereal Foods World, 27:317.
10. Loeb, J. R., Morris, N. J., and Dollear, F. G. (1949). Rice bran oil. IV.
Storage of the bran as it affects hydrolysis of the oil. 1 Am. Oil Chern. Soc.,
26:738.
11. Roberts, R. L., VanAtta, G. R., Hunter, I. R., Houston, D. F., Kester, E.
B., and Olcott, H. S. (1949). Steam blanching of fresh rough rice curbs
spoilage by fatty acids. Food Ind., 21:1041.
12. Shaheen, A. B., El Dash, A. A., and El Shirbeeny, A. E. (1975). Effect of
parboiling rice on the rate of lipid hydrolysis and deterioration of rice bran.
Cereal Chern., 52:1.
48 MCCASKILL AND ORTHOEFER

13. Sowbhagya, C. M., and Bhattacharya, K. R. (1976). Lipid autoxidation in

rice. J. Food Sci., 41:1018.
14. Randall, J. M., Sayre, R.N., Schultz, W. G., Fong, R. Y., Mossman, A. P.,
Tribelhorn, R. E., and Saunders, R. M. (1985}. Rice bran stabilization by
extrusion cooking for extraction of edible oil. J. Food Sci., 50:361.
15. Anthoni Raj, S., and Singaravadivel, K. (1982}. Reduction in free fatty acids
due to parboiling of paddy. J. Food Sci., 47:692.
16. Houston, D. F., Hunter, I. R., McComb, E. A., and Kester, E. B. (1954).
Deteriorative changes in the oil fraction of stored parboiled rice. J. Agric.
Food Chern., 2:1185.
17. Champagne, E. T., Hron, R. J., and Abraham, G. (1991}. Stabilizing brown
rice products by aqueous ethanol extraction. Cereal Chern., 68:267.
18. Kelly, V. K. (1979}. Stabilization of rice polish. U.S. Patent 4,158,066.
19. Sharp, R. N., and Timme, L. K. (1986}. Effects of storage time, storage
temperature, and packaging method on shelf life of brown rice. Cereal Chern.,
63:247.
20. McCormick, R. (1984). Prediction of rancidity essential when extending shelf
life of processed foods. Prepared Foods, (Feb. 1984}:111.
21. Shin, M. G., Yoon, S. H., Rhee, J. S., and Kwon, T. (1986}. Correlation
between oxidative deterioration of unsaturated lipid and n-hexanal during
storage of brown rice, J. Food Sci., 51:460.
22. Kroll, R. D., Dennis, M. R., and Hether, N. W. (1989). Determination of
volatile compounds in cereal foods by dynamic headspace analysis. Cereal
Foods World, 34:775.
23. American Oil Chemists Society. (1987). Official and Tentative Methods, 3rd
ed. The American Oil Chemists Society, Champaign, IL.
 4
Effect of Storage on the Physicochemical
Properties and Quality Factors of Rice
Joseph Chrastil
Agricultural Research Service, U.S. Department of Agriculture,
New Orleans, Louisiana

I. INTRODUCTION
Storage of rice is a normal step between harvest and consumption. Stor-
age conditions (temperature, time, and moisture) may differ significantly.
In the United States, rice storage is required to change certain physico-
chemical properties of the rice, such as stickiness, from a sticky to a
relatively nonsticky product after cooking. Storage temperatures, mois-
ture content, and storage times vary between 10 and 35°C, 10 and 15%
water, and 2 and 24 months, respectively. In this chapter we will report
changes in physicochemical and quality factors of rice grains that occur
during postharvest storage. These changes influence the chemical, physi-
cal, and functional qualities of rice. Storage-induced aging has both desir-
able and undesirable effects on the end product, depending on storage
conditions and rice variety.
Aging changes are the result of the physicochemical interactions among
rice components and enzymic reactions involving protein, starch, and lipid.
Generally, the outer (aleurone) layers of the rice grains are more suscepti-
ble to these changes than the inner (endosperm) layers. Storage changes
are temperature dependent. The changes probably start during field drying
of the rice and after harvest (1). Many of the physicochemical and func-
tional changes that occur during storage, such as swelling, water uptake by
cooking, cooking time, stickiness after cooking, and dough leavening, are

49
 50 CHRASTIL

TABLE 1 Changes in Reducing, Nonreducing, and Total Sugar Contents of Milled

Rice During Airtight Storage. Comparison of Changes in Outer Layer, Nucleus, and
Entire Kernel
Residual
Outer layer nucleus
Entire kernel
Moisture Temper-
Milling content ature Before After 5 Before After 5 Before After 5
(%) (%) eq storage months storage months storage months
Reducing sugars (g maltose/100 g rice, d. b.)
7.7 15.6 -20 0.50 0.64 0.08 0.12 0.15 0.16
7.7 15.6 +5 0.50 0.90 0.08 0.13 0.15 0.19
7.7 15.6 +25 0.50 1.54 0.08 0.24 0.15 0.42
7.7 15.6 +35 0.50 1.05 0.08 0.35 0.15 0.47
7.7 13.7 +35 0.50 1.35 0.08 0.15 0.14 0.36
7.7 12.9 +35 0.50 0.94 0.08 0.12 0.14 0.22
12.0 15.5 -20 0.37 0.35 0.07 0.10 0.08 0.10
12.0 15.5 +5 0.37 0.36 0.07 0.14 0.08 0.11
12.0 15.5 +25 0.37 0.69 0.07 0.13 0.08 0.20
12.0 15.5 +35 0.37 0.47 O.D7 0.14 0.08 0.17
Nonreducing sugars (g sucrose/100 g rice, d. b.)
7.7 15.6 -20 3.52 3.37 0.09 0.08 0.50 0.50
7.7 15.6 +5 3.52 3.26 0.09 0.06 0.50 0.49
7.7 15.6 +25 3.52 0.51 0.09 0.04 0.50 0.14
7.7 15.6 +35 3.52 0.32 0.09 0.02 0.50 0.05
7.7 13.7 +35 3.52 0.12 0.09 0.02 0.47 0.04
7.7 12.9 +35 3.52 1.04 0.09 0.05 0.51 0.15
12.0 15.5 -20 0.86 0.66 0.05 O.Q3 0.17 0.11
12.0 15.5 +5 0.86 0.65 0.05 0.02 0.17 0.11
12.0 15.5 +25 0.86 0.12 0.05 0.04 0.17 O.D3
12.0 15.5 +35 0.86 0.02 0.05 O.Dl 0.17 0.01
Total sugars(%)
7.7 15.6 -20 4.02 4.01 0.17 0.20 0.65 0.67
7.7 15.6 +5 4.02 4.16 0.17 0.19 0.65 0.68
7.7 15.6 +25 4.02 2.05 0.17 0.28 0.65 0.56
7.7 15.6 +35 4.02 1.37 0.17 0.37 0.65 0.52
7.7 13.7 +35 4.02 1.47 0.17 0.17 0.61 0.40
7.7 12.9 +35 4.02 1.98 0.17 0.17 0.65 0.37
12.0 15.5 -20 1.23 1.01 0.13 0.13 0.25 0.21
12.0 15.5 +5 1.23 1.01 0.13 0.16 0.25 0.22
12.0 15.5 +25 1.23 0.81 0.13 0.17 0.25 0.23
12.0 15.5 +35 1.23 0.49 0.13 0.15 0.25 0.18
Source: Ref. 17.
EFFECT OF STORAGE ON RICE 51

caused by protein-starch interactions (2-6). Other changes due to aging are

not yet fully understood.

II. PHYSICOCHEMICAL CHANGES IN RICE COMPONENTS

A. Carbohydrates
Reducing sugars (maltose) increase and nonreducing sugars (sucrose) de-
crease during storage (7-13,17) (Table 1). The degradation of carbohy-
drates to C02 is usually very small, but it may become significant at mois-
ture content >14% (12,14,15) (Table 2). The changes in free carbohydrate
content are greatly influenced by temperature. Maillard's browning reac-
tion between free carbohydrates and free amino acids may play a signifi-
cant role in C02 formation. Nonreducing carbohydrates, free fatty acids,
and germination activity are closely related (16-18).

B. Starch
Starch in rice grains is concentrated mainly in the endosperm. Total starch
content does not change significantly during storage (4,12,17,19-23), how-
ever, some small changes in starch properties have been observed (18,24).
The changes in molecular weight of starch and its components (amylose
and amylopectin), although small, may be significant (4,15,17,25,26) (Ta-
ble 3).

Table 2 Changes in C0 2 Concentration in Intergranular Air During Airtight

Storage of Milled Rice
Storage conditions
Moisture C0 2 (%)
content Temperature
(%) (OC) Dec. 13 Jan. 23 Feb. 14 March 7 MayS May29
13.0 +5 0.06 0.06 0.06 0.08 0.08 0.08
13.0 +25 0.06 0.2 0.2 0.2 0.3 0.4
13.0 +35 0.06 0.9 1.1 1.2 2.0 2.8
14.3 +5 0.09 0.1 0.1 0.1 0.09
14.3 +25 0.3 0.4 0.6 2.4 7.2
14.3 +35 10.6 12.2 10.3 13.3 13.3
15.7 +5 0.2 0.2 0.2 0.3 0.3
15.7 +25 8.3 8.9 9.4 11.2 12.3
15.7 +35 13.2 16.9 15.7 16.6 17.8

Source: Ref. 12.

52 CHRASTIL

Table 3 Molecular Weightsa of Rice Oryzenin and Starch Components

M(T/), M(s), M(A), M(u),

Rice (T/), mP x1<f s x1<f A,nm x1<f X1cf

Oryzenin
medium-grain rice
control 0.094 1.03 6.52 1.19 1.2
4°C 0.095 1.12 6.58 1.20 1.3
40°C 0.102 2.01 9.12 2.06 2.0
long-grain rice
control 0.096 1.22 6.98 1.33 1.2
4°C 0.098 1.44 7.08 1.36 1.1
40oc 0.102 2.01 9.08 2.05 2.1
Amylose
medium-grain rice
control 0.356 1.24 4.80 1.15 612 1.20
4°C 0.352 1.18 4.86 1.18 611 1.16
40oc 0.341 1.03 4.50 1.02 608 1.04
long-grain rice
control 0.375 1.55 5.40 1.45 619 1.53
4°C 0.370 1.46 5.45 1.47 618 1.48
40°C 0.346 1.10 4.66 1.09 609 1.08
Amylopectin
medium-grain rice
control 0.750 29.9 26.06 31.0
4°C 0.756 31.0 26.64 31.1
40°C 0.762 32.0 27.15 32.3
long-grain rice
control 0.720 25.2 23.70 24.8
4°C 0.723 25.6 23.80 25.0
40°C 0.757 31.2 26.67 31.2
Starch
medium-grain rice
control 0.718 24.9 23.95 23.3 25
4°C 0.721 25.3 23.97 25.4 24
40°C 0.735 27.5 24.52 26.5 26
long-grain rice
control 0.680 19.7 21.70 21.0 20
4°C 0.685 20.3 21.76 21.1 20
40°C 0.703 22.7 22.12 21.7 21
•Rice was stored for one year at 4 and 40°C. Control was postharvest rice stored at 25°C.
The average molecular weights M(71), M(s), M(A), and M(JL) were determined from viscos-
ity, sedimentation constant, absorption maxima, or light scattering measurements, respec-
tively.
Source: Ref. 4.
EFFECT OF STORAGE ON RICE 53

The average molecular weight of starch slightly increased during storage

at higher storage temperatures (25-40°C). In these experiments the in-
crease was caused mainly by the increased molecular weight of amylo-
pectin, because the molecular weight of amylose slightly decreased (see
Table 3). Thus, depending on rice variety and storage conditions, the aver-
age molecular weight of starch might increase or decrease. These changes
are small but real and are probably caused by starch synthesis or degrada-
tion within the kernel by, for example, amylases, Q-enzyme, R-enzyme,
etc.

C. Proteins
Although the proteins are more concentrated in outer layers (aleurone and/
or bran) of rice grains, significant amounts are present also in endosperm
(polished rice) (Table 4). Total protein content does not change signifi-
cantly during storage (4,12,19,21-23,27,28). On the other hand, the chemi-
cal and physicochemical properties of proteins (especially of the rice stor-
age protein oryzenin (glutelin), which constitutes more than 80-90% ofthe
total protein) change significantly during storage (12,22,27,29-35).
The resistance to proteases increased and the solubility of protein frac-
tions decreased during storage (4,12,27 ,28,36) (Table 5). Small changes of
the albumin:globulin:prolamin:glutelin (oryzenin) ratio have been found
during storage (28), especially at higher storage temperatures where a
marked decrease in amino acids was also found (Table 6) (10,12,28,30,37).
The free amino acids and carbohydrates were related to the Maillard's
nonenzymic browning as determined by color changes in rice during stor-
age (Fig. 1).
The molecular weight and the number of -S-S- bridges (4,6,35,38) in
oryzenin increased significantly during storage, especially at higher storage
temperatures. After one year of storage at 40°C the average molecular
weight of oryzenin (measured after extraction and purification) had dou-
bled in both medium- and long-grain rice varieties (see Table 3).
Simultaneously, the binding capacity of oryzenin for starch or its compo-
nents decreased (Table 7). The composition of oryzenin subunits changed
to higher molecular weight subunits during storage (Table 8). These
oryzenin-starch interactions were correlated with several functional proper-
ties of rice.

D. Enzymes
The apparent initial velocities of rice enzymes frequently decreased during
storage (12,33,34,39-45) (Table 9), but this fact was misleading because the
specific activities of many enzymes did not decrease (33 ,34) (Tables 10-12).
54 CHRASTIL

Table 4 Chemical Composition of Outer Layer, Nucleus, and Entire Kernel of

Milled Rice
Outer Entire
Constituent Unit layer Nucleus kernel Ref.
Starch % 61.86 92.00 90.68 64
Amylose % 16.12 29.85 29.46 64
Reducing sugars g maltose/100 g rice 0.50 0.07 0.12 4
Nonreducing sugars g sucrose/100 g rice 2.42 0.11 0.26 4
Total sugars % 2.92 0.18 0.38 4
Fiber % 1.47 0.22 0.28 38
Total N g N/100 g rice 2.53 1.27 1.39 71
Nonprotein N g N/100 g rice 0.04 0.018 0.019 71
Protein N g N/100 g rice 2.49 1.25 1.37 71
Albumin g (N x 5.95}/100 g rice 1.75 0.29 0.30 71
Globulin g (N x 5.95}/100 g rice 1.12 0.60 0.67 71
Prolamin g (N x 5.95)/100 g rice 0.72 0.22 0.25 71
Glutelin g (N x 5.95}/100 g rice 7.93 5.05 5.25 71
Insoluble fraction g (N x 5.95)/100 g rice 3.28 1.48 1.69 71
Free amino N mg/100 g 25.11 2.55 3.40 71
Alpha-amylase SKB units/g rice 1.0 0.07 0.1 1
Beta-amylase mg maltose/g rice 223.8 31.2 44.9 1
Protease hemoglobin units/g rice 6.0 0.6 0.9 1
Total lipids % 4.44 0.45 0.66 1
Free fatty acids % 1.34 0.15 0.21 1
Neutral fats % 2.53 0.26 0.38 1
Phospholipids % 0.57 0.04 0.07 1
Thiamine mg/100 g 0.797 0.047 0.081 64
Riboflavin mg/100 g 0.075 0.019 0.022 64
Niacin mg/100 g 9.270 0.885 1.264 64
Pyridoxine mg/100 g 1.185 0.080 0.128 43
Ash % 6.10 0.45 0.72 19
Calcium % 0.359 0.007 0.023 64
Iron % 0.028 0.001 19
Phosphorus % 1.022 0.099 0.140 64
Source: Ref. 12.

The decrease of the initial velocities could be explained by specific revers-

ible substrate binding (33,46).
Substrate-binding proteins inactivated the substrates so that they could
not be converted to products. The amount of these "blocking" proteins
increased during storage, which resulted in an apparent inhibition of the
enzymic activities when measured by initial velocities. However, the spe-
EFFECT OF STORAGE ON RICE 55

Table 5 Changes in Protein Solubility Fractions of Milled Rice During Airtight

Storage. Comparison of Changes in Outer Layer, Nucleus, and Entire Kernel
After 5 months
Before storage at +5°C at +25°C at +35°C
Albumins, %, d. b
Entire kernel 0.30 0.38 0.25 0.18
Outer layer 1.75 1.44 1.44 0.79
Residual nucleus 0.29 0.27 0.16 0.17
Globulins, %, d.b.
Entire kernel 0.67 0.57 0.59 0.45
Outer layer 1.12 0.71 0.65 0.89
Residual nucleus 0.60 0.45 0.63 0.44
Prolamins, %, d.b.
Entire kernel 0.25 0.14 0.08 0.13
Outer layer 0.72 0.19 0.19 0.21
Residual nucleus 0.22 0.10 0.10 0.11
Glutelins, %, d.b.
Entire kernel 5.25 4.90 4.81 3.74
Outer layer 7.93 8.85 7.84 6.00
Residual nucleus 5.05 4.10 4.36 3.41
Total soluble, %, d. b.
Entire kernel 6.47 5.99 5.73 4.50
Outer layer 11.62 11.19 10.12 7.89
Residual nucleus 6.16 4.92 5.25 4.13
Insoluble,%, d.b.
Entire kernel 1.68 1.87 1.98 3.11
Outer layer 3.17 3.58 4.33 6.96
Residual nucleus 1.27 2.58 1.96 2.93
% Extraction yield
Entire kernel 79.3 76.1 74.3 59.0
Outer layer 77.9 75.7 70.0 53.1
Residual nucleus 80.7 71.9 72.7 58.5
Source: Ref. 12.

cific activities and the enzyme content remained unchanged during storage.
Thus, these substrate-inhibiting proteins protected the enzymes and simul-
taneously depressed their apparent activities.

E. Lipids
Most of the lipids in rice grains are concentrated in the outer layers
(aleurone and/or bran). During aging the lipids and/or phospholipids are
 56 CHRASTIL

INFLUENCE OF INFLUENCE OF

--
TEMPERATURE MOISTURE CONTENT
L VALUE
72
~:=:c·::.c".~;:-.;:.~;
70
------.., ~
68 ~
66
b 1'-

~
en
!:::
z
a VALUE
v
?
../
::::> 2 J.---
a:
0
....1
~
-
0 0 v
u kf:.~~=-~.:-~ :-:-:-;-:-:- .

b VALUE
/.
7

5~ ;:;;~
z~~:: : : c;~: v
0 5 10 0 5 10
MONTHS

Figure 1 Color changes in polished rice during airtight storage. Stored at (· ···)
-20, (----) +5, (-·-·) +25, and(-) +35°C, respectively; (e) 12.9, (•) 13.7, and
(.6.) 15.6% moisture. L, a, b =color parameters. (From Ref. 12.)

hydrolyzed and/or oxidized to free fatty acids and/or peroxides. This causes
increased acidity and a markedly increased deterioration of taste and flavor
and production of rancid odors. Several kinds of lipases, lipoxidases, es-
terases, phospholipases, and/or phosphatases are involved in this process
(47,48).
The intermediates of the lipid oxidation are organic peroxides, which
decompose further to the final products: aldehydes, ketones, ketoacids, or
carboxylic acids. These reactions occur more frequently in the outer
(aleurone and/or bran) layers of the rice grains, because these layers con-
EFFECT OF STORAGE ON RICE 57

Table 6 Changes in the Free Amino Nitrogen Content of Milled Rice

During Airtight Storage. Comparison of Changes in Outer Layer,
Nucleus, and Entire Kernel
Storage conditions
Free amino N (mg/100 g rice, d. b.)
Moisture
content Temperature Residual
(%, d.b.) eq Outer layer nucleus Entire kernel
Undermilled rices (7.7% milling)
15.6 -20 46.0 4.3 9.8
15.6 +5 42.1 4.6 9.6
15.6 -25 35.9 6.5 10.6
15.6 +35 23.8 7.6 10.9
13.7 +35 28.0 6.9 9.6
12.9 +35 36.0 5.5 10.1
Original sample 40.0 4.5 10.9
Well-milled rices (12.0% milling)
15.5 -20 17.7 3.2 4.7
15.5 +5 16.9 3.4 5.0
15.5 +25 13.4 4.8 6.3
15.5 +35 11.6 5.3 5.8
14.2 +35 12.6 4.5 5.7
13.1 +35 13.1 3.1 4.3
Original sample 16.1 3.2 3.9
Source: Ref. 28.

tain most of the lipids. In well-polished rice grains, which contain only a
fraction of the lipids, this phenomenon is less important (49,50).
Free fatty acids and other lipid degradation products increase during
storage (10,12,15,27,32,50-59) (Table 13). Phospholipids decomposed dur-
ing storage of rice mostly to phosphatidic acid, which consequently changed
the properties ofpherosome membranes (37). Tocopherols, natural antioxi-
dants, inhibit the lipid deterioration in rice (60).
Because rice bran contains relatively large amounts of lipids and
lipolytic enzymes, its deterioration during storage is very fast (37 ,85,103-
105). Rice bran is usually used for animal feed and/or rice oil extraction
(85). Several methods have been proposed for reducing the storage deterio-
ration of bran and/or rice oil. Most of these are based on the heat or
chemical denaturation of the enzymes present in the bran (59 ,102,103-116)
(Table 14). With decreased moisture content, higher temperatures must be
used to inactivate the enzymes (117).
58 CHRASTIL

Table 7 Interactions of Oryzenin with Starch Components from Rice

Ricea Keqb n m n:mc rd

Amylose + oryzenin
medium-grain rice
control 0.041 0.265 0.107 2.48 0.997
4°C 0.039 0.258 0.106 2.43 0.998
40°C 0.029 0.196 0.101 1.94 0.998
long-grain rice
control 0.025 0.155 0.101 1.53 0.998
4°C 0.024 0.149 0.100 1.49 0.988
40°C 0.020 0.116 0.089 1.30 0.988
Amylopectin + oryzenin
medium-grain rice
control 0.025 0.264 0.115 2.30 0.998
4°C 0.023 0.259 0.117 2.21 0.999
40°C 0.016 0.113 0.064 1.76 0.975
long-grain rice
control 0.013 0.150 0.103 1.45 0.997
4°C 0.013 0.144 0.102 1.41 0.984
40oc 0.010 0.090 0.072 1.25 0.987
Starch + oryzenin
medium-grain rice
control 0.054 0.268 0.099 2.70 0.999
4oc 0.055 0.261 0.100 2.61 0.999
40°C 0.040 0.187 0.101 1.85 0.996
long-grain rice
control O.Q35 0.150 0.098 1.53 0.997
4°C 0.035 0.146 0.101 1.45 0.995
40oc 0.028 0.117 0.107 1.09 0.990

•Rice was stored for one year at 4 and 40°C. Control was postharvest rice stored at 25°C.
bKeq is the equilibrium adsorption constant.
'n:m is the oryzenin-to-starch ratio.
dr is the correlation coefficient.
Source: Ref. 4.

Rice oil is usually prepared by solvent extraction of bran, which pro-

duces higher yields than pressing alone (85,104,108,113) (see Chapter 16).
Enzyme inactivation by heating or cooking is often a part of the extraction
process. The extracted oil is refined by a variety of different methods, such
as alkali extraction of free fatty acids, phospholipids, and some gums;
filtering and centrifuging to remove waxes and gums; bleaching to improve
color; steam deodorizing to improve flavor; and low-pressure distillation to
EFFECT OF STORAGE ON RICE 59

Table 8 Peptide Subunits in Oryzenin from Postharvest and Stored Rice•

M (P)b M (S)b L (P)d L (S)d
Subunit (MW) (%) (%) pc (%) (%) pc

12,300 ± 200 <1 <1 <1 <1

14,200 ± 200 9 ± 0.4 8 ± 0.3 0.98 8 ± 0.3 7 ± 0.2 0.99
22,000 - 23,000 34 ± 1.0 31 ± 1.0 0.99 33 ± 1.0 30 ± 0.8 0.99
32,000 - 37,000 37 ± 0.5 35 ± 0.6 1.00 36 ± 0.2 35 ± 0.3 0.98
56,000 ± 300 3 ± 0.3 2 ± 0.2 0.99 3 ± 0.2 2 ± 0.3 0.98
79,200 ± 500 5 ± 0.3 6 ± 0.3 0.97 5 ± 0.1 6 ± 0.3 0.99
83,000 ± 400 4 ± 0.2 4 ± 0.2 0.25 3 ± 0.3 4 ± 0.2 0.98
91,300 ± 400 2 ± 0.2 3 ± 0.2 1.00 3 ± 0.3 5 ± 0.6 0.99
104,000 ± 500 1 ± 0.2 3 ± 0.3 1.00 1 ± 0.2 2 ± 0.2 1.00
141,000 ± 2000 <1 <1 <1 <1
168,000 ± 2000 <1 <1 <1 <1
181,000 ± 5000 <1 <1 <1 <1
202,000 ± 5000 5 ± 0.6 8± 1 0.99 7 ± 0.5 9 ± 0.5 1.00

•Rice was stored one year at 40°C.

bM(P) and M(S) are the average molecular weights of medium-grain postharvest (P) and
stored (S) rice.
cp is the P-statistic for the probability that postharvest rice was different from stored rice.
dL(P) and L(S) are the average molecular weights of long-grain postharvest (P) and stored
(S) rice.
Source: Ref. 35.

remove free fatty acids. Refined oil is usually stable when stored because it
contains only small quantities of unsaturated fatty acids and has a high
content of tocopherols (114).

F. Vitamins
Rice grains contain several important vitamins (12) (see Table 4). Most of
them are concentrated in outer layers (aleurone and/or bran). During stor-
age some vitamins are lost (13,19,27,60-62). The deterioration of vitamins
depends on storage conditions. Under normal storage conditions the loss of,
for example, riboflavin or niacin does not exceed 5%. On the other hand, the
loss of thiamine is usually much higher (up to 30%) (13,19,27,62).

III. CHANGES IN QUALITY FACTORS

A. Storage Gases and Color
Replacing air by inert gases made little difference in the composition and
physicochemical properties of rice during storage (63,64). The color of
 gs

Table 9 Changes in the Alpha-Amylase, Beta-Amylase, and Proteolytic Activities of Milled Rice During Airtight Storage.
Comparison of Changes in Outer Layer, Nucleus, and Entire Kernel

Proteolytic activity
Storage conditions Alpha-amylase activity Beta-amylase activity (hemoglobin units/g rice,
(SKB units/g rice, d. b.) (mg maltose/g rice, d. b.) d. b.)
Moisture
content Temperature Outer Residual Entire Outer Residual Entire Outer Residual Entire
(%) (OC) layer nucleus kernel layer nucleus kernel layer nucleus kernel
13.0 +5 0.65 O.D7 0.11 223.81 31.26 44.89 5.77 0.61 1.12
13.0 +25 0.47 0.08 0.13 216.00 32.13 42.17 5.36 0.57 1.08
13.0 +35 0.86 0.06 0.11 148.06 19.43 27.11 4.22 0.53 0.91
14.3 +5 - - - - - - - -
14.3 +25 0.62 0.09 0.14 192.22 11.83 25.16 4.23 0.61 1.12
14.3 +35 0.38 0.03 0.06 88.20 15.88 20.34 3.89 0.44 0.77
15.7 +5 36.87
15.7 +25 0.13 - - 30.34 0.76
15.7 +35 0.06 - - 18.63
Original rice sample 1.03 0.07 0.11 6.03 0.63 0.97
Source: Ref. 12.
n
::=
~
Vl
~
t""'
 EFFECT OF STORAGE ON RICE 61

Table 10 Kinetic Constants• of Protease and Amylase from Stored Riceb

Protease Amylase
10 months 10 months
0 month Omonth
4•c 4°C 25•c 37•c 4•c 4•c 25•c 37•c

Long Grain
Pmax•A 0.247 0.217 0.192 0.170 0.240 0.230 0.220 0.185
k, min- 1 g- 1 mL 1.3 1.0 1.0 1.1 2.7 2.7 2.4 2.2
V0 , min- 1 g- 1 mL 0.78 0.65 0.60 0.50 6.7 3.2 3.2 2.0
n 0.86 0.83 0.81 0.83 0.61 0.71 0.70 0.74
r 0.999 0.999 0.999 0.997 0.998 1.000 0.999 1.000
Medium Grain
Pmax• A 0.150 0.130 0.125 0.115 0.230 0.175 0.160 0.120
k, min- 1 g- 1 mL 1.5 1.7 2.0 1.3 3.0 3.2 3.2 3.3
V0 , min- 1 g- 1 mL 1.0 0.85 0.80 0.45 7.0 5.3 5.0 3.1
n 0.75 0.78 0.81 0.81 0.59 0.61 0.62 0.63
r 0.999 0.999 0.999 0.999 0.997 0.999 0.999 0.998
Short Grain
pmax•A 0.346 0.309 0.255 0.245 0.240 0.199 0.165 0.140
k, min- 1 g- 1 mL 1.0 1.1 1.4 1.5 1.5 1.2 1.6 1.8
V0 , min- 1 g- 1 mL 1.8 1.1 1.0 1.2 3.5 2.2 2.3 2.2
n 0.71 0.78 0.80 0.79 0.58 0.60 0.60 0.61
r 1.000 0.998 0.997 1.000 0.999 1.000 0.999 1.000
•P max is the maximum amount of product formed at equilibrium given as a decimal fraction of
1.00 (100% of product formed); k is the specific rate constant; V0 is the initial velocity; n is
the heterogeneous diffusion resistance; and r is the correlation coefficient.
bRice was cooled to 4•c, then stored for 10 months at 4, 25, and 37•c.
Source: Refs. 33, 34.

polished rice did not change significantly during low-temperature storage

but increased considerably at higher storage temperatures (3, 17,55 ,63) (Ta-
ble 15). Petroleum ether extraction improved rice color during storage
(22).

B. Water Uptake and Solids Loss

Water uptake by cooked rice grains increased during storage (3,5,10-

13,22-24,27,55,63,65-68), especially at higher storage temperatures (Ta-
ble 15), but the loss of solids during cooking decreased during storage
(12,15,55,67) (Table 16).
 62 CHRASTIL

Table 11 Kinetic Constants• of Peroxidase and Cytochrome c Reductase from Stored

Riceb

Peroxidase Cyt c reductase

10 months 10 months
0 month Omonth
4°C 4°C 25°C 37°C 4°C 4°C 25°C 37°C

Long Grain
p rrntX' A 0.481 0.210 0.205 0.165 0.136 0.139 0.093 0.037
k, min- 1 132 152 150 174 157 24 24 28
g- 1 mL
V0 ,min- 1 63 32 31 29 21 3.3 2.2 1.0
g- 1 mL
n 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
r 0.998 0.995 0.991 0.995 0.997 0.999 0.999 0.997
Medium Grain
pmax•A 0.511 0.350 0.320 0.240 0.283 0.278 0.093 0.028
k, min- 1 93 97 115 112 83 33 29 35
g- 1 mL
V0 ,min- 1 47 34 30 27 23 3.5 2.7 1.0
g- 1 mL
n 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
r 0.991 0.996 0.997 0.996 0.998 0.999 1.000 0.999
Short Grain
pmax•A 0.553 0.590 0.530 0.420 0.450 0.417 0.278 0.139
k, min- 1 82 75 85 70 84 74 36 29
g- 1 mL
V0 ,min- 1 45 44 45 29 38 31 10 4.0
g- 1 mL
n 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
r 0.997 0.998 0.999 1.000 0.998 0.995 0.992 0.997

•Pmax is the maximum amount of product formed at equilibrium given as a decimal fraction of
1.00 (100% of product formed); k is the specific rate constant; V0 is the initial velocity; n is the
heterogeneous diffusion resistance; and r is the correlation coefficient.
bRice was cooled to 4°C, then stored for 10 months at 4, 25, and 37°C.
Source: Refs. 33, 34.

C. Starch Properties
Gelatinization temperature of polished rice, determined by amylography as
a 10% slurry, did not change (12,15,23,70,71), but the supercooling tem-
perature of rice starch or rice flour gels decreased during storage (72)
(Table 17).
 EFFECT OF STORAGE ON RICE 63

Table 12 Kinetic Constants• of Phosphatase and Phospholipase C from Stored

Riceb

Phosphatase Phospholipase C
10 months 10 months
Omonth 0 month
4°C 4°C 25°C 37oc 4°C 4°C 25°C 37oc

Long-Grain
pmax•A 0.520 0.490 0.490 0.480 0.780 0.620 0.490 0.350
k, min- 1 77 87 81 76 1.6 1.9 1.9 2.0
g- 1 mL
VO,min- 1 40 42 40 36 1.3 1.1 0.91 0.70
g- 1 mL
n 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
r 1.000 0.998 0.999 0.999 1.000 0.985 0.984 0.985
Medium-Grain
pmax•A 0.560 0.450 0.450 0.440 0.570 0.380 0.270 0.190
k, min- 1 7.1 5.9 6.0 5.6 1.4 1.5 1.8 1.8
g- 1 mL
V0 ,min- 1 40 26 27 25 0.80 0.56 0.49 0.34
g- 1 mL
n 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
r 1.000 0.998 0.999 0.999 0.999 0.998 0.990 0.994
Short-Grain
pmax• A 0.546 0.455 0.455 0.445 1.100 0.985 0.930 0.809
k, min- 1 133 129 144 142 0.8 0.8 1.0 1.2
g- 1 mL
V0 ,min- 1 72 59 65 63 0.90 0.80 0.93 0.96
g- 1 mL
n 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0
r 1.000 1.000 0.998 1.000 0.999 0.999 0.999 0.998
•P max is the maximum amount of product formed at equilibrium given as a decimal fraction of
1.00 (100% of product formed); k is the specific rate constant; V0 is the initial velocity; n is
the heterogeneous diffusion resistance; and r is the correlation coefficient.
bRice was cooled to 4°C, then stored for 10 months at 4, 25, and 37°C.
Source: Refs. 33, 34.

D. Kernel Hardness and Swelling

Tensile strength and the crushing or breaking hardness of rice grains
increased during storage (19,69). Swelling of rice grains also increased
during storage, especially at higher storage temperatures (3,5) (Table
18).
64 CHRASTIL

Table 13 Changes in Lipids and Lipid Fraction Content of Milled Rice During
Airtight Storage. Comparison of Changes in Outer Layer, Nucleus, and Entire
Kernel
Storage conditions Total lipids
(%, d.b.)
Moisture
content Temperature Residual
(%) eq Outer layer nucleus Entire kernel
13.0 +5 4.43 0.47 0.67
13.0 +25 4.47 0.48 0.67
13.0 +35 4.41 0.46 0.66
14.3 +25 4.41 0.41 0.61
15.7 +25 0.64
Original rice sample 4.44 0.45 0.66
Free fatty acids(%, d.b.)
13.0 +5 1.32 0.16 0.20
13.0 +25 1.61 0.22 0.29
13.0 +35 2.14 0.28 0.38
14.3 +25 2.30 0.25 0.36
15.7 +25 0.45
Original rice sample 1.34 0.15 0.21
Neutral fats(%, d.b.)
13.0 +5 2.57 0.27 0.39
13.0 +25 2.36 0.22 0.32
13.0 +35 1.62 0.13 0.21
14.3 +25 1.74 0.12 0.20
15.7 +25 0.14
Original rice sample 2.53 0.26 0.38
Phospholipids(%, d.b.)
13.0 +5 0.54 0.04 0.08
13.0 +25 0.49 0.04 0.06
13.0 +35 0.64 0.05 0.08
14.3 +25 0.37 0.04 0.05
15.7 +25 0.05
Original rice sample 0.57 0.04 0.07
Source: Ref. 12.

The increased hardness during storage resulted in increased flour milling

yields (22) and in changes in the particle size distribution after milling (3)
(Table 19). Bulk density of rice flours and water uptake by rice flours
increased during storage, especially at higher storage temperatures (3)
(Table 18).
Table 14 Storage Characteristics of Rice Bran at 25°C After
Various Heat Treatments

Free faty acid content (%)

Treatment time Temperature
(hr) eq After 25 days After 50 days

None 54 85
1 70 18 25
3 70 6 8
1 85 25 36
3 85 4 5
1 100 5 8
3 100 4 4
1 110 4 7
3 110 4 4

Source: Ref. 102.

Table 15 Color, Water Absorption, and Dough Leavening of Flours from Stored
Ricea

Colorb Water uptake Dough leaven-

Storage (%) (%) ing (%)
temperature
Rice type eq Fresh Stored Fresh Stored Fresh Stored

Long-grain, brown 4 2.1 8.6 65 68 40 43

25 9.7 83 43
37 11.3 94 45
Long-grain, polished 4 1.5 5.4 62 67 26 28
25 5.9 81 30
37 8.4 90 35
Medium-grain, brown 4 2.9 9.0 64 68 36 39
25 11.3 80 40
37 15.4 91 42
Medium-grain, 4 1.4 6.0 73 78 43 47
polished 25 6.2 87 51
37 7.2 92 56
Short-grain, brown 4 2.8 11.8 75 80 43 47
25 12.9 82 50
37 14.0 84 57
Short-grain, polished 4 2.2 10.5 61 67 41 46
25 11.2 75 52
37 12.0 81 57

•Rice was cooled to 4°C, then stored for 10 months at the temperatures indicated.
bColor values are given as the Hunterlab "b" value.
Source: Ref. 3.

65
66 CHRASTIL

Table 16 Changes in Water Absorption and Total Solids in Residual Cooking

Liquids During Airtight Storage of Milled Rice

Storage conditions Water absorption Residual solids

(g water/100 g rice) (g/100 g rice)
Moisture Temper-
content ature Feb- Feb-
(%) eq ruary April July October ruary April July October

13.0 +5 256 269 284 291 6.4 6.6 6.7 6.6

13.0 +25 256 277 317 291 6.4 6.4 5.8 5.3
13.0 +35 256 283 307 304 6.4 5.0 4.9 4.6
14.3 +5 258 272 284 273 6.3 6.3 6.4 6.2
14.3 +25 258 269 302 6.3 5.7 5.7 5.3
14.3 +35 258 286 309 311 6.3 4.0 3.4 3.1
15.7 +5 262 270 291 287 6.1 5.8 6.3 6.4
15.7 +25 262 271 304 304 6.1 5.8 5.7 5.2
15.7 +35 262 288 302 288 6.1 3.4 2.9 2.6
Source: Ref. 12.

Table 17 Supercooling of Starch from Fresh and Stored• Rice Grains

Starch composition
Storage
Rice variety (months) Amylose(%) Amylopectin (%) .::1Tbeq
Mercury 0 17.0 83.0 7.2
6 17.5 82.5 6.4
12 17.7 82.3 6.2
Lemont 0 26.0 74.0 6.0
6 26.8 73.2 5.3
12 27.0 73.0 5.0
•Rice was stored at 40°C for the storage times indicated.
b.dT represents the extent of supercooling.
Source: Ref. 72.

E. Cooking Properties
Cooking time increased during storage (3,12,13,67,73-77) but the number
of broken grains after cooking decreased (3,5,66) (Table 20). Stickiness of
cooked rice grains decreases during storage, especially at higher storage
temperatures (2-6,22,68,73-75,78-85) (Table 20). The relative decrease in
stickiness in different rice varieties was proportional to the stickiness (3-6).
As a result, there was a leveling trend observed in different varieties after
sufficient storage time. The varietal differences in stickiness observed in
EFFECT OF STORAGE ON RICE 67

Table 18 Moisture Content and Swelling of Grains and Bulk Density of Flours
from Storeda Rice

Bulk density
Moisture con- Swelling of flour (g/
Storage tent(%) (%) cm3)
temperature
Rice type eq Fresh Stored Fresh Stored Fresh Stored
Long-grain, brown 4 11.6 11.5 20.0 21.2 0.78 0.79
25 11.0 23.1 0.81
37 11.2 25.8 0.84
Long-grain, polished 4 10.2 10.2 22.6 24.1 0.92 0.94
25 10.4 27.6 0.95
37 10.0 37.9 0.98
Medium-grain, brown 4 12.2 12.2 22.7 24.2 0.80 0.82
25 12.0 32.3 0.87
37 12.0 37.3 0.94
Medium-grain, 4 12.4 12.1 33.6 35.9 1.01 1.03
polished 25 12.1 39.3 1.04
37 12.0 46.2 1.06
Short-grain, brown 4 10.5 10.5 40.8 46.7 0.95 0.96
25 10.4 46.3 0.97
37 10.6 46.4 0.99
Short-grain, polished 4 11.0 11.2 36.1 37.9 1.21 1.22
25 11.4 41.4 1.23
37 11.9 46.7 1.23

•Rice was cooled to 4°C (fresh), then stored for 10 months at the storage temperature
indicated.
Source: Ref. 3.

rice before storage became low after storage in all varieties. Dough leaven-
ing of rice flours increased during storage, especially at higher storage
temperatures (3-6) (Table 15).
Other viscoelastic properties, as determined by amylography, changed
significantly during storage of rice (12,15,22,23,55,57,68,71,86,87). The
peak viscosity of an amylogram was influenced by lipids and their degrada-
tion products (86-92). Lipid extraction by methanol, ethyl ether, hexane,
or petroleum ether eliminated some viscoelastic differences caused by stor-
age (12,18,22,55,56,93,94), but defatting of aged rice did not change the
viscoelastic properties (56). Thus, the role of lipids in changes due to
storage is not clear (84).
Curing resulted in accelerated aging of rice grains (75,95), but its effects
were greatly diminished by lipid extraction with ether (12,15) (Table 21).
Table 19 Particle Size Distribution of Flours from Stored• Rice Grains
Ol
Storage Particle size (p,m)b 00
temperature Average
Rice type ("C) 3 8 11 15 22 30 40 60 80 size (p,m)

Long-grain, brown Control 7.0 43.0 14.1 7.0 18.3 8.7 1.4 0.5 0.1 13.5
4 7.1 42.9 14.3 7.1 18.4 8.4 1.2 0.5 0.1 13.7
25 7.2 43.2 14.5 7.2 19.1 7.0 0.8 0.5 0.1 13.4
37 7.4 44.1 14.7 7.4 10.2 5.9 0.6 0.7 0.1 11.2
Long-grain, polished Control 6.5 38.5 12.8 6.5 14.7 12.3 6.7 2.0 0.2 16.7
4 6.5 38.7 12.9 6.5 14.7 12.4 6.5 1.8 0.2 16.5
25 6.5 39.2 13.1 6.5 14.9 12.2 6.1 1.4 0.2 16.1
37 6.5 39.2 13.1 6.5 18.6 11.8 3.3 0.9 0.1 15.3
Medium-grain, brown Control 6.5 39.2 13.0 6.6 21.3 11.1 1.5 0.4 0.1 14.8
4 6.6 39.3 13.1 6.6 21.5 11.2 1.4 0.4 0.04 14.7
25 6.5 39.3 13.2 6.6 21.0 11.7 1.4 0.3 0.1 14.7
37 7.0 43.0 14.0 7.4 18.0 8.8 1.2 0.5 0.1 13.8
Medium-grain, polished Control 3.6 22.1 7.3 3.7 11.8 20.4 14.7 14.6 1.9 28.0
4 3.7 22.2 7.4 3.7 11.7 20.5 14.6 14.6 1.6 27.9
25 3.5 22.0 7.8 3.9 11.5 20.0 15.1 14.0 2.2 28.0
37 4.1 24.6 8.2 4.1 12.9 17.2 12.9 14.6 1.6 26.8
Short-grain, brown Control 7.4 42.9 14.9 7.1 18.9 7.3 0.9 0.5 0.1 13.4
4 7.3 43.0 14.8 7.2 19.3 6.8 0.9 0.4 0.1 13.3
25 7.8 49.2 15.1 7.5 15.0 4.2 0.6 0.5 0.1 12.1
37 7.6 49.4 15.0 7.7 15.2 4.2 0.4 0.5 0.1 12.1
Short-grain, polished Control 4.8 27.8 9.4 4.6 10.4 14.6 10.4 16.0 2.0 26.1
4.7 28.0 9.3 4.7 10.5 14.7 10.1 16.1 1.8 26.0 ~
4
25 4.8 27.8 9.4 4.5 10.7 14.5 10.6 16.0 2.0 26.2
:=
1::0
37 5.3 31.9 10.6 5.3 8.0 14.0 9.8 13.5 1.5 23.9
...,>
rJl

•Rice was cooled to 4°C (Control), then stored for 10 months at the storage temperature indicated.
....
~
bParticle size values are the weight percent of four fractions on top of the sieves of different particle sizes.
Source: Ref. 3.
 t!j
~
Table 20 Water Absorption, Breakdown, and Stickiness of Cooked Rice from Stored• Rice Grains ~
t!j
l.)
Water uptake Cooking time Broken grains Stickinessb ....,
Storage (%) (min) (%) (7JmaJ
temperature
0
~
Rice type CCC) Fresh Stored Fresh Stored Fresh Stored Fresh Stored Stored ....,
rJl

4 120 126 23 25 16 14 0.7 0.5 0.1 0

Long-grain, brown ~
25 134 26 12 0.5 0.1
37 138 30 8 0.4 0.1
~t!j
Long-grain, polished 4 129 134 19 20 17 14 1.4 1.0 0.4 0
25 139 22 13 0.9 0.3 z
37 140 25 11 0.6 0.2 ....l.)
~

Medium-grain, brown 4 117 123 18 20 21 0.4 0.3 0.2 t!j

25 124 20 1 0.3 0.1
37 126 25 0 0.2 0.1
Medium-grain, polished 4 123 128 14 15 20 16 15.1 10.0 4.3
25 130 17 16 8.0 2.8
37 135 20 14 5.0 0.8
Short-grain, brown 4 120 125 25 30 15 11 1.3 1.0 0.3
25 128 31 9 0.8 0.2
37 135 35 6 0.6 0.1
Short-grain, polished 4 125 131 20 25 17 14 5.2 4.0 0.8
25 134 25 10 3.0 0.4
37 138 30 3 2.0 0.2

•Rice was cooled to 4°C (fresh), then stored for 10 months at the storage temperature indicated.
bStickiness was measured as 1'/max' which is the frequency curve mode for clusters formed by cooking rice grains.
Source: Ref. 3.
0)

"'
70 CHRASTIL

Table 21 Effects of Defatting Prior to Curing on the Arnylograrn

Characteristics of Rice Flour and Rice Starch
Methanol ex-
tracted ethyl
Ethyl ether ether defatted
Undefatted rice defatted rice rice

Characteristics Uncured Cured Uncured Cured Uncured Cured

Rice flour arnylograrn

Gelatinization tern- 87.0 83.0 86.0 82.0 77.0 76.5
perature, ac
Viscosity at 94°C 525 700 525 690 740 800
Peak viscosity 635 1000 625 925 830 920
Breakdown 170 245 150 180 150 140
Viscosity at 50°C 960 1440 945 1360 1160 1250
Setback 495 685 465 610 480 470
Rice starch arnylograrn
Gelatinization tern- 69.0 69.0 69.0 70.0
perature, oc
Viscosity at 95°C 560 500 560 520
Viscosity at 50°C 1180 1200 1390 1250
Setback 400 380 490 400

Source: Ref. 12.

Accelerated aging was also accomplished by heating rice grains to l10°C in

sealed containers (66,96).
Defatting polished rice by hexane, which removed most of the lipids, did
not improve the cooking qualities (94) after storage, but improved the
flavor.

F. Odor and Flavor

Odor and flavor usually change during storage of rice even under airtight
conditions (12,42) (Fig. 2). The presence of off-flavors is more easily de-
tected in cooked samples, although the types of off-flavor are better identi-
fied in uncooked samples. Aldehydes and ketones were identified as
sources of off-flavors (22,68,97-101). The off-flavor samples also con-
tained methyl mercaptan, dimethyl sulfide, dimethyl disulfide, sulfur diox-
ide, and hydrogen sulfide (102). During storage, hydrogen sulfide, di-
methyl sulfide, and sulfur dioxide decreased, but methyl mercaptan did not
change. Most of these volatiles are lipid, amino acid, or vitamin decomposi-
tion products. For example, it was found that the typical flavor of cooked
 EFFECT OF STORAGE ON RICE 71

MILLING DEGREE OF 7.7%

3 months 5 months 10 months

-20 DD~ DD~ DO~

+5 DO~ D~~ D~~
+25 D~rl 0~11 01111
g
a:
+35 ~II II 01111 111111
12. 9 13. 7 15. 6 12. 9 13. 7 15. 6 12. 9 13. 7 15. 6
::J
~a: MOISTURE CONTENT (%)
w
0..
~ MILLING DEGREE OF 12.0%
w
1- 3 months 5 months 10 months

ODD ODD
w
(!)
<(
a:
-20
DO~
0
1-
+5
DO~ DD~ DO~
DOll
(/)

+25 DO~ ~~II

+35 D~~ ~1111 111111
13. 1 14. 2 15. 5 13. 1 14. 2 15. 5 13. 1 14. 2 15. 5
MOISTURE CONTENT (%)

Figure 2 Development of off-odors during airtight storage of polished rice. The

strength of perception is shown by the degree to which the square is filled; a blank
square represents no perceived odor and a completely filled square represents the
strongest perceived odor; a cross with a point means no data obtained. (From
Ref. 12.)

rice could be developed from an aqueous solution of cysteine and cystine in

the presence of riboflavin upon exposure to sunlight (102). Except for the
aldehydes, this flavor contained H 2S and C02 •
Well-polished rice has much better flavor than underpolished rice. This
is understandable because the outer layers of rice grain contain many more
off-flavor precursors. However, relatively small differences in polishing
might result in significant differences in off-flavor.
Eating qualities of cooked rice (appearance, cohesiveness, tenderness,
palatability, and flavor) usually undergo significant changes during stor-
age. Freshly harvested rice becomes a paste on cooking, but after stor-
age pasting is greatly reduced (7,66). Storage decreases cohesiveness,
increases the volume of cooked rice, and results in firmer texture
(3,66,75,95,96).
72 CHRASTIL

Most of these changes occur before perceptible flavor deterioration are

desirable or undesirable depending on country and consumer preferences.
For example, the preservation of high cohesiveness by cold storage is a
common practice in Japan. Palatability increases during storage, especially
at higher storage temperatures (73).

IV. CONCLUSIONS
Although there are significant differences between rice varieties in the
storage changes described above (3,73) (see Tables 15-20), the trend (in-
creasing or decreasing) in all varieties is the same. Generally, storage condi-
tions are more important than varietal differences.
Regardless of storage conditions, statistically significant correlations
were found between the equilibrium binding of oryzenin to starch (n:m)
and the following physicochemical properties and quality factors: average
molecular weights of extracted oryzenin, water intake by cooked grains,
cooking time, stickiness of cooked rice grains, and dough leavening (5)
(Figs. 3-7). The interrelationships between these physicochemical and qual-
ity factors must also result in linear correlations. Thus, for the prediction of
one quality factor, another easily determined property may be chosen (5).
These correlations do not tell us how or why these relationships occur, but

3.--------------------------------------,

on-~.
~
~ ++~
w +,+
N 2 o..q_*
~ +~
o o"-...*
0~ '*~

~ 1 ·~""
0 2 3
n:m

Figure 3 Correlation between equilibrium binding ratio (n:m) of oryzenin with

starch and average molecular weight (MW) of oryzenin in different varieties of rice.
+ = long-grain; * = medium-grain; o = short-grain. Line was calculated by
regression analysis. MW of oryzenin = 272508 - 57080.8 (n:m), where r = 0.996.
(From Refs. 2-6.)
 EFFECT OF STORAGE ON RICE 73

"'+,
1~----------~-----------------------------.

~
(!)
+~
z o&ib
i:2
0
0
u ~0
~~
>
~ 130
:.;:
<(
!i::
:::>

~
a:
w

~
1~+------------.-----------.,---------~
0 2 3
n:m

Figure 4 Correlation between equilibrium binding ratio (n:m) of oryzenin with

starch and water uptake by cooking. + = long-grain; * = medium-grain; o = short-
grain. Line was calculated by regression analysis. Water intake by cooking = 145.69
- 7.446 (n:m), where r = 0.979. (From Refs. 2-6).

~--------------------------------------,

'2
l
w
:2
i= 20
(!)
z
i:2
0
0
u

10+------------.------------.---------~
0 2 3
n:m

Figure 5 Correlation between equilibrium binding ratio (n:m) of oryzenin with

starch and cooking time. + =long-grain;* =medium-grain; o =short-grain. Line
was calculated by regression analysis. Cooking time = 29.2 - 5.35 (n:m), where r =
0.984. (From Refs. 2-6.)
 74 CHRASTIL

ro.----------------------------------------.

tl)
ff3 10
z
:;;;!
u
i=
tl)

n:m

Figure 6 Correlation between equilibrium binding ratio (n:m) of oryzenin with

starch and stickiness. + = long-grain; * = medium-grain; o = short-grain. Line was
calculated by regression analysis. Stickiness = -10.406 + 8.283 (n:m), with r =
0.999. (From Refs. 2-6.)

70

g
Cl
z 60
zw
>
<(
w
...J
J:
Cl
::::>
0 50
0

~+--------------.-------------.------------~
0 2 3
n:m

Figure 7 General correlation between equilibrium binding ratio (n:m) of ory-

zenin with starch and dough leavening. + = long-grain; * = medium-grain; o =
short-grain. Line was calculated by regression analysis. Dough leavening= 79.98-
12.012 (n:m), where r = 0.977. (From Refs. 2-6.)
EFFECT OF STORAGE ON RICE 75

we may assume that the main determining factors for the above-mentioned
physicochemical properties and quality factors of rice are the molecular
weight and structure of oryzenin and consequently the equilibrium binding
(n:m) or oryzenin to starch.

REFERENCES
1. Kester, E. B., Lukens, H. C., Ferrel, R. E., Mohammad, A., and Finfrock,
D. C. (1963). Influence of maturity on properties of Western rices. Cereal
Chern., 40:323.
2. Chrastil, J. (1989). Storage conditions affect rice stickiness. Chern. Eng. News,
67:52.
3. Chrastil, J. (1990). Chemical and physicochemical changes of rice during stor-
age at different temperatures.1 Cereal Sci., 11:71.
4. Chrastil, J. (1990). Protein-starch interactions in rice grains. Influence of stor-
age on oryzenin and starch.1 Agric. Food Chern., 38:1804.
5. Chrastil, J. (1992). Correlations between the physicochemical and functional
properties of rice.1 Agric. Food Chern. 40: 1683.
6. Zarins, Z., Chrastil, J. (1992). Separation and purification of rice oryxenin
subunits by anion-exchange and gel-permeation chromatography. J. Agric.
Food Chern., 40: 1599.
7. Screenivasan, A. (1939). Studies on quality in rice. IV. Storage changes in rice
after harvest. Indian 1 Agric. Sci., 9:208.
8. Kester, E. B., Houston, D. F., Ferrel, R. E., Hunter, I. R., and Finfrock, D.
C. (1956). Storage behavior of rice in experimental bins, 1954-55. Results of
the second year test. Rice 1, 59:24.
9. Houston, D. F., Straka, R. P., Hunter, I. R., Roberts, R. L., and Kester, E.
B. (1957). Changes in rough rice of different moisture contents during storage
at controlled temperatures. Cereal Chern., 34:444.
10. Tani, T., Chikubu, S., and Iwasaki, T. (1964). Changes in chemical properties
of rice during storage at controlled temperature. Eiyo To Shokuryo, 16:436.
11. Iwasaki, T., and Tani, T. (1967). Effect of oxygen concentration on deteriora-
tive mechanism of rice during storage. Cereal Chern., 44:233.
12. Barber, S. (1972). Milled rice and changes during aging, Rice: Chemistry and
Technology (D. F. Houston, ed.), American Association of Cereal Chemists,
St. Paul, MN, p. 215.
13. Pushpamma, E. C., Barber, S., and Benedito de Barber, C. (1979). Physico-
chemical changes in rice and jowar stored in different agro-climatic regions of
Andhra Pradesh. Bull. Grain Techno/., 17:97.
14. Zeleny, L. (1960). Effects of commercial storage on the nutrient content of
processed food. Cereal grain, Nutritional Evaluation of Food Processing (R.
S. Harris and H. von Loesecke, eds.), Wiley, New York, p. 358.
15. Barber, S. (1969). Basic studies on aging of milled rice and application to
discriminating quality factors. USDA, ARS, Foreign Research and Technical
Program Division, Project No. E-25-AMS-(9). Final Report, p. 1.
76 CHRASTIL

16. Barber, S., Benedito de Barber, C., Guardiola, J. L., and Alberola, J., (1967).
Chemical composition of rice. IV. Distribution of sugars in grains. Rev.
Agroquirn. Tecnol. Alirnentos. 7:346.
17. Barber, S., Benedito de Barber, C., and Guardiola, J. L., (1968). Distribution
of sugars in rice grains. Rev. Agroquirn. Tecnol. Alirnentos, 8:89.
18. Shoji, I., and Kurasawa, H., (1981). On viscosity characteristics of rice and rice
starch against Brabender amylography. II. Effects of protein and fat on viscos-
ity of glutinous and non-glutinous rice starches. Kaseigaku Zasshi, 32:167.
19. Kondo, M., and Okamura, T., (1937). Storage of rice. XVII. Comparative
study of unhulled rice and hulled rice in regard to the changes of its qualities
during long storage in straw bags. Ber. Ohara Inst. Agric. Bioi., 7:483.
20. Jagoe, R. B. (1954). An experiment in the storage of under-milled rice. Ma-
layan Agr. 1, 37:154.
21. Yasumatsu, K., Moritaka, S., Ishii, K., Shimazono, H., and Fujita, E. (1965).
Studies on cereals. I. Changes of chemical characteristics of polished rice
during storage. Eiyo To Shokuryo, /8:123.
22. Villareal, R. M., Ressurreccion, A. P., Suzuki, L. B., and Julia-no, B. 0.
(1976). Changes in physicochemical properties of rice during storage. Starke,
28:88.
23. Swamy, I. Y. M., Sowbhagya, C. M., and Bhattacharya, K. R. (1978).
Changes in the physicochemical properties of rice with aging. 1 Sci. Food
Agric., 29:627.
24. Sreenivasan, A. (1939). Studies on quality in rice. IV. Storage changes in rice
after harvest. Indian J. Agric. Sci., 9:208.
25. Desikachar, H. S. R., (1956). Changes leading to improved culinary proper-
ties of rice on storage. Cereal Chern., 33:324.
26. Primo, E., Casas, A., Barber, S., and Benedito de Barber, C. (1962). Factors
of quality of rice. Physicochemical characteristics of starch and its fractions.
Rev. Agroquirn. Tecnol. Alirnentos, 2:343.
27. Rao, N. M., Viswanatha, T., Mathur, P. B., Swaminathan, M., and Subrah-
manyan, V. (1954). Effect of storage on the chemical composition of husked,
undermilled and milled rice. J. Sci. Food Agric., 5:405.
28. Primo, E., Barber, S., Benedito de Barber, C., and Sanchez, L. (1968). IV.
Modification of the nitrogenous components. Rev. Agroquirn. Tecnol. Ali-
rnentos, 8:224.
29. Iwasaki, T., Shibuya, N., and Chikubu, S. (1972). Studies on rice protein. I.
Albumin and globulin in rice endosperm: their heterogeneity and difference in
subunit composition among stored, heated and fresh rice. 1 Japan Soc. Food
Sci. Techno/., 19:70.
30. Baldi, G., Fossati, G., Ranghino, F., and Fantone, G. C. (1977). Storage of
rice: Changes in protein and amino acids. Riso, 26:253.
31. Bolling, H., Hampel, G., and El Baya, A. W. (1977). Changes in physical and
chemical characteristics of rice during prolonged storage. Riso, 26:65.
32. Bolling, H., Hampel, G., and ElBaya, A. W. (1978). Studies on storage of
milled rice for a long period. Food Chern., 3:17.
EFFECT OF STORAGE ON RICE 77

33. Chrastil, J. (1990). Influence of storage on enzymes in rice grains. 1 Agric.

Food Chern., 38:1198.
34. Chrastil, J. (1991). Influence of storage on enzymes in rice grains. Rice
Biotechnol. Q., 6:11.
35. Chrastil, J., and Zarins, Z. (1992). Influence of storage on oryzenin composi-
tion in rice endosperm. J. Agric. Food Chern. (in press).
36. Fedorova, S. A., Shcherbakov, V. G., and Beloglazova, L. K. (1982).
Changes in nitrogen and protein susceptibility (to enzymic hydrolysis) of rice
with varying contents during prolonged storage. Voprosy Pitan., 5:41.
37. Aibara, S., Ismail, I. A., Yamashita, H., Ohta, H., Sekiyama, F., and Morita,
Y. (1986). Changes in rice bran lipids and free amino acids during storage.
Agric. Bioi. Chern. (Tokyo), 50:665.
38. Moritaka, S., and Yasumatsu, K. (1972). Studies on cereals. IX. Sulfhydryl
and sulfide contents of milled rice. Eiyo to Shokuryo, 25:42.
39. Noguchi, Y. (1923). Effects of storage on enzymes.1 Sci. Agric. Soc. Japan,
245:115.
40. Lozsa, A., and Koller, K. (1955). Changes during storage in the thiamine
content and in the peroxidase and catalase activity of rice in Hungary.
Agrokern. Talajtan, 4:1.
41. Bautista, G. M., Lugay, J. C., Cruz, L. J., and Juliano, B. 0. (1964). Glutamic
acid decarboxylase activity as a viability index of artificially dried and stored
rice. Cereal Chern. 41:188.
42. Yasumatsu, K., Moritaka, S., Ishii, K., Shimazono, H., and Fujita, E. (1965).
Studies on cereals. I. Change of chemical characteristics of polished rice dur-
ing storage. 1 Japan. Soc. Food Nutr., 18:123.
43. Desikachar, H. S., and Subrahmanyan, V., (1960). The relative effects of
enzymatic and physical changes during storage on the culinary properties of
rice. Cereal Chern., 37:1.
44. Matsuda, H., and Hirayama, 0. (1973). Changes of lipid components and
lipolytic acyl-hydrolase activities in rice grains during their storage. Nippon
Nogei Kagaku Kaishi, 47:379.
45. Shibuya, N., Iwasaki, T., and Chikubu, S. (1975). Lipase activity in rice
kernel. Shokuhin Sogo Kenkyusho Kenkyu Hokoku, 30:10.
46. Chrastil, J. (1988). Enzymic product formation curves with the normal or
diffusion limited reaction mechanism and in the presence of substrate recep-
tors. Int. J. Biochern., 20:683.
47. Morrison, W. R. (1978). Cereal lipids, Advances in Cereal Science and Technol-
ogy 11 (Y. Pomeranz, ed.)., American Association of Cereal Chemists, St.
Paul, MN, p. 221.
48. Chrastil, J., and Parrish, F. W. (1987). Phospholipases C and Din rice grains.
1 Agric. Food Chern., 35:624.
49. Yasumatsu, K. (1965). Studies on the effects of lipid on the rheological proper-
ties of starch. Takeda Chemical Industries, Ltd., Osaka, Japan, p. 1.
50. Barber, S., Benedito de Barber, C., and Tortosa, E. (1967). Composition and
localization of lipids in rice grains. Rev. Agroquirn. Tecnol. Alirnentos, 7:235.
78 CHRASTIL

51. Chen, W., and Huang, W. H. (1947). Acidity and quality of stored rice.
Fukien Agric. 1, 8:10.
52. Tsuchiya, T., and Kinomura, S. (1949). Change of oil during storage of rice
polishings. II. Nippon Oil Techno[. Soc., 2:30.
53. Hunter, I. R., Houston, D. F., and Kester, E. B. (1951). Development offree
fatty acids during storage of brown (husked) rice. Cereal Chern., 28:232.
54. Houston, D. F., Hunter, I. R., McComb, E. A., and Kester, E. B. (1954).
Cereal storage effects. Deteriorative changes in the oil fraction of stored
parboiled rice.1 Agric. Food Chern., 2:1185.
55. Yasumatsu, K., and Moritaka, S. (1964). Fatty acid composition of rice lipids
and their changes during storage. Agric. Bioi. Chern (Tokyo), 28:257.
56. Shibuya, N., Iwasaki, T., and Chikubu, S. (1977). Studies on deterioration of
rice during storage. II. Role of the free fatty acids in the changes of rheologi-
cal properties of cooked rice and its paste during storage of rice. Denpun
Kagaku, 24:67.
57. Shoji, I., and Kurasawa, H. (1981). On the viscosity of rice and rice starch by
Brabender amylogram. III. The viscosity changes of the brown rices (gluti-
nous, non-glutinous rices) stored for some periods in room temperature.
Kaseigaku Zasshi, 32:350.
58. Shin, M. G., Yoon, S. H., Rhee, J. S., and Kwon, T. (1986). Correlation
between oxidative deterioration of unsaturated lipid and n-hexanal during
storage of brown rice.1 Food Sci., 51:460.
59. Ohta, H., Aibara, S., Yamashita, H., Sekiyama, F., and Morita, Y. (1990).
Post-harvest drying of fresh rice grain and its effects on deterioration of lipds
during storage. Agric. Bioi. Chern (Tokyo), 54:1157.
60. Juliano, B. 0. (1966). Physicochemical data on the rice grain. Inter. Rice Res.
Inst. Tech. Bull., 6:1.
61. Kik, M. C. (1945). Effect of milling, processing, washing, cooking and storage
on thiamine, riboflavin and niacin in rice. Arkansas Univ. Agr. Expt. Sta.
Bull., 458:1.
62. Kik, M. C., and Williams, R. R. (1945). The nutritional improvement of
white rice. Bulletin 112th National Research Council, National Academy Sci-
ence, Washington, DC.
63. Pelshenke, P. F., and Hampel, G. (1967). Studies of the effect of extreme
storage conditions on white rice. Milling, 149:192.
64. Yanai, S., Ishtani, T., and Kakiuchi, Y. (1979). Influence of gaseous environ-
ment on hermetic storage of brown rice. Nippon Shokuhin Kogyo Gakkaishi,
26:25.
65. Lin, W. M., Chang, S. M., and Lii, C. Y. (1979). Studies on the changes of
physicochemical properties of rice during storage. Bull. Inst. Acad. Sin.,
26:13.
66. Desikachar, H. S. R., and Subrahmanyan, V. (1959). Expansion of new and
old rice during cooking. Cereal Chern., 36:385.
67. Primo, E., Casas, A. Barber, S., and Benedito de Barber, C. (1962). Quality
of Rice. VII. Organoleptic and pohysicochemical properties of rice grains.
Rev. Agroquirn. Tecnol. Alirnentos, 2:241.
EFFECT OF STORAGE ON RICE 79

68. Shibuya, N., Iwasaki, T., Yanase, H., and Chikubu, S. (1974). Studies on
deterioration of rice during storage. I. Changes of brown rice and milled rice
during storage.1 Japan. Soc. Food Sci. Techno/., 21:597.
69. Kunze, 0. R., and Choudhury, M.S. U. (1972). Moisture adsorption related
to the tensile strength of rice. Cereal Chern., 49:684.
70. Halick, J. V., Beachell, H. M., Stansel, J. W., and Kramer, H. H. (1960). A
note on the determination of gelatinization temperatures of rice varieties.
Cereal Chern., 37:670.
71. Barber, S., Benedito de Barber, C., and Burches, J. (1966). Gelatinization of
rice grains and doughs. Rev. Agroquim. Tecnol. Alimentos, 6:99.
72. Chrastil, J. (1991). Influence of storage on supercooling of rice starch and rice
flour gels. J. Agric. Food Chern., 39:1729.
73. Irwin, M. I. (1959). Cooking quality of milled white rice. Arkansas Farm Res.,
8:12.
74. Tani, T., Chikubu, S., and Iwasaki, T. (1964). Changes of quality of husked
rice during low temperature storage. 1.1 Japan Soc. Food Nutr., 16:436.
75. Normand, F. L., Hogan, J. T., and Deobald, H. J. (1964). Improvement of
culinary quality of freshly harvested rice by heat treatment. Rice J., 67:7.
76. Pushpamma, P., and Reddy, M. U. (1979). Physicochemical changes in rice
and jowar stored in different agro-climatic regions of Andhra Pradesh. Bull.
Grain Techno/., 17:97.
77. Ramarathnam, N., and Kulkarni, P. R. (1984). Changes during storage of
rice. Z. Ernahrungswis., 23:143.
78. Sreenivasan, A. (1939). Storage changes in rice after harvest. Indian 1 Agric.
Sci., 8:208.
79. Moritaka, S., Sawada, K., and Yasumatsu, K. (1971). Studies on cereals. VII.
Relation between lipid content of milled rice and deterioration of rice flavors
during storage. Eiyo To Shokuryo, 24:474.
80. Hwangbo, J. S., and Lee, S. R. (1976). Changes in eating quality and lipid com-
ponents of Tongil rice variety in storage. Korean 1 Food Sci. Techno/., 8:74.
81. Okabe, M. (1979). Texture measurement of cooked rice and its relationship to
the eating quality.1 Texture Studies, 10:131.
82. Perez, C. M., and Juliano, B. 0. (1981). Texture changes and storage of rice.
1 Texture Studies, 12:321.
83. Tsugita, T., Ohta, T., and Kato, H. (1983). Cooking flavor and texture of rice
stored under different conditions. Agric. Bioi. Chern. (Japan), 47:543.
84. Juliano, B. 0. (1985). Criteria and tests for rice grain qualities, Rice: Chemis-
try and Technology (B. 0. Juliano, ed.), American Association of Cereal
Chemists, St. Paul, MN, p. 443.
85. Juliano, B. 0. (1985). Rice bran, Rice: Chemistry and Technology (B. 0.
Juliano, ed.), American Association of Cereal Chemists, St. Paul, MN, p. 1.
86. Kester, E. B. (1960). Rice investigations at Western Regional Research Labo-
ratory. Rice J., 63:28.
87. Yasumatsu, K., Moritaka, S., and Karinuma, T. (1964). Effect of the change
during storage in lipid composition of rice on its amylogram. Agric. Bioi.
Chern. (Tokyo), 28:265.
80 CHRASTIL

88. Glabe, E. F. (1939). Some factors influencing the viscosity of rice flour
suspensions. Cereal Chern., 16:661.
89. Kerr, R. W. (1950). Chemistry and Industry of Starch. Academic Press, New
York.
90. Mitchell, W. A., and Zillman, E. (1951). The effect of fatty acids on starch
and flour viscosity. Trans. Am. Assoc. Cereal Chern., 9:64.
91. Gray, V. M., and Schoch, T. J. (1962). Effects of surfactants and fatty ad-
juncts on the swelling and solubilization of granular starches. Stiirke, 7:239.
92. Yasumatsu, K., and Moritaka, S. (1964). Changes of characteristics of starch
during gelatinization in the presence or absence of fatty acids. J. Food Sci.,
29:198.
93. Schoch, T. J. (1942). Non-carbohydrate substances in the cereal starches. 1
Am. Chern. Soc., 64:2954.
94. Moritaka, S., Sawada, K., and Yasumatsu, K., (1971). Studies on cereals.
VI. Effect of fat extraction on rice quality and storage deterioration of
defatted rice. Eiyo To Shokuryo, 24:457.
95. Desikachar, H. S. R., and Subrahmanyan, V. (1957). The curing of freshly
harvested paddy. I. Principles of curing. J. Sci. lnd. Res. (India), 16A:365.
96. Bhattacharya, K. R., Desikachar, H. S. R., and Subrahmanyan, V. (1964).
Curing of freshly harvested rice by heat treatment. Indian 1 Techno/., 2:378.
97. Yasumatsu, K., Moritaka, S., and Wada, S., (1966). Studies on cereals. V.
State flavor of stored rice. Agric. Bioi. Chern. (Tokyo), 30:483.
98. Mitsuda, H., Yasumoto, K., and Kimikazu, 1., (1968). Analysis of volatile
components in rice bran. Agric. Bioi. Chern. (Tokyo), 32:453.
99. Mitsuda, H., Kawai, F., and Yamamoto, A. (1972). Underwater and under-
ground storage of cereal grains. Food Techno/., 26:50.
100. Tsuzuki, E., Tanaka, K., and Shida, S. (1981). Studies on the characteristics
of scented rice. VIII. Effect of storage conditions on some fatty acid composi-
tions and flavor components in rice grains. Kenkyu Hokoku Miyazaki
Daigaku Nogakubu, 28:31.
101. Tani, C. Y., Ohta, T., and Kato, H. (1983). Cooking flavor and texture of
rice stored under different conditions. Agric. Bioi. Chern. (Tokyo), 47:543.
102. Obata, Y., and Tanaka, H. (1965). Studies on the photolysis of L-cysteine
and L-cystine. Formation of the flavor of cooked rice from L-cysteine and L-
cystine. Agric. Bioi. Chern. (Tokyo), 29:191.
103. Desikachar, H. S. R. (1977). Preservation of by-products of rice milling,
Proceedings of Rice By-Products Utilization, Vol. 2 (S. Barber and E.
Tortosa, eds.), Inst. Agroquim. Tecnol. Aliment., Valencia, Spain, p. 1.
104. Yokochi, K. (1977). Rice bran processing for the production of rice bran oil
and characteristics and uses of the oil and deoiled bran. Proceedings of Rice
By-products Utilization, Vol. 3 (S. Barber and E. Tortosa, eds.), Inst.
Agroquim. Tecnol. Aliment., Valencia, Spain, p. 1.
105. Ryu, C. H., and Cheigh, H. S. (1980). Fractionation of rice bran lipid and
storage effects on bran lipid composition. Korean 1 Food Sci. Techno/.,
12:278.
EFFECT OF STORAGE ON RICE 81

106. Loeb, J. R., Morris, N. J., and Dollear, F. G. (1949). Rice bran oil. IV.
Storage of bran as it affects hydrolysis of the oil. J. Am. Oil Chern. Soc.,
26:738.
107. Srimani, B. N., Chattopadnyay, P., and Bose, A. N. (1977). Stabilization of
rice bran. I. Direct measurement of the lipase activity in rice bran and the
methods for the inactivation of the same. Proceedings of Rice By-products
Utilization, Vol. 2 (S. Barber and E. Tortosa, eds.), Inst. Agroquim. Tecnol.
Aliment., Valencia, Spain, p. 1.
108. Yokochi, K. (1978). Efficiency test with a new type of stabilizer in Burma.
Rice Report 1977 (S. Barber, H. Mitsuda, H. S. R. Desikachar, and E.
Tortosa, eds.), Inst. Agroquim. Tecnol. Aliment., Valencia, Spain, p. 1.
109. Hawkey, R., and Hawkey, T. (1978). Socioeconomic feasibility study of a rice
bran stabilization unit for Thailand and the Philippines. United Nations In-
dustrial Development Organization, Vienna, Austria.
110. Sayre, R.N., Saunders, R. M., Enochian, R. V., Schultz, W. G., and Beagle,
E. C. (1982). Review of rice bran stabilization systems with emphasis on
extrusion cooking. Cereal Foods World, 27:317.
111. Schultz, E. B., and Morgan, R. P. (1979). Appropriate technology for
village-level rice bran processing and utilization in developing countries.
Report CDT 79/1. Center for Development Technology, Washington Univer-
sity, St. Louis, MO, p. 1.
112. Barber, S., and Benedito de Barber, C. (1980). Rice bran: Chemistry and
technology, Rice: Production and Utilization (B.S. Luh ed.), Avi Pub!. Co.,
Inc., Westport, CT, p. 1.
113. Chang, S.C., Saunders, R. M., and Luh, B.S. (1980). Rice oil: Chemistry
and technology, Rice: Production and Utilization (B.S. Luh ed.), Avi Publish-
ing Co., Inc., Westport, CT, p. 1.
114. Cornelius, J. A. (1980). Rice bran oil for edible purposes. A review. Trop.
Sci., 22:1.
115. Barber, S., and Benedito de Barber, C. (1984). Rice bran: An under-utilized
raw material. United Nations Industrial Development Organization, Pub!.
E.84.II.B.3. U.N., New York, p. 1.
116. Randall, J. M., Sayre, R. M., Schultz, W. G., Fong, R. Y., Mossman, A. P.,
Tribelhorn, R. E., and Saunders, R. M. (1985). Rice bran stabilization by
extrusion cooking for extraction of edible oil. J. Food Sci., 50:361.
117. Rothe, M. (1967). Inactivation of rice bran enzymes by heating and stabiliza-
tion of brown rice, International Problems of Modern Cereal Processing and
Chemistry. Part I. Milling, Inst. Getreideverarbeitung, Bergholz-Rehbriicke,
Germany, p. 1.
 5
Breeding for Rice Quality
KentS. McKenzie
California Cooperative Rice Research Foundation, Biggs, California

I. INTRODUCTION
The predominant cultivated species of rice ( Oryza sativa L.) was probably
first domesticated in Southeast Asia (1 ). Archeological evidence of rice culti-
vation in China 7000 years ago has been reported (2). Three ecogeographic
races of 0. sativa have evolved. Indica is the tropical race, Japonica is the
temperate race, and Javanica, believed to have evolved in Indonesia, is inter-
mediate between Indica and Japonica (3). In addition to the partial sterility
oflndica X Japonica hybrids, climatic adaptation, and morphological charac-
teristics, certain general quality characteristics have been associated with the
Indica and Japonica races. Indica types generally have long slender grains
and cook dry and flaky, whereas Japonica types have short round grains and
cook moist and sticky. Hybridization between types and selection for certain
grain shape and quality characteristics have made this classification of qual-
ity less definitive and useful. Genetic studies of isozyme polymorphisms indi-
cated that modern U.S. rice cultivars were grouped with the classical Japon-
ica and Javanica types and were distinct from the classical Indicas (4,5).
However, the grain shape and cooking characteristics of U.S. long-grain
cultivars, as will be discussed later, are not typical of classical Japonicas.
Rice was introduced to North America in the early 1600s, and by about
1690 commercial rice production was firmly established in South Carolina
(6,7). In the late 1800s rice became an established crop in southwest Louisi-

83
84 MCKENZIE

Table I Average Annual Rice Production in the United States 1986-1990

Area Grain type (%)
Yield
State (ha x 1000) (%) (kg/ha) Long Medium Short
Arkansas 452 43 5940 89 11 <1
Louisiana 195 18 5130 65 35
California 158 15 8460 8 81 11
Texas 132 13 6690 98 2
Mississippi 92 8 6090 99 1
Missouri 30 3 5720 97 3
Florida• 3 <1 5040 100
Total (Avg) 1062 100 (6270) 79 19 2
•Estimated values only.
Source: Compiled from annual crop summaries 1986-1990 crop Reporting Board, Statisti-
cal Research Service, USDA.

ana and spread to Texas, Arkansas, and California by the early 1900s.
Today, there are three major rice-producing regions in the United States:
the Mississippi Delta (Arkansas, Mississippi, northeast Louisiana, and
southwest Missouri), the Gulf Coast (southwest Louisiana, and southeast
Texas), and the Central Valleys of California. A relatively small acreage of
rice is also grown commercially in Florida. Table 1 summarizes some of the
more recent rice production statistics for the United States.
Rice is a semiaquatic crop and is adapted to growing in flooded soil
conditions. It is grown under the more productive "lowland" (versus "up-
land") conditions where fields are under a shallow flood for most of the
growing season. Rice is direct seeded in the United States (drilled, broad-
cast, or water-seeded) in contrast to transplanting, which predominates in
Asia. It is an annual crop, although plants can be maintained for many years
by vegetative propagation. Rice is sensitive to cool temperatures, which
limit its production areas and growing season. Ratoon or second cropping is
practiced in the Gulf Coast region, which involves fertilizing and reflooding
the rice stubble of early-maturing varieties harvested in July or early Au-
gust. Yields and quality of this second crop are generally low, but input
costs are reduced and ratooning is a popular practice in some areas.

II. BREEDING OBJECTIVES AND METHODS

A. Breeding Programs
In 1899, the U.S. Department of Agriculture (USDA) introduced Kyushu
rice from Japan and distributed it for on-farm testing in southwest Louisi-
BREEDING FOR RICE QUALITY 85

ana and Texas and formally began rice improvement in the United States
(8). Introduction of varieties from other parts of the world represented the
first step in rice improvement, and many of these introductions were used
directly or selections were made from original introductions. These early
varieties are the ancestors of our modern U.S. cultivars, and they certainly
had a major role in the evolution of U.S. rice grain quality. Dilday (9)
recently examined the ancestral relationships of the rice cultivars devel-
oped by the U.S. public rice-breeding programs. The report indicated that
all parental germplasm can be traced back to 22 plant introductions in the
southern region and 23 plant introductions in California. Calculations also
showed that current rice cultivars are closely related. This lack of genetic
diversity is considered undesirable because it increases the potential ge-
netic vulnerability of the crop. However, it has allowed the continued
development of improved cultivars with required quality characteristics
that are similar to and consistent with older cultivars and established mar-
ket types. Considerable effort is made to introduce foreign germplasm into
rice-improvement programs. Breeding experience has shown that recover-
ing desired rice quality characteristics is not always easy, especially when
using new introductions or a different germplasm base.
In 1951, an agronomically superior long-grain cultivar, Century Patna
231, was released but proved to be unacceptable for traditional long-grain
cooking and processing (10). To help deal with this problem, a rice quality
research laboratory was established at Beaumont, Texas, to perform specific
physicochemical tests to guide rice breeders in developing new cultivars with
desirable cooking and processing characteristics (11). This illustrates the
importance of quality in U.S. rice-breeding programs and marketing.
Cooperative rice-breeding programs were established between the
USDA and the state agricultural experiment stations starting in 1909.
Breeders were established at state agricultural experiment stations in Ar-
kansas, California, Louisiana, and Texas by the USDA-Agricultural Re-
search Service (USDA-ARS). A state rice-breeding program funded en-
tirely by state rice growers was established in California in 1969. The
agricultural experiment stations have assumed responsibility for state
breeding programs in Arkansas, Louisiana, Mississippi, and Florida in
recent years. Rice-breeding programs and activities have increased mark-
edly in the last 20 years. Over 50 new rice cultivars have been released
since 1970, and they have played a primary role increasing U.S. rice
yields. Public rice-breeding programs are located at Rice Research &
Extension Center, Stuttgart, AR; Rice Experiment Station, Biggs, CA;
Rice Research Station, Crowley, LA; Texas Agricultural Research & Ex-
tension Center, Beaumont, TX; Delta Branch Experiment Station, Stone-
ville, MS; and Everglades Research and Extension Center, Belle Glade,
86 MCKENZIE

FL. Public cultivars have accounted for more than 90% of the U.S. acre-
age. Private breeding programs are present in several states, and some are
relatively young programs. The major private programs include Busch
Agricultural Resources, Inc., Pleasant Grove, CA, and Jonesboro, AR;
N. F. Davis Drier & Elevator, Inc., Firebaugh, CA; Rice Researchers,
Inc., Glenn, CA, and RiceTec, Alvin, TX.

B. Breeding Methods
Rice is a member of the grass family and has a panicle inflorescence with
one floret per spikelet. It has a perfect flower (both male and female),
which consists of two interlocking glumes (hulls), six anthers, and a single
ovary. Rice is naturally self-pollinating, and rice growers in the United
States use pure-line cultivars. Hybrid rice is grown extensively in China and
has been under experimental evaluation in the United States for several
years.
Pedigree breeding is the predominant method used in rice improvement.
It may include modifications including backcrossing and the bulk method.
The first step involves the selection of parents for artificial hybridization.
The parents may be introductions, old cultivars, breeding lines, or muta-
tions with certain desired characteristics. The designated female parent is
emasculated by removing the anthers and fertilized with pollen from the
desired male (12,13). The resulting F 1 seed is grown to produce the F2 seed,
which is usually planted at a low seeding rate for selection by the breeder.
The latter is the first segregating generation, and individual panicles are
selected for advancement and reselection. The F2 population size may vary
from 1,000 to 10,000 or more individuals. Selections are usually made on
the basis of visual characteristics. Selected panicles are grown in rows for
advancement and reselection for several more generations. The rice contin-
ues to inbreed by natural self-pollination. Genes become fixed (homozy-
gous) for the many characters, genetic segregation decreases, and the prog-
eny rows breed true. From about the F4 generation on, promising selections
are included in more advanced screening tests for agronomic and quality
characteristics. The outstanding few lines that survive this continuing
screening, selection, purification, and advancement process are increased
for release as pure-line cultivars to growers through seed-production pro-
grams. Winter nurseries in Puerto Rico or Hawaii are used to hasten
cultivar development by allowing the growing of two or more generations
per year of the most promising material. Development time from the initial
cross to release to growers is often 10 years or more.
Induced mutation has made a major contribution to cultivar develop-
ment in California (14). Adapted cultivars and germplasm were irradiated
BREEDING FOR RICE QUALITY 87

and selections made for desirable mutants for changes in plant height,
maturity, and even grain characteristics. This technique had the advantage
of being a relatively rapid and simple method, which maintained the
agronomic adaptation and quality characteristics of parental material.
Calrose 76 (15) was a semidwarf version of Calrose, the old established tall
medium-grain cultivar and conventional market type in California. In-
duced mutants were used directly, but their contribution as parents in
breeding new cultivars has been even more important. Mutation breeding
techniques were also used to generate the semidwarf and early maturing
cultivars M-401 and M-203 from the tall proprietary premium quality
medium-grain cultivar Terso (16,17). The proprietary cultivar KRM-2 was
developed using mutation breeding on another tall proprietary premium
quality medium-grain cultivar, Kokuhorose, by Rice Researchers, Inc. In-
duced mutation also has been used to develop cultivars with waxy endo-
sperms and changes in kernel shape (18,19).

C. Breeding Objectives
Each breeding program has its own rice-improvement objectives, and its
project objectives reflect the needs of the production area. Factors that
influence production, grain quality, and market demands determine the
profitability of rice production and are major considerations in establishing
breeding objectives. Objectives will vary among programs and growing
regions, but some are common to rice-improvement programs.
Grain yield is one of the primary breeding objectives. Yield evaluation is
one of the main activities in the later stages of screening breeding material,
and many genetic and environmental factors determine yield. Selection is
done for numerous characters contributing to yield potential including
panicle size, panicle number, grain size, plant height, and straw strength.
Yield is frequently negatively correlated with quality, and the breeder must
compromise between those two important objectives. Very high yielding
lines often exhibit poor grain quality characteristics, making them unaccept-
able for release as cultivars. Conversely, materials with high or unique
quality characteristics usually are poorly adapted or agronomically inferior
and therefore unacceptable.
Disease and insect resistance are other important characteristics, and
considerable efforts are made by breeders, pathologists, and entomologists
in screening, selecting, and developing cultivars resistant to the diseases and
insects that attack rice in the United States. Diseases and insect pests are a
greater problem in the southern rice-producing regions than in California.
Diseases and insect pests usually have detrimental effects on quality as well
as yield, thus pest resistance is a doubly desirable feature. Unfortunately,
88 MCKENZIE

sources of resistance usually have undesirable agronomic and quality charac-

teristics. As in the case of yield, breeders must struggle to recover the desired
resistance in combination with acceptable quality characteristics.
Rice may face a number of environmental stresses that can adversely
affect grain yield and quality. Cool temperatures at seeding and stand
establishment stages are often encountered in rice production. Improving
seedling vigor is an objective in many breeding programs. Cool tempera-
tures prior to flowering (during pollen microsporogenesis) may induce ste-
rility (blanking) in rice in California, hence resistance to blanking is a
major breeding objective there. This is also the primary reason cultivars
grown in the southern United States are not grown commercially in Califor-
nia, i.e., cultivars from the southern United States are susceptible to cold.
The abiotic disease "straighthead," which causes the panicles to be sterile
and distorted, is a problem in some areas of the southern rice region, and
resistance to this disease is desirable. Cultivars also must be tolerant to the
herbicides needed to control weed pests.
"Every variety is a compromise" is a saying used by plant breeders
regarding the end product of their labors, a new cultivar. Rice breeders set
objectives, select parents, hybridize, select, and test their material. The
ultimate goal is to recover the desirable features or enhancements of the
parents. Seldom is the ultimate goal achieved because of the many limita-
tions ranging from time and cost factors to genetic barriers. Balances are
made between the traits of importance. If warranted, the new cultivar is
released and the recurrent breeding process continues. More in-depth dis-
cussions of rice breeding in the United States and other countries are
available in the literature (12,13,20-23).

III. BREEDING FOR QUALITY CHARACTERISTICS

Rice quality must certainly be considered one of the primary rice-breeding
objectives in the United States. Kernel size, shape, and translucency and
acceptable milling, cooking, and processing characteristics are needed to
satisfy the requirements for the export markets and the diverse and increas-
ing domestic markets. Quality characteristics are considered and evaluated
by the breeders from the initial selection of parents through the final deci-
sion to release an experimental line as a new cultivar. Quality evaluations
begin with selection for simple visual characteristics and advance to physico-
chemical and milling tests. Quality information is collected on breeding
lines over many generations and several years of testing. Quality testing is
typically done during the winter months when there is no crop in the field.
Evaluation and selection for kernel size and shape and milling are con-
ducted by the breeding program with more sophisticated physicochemical
BREEDING FOR RICE QUALITY 89

tests typically performed as a service for the public breeding programs by

the USDA-ARS Rice Quality Research Laboratory, Beaumont, TX. Ad-
vanced lines under consideration for release as new cultivars may be pro-
vided to marketing organizations for evaluation. Information on the yield
performance, physical, milling, cooking, and processing characteristics of
public cultivars is reported in registration articles in Crop Science.

A. Market Types
Rice in the United States is classified and marketed by grain (kernel) size
and shape. The three primary market classes are long-, medium-, and
short-grain types (Fig. 1). Table 2 contains the grain size and shape charac-
teristics for the current market types. In addition to the traditional long,
medium, and short grains, there is a general category termed special pur-
pose types. This would include aromatic, glutinous (waxy) rices, and rices
developed for ethnic and industrial uses. This is a relatively small portion of
the U.S. production, however, interest and breeding efforts are increasing
in this area. Historically, and through planned breeding, each of these
market types is associated with specific cooking, eating, and processing

Figure 1 U.S. long-, medium-, and short-grain rices shown in the rough (paddy),
brown, and milled rice forms.
90 MCKENZIE

Table 2 Grain Dimensions of Traditional U.S. Market Classes of Rice

Length/Width ratio
Grain Brown rice length
type (mm) Rough Brown Milled
Long 6.61-7.5 3.4:1 and more 3.1:1 and more 3.0:1 and more
Medium 5.51-6.6 2.3:1-3.3:1 2.1:1-3.0:1 2.0:1-2.9:1
Short :5':5.5 2.2:1 and less 2.0:1 and less 1.9:1 and less
Source: Refs. 24, 25.

characteristics and product uses (8,24). Several physicochemical tests are

used by breeders to identify and select for the proper cooking characteris-
tics in their breeding programs. The tests include determination of appar-
ent amylose content (26-28), alkali spreading value of whole kernel milled
rice soaked in dilute alkali as an indicator of gelatinization temperature
(29), amylographic gelatinization temperature (30), amylographic gela-
tinization and paste viscosity characteristics (31), protein content (32), and
parboil-canning stability (33). Table 3 contains a summary of the physico-
chemical characteristics of U.S. rice cultivars and will be used in the discus-
sion of breeding for different market classes.

B. Long Grains
The majority of rice-breeding activity in the United States has been directed
toward the traditional long-grain market. These types occupy the majority of
U.S. acreage (see Table 1), and this market class is the major focus of the
rice-breeding programs in the southern United States. California, with its
need for cold-tolerant cultivars, has traditionally produced medium- and
short-grain rices. Long-grain cultivars and germplasm from the southern
United States generally perform poorly in California. Recent breeding ef-
forts have resulted in the release of an adapted long-grain cultivar, L-202
(36), for that environment. Because of the importance and requirements for
this market class, breeding and screening work for long-grain quality has
received major emphasis in the United States. Proper kernel size and shape
are requirements for long grains as well as the other types. Kernel length,
width, and thickness are under quantitative genetic control and have high
heritability, making these characteristics relatively easy to select for by sim-
ple visual screening. A wide combination of kernel sizes and shapes will
occur in segregating generations depending on the parents involved in the
cross. If there are large differences in kernel size and shape in the parents,
recovery of the desired kernel size and shape may be more difficult. In
crosses between long grains and short grains, for example, the frequency of
 Table 3 Average Physicochemical, Cooking, and Processing Characteristics of U.S. Rice when Grown in Southern Locations 1985-1988

Alkali Amylographic viscosity (Brabender units) Parboiling

spreading Gelatinization canning
Quality Amylose score temp./type Hotpaste Cool paste Break- Protein stability
Cultivar type (%)• (avg) (OC)b Peak 15 min/95°C 50°C down' Setbackct (%)• (%loss)

[ Gul~unt
Lemont
]
Traditional
southern
21.8
22.7
3.9
3.4
70.5/I
70.3/I
730
750
400
390
790
780
-330
-360
60
20
8.9
8.6
23
24
New bonnet long-grain 22.9 3.4 70.6/1 690 380 720 -310 40 8.6 24
Tebonnet types 22.7 3.4 70.5/1 780 430 780 -360 0 8.7 25
L-202 California 26.5 3.3 71.6/1 460 310 790 -150 340 9.4 23
Rexmont Newrex 26.0 3.7 70.7/I 810 500 1060 -310 250 9.4 19

Mars Southern 15.2 6.2 64.0/l 800 380 650 -410 -150 6.9 37
Rico 1 Southern 16.7 6.3 64.3/I 760 380 650 -380 -110 7.5 35
M-201 Calrose 14.2 6.7 -/I 870 390 620 -490 -250 8.6 39
M-2021 Calrose 16.1 7.0 61.2/I 600 300 440 -300 -160
M-401 Premium 17.1 7.0 59.7/l 480 260 450 -220 -30
Kokuhorose Premium 17.3 7.0 60.8/I 590 310 500 -280 -90
S-201 Short grain 16.7 6.0 64.8/I 760 360 640 -410 -120 7.5 37

Della Aromatic 22.5 3.6 69.2/l 690 380 730 -310 40 9.5 27
A-301 Aromatic 21.7 3.3 70.9/I 620 350 670 -270 -50 9.7 30
Jasmine 85 Aromatic 15.6 6.5 -/L 850 410 760 -440 -90 8.9
Calmochi-101 Waxy 0 5.9 61.7/L 320 200 270 -130 -30 8.8 46
TOR0-2 Toro 17.2 6.1 64.6/L 730 370 630 -360 -100 8.9 38

•Milled rice apparent amylose 12% moisture basis.

bf=intermediate, L=low.
'Breakdown viscosity=peak - hotpaste.
ctSetback viscosity=peak - coolpaste.
•Brown rice % N x 5.95 dry basis.
1 M-202, M-401, and Kokuhorose samples from CA 1990 only.

Source: Refs. 34, 35.

92 MCKENZIE

acceptable long grains or short grains in the offspring is very low. Intense
selection, large populations, and backcrossing may be needed to achieve
the desired kernel size and shape. Because of these and other quality
problems, the number of these types of crosses made in breeding programs
may be limited. Visual selection for kernel size and shape can begin in the
early generations by making selections in the field or laboratory. As mate-
rial becomes more advanced and promising, kernel measurements may be
taken to select among siblings. This is a slow and tedious task, typically
done by hand. Image analysis systems have been developed to aid in kernel
size and shape determinations (37). Grain dimensions are affected by envi-
ronmental factors such as temperature. Grain length may be reduced signifi-
cantly in low-temperature environments. Table 4 contains milled rice ker-
nel dimensions for some U.S. long-grain cultivars.
Other aspects of the size and shape of the kernel beside length and width
are also important. Larger kernels are often desired by the market for
appearance or processing, however, increases in kernel length may increase
breakage leading to reduced milling yields. Processing conditions for pro-
ducing parboiled and quick cooking rice products are also affected by
kernel size, even within long-grain types, and must be adjusted. Curvature
or distortion of the kernel as well as variability in size among grains on the
panicle are undesirable characteristics. Selection may be made for a straw

Table 4 Milled Rice Grain Dimensions of Some U.S. Long-, Medium-, and
Short-grain Rice Cultivars
Milled rice kernel
dimensions (mm) Length/ Kernel
Width weight
Cultivar Origin Type• Length Width Thickness ratio (mg)
Gulfmont Texas L 6.7 2.2 1.6 3.1 18
Lemont Texas L 6.8 2.2 1.7 3.0 19
Newbonnet Arkansas L 6.5 2.0 1.6 3.3 16
Tebonnet Arkansas L 6.6 2.0 1.6 3.2 17
L-202 California L 7.2 2.2 1.7 3.5 20
Rexmont Texas L 6.7 2.1 1.6 3.2 17
M-201 California M 5.8 2.6 1.8 2.3 21
Mars Arkansas M 5.8 2.5 1.8 2.3 19
Mercury Louisiana M 5.8 2.5 1.8 2.4 19
Rico 1 Texas M 5.4 2.6 1.8 2.1 19
S-201 California s 5.3 3.0 2.0 1.8 23
•Grain type where L = long-grain, M = medium-grain, and S = short-grain.
Source: Refs. 35, 36.
BREEDING FOR RICE QUALITY 93

hull color in preference to gold or other colors, which might effect the final
appearance of the rice. Colored bran types used in ethnic and gourmet
markets (red-, purple-, and amber-pigmented aleurone layers) are unac-
ceptable in all traditional U.S. market types. Selections are made for kernel
translucency and freedom from chalkiness or "white belly." High degree of
whiteness of the milled kernel is desirable. Whiteness meters are used by
the milling industry to measure color and may be of use in screening for this
characteristic. Cooking and processing characteristics are not determined
by the size and shape of the kernel but through planned breeding are
associated with grain size and shape, i.e., long-, medium-, and short-grain
cooking types.
Traditional southern U.S. long grains cook dry and fluffy with the ker-
nels remaining separate when steamed or boiled. The physicochemical
(cooking quality) characteristics of U.S. long-grain cultivars have been
summarized by Webb et al. (24) and are presented in Table 3. They are
characterized by (a) intermediate amylose content (21-24% ), (b) a moder-
ate alkali spreading score (indicating intermediate gelatinization tempera-
ture), (c) intermediate- to high-amylographic gelatinization temperature,
(d) intermediate amylographic paste viscosity peak height, relatively high
breakdown viscosity, and a slight positive set back, and (e) relatively low
parboil-canning stability solid loss with little splitting and fraying of edges
and ends of canned kernels. The California long-grain cultivar L-202 has
similar physicochemical characteristics, except it is 2% higher in amylose
content than the traditional southern cultivars, gelatinization temperature
is slightly lower, it has a weaker amylographic viscosity profile, and cooking
and processing do not exactly match the traditional U.S. long grains.
Amylose content is controlled by one or two major genes with modifiers
(38,39). Selection for amylose content can be done in early generations, if
necessary, as in the case of widely different parents. Ifthe parents are tradi-
tionallong-grain types, tests for amylose content may be delayed until the
advanced generations. Gelatinization temperature is evaluated by alkali
spreading scores, and this characteristic is controlled by a few genetic fac-
tors, is highly heritable, and screening can be done effectively in the early
generations (39,40). Amylose content and gelatinization temperature are
influenced by environmental conditions, especially temperature. Amylose
content and gelatinization of the same rice grown in different regions will
vary. For example, California-grown rice will typically have higher amylose
content and lower gelatinization temperature than the same rice cultivar
grown in Texas. Differences in physicochemical reaction are also observed in
rices planted at different dates and also in first crops and ratoon crops.
The use of amylose content and alkali reaction score has generally been
quite effective in selecting for the traditional long-grain market type in the
94 MCKENZIE

breeding programs in the southern United States. This may have been in
part due to the core germplasm pool used for cultivar development in that
region. The experience in California long-grain breeding has shown that
selecting for amylose content and alkali reaction is not sufficient to recover
adapted long grains with traditional long-grain quality (41). Micro-cooking
tests have been used at the Rice Experiment Station, Biggs, CA, to screen
several thousand rows annually for cooking quality. Amylographic viscosity
testing is also being used to help in screening. L-202 and related materials
are being used extensively for their agronomic characteristics by breeders
in the southern region. Consequently, material with that parentage is being
screened more intensively for long-grain quality including measuring amylo-
graphic viscosity.
In the early 1980s a new type of long-grain rice that cooks drier and
fluffier combined with an improved canning stability and low washout
losses in processing became available with the release of the cultivar
Newrex (42). The "Newrex quality" found in Newrex and Rexmont (43)
differs from traditional southern long-grain types by having (a) lower
amylographic breakdown viscosity and higher setback viscosity, (b) signifi-
cantly lower solids loss in processing, indicating less canned kernel splitting
and fraying, and with improved grain stability and integrity, (c) a 2-4%
higher amylose content, and (d) a slightly firmer cooked kernel texture
(44,24). This superior long-grain quality is a quality objective in several
long-grain rice-breeding programs. Several genetic factors are involved,
and the inheritance of this quality type is under investigation. Screening for
high amylose content (25-28% ), intermediate gelatinization temperature,
and the Newrex amylographic viscosity profile is used in selection for
Newrex quality. Newrex and Rexmont are grown on a limited acreage.
However, this improved processing type may become more common in
future long-grain cultivars as milling quality and field yields are improved.

C. Medium and Short Grains

Medium-grain cultivars occupy a much smaller part of the U.S. acreage
than long grains, except in California, where medium grains account for
about 90% of the acreage (see Table 1). Short-grain cultivars are essentially
only grown in California, and the acreage has declined in recent years. Just
as long-grain breeding and quality have been emphasized in the southern
United States, medium-grain breeding and quality have been stressed in
California. The southern U.S. long-grain types are recognized in the mar-
ket for their long-grain quality characteristics, and the Calrose medium
grains grown in California are recognized for their particular quality charac-
teristics. Kernel size and shape for medium and short grains must fall in the
BREEDING FOR RICE QUALITY 95

defined range of dimensions for the market type (see Table 2). Kernel size
and shape for medium- and short-grain cultivars do show notable differ-
ences. Calrose-type medium grains like M-201 (45) have larger kernels
than Mars (46), the predominant medium grain grown in the southern
United States. Rico 1 (47), a new southern medium grain has a different
medium-grain kernel shape than Mars. S-201 (48), the major California
short grain, has a large round kernel with a distinctive white belly on the
kernel. Kernel dimensions of these cultivars are presented in Table 4.
Kernel translucency and freedom from chalkiness and white belly are
very important characteristics in medium and short grains, and these grain
types often show such undesirable kernel characteristics. With increasing
kernel size, there is often an increase in chalkiness and/or white belly.
Screening for shape, translucency, and freedom from white belly is initiated
in the F2 generation in California (49). Brown rice samples from panicles
selected from the field are examined each generation in the laboratory
during the winter. Translucency is also selected on the basis of milled
samples in later generations. In addition to these characteristics, selections
are discarded for kernel distortion (notches, grooves, and twists), nonuni-
formity of kernels, and high frequencies of green and immature kernels.
Recovering translucent short grains with large kernel size similar to S-201
has been particularly difficult. Selections for translucency have tended to
give reduced kernel size and oblong shape in comparison to S-201. Short
grain cultivars like S-101 (50) with smaller, more translucent kernels have
been released, however, a larger translucent kernel would be desirable.
Mutation-breeding techniques were used by Hu (18) to develop Valencia 87
from the proprietary cultivar Calpearl, which shows an improved short-
grain kernel.
Traditional U.S. medium and short grains cook moist and sticky when
steamed or boiled. These cultivars are characterized by (a) a low amylose
content, (b) a high alkali spreading score (indicating a low gelatinization
temperature), (c) a low gelatinization temperature, (d) relatively low
amylographic paste viscosity, and (e) high solid loss in parboil-canning
stability with extensive splitting and fraying of kernel edges and ends (24).
Table 3 contains a summary of the physicochemical characteristic of some
U.S. medium-grain cultivars.
The Calrose type of medium-grain rice grown in California is preferred
over the southern medium grains by some ethnic groups (Japanese and
Korean) and for manufactured products. In other uses both the southern
and California types are acceptable. These ethnic groups can distinguish
between these types in taste tests, and the reasons for preferences in some
manufactured products for Calrose types are not known. Unidentified qual-
ity characteristics found in the California germplasm are probably a major
96 MCKENZIE

contributing factor. California-grown rice usually shows slightly higher

amylose content and lower gelatinization temperatures than rice grown in
the southern United States, suggesting that the growing environment may
be important.
Quality selection in traditional medium- and short-grain breeding would
usually focus on grain characteristics. Physicochemical tests are not per-
formed in early generations unless a different quality type is in the pedi-
gree. Physicochemical tests are used in the later stages on advanced lines.
Advanced experimental lines (of all grain types) under consideration for
release as a cultivar are often provided to marketing organizations, proces-
sors, or independent quality evaluators. Evaluation sheets may be used to
make comparisons with current cultivars in cooking, processing, and taste
panel evaluations.
California also produces rices referred to as "premium quality medium
grains." These are recognized for their cooking quality and are preferred
by certain ethnic groups, especially Japanese and Korean customers. The
appearance, texture, and taste of this rice is judged to be superior to other
medium grains by these groups. Generally these rices are very glossy after
cooking, sticky but have a smooth texture, and they remain soft as they
cool. Aroma and taste are also cited as important features. M-401 (public)
and Kokuhorose (proprietary) are the predominant cultivars used in this
market. M-401 and the early maturing M-203 were mentioned earlier as an
example of the use of mutation breeding for rice improvement. Rice Re-
searchers, Inc. has a major emphasis in this market type and has developed
several improved versions in its KR series of releases. Busch Agricultural
Resources, Inc. has recently released a proprietary premium quality me-
dium grain, SP-411. There has been considerable expansion in breeding for
this premium quality market in both the public and private sector in recent
years. Breeding efforts are also under way to develop premium quality
short-grain cultivars. These types would be similar to premium quality
medium grains and/or the high quality Japanese short-grain cultivars, like
Koshihikari. The high quality Japanese types have cooking characteristics
similar to premium quality medium grains.
Developing improved premium quality rice is a formidable breeding
challenge. The cooking characteristics of these types are not identifiable
with the physicochemical tests used for rice mentioned earlier. Screening
becomes a particular problem since the cooking tests used on breeding lines
are laborious and variable. Expert taste panels are needed to evaluate
advanced materials. The premium quality characteristics are certainly con-
trolled by a number of genetic factors, in addition to environmental fac-
tors, making identification and transfer of these quality characteristics very
difficult. The breeder is faced with the challenge of matching the cooking
BREEDING FOR RICE QUAUTY 97

characteristics of a standard like Kokuhorose, where subtle differences in

cooking quality may cause rejection by the marketing organization. If the
quality is satisfactory, a selection must then show some improved agro-
nomic feature to justify its release. Thus, replacing a premium quality
cultivar is difficult, and some market standards like Kokuhorose in Califor-
nia and Koshihikari in Japan have been grown for several decades.

D. Special Purpose Types

A small percentage of the U.S. rice area is planted with special purpose
types that have quality characteristics that distinguish them from the tradi-
tional U.S. market classes. They are often grown on a contract basis with a
marketer and cannot be blended with the traditional market classes. Sev-
eral are sold in specialty markets and used in ethnic foods. The agronomic
performance of these cultivars is usually below conventional cultivars. Inter-
est in these specialty rices has increased recently and many breeding pro-
grams include some work on these types, although this amounts to only a
small portion of the total rice-breeding effort. Table 3 contains the result of
physicochemical tests on some of the special purpose cultivars. Aromatic
rice is probably the most familiar of these types. This rice has a "pop-
cornlike" aroma and is marketed under various trade names. Della (51), a
long grain from Louisiana, is grown commercially and is the donor parent
for several of the newer cultivars like Texmati (proprietary), A-301 (52),
and Dellmont (53). The physicochemical characteristics of these cultivars
are generally similar to the traditional long grains with the addition of the
aromatic characteristic (Table 3). Jasmine 85 (54) is a low amylose long-
grain aromatic introduction, which was released to fill an ethnic market
using imported Thai Jasmine. Its quality and market evaluations are in
progress. A long-grain aromatic rice with colored bran, Wehanni, is mar-
keted in the brown rice specialty market by Lundberg Family Farms,
Richvale, CA. Basmati rice is another major aromatic type, which is native
to Pakistan, India, and the Middle East. This is fine-grained aromatic rice
which cooks dry and fluffy and shows extreme kernel elongation during
cooking. There are no true basmati rice cultivars in commercial production
in the United States, as they are poorly adapted and have low grain and
milling yields. However, several breeding programs are working to develop
basmati types suitable for production in the United States.
The aromatic character is relatively easy to identify in breeding lines.
The scent can be detected in leaf tissue (55-57), raw kernesl (58,59), and,
of course, cooked rice. The inheritance of the aromatic characteristics has
been reported to be controlled by a single gene in crosses with Della (60)
and as many as three genes in other studies (59,61). Differences in intensity
98 MCKENZIE

and aroma exist. Research has indicated that 2-acetyl-1-pyrroline is a major

aroma component (62) in aromatic rice. Research on the aromatic charac-
teristics of rice is currently being investigated by researchers at Louisiana
State University. Kernel elongation appears to be independent of aroma, is
difficult to transfer, and is controlled by several genetic factors (63). Stud-
ies on kernel elongation in some crosses indicate that relatively few genes
may be involved (C. N. Bollich, personal communication).
Another important special purpose rice is the glutinous or waxy type
that is sometimes referred to as "sweet" or "mochi." Waxy rice starch is
essentially all amylopectin, making the kernels white and opaque. Waxy
rice may be used as whole milled kernels or flour. It cooks very sticky and is
used in a variety of ethnic foods including mochi cake and crackers. Waxy
rice flour is also used in processed foods because of its processing character-
istics. Waxy rice flour is resistant to liquid separation during freezing and
thawing (64,65) and is used in frozen food sauces and gravies.
In the United States, waxy rice is commercially grown in California under
a contract with a processor or marketer. The major public cultivar is
Calmochi-101 (66). Mochigome is an old proprietary variety, introduced
from Japan, that is preferred for traditional mochi cake. N.F. Davis Drier,
Inc. has in production two recently waxy cultivars, NFD-108, and NFD-109.
These cultivars are all short-grain types with low gelatinization tempera-
tures. The waxy characteristic is not restricted to short grains, and a number
of waxy types with long or medium grains as well as differing quality charac-
teristics can be found throughout the world. The waxy characteristic is con-
trolled by a single recessive gene. It is easy to visually identify a waxy kernel
in brown or milled rice, and thus waxy types can be readily identified in
segregating breeding lines. The waxy endosperm can be verified by the
distinctive brown (in contrast to blue) color produced when treated with an
iodine solution. Different waxy genes as well as modifying genetic factors are
present in rice, and presence of a waxy endosperm does not guarantee the
rice will meet market requirements. This again indicates that rice quality is
very complex and not well understood. Calmochi-101 is used to produce
mochi cake but is considered less desirable than Mochigome. The preferred
quality characteristics for waxy rices in their different uses is not known.
Work has been carried out by Bean et al. (67) characterizing the cooking
behavior of different waxy varieties. Work on developing waxy cultivars is
continuing in California and to a lesser extent elsewhere in the United States.
There is a small local market in southwest Louisiana that prefers a low
amylose long grain that cooks soft and sticky similar to traditional medium
and short grains. The founding cultivar was Toro; a more recent cultivar is
TOR0-2 (68). There is some breeding interest in this type of soft cooking
long grain for certain export and domestic markets.
BREEDING FOR RICE QUALITY 99

E. Milling

Milling yield largely determines the economic value of rough rice, thus
making this a very important rice quality characteristic. It is measured as a
percent value for head (whole kernel milled rice) and total (whole and
broken kernel milled rice) on a given lot of rice. Head rice sells for a much
higher price than broken kernels, and achieving high head rice yields is a
mutual goal in commercial production and cultivar development. Improved
milling yields have been the primary improved feature of some new culti-
vars (69). In some cases losses in head rice yield in improved higher yield-
ing cultivars have occurred (70). U.S. rice-breeding programs invest consid-
erable effort in screening and selecting their breeding material for milling
yield.
Milling yield is determined by both environmental and genetic factors
and is a very complex characteristic. Many environmental factors are
known to influence milling quality. Harvest moisture (70-72), field drain-
age (73), fluctuations in relative humidity and temperature (74), nitrogen
fertility, ripening, disease or other stresses, combine adjustments (72), and
grain processing can have major impact on milling yield. In screening breed-
ing material, the breeder will try to control environmental factors, includ-
ing the use of replicated tests on advanced materials. Genetic differences in
milling characteristics of cultivars are widely recognized, and there is con-
tinuing interest in research on this rice quality characteristic. Screening for
milling yield usually focuses on head rice. Rapid visual screening for per-
cent broken kernels in brown rice samples is used in California beginning in
the F2 generation. A high correlation between whole kernel brown rice
after shelling and head rice has been reported (75). In evaluating brown
rice samples for kernel and translucency, lines that show high or low per-
cent broken kernels are easily identified. As breeding lines advance, actual
milling yields are determined on a small portion of the seed harvested from
rows. Efforts are made to harvest each line within the optimum harvest
moisture range conducive to good milling yields. This would be 18-22%
and 21-24% for the southern states and California, respectively. Harvest
moisture on each sample may be determined, and the samples are air dried
to storage moisture (12-14%) for milling at a later time.
Standard milling procedures, used by public breeders, have been devel-
oped in coordination with the USDA-ARS Rice Quality Research Labora-
tory. A 125-gram sample is shelled and milled with a McGill No.2 sample
mill for 30 seconds with a defined weight and arm length (depending on the
grain type). Broken milled kernels are graded out, and a head and total
milled rice yield is determined. This general procedure is used for determin-
ing milling yields on more advanced lines in replicated yield tests.
 100 MCKENZIE

In more advanced yield testing stages, individual rows of experimental

entries may be grown in the breeding nurseries for milling determination.
Duplicate rows may be used to obtain replicate sampling or to allow har-
vest of an optimum and low harvest moisture in the hopes of identifying
entries that maintain high milling yields at suboptimal harvest moisture. On
the most advanced breeding material, milling samples may be collected
twice a week for several weeks. This allows milling yield determinations
over a range of harvest moistures and provides information on the milling
performance of the line, including the optimum harvest moisture for head
rice. Figure 2 shows the results of harvest moisture milling studies on two
very early medium-grain cultivars. M-103 shows higher and more stable
head rice yields than M-101. Milling yield data are collected over several
years and in different locations in the breeding and testing process. Gravois
et al. (76) has recently published a study of genotype and environment
effects on head rice yields that focuses on this area of quality evaluation.
Because of the complexity of milling and the influence on it of both
environmental and genetic factors, researchers are trying to identify major
traits controlling milling yield. Fissuring of the kernel has been studied
extensively by Kunze (74) and others (77, 78). Varietal differences in resis-
tance to breakage exist, and screening techniques to select for resistance to
kernel fissuring as a head rice selection tool have been proposed and stud-
ied (74,78). Studies at the International Rice Research Institute (IRRI)
Los Banos, The Philippines, have investigated the presence of grains in rice
panicles which have a higher volume and weight. These high density grains
produce higher milling and head rice recovery (79). Varietal differences for

80
70
........ 60
....,
114!

u 50
w
0::
Cl 40
L5
:c 0 - M-101, R2 =0.93**
30
e -- M-103, R2 =0.63**
20
10
10 15 20 25 30 35
HARVEST MOISTURE (%)

Figure 2 Effect of harvest moisture on head rice yields of M-101 and M-103 in
1988 and 1989, Rice Experiment Station, Biggs, CA.
BREEDING FOR RICE QUALITY 101

the number of high density grains per panicle exist, and increasing the
number of high density grains may improve head rice yields. Jong-
kaewwattana (80) conducted a comprehensive study of the factors influenc-
ing rice-milling quality in some California cultivars. The study, which fo-
cused on panicle and kernel characteristics, indicated that selection for
uniformity in grain characteristics (such as grain size, weight, and density)
should have positive impact on head rice yield. Heritability studies on
panicle length, panicle moisture content, grain length, width, weight, vol-
ume, and density are in progress (81). Preliminary results show these traits
to have high heritability estimates, while the heritability estimates for head
yield, as expected, were lower. In another study of a cross involving the
long-grain cultivars Lemont and Leah, heritability of head rice yield was
estimated to be 0.29 (82).

IV. FUTURE DIRECTIONS

Research on rice is enjoying a relatively high level of interest in the United
States. Rice-breeding programs have undergone considerable expansion in
both the public and private sectors in the past decade. The Rice Technical
Working Group, a group of rice researchers that meets biennially, is con-
tinuing to grow in membership and participation. Research in biotech-
nology has become extensive due to the successes using rice, international
importance of the crop, and research funding. Rice consumption has in-
creased in the United States, and rice is being viewed as a nutritional,
healthful food by the public. Some new products using rice are also moving
into the market place, all suggesting a very positive future for rice in the
United States (see Chapter 19).
There also are a number of concerns to mention in looking to the future.
Agriculture is under intense scrutiny in terms of its impact on the environ-
ment and government support programs. Improved cultivars in combina-
tion with pesticides and chemical fertilizers have made possible the high
quality and productivity found in modern U.S. rice production. Their use is
becoming tightly regulated and restricted, although the United States is
fortunate not to have many of the important rice pests found elsewhere.
Several products have been eliminated due to environmental risks and
concerns. Development and registration of new replacement chemicals is
difficult and expensive, especially for a relatively small crop like rice in the
United States. Water is another area of concern-both the amount needed
for the crop as well as the fate of any residual materials in the water used in
rice production. Price supports for rice and other crops are under continu-
ing revision, and support for these programs is declining. Rice growers
have a good track record for funding rice research. However, budget restric-
102 MCKENZIE

tions faced by universities, the USDA, and the growers themselves will
affect the extent of future research activities.

A. New Techniques
Biotechnology is recognized as the major new area in agricultural research.
Because of the new and theoretical potential of this field, some unrealistic
or overly optimistic expectations in terms of the quick and direct applica-
tion of this technology to rice breeding have occurred. The technology, in
its different forms, will contribute to rice improvement in the short term
and even more so in the long term. It will greatly increase our understand-
ing of the plant, enhance rice-improvement programs rather than replace
them, and should not be viewed as a genetic panacea. An extensive amount
of work using rice is continuing both nationally and internationally.
Biotechnology in the United States can be grouped as either direct support
of a breeding program or more basic genetic research.
The Louisiana State University Rice Research Station (RRS) in Crowley,
LA, has developed a large tissue culture program to support cultivar develop-
ment at that facility. Early in the program the focus was on developing useful
somaclonal mutants for direct use or in crossbreeding. More recently the
focus of direct breeding support is in the area of anther culture. This is a
technique that allows the regeneration of true breeding lines (doubled hap-
loids) from anthers of rice. The primary advantage of this method is to
hasten the conventional breeding process, which requires several genera-
tions after a cross to have lines reach a true breeding stage (homozygosity)
for evaluation and testing. RRS is regenerating about 10,000 plants annually
in their program. The USDA-ARS and Texas A&M University have re-
cently started an extensive biotechnology program at the Texas Agricultural
Research and Extension Center, Beaumont, TX. Anther culture and genetic
studies in support of rice improvement are in progress there. Preceding this
research expansion, a new rice cultivar Texmont (83), which was produced
using anther culture techniques, was released by Texas A&M and the
USDA-ARS in 1990. Anther culture is also being used in other public and
private programs. Anther culture may become a common tool used for rice
improvement, but some limitations exist. The process is labor intensive and
requires at least some laboratory facilities and technical support. Regenera-
tion rates are typically low ( <10% ), and some germplasm, like Indica types,
is very difficult to regenerate. Research advances that improve these aspects
of anther culture will expand its use in rice improvement.
More basic research is being conducted by many universities in U.S.
production areas like Louisiana State University, Texas A&M University,
and the University of California, Davis. Because of international interest
BREEDING FOR RICE QUALITY 103

and grants, rice research is underway at other universities like Cornell,

Wisconsin, Georgia, Purdue, Stanford, and the USDA Plant Gene Expres-
sion Center, Albany, CA.
Rice has become a popular plant species for molecular genetic study and
manipulation and is being used as a model system for monocots (V. Walbot,
personal communication). This is because rice is relatively easily trans-
formed. Protoplasts can be regenerated into plants in a reliable fashion,
and much of rice tissue is highly embryogenic, lending itself to recovery of
transgenic plants through "ballistic" methods of genetic insertion. Work by
these laboratories is more basic in nature or may focus on problems of
somewhat less importance in the United States. In the future it will contrib-
ute to a much better understanding of rice, improvements in technology, as
well as cultivar development. This work in molecular biology is beyond the
scope of this chapter. Research in this area is so extensive and evolving so
rapidly that The Rockefeller Foundation is publishing a quarterly report,
Rice Biotechnology Quarterly, to hasten progress.
One particular area to note is the genetic research by the group at
Cornell (Ithaca, NY) on mapping rice genome using restriction fragment
length polymorphism (RFLP) mapping. This technique is of particular in-
terest to breeders and geneticists as a tool for identifying important genetic
characters in segregating populations. This would be very useful for charac-
teristics that are currently difficult to identify allowing improved selection
effectiveness and hastening cultivar improvement. Tanksley et al. (84) have
published a discussion of this subject. Other sophisticated mapping tech-
niques are under investigation at the University of California, Davis, and
correlations of molecular markers and quality factors are being sought by
researchers in Georgia and Texas.
Research using biotechnology methods to directly influence quality was
conducted by Shaeffer and Sharpe (85) on rice protein and amino acid
content. Endosperm protein mutants in rice were recovered by biochemical
selections with inhibitory levels of lysine and threonine from tissue culture.
These mutants have been studied extensively in regard to the nature of the
protein and amino acid (lysine and methionine) increases and endosperm
chalkiness. Their research indicated that improved lysine types can be gener-
ated in a predictable manner using biochemical selection, some high lysine/
high protein types with low chalkiness were recovered, and this material can
be used in rice-improvement programs for nutritional quality improvement
(G. W. Shaeffer, personal communication). Rice plants can be successfully
regenerated from protoplasts (single cells), an important step for the use of
genetic engineering in transferring genetic characters into rice. This has not
been the case in some other important crop species. Transfers may include
genes from other species or new synthesized genes, which could enhance the
104 MCKENZIE

quality, starch characteristics, or nutritional value of rice. Transgenic rice

plants are being produced by a number of research groups. Such techniques
may yield new or improved quality types for future rice cultivars. It is also
hoped that genetic engineering techniques will allow improvements of rice
cultivars by increasing productivity, tolerance to stress, and broader adapta-
tion in material with desirable quality characteristics.

B. Quality and New Types

Although breeders use a number of physical criteria and physicochemical
tests to screen and identify desired rice grain quality, there is still much that
is not understood. Breeding for rice quality would benefit greatly from
further research and the development of new techniques in this area. Physi-
cochemical tests used to screen cooking and processing behaviors of rice
are not always indicative in breeding material with new or different
germplasm bases, suggesting that other factors are controlling the quality.
The nature of rice starch, proteins, endosperm construction, and aromatic
and flavor components are of interest in breeding for quality. In some
quality types (like the premium quality medium grains in California), no
rapid accurate physicochemical criteria or methods exist to identify these
types. A "rice taste analyzer" has been developed by Satake Engineering
Co., which can quantitatively measure the taste of rice as a digital number
from 1 to 100. The system is based on near-infrared reflectance (NIR) and
is correlated with Japanese quality standards (amylose, protein, moisture,
and fatty acids) and taste panels results. NIR techniques are being studied
by the USDA Western Research Laboratory, Albany, CA. The develop-
ment of new tests and the use of new tools and machinery could be very
beneficial to rice improvement. A better understanding of rice quality and
the role that environmental conditions play would also be helpful in rice
improvement, processing, and utilization.
Interest by breeders and marketing organizations has increased in recent
years in new market types of rice. These types may be for specialty and
ethnic markets or rice designed for special processing uses. Brown rice is
marketed by several organizations. The appearance of the brown rice is
important, and the breeder selects for traits such as color and the absence
of immature or green seeds. Proprietary varieties of colored bran (red,
purple) and aromatic brown rice are successfully marketed by the Lund-
berg Family Farms. Large seeded medium- and short-grain European culti-
vars (with white belly or high degree of chalkiness) like Arborio demand
premium prices and are popular in dishes like risotto and paella. A limited
amount of breeding for these markets is being conducted. Figure 3 shows
some specialty types of rice.
BREEDING FOR RICE QUALITY 105

Figure 3 Specialty types of rice with very large kernels (Arborio), waxy endo-
sperm, or colored bran in comparison to the premium quality medium-grain
cultivar M-401.

A large volume of broken rice is used in brewing and other processes

where head rice yield is of minor importance. Busch Agriculture Re-
sources, Inc. has identified this as one of their breeding interests. Market-
ers frequently express interest in developing a very high yielding cultivar
where quality characteristics are of minor importance. LA 110 (86) was
developed as an industrial purpose variety. However, it was never exten-
sively grown. In practice, most breeders will keep, incorporate, and use
any exceptionally high yielding lines in their program. Thus, the question is
whether the yield advantage is large enough to compete with a broken rice
by-product at a much lower price. Ifdemand for broken milled rice exceeds
supply and a higher stable price develops, such high yielding, low milling
cultivars may appear in the future.
106 MCKENZIE

Breeding programs at Arkansas and Texas are examining the potential

for long-grain types with high amylose contents and low gelatinization
temperatures characteristic of some Asian or South American cultivars.
Processors have expressed interest in these types for parboiling and other
processed products. These types have characteristically been discarded in
the selection process but may appear as future special purpose cultivars.
Some breeding work has been initiated to develop very high amylose
types. An amylose extender mutant developed by K. Okuno in Japan was
used to develop a very high amylose lineae (36M4) at the IRRI (87). This
line was introduced to the United States and apparent amylose content of
samples grown in the United States and analyzed at the USDA-ARS Rice
Quality Research Laboratory was 40% amylose, approximately 10%
higher amylose than any other known rice (B. D. Webb, personal communi-
cation). Analysis also showed that milled rice of ae (36M4) had a total
dietary fiber (TDF) content of ca. 7.5% in comparison to 0.5% for conven-
tional U.S. rice cultivars (88). The high TDF makes this of interest for a
natural, relatively high fiber, rice-based food for health- and calorie-
conscious consumers. Other uses of very high amylose types might include
extruded noodles, food for diabetics (87), or use in amylose films for pack-
aging or other industrial processes.

REFERENCES
1. Chang, T. T. (1976). Rice, Evolution of Crop Plants (N.W. Simmonds, ed.),
Longman Group Ltd., London, p. 98.
2. Chang, K. C. (1981). In search of China's beginnings: New light on an old
civilization. Am. Sci., 69(2):148.
3. Chang, T. T. (1976). The origin, evolution, cultivation, dissemination, and
diversification of Asian and African rice. Euphytica, 25:425.
4. Glaszmann, J. C. (1986). A varietal classification of Asian cultivated rice
based on isozyme polymorphism, Rice Genetics, Proc. Int. Rice Gen. Symp.,
IRRI, Manila, Philippines, p. 88.
5. Glaszmann, J. C. (1986). Varietal classification of rice based on isozyme
polymorphism-focus on U.S. rices, Proc. 21st Rice Tech. Work. Group Hous-
ton, TX, Feb. 24-26, 1986. Texas Agr. Exp. Sta., College Station, p. 23.
6. Adair, C. R. (1972). Production and utilization of rice, Rice Chemistry and
Technology (D. F. Houston, ed.), Am. Assoc. Cereal Chemists. St. Paul, MN,
p.l.
7. Holder, S. H., Jr., and Grant, W. R. (1979). U.S. Rice Industry. Economics,
Statistics, and Cooperative Service. AER No. 433. Washington, D.C.
8. Adair, C. R., Bollich, C. N., Bowman, D. H., Jodon, N. E., Johnston, T. H.,
Webb, B. D., and Atkins, J. G. (1973). Rice breeding and testing methods in
BREEDING FOR RICE QUALITY 107

the United States, Rice in the United States Varieties and Production, Agricul-
tural Handbook No. 289, USDA-ARS, Washington, D.C.
9. Dilday, R. H. (1990). Contribution of ancestral lines in the development of
new cultivars of rice. Crop Sci., 30:905.
10. Webb, B. D. (1975). Cooking, processing and milling qualities of rice, Six
Decades of Rice Research in Texas, Res. Monograph No. 4. Texas Agr. Exp.
Sta., College Station, p. 97.
11. Webb, B. D., and Stermer, R. A. (1972). Criteria for rice quality, Rice Chemis-
try and Technology (D. F. Houston, ed.), Am. Assoc. Cereal Chemists, St.
Paul, MN, p. 102.
12. Jennings, P. R., Coffinan, W. R., and Kauffman, H. E. (1979). Rice Improve-
ment, Int. Rice Res. lnst., Los Banos, Philippines.
13. McKenzie, K. S., Bollich, C. N., Rutger, J. N., and Moldenhauer, K. A. K.
(1987). Rice, Principles of Cultivar Development, Vol. 2, Crop Species (W. R.
Fehr, ed.), Macmillan Publ. Company, New York, p. 487.
14. Rutger, J. N. (1990). Mutation breeding of rice in California and the United
States of America. Plant Mutation Breeding for Crop Improvement, Proc. of
Int. Symp. on the contribution of plant mutation breeding to crop improve-
ment, June 18-22, 1990. IAEA/FAO, Vienna, p. 155.
15. Rutger, J. N., Peterson, M. L., and Hu, C. H. (1977). Registration of Calrose
76 rice. Crop Sci., 17:978.
16. Carnahan, H. L., Johnson, C. W., Tseng, S. T., and Brandon, D. M. (1981).
Registration of 'M-401' rice. Crop Sci., 21:686.
17. Carnahan, H. L., Johnson, C. W., Tseng, S. T., Oster, J. J., Hill, J. E., and
Brandon, D. M. (1989). Registration of 'M-203' rice. Crop Sci., 29:1089.
18. Hu, C. H. (1988). Family tree of California Japonica rice varieties with special
focus on cultivars developed from private firms, Proc. 22nd Rice Tech. Work.
Group, Davis, CA, June 26-29, 1988. Texas Agric. Exp. Sta., College Sta-
tion, p. 31.
19. Carnahan, H. L., Johnson, C. W., Tseng, S. T., Oster, J. J., and Hill, J. E.
(1979). Registration of 'Calmochi-201' rice. Crop Sci., 19:746.
20. Chang, T. T., and Li, C. E. (1980). Genetics and breeding, Rice: Production
and Utilization (B.S. Luh, ed.), AVI, Westport, CT, p. 87.
21. Toriyama, K. (1979). National program of rice breeding in Japan. lap Agr.
Res. Quart., 13(1):1.
22. International Rice Research Institute (IRRI). 1972. Rice Breeding, IRRI, Los
Banos, Philippines.
23. Poehlman, J. M. (1987). Rice breeding, Breeding Field Crops, 3rd ed. AVI,
Westport, CT, p. 340.
24. Webb, B. D., Bollich, C. N., Carnahan, H. L., Kuenzel, K. A., and Mc-
Kenzie, K. S. (1985). Utilization characteristics and qualities of United States
rice, Rice Grain Quality and Marketing, IRRI, Manila, Philippines, p. 25.
25. Federal Grain Inspection Service (1990). Federal Grain Inspection Service
Rice Inspection Handbook. United States Department of Agriculture, Wash-
ington, D.C.
108 MCKENZIE

26. Williams, V. R., Wu, W. T., Tsai, H. Y., Bates, H. G. (1958). Varietal differ-
ences in amylose content of rice starch. J. Agric. Food Chern., 6:47.
27. Juliano, B. 0. (1971). A simplified assay for milled-rice amylose. Cereal Sci.
Today, 16(10):334, 340.
28. Webb, B. D. (1972). A totally automated system of amylose analysis in whole
kernel milled rice, Proc. 14th Rice Tech. Work. Group, Davis, CA, June 20-
22, 1972. Texas Agr. Exp. Sta., College Station, p. 62.
29. Little, R. R., Hilder, G. B., and Dawson, E. H. (1958). Differential effect of
dilute alkali on varieties of milled rice. Cereal Chern., 35:111.
30. Halick, J. V., Beachell, H. M., Stansel, J. W., and Kramer, H. H. (1960). A
note on determining gelatinization temperature of rice varieties. Cereal
Chern., 37:670.
31. Halick, J. V., and Kelley, V. J. (1959). Gelatinization and pasting characteris-
tics of rice varieties as related to cooking behavior. Cereal Chern., 36:91.
32. American Association of Cereal Chemists, Inc. (AACC) (1962). AACC Ap-
proved Methods (formerly Cereal Laboratory Methods, Method 46-13, 7th
ed., St. Paul, MN.
33. Webb, B. D., and Adair, C. R. (1970). Laboratory parboiling apparatus and
methods of evaluating parboiled canning stability of rice. Cereal Chern., 47:708.
34. Webb, B. D., Bollich, C. N., Jackson, B. R., Kanter, D. G., Linscombe, S.
D., Moldenhauer, K. A. K., Tseng, S. T., and Peterson, H. D. (1989). Evalua-
tion of rice quality compounds for named varieties grown in uniform perfor-
mance tests in Arkansas, Louisiana, Mississippi, Texas, and California (1986-
1988). Cooperative rice quality report 1981-1, Beaumont, TX.
35. Webb, B. D., Bollich, C. N., Jackson, B. R., Kanter, D. G., Linscombe, S.
D., Moldenhauer, K. A. K., Tseng, S. T., and Peterson, H. D. (1988). Evalua-
tion of rice quality components for named varieties grown in uniform perfor-
mance tests in Arkansas, Louisiana, Mississippi, Texas, and California (1985-
1987). Cooperative rice quality report 1988-1, Beaumont, Texas.
36. Tseng, S. T., Carnahan, H. L., Johnson, C. W., Oster, J. J., Hill, J. E., and
Scardaci, S. C. (1984). Registration of 'L-202' rice. Crop Sci., 24:1213.
37. Paige, D., Geng, S., and Jongkaerwattana, S. (1991). Apparatus for automatic
measurement of kernel weight, length, and thickness. Crop Sci., 31:1314.
38. Bollich, C. N., and Webb, B. D. (1973). Inheritance of amylose in two hybrid
populations ofrice. Cereal Chern., 50:631.
39. McKenzie, K. S., and Rutger, J. N. (1983). Genetic analysis of amylose con-
tent, alkali spreading score, and grain dimensions in rice. Crop Sci., 23:306.
40. Tseng, S. T., and Webb, B. D. (1984). A possible explanation for trans-
gressive segregation for gelatinization temperature in certain rice crosses,
Proc. 20th Rice Tech. Work. Group, Lafayette, LA, June 25-27, 1984. Texas
Agr. Exp. Sta., College Station, p. 27.
41. Tseng, S. T., Webb, B. D., Bollich, C. N., and Brandon, D. M. (1988).
Genetic and environmental influence on long-grain cooking quality, Proc.
22nd Rice Tech. Work. Group, Davis, CA, June 26-29, 1988. Texas Agr. Exp.
Sta., College Station, p. 33.
BREEDING FOR RICE QUALITY 109

42. Bollich, C. N., Webb, B. D., Marchetti, M. A., and Scott, J. E. (1980).
Registration of 'Newrex' rice. Crop Sci., 20:285.
43. Bollich, C. N., Webb, B. D., Marchetti, M. A., and Scott, J. B. (1990).
Registration of 'Rexmont' rice. Crop Sci., 30:1160.
44. Webb, B. D., and Bollich, C. N. (1980). Cooking and processing quality
characteristics of Newrex rice, Proc. 18th Rice Tech. Work. Group, Davis,
CA, June 17-19, 1980. Texas Agr. Exp. Sta., College Station, p. 61.
45. Carnahan, H. L., Johnson, C. W., Tseng, S. T., Oster, J. J., and Hill, J. E.
(1982). Registration of 'M-201' rice. Crop Sci., 22:1087.
46. Johnston, T. H., Wells, B. R., Marchetti, M. A., Lee, E. N., and Henry, S. E.
(1979). Registration of 'Mars' rice. Crop Sci., 19:743.
47. Bollich, C. N., Webb, B. D., Marchetti, M. A., and Scott, J. B. (1990).
Registration of 'Rico 1' rice. Crop Sci., 30:1161.
48. Carnahan, H. L., Johnson, C. W., Tseng, S. T., and Brandon, D. M. (1980).
Registration of 'S-201' rice. Crop Sci., 20:551.
49. Johnson, C. W. (1988). Procedures for breeding high quality short-and
medium-grain rices for California, Proc. 22nd Rice Tech. Work. Group Davis,
CA, June 26-29, 1988. Texas Agr. Exp. Sta., College Station, p. 34.
50. Johnson, C. W., Carnahan, H. L., Tseng, S. T., Oster, J. J., Hill, J. E., Rutger,
J. N., and Brandon, D. M. (1989). Registration of 'S-101' rice. Crop Sci.,
29:1090.
51. Jodon, N. E., and Sonnier, E. A. (1973). Registration of 'Della' rice. Crop
Sci., 13:773.
52. Tseng, S. T., Carnahan, H. L., Johnson, C. W., Oster, J. J., and Hill, J. E.
(1987). Registration of 'A-301' rice. Crop Sci., 27:1310.
53. Bollich, C. N., Hung, H. H., Webb, B. D., Scott, J. E., and Marchetti, M.A.
(1992). Registration of 'Dellmont' rice. Crop Sci. (submitted).
54. Bollich, C. N., Jackson, B. R., Khush, G. S., Webb, B. D., Scott, J. E.,
and Marchetti, M. A. (1991). Registration of 'Jasmine 85' rice. Crop Sci.
(submitted).
55. Choudhury, D., and Gosh, A. K. (1978). Evaluation of agronomic and physi-
cochemical characteristics of fine and scented rice varieties. Indian J. Agric.
Sci., 48:573.
56. Ghose, R. L. M., and Butany, W. T. (1952). Studies on the inheritance of
some characters in rice (Oryza sativa L.). Indian Genetics Plant Breed.,
12:26.
57. Kadam, B. S., and Palankar, V. K. (1938). Inheritance of aroma. Chrom.
Bot., 4:32.
58. Dhulappanavar, C. V. (1976). Inheritance of scent in rice. Euphytica, 25:659.
59. Reddy, P. R., and Sathyanarayanaiah, K. (1980). Inheritance of aroma in rice.
Indian J. Plant Breed., 40:327.
60. Berner, D. K., and Hoff, B. J. (1986). Inheritance of scent in American long-
grain rice. Crop Sci., 26:876.
61. Tomar, J. B., and Nanda, J. S. (1983). Inheritance of aroma in rice (Oryza
sativa L.). Madras Agric. J., 70:491.
110 MCKENZIE

62. Buttery, R. B., Ling, L. C., Juliano, B. 0., and Turnbaugh, J. G. (1983).
Cooked rice aroma and 2-acetyl-l-pyrroline. J. Agric. Food Chem., 30:826.
63. Khush, G. S., and Juliano, B. 0. (1985). Breeding for high-yielding rices of
excellent cooking and eating qualities, Rice Grain Quality and Marketing,
IRRI, Manila, Philippines, p. 61.
64. Deobald, H. J. (1980). Rice flour, Rice Chemistry and Technology (D. F. Hous-
ton, ed.), American Association of Cereal Chemists, St. Paul, MN, p. 264.
65. Luh, B.S., and Liu, Y. K. (1980). Rice flours in baking, Rice: Production and
Utilization. AVI Westport, CT, p. 470.
66. Carnahan, H. L., Johnson, C. W., Tseng, S. T., Oster, J. J., and Hill, J. E.
(1986). Registration of 'Calmochi-101' rice. Crop Sci., 26:197.
67. Bean, M. M., Esser, C. A., and Nishita, K. D. (1984). Some physicochemical
and food application characteristics of California waxy rice varieties. Cereal
Chem., 61:475.
68. McKenzie, K. S., Jodon, N. E., Brandon, D. M., Rush, M. C., Robinson, J.
F., and Miller, M. F. (1984). Registration of 'TOR0-2' rice. Crop Sci.,
24:1212.
69. Moldenhauer, K. A. K., Gravois, K. A., Norman, R. J., Lee, F. N., Helms,
R. S., Wells, B. R., Dilday, R. H., and Rohman, P. C. (1991). Registration of
'Millie' rice. Crop Sci., 31:229.
70. Geng, S., Williams, J. F., and Hill, J. E. (1984). Harvest moisture effects on
rice milling quality. Calif. Agric., 38: 11.
71. Morse, M.D., Lindt, J.P. Oelke, E. A., Brandon, D. M., and Curley, R. G.
(1967). The effect of grain moisture at time of harvest on yield and milling
quality of rice. Rice J., 70:16.
72. Dilday, R. H. (1989). Milling quality of rice: Cylinder speed vs. grain-
moisture content at harvest. Crop Sci., 29:1532.
73. Tseng, S. T. (1986). Influences of grain type and field draining on grain mois-
ture and milling yield, Proc. 21st Rice Tech. Work. Group, Houston, TX, Feb.
24-26, 1986. Texas Agr. Exp. Sta., College Station, p. 32.
74. Kunze, 0. R. (1985). Effect of environment and variety of milling qualities of
rice. Rice Grain Quality and Marketing, IRRI, Manila, Philippines.
75. McKenzie, K. S. (1990). Relation of whole kernel brown rice yields to head
rice yields, Proc. 23rd Rice Tech. Work. Group, Biloxi, MS, Feb. 26-28, 1990.
Texas Agr. Exp. Sta., College Station, p. 39.
76. Gravois, K. A., Moldenhauer, K. A. K., and Rohman, P. C. (1991). Genetic
and genotype x environment effects for rough rice and head rice yields. Crop
Sci., 31:907.
77. Velupillai, L., and Pandey, J.P. (1990). The impact of fissured rice on milling
yield. Cereal Chem., 67:118.
78. Srinivas, T., and Ehashuyan, M. K. (1985). Effect of variety and environment
on milling quality of rice, Rice Grain Quality and Marketing, IRRI, Manila,
Philippines, p. 49.
79. Venkateswarlu, B., Parao, F. T., and Vergara, B. S. (1987). Occurrence of
good and high density grains on rice panicles. SABRAO J. 19:1-6.
BREEDING FOR RICE QUALITY 111

80. Jongkaewwattana S. 1990. A comprehensive study of factors influencing rice

( Oryza sativa L. ). Ph.D. dissertation, University of California, Davis.
81. Geng, S., and MacGregor, R. (1991). Methods and technology development
for head rice improvement in California-heritability investigation for three
California grain types, Annual Report: Comprehensive Rice Research. 1990.
University of California, USDA, California Rice Res. Board, Davis, CA, p. 1.
82. Jodari-Karimi, F., and McKenzie, K. S. (1986). Milling performance of
Lemont/Leah F3 lines, Proc. 21st Rice Tech. Work Group, Houston, Feb. 24-
26, 1986. Texas Agric. Exp. Stat., College Station, p. 32.
83. Bollich, C. N., MacGill, C., Webb, B. D., Marchetti, M.A., and Scott, J. E.
(1991). Registration of 'Texmont' rice. Crop Sci. (submitted).
84. Tanksley, S. D., Young, N. D., Paterson, A. H., and Bonierbale, M. W.
(1989). RFLP mapping in plant breeding: New tools for an old science.
Biotechnology, 7:257.
85. Shaeffer, G. W., and Sharpe, Jr., F. T. (1990). Modification of amino acid
composition of endosperm proteins from in-vitro selected high lysine mutants
in rice. Theor. Appl. Genet., 80:841.
86. Mcilrath, W. 0., Jodon, N. E., Sonnier, E. A., Trahan, G. J., and Marchetti,
M. A. (1979). Registration of 'LAllO' rice. Crop Sci., 19:744.
87. Juliano, B. 0., Perez, C. M., Kaishik, R., and Khush, G. S. (1990). Some
grain properties of IR36-based starch mutants. Starch/Starke, 42:256.
88. Kaushik, R. P., and Khush, G. S. (1991). Genetic analysis of endosperm
mutants in rice Oryza sativa L. Theor. Appl. Genet., 83: 146.
 6
New Methods for the Evaluation of Rice
Quality and Related Terminology
David E. Kohlwey
Riviana Foods, Inc., Houston, Texas

1. INTRODUCTION
There are few methods in the cereal literature that are specific to rice.
Methods used in the food industry are adapted from other cereal products.
Some procedures, such as moisture determinations, are taken directly from
standard methods. Others, such as Kjeldahl protein determinations, have
minor changes. For example, the conversion factor for rice protein is 5.95
instead of 6.25, a value used for a variety of other proteins. This chapter
gives an industrial view of the state of quality evaluation, where it is going,
and what is useful.
Rice is a unique grain. Unlike most other cereals, it is usually eaten as a
cooked intact grain. As a result, the industry has evolved a set of tests that
are specific for examining intact grains. The visual appearance of the grains
is the first quality characteristic that the consumer observes. The grades
assigned by the United States Department of Agriculture-Federal Grain
Inspection Service (USDA-FGIS) revolve around this first quality impres-
sion. Two additional quality profiles relate to cooking performance. First,
evaluations from breeding stations screen for traits, which relate to cooking
quality and appearance. Second, processors test the suitability of the mate-
rial as an ingredient in their products. Another area of examination con-
cerns the functional properties that make rice a good industrial ingredient.

113
114 KOHLWEY

The final evaluation is done by consumers. When the product appeals to

consumers, they select it with repeat purchases.
The segments of the rice-producing and -processing industry tend to oper-
ate independently of each other. Many of the products and/or concepts of
what good quality products are have evolved along with the region or culture
that consumes them. The definition of quality appears to be culturally re-
lated (1). The rice industry still has much to understand and turn into a
formal scientific system. Unfortunately, much of what is known is based on
subjective evaluation. Subjective evaluation by trained workers is often the
fastest or best quality control method. The following sections will discuss
what quality tests are currently used, areas of growth or change, new tools
for examining quality, and the vocabulary associated with these methods.
Today the United States is entering a period where personal computers
and factory-based programmable controllers are found at all levels of the
processing industry. Many of the most important developments of the
1980s (and which are still evolving) are a response to having these new
tools. They pay their way by increasing yield and quality. Many of the
developments in the future will revolve around on-line sensors and meth-
ods of analysis that permit the programmable controllers to do their job.
Control principles that survive will be reliable, user-friendly, and correlate
to or actually be valid quality parameters.

ll. EXISTING METHODS

A. Breeder Tests
Rice breeders play an important role in producing quality rice. The farmers
are the initial customer of the breeders. As a result, the first target of the
breeding program has been improving yield and disease resistance. The
second target of the breeding program has been to maintain the cooking
quality of the crop. Tests used in the breeding program are aimed at evaluat-
ing small quantities of grain that have progressed past the initial selection
process. Breeder tests evaluate appearance and texture (2-5). Some
special-purpose tests exist for monitoring flavor. Flavor normally is com-
pared to a reference sample by a trained panel or individual.
Examples of breeder tests are listed in Table 1. These procedures primar-
ily screen secondary properties of the rice, which have been correlated to
the acceptability of the cooked product. There are cases where varieties
have been developed only to find that the specific cross does not meet the
textural, appearance, or processing expectations of the targeted users. This
problem has been minor in the United States since new variety releases
tend to carry a significant portion of the genes of the previous dominant
NEW EVALUATION METHODS 115

Table 1 Typical U.S. Breeder Tests•

Test Purpose (information gained)
Alkali spreading value How thick a paste is-differentiates high
amylose types
Alkali value Ease of cooking (in alkali)-correlates to
gelatinization temperature
Amylograph or RVA Cooked texture and breakdown, gela-
tinization temperature
Amylose Correlates to cooked rice firmness
Aromatic flavor Scented type relative to standard type
Cooking rate Optimum cooking time
Cooked grain length Elongation potential of the variety
Firmness Texture of the cooked product; eating quality
Gelatinization temperature Ease of cooking
Grain size Length, width, and height-cultural expecta-
tion related to size and shape
Grain appearance Cosmetic acceptance
Parboil canning loss Suitability for canning
Protein Ease of cooking, nutrition
•The equipment used and test needs of some foreign breeders are different. Reference 2
contains details about special tests used in other breeding programs.

variety. The breeders have done an excellent job of developing the vari-
eties that farmers plant. The United States has a uniform and disease-
resistant source of rice with which to work.
Processors have been remiss in identifying their needs and guiding breed-
ing developments. This can create havoc in the system. In 1985, the domi-
nant Gulf Coast varieties of Lebonnet and Tebonnet were replaced by
Lemont and Newbonnet in one crop year. The sharp transition had rice
processors complaining when the new crop was harvested. Processors didn't
know how to use them. They claimed that Lemont was oily milling. Having
worked with these varieties, we can now say that the Lemont kernel is larger,
which results in a longer cooking time. Lemont can be harder to cook (but
not always) than New bonnet since it has a higher gelatinization temperature.
Processors still have not written specifications for the crop that they pur-
chase. They are not likely to. They are again waiting for the next surprise.

B. Federal Grain Inspection Service Standards

To most processors in the United States, rice quality means compliance
with USDA-FGIS standards (6,7). The USDA-FGIS grading system is the
building block upon which the rest of the U.S. marketing system is built.
116 KOHLWEY

Table 2 Selected Grade Factors for The Federal Grain Inspection Service
Grading factor• Purpose (information gained)

Broken kernels Cosmetic, indicates processing quality

Chalky kernels Cosmetic, related to nonuniform cooking
Damaged kernels Cosmetic, related to farm practices
Foreign material Indicates poor processing or storage
Heat-damaged kernels Related to poor storage or processing
Infestation Indicates poor sanitation
Moisture Indicates possible instability
Paddy kernels Indicates poor cleaning
Related material Indicates poor cleaning
Red rice Related to poor farming practices
Objectionable seeds Related to poor cleaning or farming practices
Objectionable odor Suggests poor consumer acceptance
Smutty kernels Indicates a milling loss
Undermilled kernels Cosmetic, sometimes intentionally done
Ungelatinized kernels Cosmetic, nonuniform cooking
•Traits graded for are not forbidden, only quantified.
Source: Refs. 5, 6.

USDA-FGIS grades define the economic basis for transferring rice from
one processor to the next. It eventually results in the apparent quality of
the package that the consumer will cook and hopefully enjoy. This is not to
imply that the consumer considers this quality.
Several of the defects that are classified by USDA-FGIS are listed in
Table 2. They are appearance-, health-, and filth-related. Most of the de-
fects are visual and pose no health threat. The purpose of USDA-FGIS
standards is not to eliminate these defects, but to aid in establishing a
market value based on named class and defect level.
The USDA-FGIS grading system has worked reasonably well for several
reasons: (a) the U.S. crop is quite uniform; (b) the top four varieties-histori-
cally have been 70-80% of the U.S. crop; and (c) the dominant varieties
have been selected such that they are compatible with the American textural
expectation for the respective grain shape. The USDA-FGIS standards are
aimed at properties that can be controlled with farming practices and/or
mechanical processing. This basic level of quality is sensible since it grades
for quality parameters that can be controlled early in the production chain.

C. Quality Tests for Value-Added Table Rice

Besides breeder tests and USDA-FGIS grades, rice is also evaluated by
producers of value-added table rice products. The United States has several
NEW EVALUATION METHODS 117

processors who cook rice to make a consumer product. These companies

cook rice by parboiling, canning, precooking (instantizing), or cooking and
then freezing. These processors have a set of tests that are used to control
the quality of the cooked product. Frequently the quality requirements for
value-added table rice are higher, than USDA-FGIS grade minimums.
Normally, a processor who sells value-added table rice starts an evaluation
sequence by preparing the product using a labortory version of the package
directions. Some of the extra tests are appearance, flavor, cooked product
yield, firmness, color, density, and stability or sterility of the product. A
subjective evaluation frequently is used on hard to measure traits such as
flavor. Some facilities establish secondary properties as quality control
points. For example, moisture is frequently used to evaluate stability of a
dry product. In reality, stability is a function of water activity.

D. Quality Tests for Ingredient Rice

Rice is also used as an ingredient in a variety of products other than table
rice. The quality evaluation methods used with ingredient rice are as varied
as the final products that are produced. Frequently the form of the rice is
modified. A few examples of such rice products are a highly puffable
starch, a crispness-inducing agent in batters, a fat substitute, a hypo-
allergenic starch and protein, and a fermentable substrate with a bland
flavor. In each product, rice is used for a specific function. A processor
might look at free fatty acids in rice oil as a negative characteristic if they
are concerned with flavor, yet consider them a positive characteristic if
interested in cooking stability (thickening power). A batter manufacturer
may not want their starch to gelatinize, while an instant beverage producer
may attempt to develop the highest level of gelatinization possible.
The type of testing of interest to ingredient rice processors varies dra-
matically. The properties that are of importance are directly related to a
specific rice product. They are generally not related to the properties that
are important for table rice quality. The tests used by rice ingredient manu-
facturers are normally borrowed from methods used with other, nonrice
ingredients. The focus of this chapter is on the evaluation of rice as intact
kernels. Ingredient rice products and their evaluation is best dealt with
relative to the function they fill in a specific end product.

ill. NEW METHODS

American tastes are becoming more diversified and convenience oriented.
In the United States, we have seen the introduction of foreign rice products
along with the novel rices that are ingredients in these products. Many of
118 KOHLWEY

these products do not fit traditional American crop goals. Much of the
growth in rice consumption in the 1980s was in specialty rices (8,9). Exam-
ples of these are scented, elongating, waxy, sticky, dry cooking, quick
cooking, and instant types. The rice marketers, growers, and processors
have endeavored to fill these new needs. The U.S. rice industry needs to
adapt and improve these non-U.S. varieties and the food products made
from them.
The forces at work on the U.S. crop also exist in other countries. Many
regional tastes exist. They are beginning to be blended. Scandinavians
consider good rice to be a breakfast porridge made by simmering round
grain rice with milk. Germans prefer U.S. quality. Italians like anal dente
(firm textured) risotto made with an arborio (japonica) rice. However,
Italians are increasingly planting long grain (indica) types. Asians also are
trying different rice products. They traditionally prefer a soft, more adhe-
sive, yet nonstarchy rice. This preference is changing. In Japan, it is possi-
ble to find new products such as rice pilaf, which uses indica rice.
Asians have evolved a multitude of unique specialty non-table rice
products. They range from sake to puffed rice to rice noodles. These prod-
ucts are making an entry into other parts of the world. At some point U.S.
processors will become interested in these products as they consider poten-
tially lost sales revenues. They will need the raw feed stock(s) for making
these foreign products. They also need an understanding of good quality. In
some cases the quality bias will be different. As an example, Japanese
consumers of rice crackers have a low tolerance for adhesiveness and gum-
miness traits. The American snack consumer, however, will not always
reject or even detect these traits. They are accustomed to eating several
types of extruded cereals and snacks, which have enough adhesiveness to
create tooth pack.
The need for improved understanding of foreign rice is increasing. The
world's rice germplasm is becoming mixed to improve global crop yields.
We are also looking at the new world of biotechnology. Gene transfer will
make it possible to put traits into the grain that do not exist now. We did
not or could not tell the breeders how to select for the varieties of the last
decade. How will we tell the biotechnologists what traits we want in rice
during the next decade?

A. Cooking
Much of the quality of a rice product is dependent on the way the product is
cooked. This is a key point since different methods of cooking will produce
different textures and appearances. A well cooked rice by U.S. standards is
probably too soft for most Europeans and not sticky enough for most
NEW EVALUATION METHODS 119

Orientals. Significant differences in what is considered good quality exist in

other parts of the world. It is surprising that the U.S. melting pot of cul-
tures has historically been a relatively homogeneous market.
Two basic test cooking methods exist: (a) cooking in excess water and
(b) cooking with measured amounts of rice and water. The rice may either
be simmered in water or with instantized and some precooked products,
brought to a boil, and then removed from the heat. With either cook
method, the length and type of cooking test method should mirror the
package directions for the product. Several products have made it to the
U.S. market that specify microwave cooking or give it as an alternative
cooking method. Micmwave energy for heating is not the preferred stan-
dard. A microwave cook test is hard to standardize and translate to a lab
test. Consumers have many sizes of ovens, and decreased output of micro-
wave power occurs as an oven ages.
The dominant industrial test cooking method in the United States is
excess water cooking. The excess water method is easy to reproduce and
reflects instructions that exist on a number of American and European
products. The procedure can be performed on a stove top (preferably gas)
or with an electric precision cooker. The precision cooker is used to im-
prove reproducibility among different users. In the precision cooker, time
is held constant and the power input is adjusted (10). The cooker can be
calibrated by measuring the loss of water mass when a cooking cycle is run.
The power source, a rheostat, is then adjusted to bring the cooker into a
standard power input range. This method of cooking reflects the rate at
which the rice cooks.
The second test method, cooking with exact ratios of rice and water, is
more popular in breeding labs. The method can be adapted to smaller
samples and to recipes used with local consumers. Exact ratio cooking is
typically done in a Japanese rice cooker. In principal, a Japanese rice
cooker inputs heat at rated power until the free water is consumed. This is
detected when a set temperature above boiling is reached (11,12). This
means that time is variable. For this test method, it is assumed that all the
water is absorbed or used in steaming. As a result a target yield will be
reproduced.
Rice quality depends upon the target against which one cooks. The
target varies significantly in specialty rices by locality, product and the
manufacturer's desire for market uniformity or marketable difference. The
product should be prepared using package directions. The acceptance/
rejection criteria can be on a yield or firmness basis. Breeders and proces-
sors need better methods that quantitate the possible differences. They also
need to be able to evaluate causes of differences. Texture analysis fills part
of this need.
120 KOHLWEY

B. Texture
This section considers the measurement of rice texture and the types of
textural properties that can be measured (13-17, L.J. Malcomson, personal
communication). Szczesniak (15) developed a series of basic textural terms
that can be used to evaluate food texture. These basic food terms have
been expanded to include ranges and conditions that are applicable to rice
(18) (Table 3). These textural properties or universal texture profile can be

Table 3 Texture Definitions Related to Cooked Rice

Firmness Force required to compress cooked rice between the molar

teeth on the first chew. The degree can range from soft (low),
firm (middle), to hard (high) degree of firmness. (Also called
hardness in texture literature.)
Adhesiveness Force required to remove cooked rice that adheres to the
mouth (during eating), itself, and serving utensils. The de-
gree can vary from flaky (low) to sticky (high).
Springiness Degree to which cooked rice returns to its original shape once it
has been compressed between the teeth. The degree can vary
from viscuous flow (low) to rubbery (high).
Cohesiveness Internal force holding a grain together before it breaks, when
compressed between the teeth. The degree can vary from
mushy (low), tender (middle), to leathery (high) or brittle in
al dente cooked rice (high).
Chewiness Length of time (in seconds) required to masticate cooked rice at
a constant rate of force application to reduce it to a consis-
tency suitable for swallowing.
Fracturability Force with which cooked rice crumbles. A high degree would
be a rice with a high degree of hardness and a low degree of
cohesiveness.
Gumminess Denseness that persists throughout mastication; the energy re-
quired to disintegrate cooked rice to a state ready for swallow-
ing. This term is a composite of hardness and cohesiveness.
Degree would range from mealy (low) to gummy (high).
Starchiness Describes the type of surface moistness. The condition can
range from dry and flaky (low) to wet and starchy (high). The
center of the grain will typically have a harder texture and
contribute a raw flavor.
Tooth pack Relates to bits of cooked rice sufficiently adhesive and gummy
that they lodge in the cusps of the molar teeth. This is primar-
ily a defect with extrusion-cooked rice, but can be found in
precooked rice that is not fully cooked.
Source: Refs. 15, 17, 18, L. J. Malcomson, personal communication.
NEW EVALUATION METHODS 121

measured by compressing and releasing a product in a cyclical chewing

pattern. Initially, a Texturometer and later an Instron universal testing
machine was used to perform this procedure.
The complete universal texture profile has not been implemented in the
rice industry. Mossman et al. (19) described a procedure specific for measur-
ing rice adhesiveness. This procedure contains lengthy instructions such as
aging the cooked sample for one hour at room temperature. The aging step
gives excellent reproducibility. However, the texture of rice is time sensi-
tive. The results change as the product ages. Does aging generate an arti-
fact that does not reflect your consumer's behavior? For U.S. consumers, it
frequently does. One case where consumers might see aged rice is on the
steam table. Steam tables can make the product more firm and rubbery or
adhesive and diminish the surface appearance. The goal of a test has to be
to reflect what the consumer will find when he or she is doing the final
evaluation.
Several instruments can run a universal texture profile. The
Texturometer is in use in Japan to generate a chewing cycle similar to the
chewing cycle developed by Szczesniak (20). Several vendors have intro-
duced instruments capable of generating a universal texture profile. One
such instrument, the universal testing machine, is pictured in Figure 1.
Figure 2 is an example of two universal texture profiles of U.S. Lemont
variety white rice. When cooked using package directions, the white rice
has significant adhesiveness, and the precooked parboiled rice does not.
The firmness and other traits are similar. These profiles were generated
with 2 g of freshly cooked rice. The piston starts at 5.0 mm of clearance
above the base. It is brought down toward the sample at a rate of 10 mm/
min to a clearance of 0.5 mm. It is raised at a rate of 10 mrnlmin to 2.0 mm
and then a repeat cycle is performed. The main points to note are the
firmness (resistance to compression/flow), adhesiveness (force holding the
piston to the table as it is lifted), and springiness (the degree of symmetry
of the compression strokes).
For reproducibility and convenience, the universal testing machine
should be computer controlled with an automatic cycle. The scan can then
be recorded, and finished results can be automatically calculated. The main
disadvantage to producing texture profiles is that they are time-consuming
to generate. The main advantages of a texture profile are that a very small
sample can be tested and several texture parameters can be calculated from
the same test. A major source of error in texture profile analysis is the
accuracy of piston positioning.
Ifone is approaching textural choices from a breeder or researcher point
of view, it is possible to see major changes in a texture profile. It is also
possible to see major differences when comparing rice (0. sativa, L.) to
122 KOHLWEY

Figure 1 Instron Universal Testing Machine with piston and anvil compression
fittings. (Courtesy of Instron USA Inc., Canton, MA.)

other grains such as wild rice, pasta (wheat), or bulgur. U.S. rice processors
adjust their product to give the consumer both consistent yield and degree
of firmness when the product is cooked using the package directions. Con-
sumers who limit themselves to processed U.S. rice varieties will primarily
see differences in firmness (peak force). The product will normally not be
excessively rubbery (high springiness) or adhesive. In contrast, the meth-
ods that come from Japan address adhesiveness as one of the most impor-
tant texture traits. Notable springiness and cohesiveness are found with
some European rice products.
In the U.S. rice industry, the primary method used to measure firmness
is the Kramer shear cell (21, P. Twigg, personal communication). Figure 3
shows a Kramer shear cell on a Food Technology Corporation (FTC) shear
press. The Kramer CS2 cell is a metal box and cover with 13 grooves and a
matching set of blades. The blades penetrate through the rice contained in
the box part of the cell. The test can be run on a shear press, universal
testing machine, or similar force-measuring device. The procedure for the
NEW EVALUATION METHODS 123

Clearance - mm. 0.5 2.0 0.5

80 -Success Brand-
Precooked
70 Parboiled Lemont
·---White - Lemont
60
~ 50
....
~40
C)
~ 30

o~.------

° . . . . . .1"=o........................
-1 ~ 20~ ........... """': ':30~
........... -' !-4o: -'"'-'- '- '-!: :s~o '-'-'-'-'""""'=s:=o.w-LJ
Cycle time - Seconds

Figure 2 A texture profile of white and precooked, parboiled Lemont variety

rice developed with the Instron Universal Testing Machine. Both rices were cooked
following package directions. (Courtesy of Riviana Foods, Houston, TX.)

test is as follows: One hundred grams of cooked rice are placed in the celL
The cell (box) is closed and slid into the blades, which hang from a load
cell. The blades are forced down through the sample, and the peak force is
recorded as the shear value. The test is sufficient for measuring firmness.
Cooking time, holding time, and sample size must be accurately controlled,
since they are major sources of error.
An alternate method to measure firmness is used in the U.S. breeding
program: the Pabst texture tester (2, R. Pabst, personal communication).
The Pabst cell is a smaller version of the Kramer shear cell. The advantage
of the Pabst instrument is that it consumes less rice than the Kramer shear
cell (10 g cooked weight). This increases the variability on single repetition
tests. For industrial testers, the amount of rice is not a problem. They use
the larger 100-g sample to have an averaging effect that reduces the error.
Recently the European Community (EC) (22) established definitions to
be used for the evaluation of indica versus japonica varieties of rice. These
are the first official methods to use parts of the Szczesniak terminology (see
Table 3). They use the adhesiveness term and call it "glutinosity." The
firmness character is measured as "consistency." The adhesiveness test is
run by compressing 2 g of cooked rice under a piston and then recording
124 KOHLWEY

Figure 3 Texture press with a CS2 Kramer shear cell. (Courtesy of Food Technol-
ogy Corporation, Rockville, MD.)
NEW EVALUATION METHODS 125

the work [force (g) x distance (mm)] required to pull the piston away from
the sample. The EC procedure measures firmness with a modified 50-mm2
Ottawa test cell. This is an extrusion-type cell where a piston is forced
down on a sample, pushing it through a perforated plate. The mean force
during extrusion is used as the result. The Kramer and Ottawa results
relate to firmness. The cycle time per test is similar. The practical differ-
ences between the methods is that the Ottawa cell is a more durable type of
cell and cleaning is not as critical.
Not all textural terms are monitored. This reflects each culture's needs
seen with locally available rice. The traits of cohesiveness, springiness,
chewiness, and gumminess exist, however, they are of concern primarily
when radically different types of rice are introduced to a new area or with
processing that changes these textural characteristics. Normally the con-
sumer is not exposed to variables in these areas. Fracturability is not a
significant property with warm cooked rice. The firmness measurements
obtained with the Ottawa, Pabst, and Kramer cells and those obtained as a
universal textural profile can be correlated with each other. The relevant
property producing firmness in cooked milled rice is resistance to flow
(viscous flow). Changing the sample to cooked brown rice which has the
bran layer adds a true shear force. Flow and viscosity measurement will be
discussed in the following section.
A final comment on the need for texture measurement needs to be
made. For most processors and consumers, a simple subjective cook and
taste test is all that is done. Mechanical texture analysis attempts to quan-
tify these subjective chewing (taste) impressions. Measuring firmness and
other texture traits consistently is difficult. Ask yourself, "What is possible
to control?" and "How much do I need to measure?"

C. Viscosity
The standard method to examine thickening ability and gelatinization of
rice has been the Brabender Visco-Amylograph. Figure 4 shows a new tool
that is taking over this task. It is the Rapid Visco Analyzer (RVA) (23,24).
With both instruments, the gelatinization temperature of rice starch can be
measured. In addition, the cooking quality of raw rice and the degree and/
or type of cooking or processing of the grain can be estimated. These
instruments operate by cooking a slurry of flour and water. The tempera-
ture and corresponding viscosity are recorded as the slurry is heated.
Both instruments can measure viscosity versus temperature (time).
When operated properly, either viscogram can accurately determine transi-
tion temperatures and find steady-state viscosities of the flour slurry before
it is heated, during cooking, and after the slurry is cooled. The flour used
126 KOHLWEY

Figure 4 Rapid Visco Analyzer viscosity-measuring device. (Courtesy of Foss

Food Technology Corp., Eden Prairie, MN.)
NEW EVALUATION METHODS 127

on either instrument needs to have a controlled particle size. Large-particle

flours shift viscosity profiles to higher temperatures. Minus fifty mesh mate-
rial is suitable. Both instruments should have the solids level in the cooked
slurry adjusted to compensate for sample moisture. The RVA can scale its
output as centipoise or to approximate the arbitrary Brabender units. The
RVA generates a higher shear than the amylograph. As a result, the stabil-
ity of cooked pastes appears to be less.
The older amylograph has limited use with breeders. It uses 50 g of dry
matter, versus 3-4 g used with the RVA. The RVA can produce a quick
scan in 9 minutes. A 2° per minute scan produces accurate transition tem-
peratures but takes 39 min. In contrast, a complete amylograph scan takes
75-90 min (2). Other advantages of the RVA are: (a) the sample contact
parts are disposable, (b) a computer analysis program exists to read the
scan, and (c) the computer-controlled RVA-C3 can be programmed for
multiple heating, holding, and/or cooling cycles. New models of both de-
vices are available, which are adapted for computer control.
It should be mentioned that viscosity analysis is also used in some ingredi-
ent rice methods. Figure 5 shows RVA viscosity profiles, which are gener-
ated with several rice flours at 11.67% dry solids (3.5 g of flour, dry basis,
slurried with distilled water to make 30 g). The initial viscosity increases with
the degree of gelatinization. A raw flour will have a relatively flat profile
until the gelatinization temperature is reached, at which time the viscosity
will increase. The RVA paddle produces more shear than an amylograph. As
a result, the relative effect of poor cooking stability is reflected by a greater
precentage decrease in the hot and cool paste viscosity.
One problem common to both devices is that a skilled operator is re-
quired to fully interpret a scan. There are multiple sample factors that can
cause similar effects in a scan (25-29). For routine quality control, measure-
ment of the initial, peak, and cool paste viscosities are sufficient.

D. Calorimetry
A new family of instruments that can determine starch gelatinization tem-
peratures, the amount of energy required to gelatinize starch, and the
extent of starch gelatinization are the Differential Scanning Calorimeter
(DSC) and the Differential Thermal Analyzer (DTA). The DSC measures
the rate of energy input required to maintain a set rate of temperature
increase. The DTA measures the difference in temperature relative to a
standard mass being heated at the same time. Calorimetry has potential
usefulness early in the selection process of breeding programs to select
differences in cooking properties between varieties (30-33). Normally,
only a few milligrams of rice flour/starch and water are needed to perform a
 128 KOHLWEY

-...-L-100
Extruded Rice
Rice Gel" 100
····RP100 Parboiled
-RL 100 White Rice
90
CJ
0

w
80 a:
.....>-
:::)
1-
·c;; <(
a:
0 70
w
0 Q.
C/} :!E
w
5 1o2 1-
.•
80.8'
60

:50.4
······+···· 50

40
0 5 10 15 20 25 30 35
Minutes
Figure 5 Rapid Visco Analyzer viscosity profiles of RL-100 raw flour, RP-100
parboiled flour, L-100 Rice Gel, and extrusion-cooked, highly gelatinized rice flour.
(Courtesy of Riviana Foods, Houston, TX.)

scan. The energy consumed in the gelatinization process is measured. Fig-

ure 6 shows DTA scans of three varieties of rice. Figure 7 shows the
Brabender viscosity profiles for the same rices. The low gelatinization tem-
perature (GT), high amylose variety (Gui Chow) gelatinizes at 80°C (Figs.
6 and 7). The high GT, high amylose variety (Century Patina 231) gelati-
nizes at 94°C (Figs. 6 and 7). The dominant U.S. varieties (such as Lemont)
typically have intermediate GT and amylose contents (Figs. 6 and 7). Poten-
tial problems, such as processing differences in new varieties, could be
identified before variety release.

E. Taste/NIR
An interesting new tool has been developed to give an index of taste in
rice-the Satake Rice Taste Analyzer (34). It is a near infrared (NIR)
analyzer that reads the levels of carboyhydrate, protein, moisture, free
fatty acids, and "unidentified other components" in rice. Work done at
Riviana Foods suggests that the "unidentified components" may include
NEW EVALUATION METHODS 129

~
w
zw
w -GuiChow
>
~w Gelatinization
a: Regions

50 70 90 110
degrees C

Figure 6 Differential thermal analysis energy scans of three different cooking

types of rice at approximately 9% solids. Gui Chow: high amylose, low gela-
tinization temperature (GT); Century Patina 231: high amylose, high GT; and
Lemont: intermediate amylose and intermediate GT. (Courtesy of USDA-ARS,
Beaumont, TX, and Riviana Foods, Houston, TX.)

water activity. The test consists of grinding a small sample, approximately

40 g, tempering it for a hour, and then taking a reading in the NIR ana-
lyzer. The results are computer analyzed and give a predicted taste value.
At this time the device is available only in Japan.
The concept of good rice taste in Japan is different from much of the
United States. The taste analyzer is an aid in blending old rice (poor
quality) with fresh rice (good quality). The American application of NIR is
twofold. It could give a relative age index (new crop vs. old crop) for rice
going to rice cakes or other applications that require puffing. NIR could
also give a relative age index, which would relate to the free grain trait that
many American consumers expect. NIR is a rapid, excellent tool for rice
analysis. Care needs to be taken to assure that the sample analysis proce-
dure is uniform. For example, wear on the grinding plate in the mill or
tempering for random time periods will change the reading.

F. Vision System
Many defects in rice are appearance characteristics. For example, appear-
ance defects account for 13 of the 15 USDA-FGIS identified grading fac-
tors listed in Table 2. They are presently counted or enumerated manually
 130 KOHLWEY

1/) 800
:!:!
c:
::l
lD
"'0
c: 600
Q)

...
.0
al
CD
~ 400
"iii
8
1/)
5
200

0 6-====----~~~~--~--~~~~
50 55 60 65 70 75 80 85 90 95 minutes
degrees C

Figure 7 Brabender viscosity profiles of three rice varieties at 9% solids. Gui

Chow: high amylose, low gelatinization temperature (GT); Century Patina 231;
high amylose, high GT; and Lemont: intermediate amylose and intermediate GT.
(Courtesy of USDA-ARS, Beaumont, TX, and Riviana Foods, Houston, TX.)

by graders who pick them by hand. The advent of inexpensive high speed
computers and frame grabber (digitizing) boards makes it possible to auto-
mate routine appearance analysis. The name of this type of computer-
based package is a vision system. A vision system consists of a video cam-
era or line scanner which feeds a signal to a frame grabber board in a
computer. The computer then runs the digitized image through a series of
software filters to categorize the items that the camera "sees." Figure 8
shows an example of such a system.
The frame grabber board is a recent development (R. Meyer, personal
communication). The IBM-compatible black and white frame grabber
board became available in 1988. The color version became available in late
1990. The chief limit on present systems is that most cameras and frame
grabbers are limited to 760 x 1040 pixels of resolution. This limits the
sample to 10 g or less. Sample size depends on the size of the trait being
examined. To examine larger samples, multiple repetitions need to be
done. Also, the imaged sample should have all grains separated and not
touching. A mechanical system for presenting the grain is not available as
an off-the-shelf item. Camera-based vision systems have enough flexibility
NEW EVALUATION METHODS 131

Figure 8 A vision system being used to examine rice. (Courtesy of Meyer Instru-
ments, Houston, TX.)

to analyze both cooked and dry rice. Several vendors who sell components
and turnkey systems are listed in the appendix. Software is the heart of the
system. It determines what hardware is compatible, what can be counted or
seen from an image, and how easy the system is to use. As a result, the
vendors listed in the appendix are mainly software producers.
A second type of vision system operates as a line scanner (similar to a
facsimile machine). It views single grains as they travel on a belt or slide.
Satake has developed a line scanner based system in Japan (R. Bashiki,
personal communication). It can identify defects ranging from peck to
cracked grains. The U.S. prototype is undergoing refinement to permit it to
judge American quality factors. The Satake vision package includes a me-
chanical kicker to segregate different classes into separate bins. The main
limits with this system are: (a) it can only be used on dry rice, (b) it
operates in black and white, and (c) the type of analysis performed cannot
be varied. The advantage of the line scanner is that it can check larger
samples with greater resolution.
Features to look for in a vision system are: (a) ease of use, (b) ability to
change or recalibrate selection criteria, (c) macros that permit repetitive
132 KOHLWEY

analysis, (d) compatibility with other hardware, and (e) subpixel calculated
resolution. Lighting is a critical factor in image analysis. It can produce
artifacts or help to visualize specific parts of the image. These systems are
close to the point where they can automate visual analysis.

G. Data Systems
In the introduction, computers and programmable controllers were men-
tioned. The tool that is having the greatest impact on the rice industry is the
computer. The tools described so far in this chapter generate new informa-
tion. Most of them exist only because of computer drivers. On-line sensors
connected to programmable controllers have reduced the need for labor to
produce products. Personal and minicomputers permit handling of numer-
ous data and deciding what is useful and real. The ability to automate
testing and data collection is even improving the efficiency of old testing
methods. An industrial example of this is a computer-driven robot which
can perform, unattended, chromatographic vitamin analyses. Another ex-
ample has been the automated determination of amylose in the U.S. breed-
ing program.
One major problem with the current family of computers and computer-
driven equipment is that data and software drivers are not always compati-
ble. If a computerized laboratory were set up with the instruments and
methods described, there would be separate software drivers for each piece
of equipment. This requires operators who are skilled in multiple software
systems as well as analytical devices. Some instruments are so user-
unfriendly that they need a computer programmer to develop software to
make the instrument compatible with a particular device driver. Any per-
son or organization setting up and staffing a laboratory on a budget should
demand compatibility.
One area where compatibility of laboratory instruments is occurring is
with multitasking software systems. It is desirable for lab equipment to be
able to operate in a multitasking environment. Single-task computers re-
duce the productivity of their operators. At this time, several software
programs are multitasking compatible. Some robotic chromatography,
calorimetry, and vision systems use a multitasking environment. An alter-
nate method for compiling results is to feed data to a separate computer-
driven Laboratory Information Management System (LIMS).

H. Other New Methods

Methods are needed to explain the role that starch and its interaction with
water plays during cooking. At present, published research is not available
to answer the question, "What happens, on a molecular level, during the
NEW EVALUATION METHODS 133

cooking of rice?" Since rice is eatean as a whole grain, the location and rate
of movement of water is important. Multiple molecular and macro-
molecular events occur during rice cooking (12,26). They need to be exam-
ined together to be understood. Also, water activity influences the cooking
characteristics of the rice. Rice shifts to a lower equilibrium moisture curve
as the temperature is raised (35-37). This occurs with a decrease in adhe-
siveness. When growing rice in a cool climate, the apparent amylose con-
tent increases. High amylose correlates with firm, flaky rice, yet a crop
grown in a cool climate has less firmness and more adhesiveness. A genetic
component also exists. Rexmont and L-202 both are high amylose types,
yet L-202 is significantly softer than Rexmont. Little is known about the
fine structure of starch. How do water, starch, lipid, and protein in rice
interact to produce different textures after cooking?
Magnetic resonance imaging (38,39), electronic spin resonance (40, and
gel permeation chromatography (41 ,42) are a few of the new methods that
will help answer the questions posed in the preceding paragraph. To be
effective they need to help explain results generated by the other methods
described in this chapter.

IV. CONCLUSIONS
The rice industry has few methods that are specific to rice. Our understand-
ing of our crop is primitive. Many of the things that we know from experi-
ence have yet to be explained. We have tended to view quality in terms of
how many pounds of clean rice we can sell to the consumer. The resulting
variability in cooked quality may be a major reason why per capita rice
consumption in the United States is low.
The rice industry has new tools available that permit the examination of
external and internal characteristics of the grain. Future research needs to
develop predictive models to explain quality. This will enable us to effi-
ciently breed new varieties and effectively control processes for this unique
grain.

Appendix A Partial List of Vendors of Quality Measurement Equipment•

Equipment Vendor

Cooking
Precision Cooker Precision Engineered Products
Houston, TX
Thermal Analysis-Flow Properties
Visco-Amylograph C. W. Brabender Corp.
South Hackensack, NJ
134 KOHLWEY

Appendix Continued
Equipment Vendor

TMA (Thermal Mechanical Analyzer) DuPont Instrument Systems

Wilmington, DE
RVA (Rapid Visco-Analyzer) Foss Food Technology Corp
Eden Prairie, MN
TMA (Thermal Mechanical Analyzer) Shimadzu Scientific Instruments
Columbia, MD
Thermal Analysis-Energy
DSCb DuPont Instrument Systems
Wilmington, DE
DSCandDTN Mettler Instrument Corp.
Highstown, NJ
DSC Perkin-Elmer Corp.
Pomona, CA
DSCandDTA Shimadzu Scientific Instruments
Columbia, MD
Texture
Universal Testing Machine Instron Corporation U.S.A.
Canton, MS
Shear Press Food Technology Corporation
Rockville, MD
Universal Testing Machine Lloyd Instruments
King of Prussia, PA
Texturometer Seiben Hiko Ltd.
Tokyo, Japan
Pabst Texture Tester Precision Engineered Products
Houston, TX
Universal Texture Analyzer Texture Technologies
Scarsdale, NY
Taste/Cooked Character (NIR)
Rice Taste Analyzer Satake U.S.A. Inc.
Houston, TX
Color
Colorimeter Hunter Lab Inc.
Reston, VA
Milling Degree Meter Satake U.S.A. Inc.
Houston, TX
Vision Systems/Components
Software American Inovision
San Diego, CA
Software Automatix Inc.
Billerica, MA
NEW EVALUATION METHODS 135

Appendix Continued
Equipment Vendor
Software BioScan Inc.
Edmonds, WA
Software Jan del Scientific
Corte Madera, CA
System Olympus
Lake Success, NY
System Zeiss
Thornywood, NY
Camera Photometries
Tucson, AZ
•A reference to a company or product name does not imply approval or recommendation
of the product to the exclusion of others that may be suitable. The list of vendors is
intended to only be a starting point in equipment sourcing.
bDSC = Differential Scanning Calorimeter.
eDTA = Differential Thermal Analyzer.

REFERENCES
1. Efferson, J. N. (1985). Rice quality in world markets, Proceedings of The
Workshop on Rice Grain Quality and Marketing, International Rice Research
Institute, Los Banos, Philippines, p. 1.
2. Juliano, B. 0. (1982). An international survey of methods used for evaluation
of the cooking and eating qualities of milled rice, IRRI Research paper No. 77.
3. Juliano, B. 0., et al. (1981). International cooperative comparison of instru-
ment methods for cooked rice texture. J. Texture Stud., 12:17.
4. Webb, B. D. (1980). Rice quality and grades, Rice Production and Utilization
(B. S. Luh, ed.), AVI, Westport, CT, p. 543.
5. Juliano, B. 0. (1985). Criteria and tests for rice grain qualities, Rice Chemis-
try and Technology (B. 0. Juliano, ed.), American Association of Cereal
Chemists, St. Paul, MN, p. 443.
6. Federal Grain Inspection Service. (1989). United States standards for rice,
Federal Grain Inspection Service.
7. (1992). United States standards for rice, Fed. Reg., 57:2482.
8. Dziezak, J.D. (1991). Romancing the kernel: A salute to rice varieties. Food
Techno[., 45:74.
9. Cramer, G. L., Taylor, E. L., and Wailes, E. J. (1990). Growth in U.S.
aromatic rice consumption. Arkansas Farm Res., (January):
10. Pabst, R. (1977). Precision Cooker Instruction Manual. PEP Inc., Houston,
TX.
11. Onishi, M., Ito, K., and Asahi, M. (1987). Method for cooking rice. U.S.
Patent 4,670,282.
136 KOHLWEY

12. Onishi, M., Ito, K., and Asahi, M. (1987). Apparatus for cooking rice substan-
tially under atmospheric pressure. U.S. Patent 4,672,179.
13. Friedman, H. H., Whitney, J. E., and Szczesniak, A. S. (1963). The
texturometer-a new instrument for objective texture measurement. J. Food
Sci., 28:390.
14. Szczesniak, A. S. (1963). Objective measurements of food texture. l Food
Sci., 28:410.
15. Szczesniak, A. S. (1963). Classification of textural characteristics. l Food
Sci., 28:385.
16. Larmond, E. (1976). The texture profile, Rheology and Texture in Food Qual-
ity (De Man, Voisey, Rasper, and Stanley, eds.), AVI, Westport, CT, p. 546.
17. Mioche, L., Auroy, P., Lepetit, J., and Compagnon, D. (1991). Oral percep-
tion of hardness in viscoelastic products. J. Texture Stud., 22:333.
18. Kohlwey, D. E., Sharp, R. L., McCaskill, D. and James C. (1989). Proposed
rice quality terminology and definitions. Presented before the Rice Milling
and Quality Committee, 1989 American Association of Cereal Chemists An-
nual Meeting, Washington, D.C.
19. Mossmann, A. P., Fellers, D. A., and Suzuki, H. (1983). Rice stickiness: I.
Determination of rice stickiness with an instron tester. Cereal Chern., 60:286.
20. Tsuji, S. (1981). Texture measurement of cooked rice kernels using the
multiple-point menstration method. J. Texture Stud., 12:93.
21. Abraham T. E., Malfait, J. L., and White, A. J. (1990). Process for producing
instantized parboiled rice. U.S. Patent 4,952,416.
22. (1988). Methods of analysis, II. Qualitative characteristics; Ila. Glutinosity;
lib. Consistency. Off. l Eur. Communities, No. L 230/10 Annex I.
23. Welsh, L.A., Blakeney, A. B., and Bannon, D. R. (1983). Modified R.V.A.
for rice flour viscometry. Unpublished method, Yanco Agricultural Institute,
Yanco, N.S.W., Australia.
24. Deffenbaugh, L. B., and Walker, C. E. (1989). Comparison of starch pasting
properties in the brabender viscoamylograph and the rapid visco-analyzer.
Cereal Chern., 66:493.
25. Dolan, K. D. and Steffe, J. F. (1990). Modeling rheological behavior of gelati-
nizing starch solutions using viscometry data.l Texture Stud., 21:265.
26. Kohlwey, D. E. (1992). Rice viscograms: What traits are we breeding for?
Presented at the 25th Rice Technical Working Group, Little Rock, AR.
27. Rasper, V. (1982). Theoretical aspects of amylography, The Amylograph
Handbook (W. C. Shuey and K. H. Tipples, eds.), American Association of
Cereal Chemists, St. Paul, MN, p. 1.
28. Sandhya Rani, M. R., and Bhattacharya, K. R. (1989). Rheology of rice-flour
pastes: Effect of variety, concentration, and temperature and time of cooking.
l Texture Stud., 20: 127.
29. Saqdhya Rani, M. R., and Bhattacharya, K. R. (1989). Slurry viscosity as a
possible indicator of rice quality. J. Texture Stud., 20:139.
30. Nakazawa, F., et al. (1984). Thermal equilibrium state of starch-water mixture
studied by differential scanning calorimetry. Agric. Bioi. Chern., 48:2647.
NEW EVALUATION METHODS 137

31. Biliaderis, C. G., Page, C. M., Maurice, T. J., and Juliano, B. 0. (1986).
Thermal characterization of rice starches: A polymeric approach to phase
transitions of granular starch.l Agric. Food Chern., 34:6.
32. Normand, F. L., and Marshall, W. E. (1989). Differential scanning calorim-
etry of whole grain milled rice and milled rice flour. Cereal Chern., 66:317.
33. Wirakartakusumah, M. (1981). Kinetics of starch gelatinization and water
absorption in rice, Ph.D. thesis, University of Wisconsin, Madison, WI.
34. ( 1990). Rice taste analyzer (Satake sales literature), Satake Ltd., Tokyo, Japan.
35. Fellers, D. A., Mossmann, A. P., and Suzuki, H. (1983). Rice stickiness. II.
Application of an instron method to make varietal comparisons and to study
modification of milled rice by hot-air treatment. Cereal Chern., 60:292.
36. Gras, P. W., et al. (1989). A quantitative study of the influences of tempera-
ture, water activity and storage atmosphere on the yellowing of milled rice. J.
Cereal Sci., 9:77.
37. Kohlwey, D. E. (1991). How milling occurs in rice-A physical chemistry
model. Presented at the 76th annual meeting of the American Association of
Cereal Chemists, Seattle, WA.
38. Hyesook, L., Sobczynska, D., and Setser, C. (1989). 0 17 NMR studies on
wheat starch-sucrose-water interactions with increasing temperature. Pre-
sented at the Institute of Food Technologists 1989 annual meeting, Dallas TX.
39. Song, H., and Litchfield, J. B. (1990). Nuclear magnetic resonance imaging of
transient three-dimensional moisture distribution in an ear of corn during
drying. Cereal Chern., 67:580.
40. Johnson, J. M., Davis, E. A., and Gordon, J. (1990). Lipid binding of modi-
fied corn starches studied by electron spin resonance. Cereal Chern., 67:236.
41. Mahanta, C. L., and Bhattacharya, K. R., (1989). Thermal degredation of
starch in parboiled rice. Starch/Starke, 41:91.
42. Griffin, V. K., Hamaker, B. R., Russell, J., and Anderson, A., (1992}. Influ-
ence of protein on rice starch gelatinization. Presented at the 25th Rice Techni-
cal Working Group, Little Rock, AR.
 7
Degree of Milling
James I. Wadsworth
Agricultural Research Service, U.S. Department of Agriculture,
New Orleans, Louisiana

I. INTRODUCTION
The purpose of rice milling is to produce an edible polished or white rice
product from harvested, dried rough rice. In the rice industry, milling can
refer to either the overall operations of a rice processing plant-cleaning,
shelling, bran removal, size separation, etc.-or simply to the one opera-
tion concerned with the removal of the bran or outer layers from brown
rice to produce a whole grain white rice product. For this discussion, we are
using the term rice milling to describe the latter process.
The brown rice kernel (caryopsis) consists of bran (several histologically
identifiable soft layers and the soft embryo, or germ) surrounding the hard
starchy endosperm (milled rice kernel). Three distinct layers make up the
caryopsis coat: the pericarp (consisting of epicarp, mesocarp, and endo-
carp), the seed coat or tegmen (consisting of spermoderm and perisperm),
and the aleurone layer, which encloses both the starchy endosperm and the
embryo. The weight distribution of the various parts of the rice caryopsis is
as follows: pericarp (1-2%), seed coat and aleurone (5%), embryo (2-
3% ), and endosperm (89-91%) (1). The starchy endosperm is the whitest
portion of the caryopsis. Generally, the whiter the milled rice, the more
value it has in the market place. Consequently, the ultimate objective of
milling is to leave the starchy endosperm free of its coverings, including the
embryo.

139
140 WADSWORTH

The term "degree of milling" refers to the quantity of bran and polish
removed from the brown rice during the rice milling operation. The degree
of milling is usually expressed quantitatively as a percentage of the rice
kernel or qualitatively by various descriptive terms such as "well-milled."
Degree of milling is an important factor related to the quality and nutri-
tional value of the milled rice and to the economic return to the rice
processor. A practical and reliable method for determining and controlling
the degree of milling of rice has long been wanted by the rice industry and
trade and research communities. This chapter will discuss the importance
of degree of milling in rice processing, review the literature for techniques
that have been used to assess rice degree of milling, and discuss in detail a
method using near infrared reflectance (NIR) spectroscopy that is suitable
for on-line determination and control of degree of milling.

II. ECONOMIC ASPECTS

A. Grading
In the United States, the degree of milling is one of the grading factors that
determine the overall quality grade assigned by the USDA to milled rice
(2). For each grade level there are set maximum limits for objectional
factors such as broken kernels, weed seeds, damaged kernels, chalky ker-
nels, etc. The maximum allowable limits of the objectionable factors in-
crease with decrease in degree of milling so that the less well-milled rices
are of generally lower quality and bring a lower price on the market. From
an economic standpoint, the more bran removed during milling, the less
milled product the miller has left to sell. The rice miller does not want to
remove more bran than is necessary to achieve the maximum quality grade
for each rice lot. Thus, it is very important that the miller is able to measure
and control the degree of milling during processing.

B. Head Yield
Head rice (unbroken kernel) yield is another very important economic
factor for marketing rice; broken rice sells for approximately one half the
price of whole kernel rice. The degree of milling achieved in a rice mill can
be adjusted by changing either the milling time or the milling pressure. As
a general rule, head rice yields decrease as degree of milling increases (3).
Autrey et al. (4), Rhind (5), Hogan (6), and Bajaj and Sidhu (7) all have
reported results showing that both degree of milling and breakage in-
creased as milling time was extended. Matthews and Spadaro (8) investi-
gated the effects of mill settings in a friction-type rice mill on breakage and
degree of milling. They reported that any changes in mill settings that
DEGREE OF MILLING 141

increased degree of milling also increased breakage. Consequently, the rice

miller should adjust the degree of milling for each rice lot in order to
balance quality grade and head rice yield to maximize profit.

C. Rice Moisture Content

Rice moisture content, head rice yield, and degree of milling are intimately
related. Webb and Calderwood (9) investigated the effect of moisture con-
tent of rice on the degree of milling and head rice yields obtained with
standard milling conditions. They also determined head yields for rice
samples with different moisture contents that were milled to the same
degree of milling by adjusting milling conditions. Several long- and
medium-grain rice varieties were evaluated at moisture contents ranging
from 6 to 18%. Representative samples of each variety at each moisture
level were milled using different mill pressure settings regulated to produce
rice of four (under- to overmilled) degrees of milling. Degree of milling was
determined on each sample by several methods.
The degree of milling obtained showed that low moisture rice (6-10%)
was considerably more resistant to milling at each mill setting than were
samples of high moisture rice (14-16%). Furthermore, compared to high
moisture rice, low moisture rice required markedly greater milling pressure
settings to obtain a comparable "well-milled" sample. Head rice yield from
the low moisture rice milled at standard mill settings was higher than that
from high moisture rice. However, at this mill setting the low moisture
samples were grossly undermilled in contrast to the well-milled rice ob-
tained with the high moisture samples. When milling pressures were in-
creased to obtain an equivalent degree of milling in the low moisture rice,
the head yields obtained were lower than those of the high moisture sam-
ples. The highest head rice yields were obtained in the high moisture sam-
ples, and this occurred at the lightest milling pressure setting required to
obtain a well-milled degree of milling. At comparable degrees of milling,
head rice yields from the high moisture rice ranged from one to three
percentage points greater than the head yields from the low moisture rice.
For more detailed information on the effects of rice moisture content and
mill settings on milling results, see Chapter 15.

ill. QUALITY AND NUTRITIONAL ASPECTS

A. Composition
The outer layers of the caryopsis and embryo differ substantially from the
starchy endosperm in chemical composition and nutritive value. Table 1
shows the proximate composition of brown rice and the various milling
142 WADSWORTH

Table 1 Range of Proximate Composition of Brown Rice, Well-

Milled Rice, Rice Bran, and Rice Germ (percent dry basis)
Constituent Brown Milled Bran Germ
Protein 8.3-9.6 7.3-8.3 13.2-17.3 17.7-23.9
Crude fat 2.1-3.3 0.4-0.6 17.0-22.9 19.3-23.8
Crude fiber 0.7-1.2 0.3-0.6 9.5-13.2 2.8-4.1
Crude ash 1.2-1.8 0.4-0.9 9.2-11.5 6.8-10.1
Starch 77.2 90.2 16.1 2.4
Dietary fiber 19.1 4.5 27.6-33.3
Source: Ref. 10.

fractions (10). A typical brown rice contains 8% protein, 66% starch, 1%

crude fiber, 4% neutral detergent fiber, 3% fat and 1% ash (wet basis, 14%
moisture). During production of well-milled rice, approximately 22% of
the kernel's protein, 5% of its starch, 89% of its crude fiber, 80% of its
neutral detergent fiber, 83% of its fat, and 68% of its ash end up in the bran
by-products (11).
Table 2 shows the amounts of various minerals and vitamins that are
removed from brown rice during milling to produce a well-milled white rice
(12). A large percentage of these nutrients are lost. Of particular impor-
tance is the 87% reduction in thiamine content; thiamine deficiency has
been the cause of outbreaks of the disease beriberi in some developing
countries where rice is a major food (13).

B. Quality Aspects
The majority of rice consumers prefer well-milled rice. Brown rice, even
though it contains more protein, vitamins, minerals, and lipids than white
rice, is the choice of a relatively small percentage of the consumers. Part of
the reluctance by consumers to eat brown rice is due to the chewier texture
and strong flavor associated with the bran. Brown rice also requires a
cooking time of about 45 minutes compared with approximately 20 minutes
for white rice.
Between the extremes of brown rice and well-milled rice is rice that has
been lightly milled or undermilled, thus retaining part of the nutrition in
the bran. In developing nations where rice is the staple food, the govern-
ments, for public health reasons, have encouraged the production and
consumption of undermilled rice to improve the nutritional value of rice
and make the maximum possible use of their rice resources (14). Three to
four percent degree of milling disrupts the bran sufficiently for rapid water
penetration and a cooking time comparable to well-milled rice. These fig-
DEGREE OF MILLING 143

Table 2 Nutrient Content and Losses During Milling for Seven

Rice Varieties• from the International Rice Research Institute
Property Brown rice Milled rice Loss(%)

Weight(% of brown rice) 100 89.9 10

Crude ash (%) 1.56 0.66 62
p (%) 0.38 0.21 50
K(%) 0.26 0.14 52
Mg (%) 0.13 0.058 60
AI(%) 0.006 0.006 10
Ca (%) 0.010 0.006 46
s (%) 0.15 0.14 16
Si (%) 0.041 0.015 67
Cl (%) 0.034 0.031 18
Mn (~-tg/g) 16.6 8.4 55
Fe (~-tg/g) 12.4 9.0 35
Cu (~-tg/g) 5.7 4.7 19
Zn (~-tg/g) 22.4 16.4 34
Thiamine (~-tg/g) 3.2 0.46 87
Riboflavin (~-tg/g) 0.98 0.50 54

•IR65, IR24, IR64, IRS, IR36, and IR72.

Source: Ref. 12.

ures contrast with the 8-10% degree of milling for well-milled rice that is
consumed in most Western countries.
A serious disadvantage of lightly or undermilled milled rice is its ten-
dency to become rancid during storage. Noticeable rancid flavors will have
developed after only 2 months of storage in hot and humid climates. While
this would be a serious problem for Western marketing practices, in most
developing countries the rice is consumed within 2 months of milling, so
that there are no practical difficulties encountered. Piggott et al. (15) inves-
tigated the effects of undermilling and subsequent storage on eating and
cooking qualities of rice using both chemical and sensory evaluation meth-
ods. Changes in sensory properties were related to chemical changes taking
place. Kwon and Jeon (16) reported on the effects of degree of milling and
long-term storage on chemical and eating quality changes in Korean rices.
Pehu and Siddiq (17) evaluated 110 varieties of Indian rice with respect to
the natural variability for quantitative and qualitative losses related to
degree of milling.
Srinivas and Dasikachar (18) studied the removal of bran and aleurone
layers at different locations on the surface of an Indian rice variety
(Bangara Sanna) during progressive polishing. The peripheral bran layers
144 WADSWORTH

were removed before the aleurone at all stages of degree of milling.

Histological data revealed that at all stages of degree of milling studied (1-
6%), the abrasion is more severe on the protruding ridges than on the
grooves and greater on the ventral region than on the dorsal. Similarly,
bran (pericarp and tegmen) and aleurone removal were greater in the
median portion than in the corresponding portions at the tapering distal
end of the grain. Residual bran and aleurone layers are present in the
grooves and dorsal pit even at 6% degree of milling.
McGaughey (19), as a result of severe insect infestations that occurred in
rice in export marketing channels, evaluated the effect of degree of milling
on insect development in stored milled rice. The results suggest that degree
of milling affects insect development in stored rice. This is in general
agreement with observations by Pingale et al. (20) that insects prefer brown
and lightly milled rices over well-milled and rough rices.

C. Deep Milling
The removal of additional peripheral layers from well-milled rice (endo-
sperm) is termed deep milling or overmilling. Because of the hardness of
the rice endosperm, deep milling must be accomplished with abrasive type
mills. Research on deep milling (6) has shown that: (a) there is a heteroge-
neous distribution of protein and other major noncarbohydrate nutritional
constituents in the milled rice kernel; (b) below the bran coat and aleurone
cells, which are removed during conventional milling, there exists a periph-
erallayer unusually high in protein, minerals, lipids and vitamins; (c) the
high protein and naturally enriched layer may be removed by presently
available commercial machinery in the form of a finely divided rice flour,
which may be considered an extension of the regular milling operation; and
(d) the intact residual kernels, although reduced slightly in the nonstarch
nutrients, are essentially an attractive white and well-milled rice suitable
for regular table and food uses.
Normand et al. (21) determined the quantities of nutrients and the
protein amino acid patterns in various high protein rice flours produced by
deep milling (Tables 3 and 4). They reported that the deep-milled flours,
which were removed as 12 fractions by successively milling layers off of
commercially milled rices, consisted of approximately 18% of the original
kernel weight as finely divided flour with a 22 to 12% protein content. Fat,
thiamine, riboflavin, niacin, phosphorus, and calcium similarly decreased
progressively from outside layers towards the center of the kernel. Starch
and amylose, by contrast, progressively increased towards the center of the
rice kernel. The increase (approximately 19%) in amylose content of the
rice starch from outer to inner layers reflects a significant difference in rice
DEGREE OF MILLING 145

starch composition. The data indicated no essential differences in amino

acids content of the rice protein except for a decrease in tryptophan toward
the center of the kernel.
Sensory evaluation of the residual kernels remaining after different de-
grees of deep milling showed an increase in cooking quality preference with
increasing degrees of milling up to a maximum, after which further milling
gave a loss of quality (6). Cooked rice samples were evaluated for cohesive-
ness (i.e., lack of tackiness), integrity of the kernels (i.e., degree of frag-
mentation), color, flavor, and texture by a taste panel. An optimum prefer-
ence was expressed by the panelists for those residual kernels resulting
from between approximately 4 and 9% weight removal by overmilling.
Kennedy et al (22) and Kennedy and Schelstraete (23) evaluated the
biological value of deep-milled rice flour using rat feeding trials. The pro-
tein efficiency ratios determined on 20 fractions of rice were found to be
2.03 ± 0.20 adjusted to 2.5 for casein. The results show that the deep-
milled flours, in addition to containing more protein, had a balance of
amino acids as good as that of the original rice. Bajaj and Suhid (7 ,24) and
Bajaj et al. (25) reported on the effects of extending milling on quality
characteristics and nutrient composition for several varieties of Indian rice.

Iv. METHODS TO ESTIMATE DEGREE OF MILLING

Rice millers must be able to determine degree of milling in order to pre-
cisely control it. A number of investigators have been concerned with the
development of methods for evaluating degree of bran removal. These
methods may generally be classified into two main groups: a) those estimat-
ing the quantity of either removed bran or residual bran; and b) those
measuring the effects of the removal of the outer layers of the rice caryopsis
on either the chemical composition or the physical characteristics of the
final milled product (26).

A. Methods Estimating Removed or Residual Bran

1. Removed Bran
Determination of degree of milling as the percentage by weight of brown
rice removed as bran during milling is a common technique in the research
laboratory where small samples are milled on a batch basis in a single mill.
This technique would be difficult to implement at a commercial rice mill
where the equipment operates continuously.
Estimating degree of milling by measuring removed bran is suitable
when precisely controlled and repeatable milling conditions are available.
From a scientific standpoint, this method gives very reproducible results,
Table 3 Weight and Nutrient Content of Successively Removed Fractions of
Conventionally Milled Rice (dry weight basis)
Successive fraction removed

Component 1 2 3 4 56

% of rice in fraction 2.07 2.33 1.94 1.65 1.43 1.38

Cumulative % removed 4.40 6.34 7.99 9.42 10.8
% N in fraction 3.66 3.74 3.67 3.48 3.25 3.00
% protein in fraction 21.78 22.25 21.84 20.71 19.34 17.85
% starch in fraction 60.27 63.28 70.59 73.24 76.67 80.03
% amylose in starch 25.57 26.49 29.11 31.02 29.45 31.26
% lipids in fraction 5.77 4.26 2.40 1.29 1.06 0.78
Thiamine (mg/lb) 4.065 3.237 2.368 1.353 0.872 0.540
Riboflavin (mg!lb) 0.402 0.327 0.343 0.295 0.237 0.162
Niacin (mg/lb) 48.09 36.67 25.08 16.87 12.62 9.30
Phosphorus(%) 1.236 0.835 0.666 0.379 0.263 0.166
Calcium(%) 0.463 0.301 0.157 0.080 0.036 0.016

Source: Ref. 21.

Table 4 Amino Acid Composition of Protein in Successively Removed

Fractions of Conventionally Milled Rice (dry weight basis)

Successive fraction removed

Amino Acid 1234 56
Lysine 3.24 3.16 2.98 3.01 2.90 2.86
Histidine 2.75 2.62 2.42 2.43 2.28 2.43
NH3 1.76 1.78 1.75 2.09 2.05 1.83
Arginine 8.69 8.68 8.48 8.77 8.12 8.50
Aspartic acid 8.76 8.82 8.52 8.49 8.29 8.90
Threonine 3.52 3.57 3.44 3.40 3.35 3.50
Serine 5.35 5.47 5.21 5.09 5.11 5.30
Glutamic acid 18.19 18.11 17.79 16.85 17.29 18.57
Proline 4.28 4.36 4.18 4.36 4.16 4.37
Glycine 4.42 4.39 4.14 4.10 4.08 4.29
Alanine 5.54 5.51 5.30 5.18 5.22 5.28
Valine 5.39 5.27 5.34 5.41 5.39 5.49
Isoleucine 3.56 3.50 3.62 3.69 3.65 3.80
Leucine 7.63 7.92 7.80 7.85 7.76 8.03
Tyrosine 5.51 5.57 5.64 5.56 5.52 5.76
Phenylalanine 5.02 5.21 5.06 5.11 5.05 5.27
Tryptophan 1.15 1.45 1.22 1.07 1.02 1.04
Cystine 1.89 1.85 1.97 1.89 1.74 1.78
Source: Ref. 21.

146
 Successive fraction removed
Whole Residual
7 8 9 10 11 12 kernel kernel
1.27 1.28 1.37 1.49 1.68 0.98
12.07 13.35 14.72 16.21 17.89 18.87
2.77 2.54 2.36 2.17 2.03 2.00 1.33 0.86
16.48 15.11 14.04 12.91 12.08 11.90 7.91 5.12
81.35 82.14 85.49 88.17 89.83 90.10 90.68 94.29
34.01 35.88 32.07 30.44 35.08 32.26 31.24
0.22 0.22 0.22 0.11 0.22 0.17 0.23 0.00
0.390 0.352 0.213 0.172 0.098 0.368 0.065
0.163 0.106 0.092 0.074 0.065 0.053 0.099 0.000
6.97 4.90 3.10 3.00 3.00 5.73 2.94
0.160 0.115 0.096 0.090 0.088 0.085 0.140 0.028
0.001 0.000 0.000 0.000 0.000 0.000 0.023 0.000

Successive fraction removed

Whole Residual
7 8 9 10 11 12 kernel kernel

2.76 2.96 2.87 3.05 3.11 2.92 3.52 3.71

2.09 2.28 2.18 2.26 2.35 2.13 2.62 2.64
1.88 2.00 2.21 2.12 2.35 2.00 2.42 2.35
7.95 8.42 8.27 8.57 8.36 8.16 8.58 8.28
8.87 8.95 8.90 9.18 8.85· 8.97 8.93 9.13
3.49 3.48 3.51 3.61 3.56 3.48 3.61 3.63
5.27 5.24 5.29 5.34 5.28 5.14 5.18 5.18
18.09 18.12 18.30 18.00 18.00 17.50 17.09 15.58
4.45 4.49 4.47 4.85 4.45 4.39 4.54 4.42
4.20 4.25 4.23 4.34 4.36 4.25 4.48 4.62
5.28 5.30 5.35 5.40 5.46 5.34 5.51 5.66
5.59 5.59 5.62 5.71 5.76 5.63 5.58 5.48
3.86 3.81 3.78 3.94 3.88 3.87 3.92 3.67
8.06 8.21 7.93 8.27 8.13 8.06 7.95 7.67
5.73 5.69 5.62 5.76 5.70 5.39 5.49 5.28
5.22 5.38 5.20 5.40 5.32 5.08 5.15 4.90
1.06 1.09 1.09 1.07 1.02 0.91 1.32 1.32
1.73 1.77 1.75 1.61 1.88 1.90 1.75 1.68

147
148 WADSWORTH

since it is based on weighing which is extremely precise. However, if the

samples are processed in different mills or with different conditions, even
with equal amounts of bran removed, they will generally contain different
proportions of pericarp, tegmen, aleurone, and starchy endosperm. The
milled samples may not be comparable with regard to their whiteness or
nutritive value. This is a serious limitation of great concern to those inter-
ested in duplicating a given milling process or in comparing the effects of
various milling conditions.
2. Residual Bran
Direct visual estimation by the naked eye or with magnification, and by
comparison with standard samples, is the method most frequently used
throughout the world. This is the official method used by the USDA Fed-
eral Grain Inspectors Service (2), which assigns qualitative descriptors for
degree of milling by observation and hand-picking and by comparison with
interpretive line samples. The qualitative degree of milling categories as-
signed by the USDA are:
1. Well milled
2. Reasonably well milled
3. Lightly milled
4. Undermilled
Although qualitative evaluation can be performed with experience and
skill, this practice provides much chance for subjective error. The need for
interpretive standards and their preservation is another disadvantage.
Visual estimation of bran is facilitated by qualitative differential dye
staining. Several dye-staining techniques have been proposed: congo red
and methylene blue (27), Sudan III (28), eosin and methylene blue (29),
and alkaline alcohol (27 ,30). Treatment of rice kernels with appropriate
reagents accentuates the presence of residual bran, which stains differently
than the endosperm. Nonetheless, some degree of subjectivity is used in
interpreting the results.
Quantitative tests have been developed by extracting bound stains with
appropriate solvents followed by colormetric determination of color inten-
sity (27 ,30). One related method uses alkaline alcohol bran pigment extrac-
tion followed by spectrometric measurement of absorbance (30). Knowing
the absorbance value of the brown rice and of an unknown sample, the
degree of milling of the latter can be read from a single graph. The method
would be applicable only for rice varieties that have a common absorbance
curve.
An iodine test can also be used. Kernels of rice are immersed in an
iodine solution, and the residual bran on a kernel stands out against a blue
DEGREE OF MILLING 149

background, thus revealing its approximate degree of milling. Bran, which

becomes yellow, protects the underlying amylose from becoming blue.
Final judgment is still subjective. Objective evaluation of results has been
attempted using color meters. However, the amylose content varies widely
among rice varieties, and this will affect the intensity of the iodine color
reaction (31).
If the degree of milling is defined in terms of the FAO model system
(32), reliable discrimination between the inner and outer bran layers is
necessary. The outer and inner layers differ substantially in their chemical
composition. The outermost layers are fibrous, whereas the inner ones and
the germ are rich in proteins and lipids. The New May-Grunwald (NMG)
selective dye-staining method (29) can be applied. It uses a mixture of eosin
yellow and methylene blue dissolved in methyl alcohol. Each part of the
rice grain stains a different color: outer bran layer stains green; inner bran
layer and germ stain blue; and endosperm stains pink. The degree of mill-
ing is judged subjectively by comparing the colors of the test sample with a
standard that has been stained simultaneously in the NMG solution (Fig.
1). The main disadvantages are the need for standards and the subjective
evalutation of results.
NMG differential dye staining has been developed into an objective
procedure by Barber and Benedito de Barber (26), who defined a colored
bran balance (CBB) index. The CBB is defined as the proportion of kernel
surface covered by bran. After NMG staining, magnified plane images of
the stained kernels are produced using either a projected image or photog-
raphy. A photograph of the stained kernels has the additional advantage of
recording and storing of the data. Both the area occupied by the bran (blue

Figure 1 Photograph of samples from one rice lot that had been milled to various
degrees of milling and stained with New May-Grunwald reagent. Line A is the brown
rice control (DOM = 0.0%) and Line E is very well milled (DOM = 11.8% ).
150 WADSWORTH

and green portions) and the total kernel area are measured by planimetry.
CBB values range from 100 for brown rice, which is assumed to be com-
pletely covered with bran, to 0 for completely milled rice, which is assumed
free of bran. The actual CBB values obtained with commercial samples of
rice that had been well milled ranged from 3.0 to 5.5.
For undermilled or lightly milled rice, the bran layers from every kernel
and from every surface of a given kernel are not removed uniformly. Compar-
ison of the CBB values from a number of individual kernels gives an accurate
evaluation of the homogeneity of milling for the rice. Using a computerized
color image analyzer to process either the stained kernels or a color photo-
graph would greatly reduce the time required to measure the colored areas,
to calculate the CBB, and to estimate the milling homogeneity.
Webb (33) evaluated the NMG technique for estimating degree and
uniformity of milling for the U.S. Varietal Improvement Program. Some
applications of the test in the rice breeding program were: (a) to accurately
judge degree of milling in order to arrive at the correct settings for labora-
tory mills; (b) to identify varietal differences in milling response among
varieties milled under constant conditions as this would aid the breeder in
developing cultivars least resistant to milling and in eliminating those types
that are most resistant to milling; (c) to provide rice breeders a means of
identifying selections and varieties with minimal lateral and dorsal grooves
in the kernel as these kernels require harder milling (resulting in increased
breakage) to remove the bran streaks; (d) to investigate the effect of
overdrying of rice on degree of milling and on milling yields (overdried rice
appears to be more resistant to milling) and to determine the laboratory
mill settings needed to effect the same degree of milling in overdried rice as
in normal moisture rice; and (e) to estimate the extent of brown rice bran
damage from the laboratory shelling operation so that sheller settings can
be adjusted to minimize this damage.

B. Methods Estimating Chemical or Physical Properties

1. Chemical Composition
As discussed above, bran layers differ in their chemical composition. Most
of the constituents show a decreasing concentration gradient toward the
inner regions of the kernel. Methods based on the determination of a
constituent that progressively decreases with milling have been developed.
Measurements of the degree of milling through analysis of constituents that
occur predominantly in the bran include assays for surface lipid (4,34-36),
ash and protein (35), thiamine (37-39), and phosphorous (39).
These methods generally require carefully controlled and time-consu-
ming procedures to obtain reproducible results and are not suitable for on-
DEGREE OF MILLING 151

line control. For them to be universally applicable, the contents of the con-
stituents analyzed or the relationship between brown rice and milled rice
composition should be fairly constant. In actuality, there is considerable
variation among rice varieties and even among rice lots of the same variety
depending upon growing location, environmental conditions, and agro-
nomic methods employed during the growing season. Also, processing tech-
niques such as parboiling, which bring about migration of constituents within
the kernel, can cause unsatisfactory results.
2. Physical Characteristics
Physical methods for determining the degree of milling have generally
utilized the optical characteristics of the kernel. These methods depend
upon the reflection of light from and the transmission of light through the
milled rice at selected wavelengths. Kik (40), Angladette (41), Stermer et
al. (42), and Johnson (43) have developed optical devices for measuring the
degree of milling.
Although these methods are quite accurate for a single variety grown in
a particular area, they fail in wide use because color influences the measure-
ments. Another difficulty encountered in the use of optical devices is the
presence of chalky kernels in the rice to be measured. The chalky kernels
are not optically translucent and thus interfere with accurate measure-
ments. This drawback can be remedied by removing the chalky kernels by
photoelectric sorting machines before making the measurement of degree
of milling. However, this extra, time-consuming step would not be suitable
for on-line use. A method developed by Stermer (44) used the far red (660
nm) and the near infrared (850 nm), which reduced some of the problems
caused by color. Another physical method is based on the total lipid con-
tent of the rice measured by nuclear magnetic resonance (45). However,
the samples had to be dried to 5% moisture content before the NMR
measurements could be taken. Velupillai and Pandey (46) measured the
color of the rice with a tristimulus colorimeter system and correlated the
color measurements with the degree of milling.
In general, with the above techniques, the correlations within a given
rice lot are excellent. However, the application of these methods to differ-
ent rice lots has not been acceptable due to the variability in composition
and color among the rice lots. Other practical difficulties (measurements
are affected by broken kernels, abnormal kernels, and moisture content)
have prevented the wide acceptance of these devices by the rice-milling
industry. The industry needs a fast, reliable technique for measuring degree
of milling that is not affected by normal variations in the composition of
various rice lots and that can be employed for on-line control of the rice-
milling machinery.
152 WADSWORTH

V. NEAR INFRARED REFLECTANCE SPECTROSCOPY

NIR spectroscopy has proven to be a rapid method for analyzing the major
chemical constituents of many grains and oilseeds. Not much has been
published on the use of NIR with rice. Iwomoto et al. (47) reported on
using NIR to analyze rice for protein and amino acids. Their procedure
used rice samples that had been ground into a powder, which is the tradi-
tional method for NIR sample preparation. Hopkins and Keneaster (48)
measured surface lipids content of rice with NIR and correlated these
values with degree of milling. Kao (49,50) reported on the use of NIR for
estimating degree of milling as a grading tool for the Federal Grain Inspec-
tion Service. Kao also used ground samples in his research. His results
indicated that the NIR technique was not completely satisfactory for this
purpose. This was probably due to two reasons: the variability in composi-
tion of brown rice from which the milled rice was produced and the subjec-
tive method of rice grading with which the NIR results were being com-
pared. Also, the technique used ground rice samples, which would make it
less than ideal for on-line application.
Wadsworth et al [51] reported on a method for measuring the degree of
milling of rice using near infrared (NIR) spectroscopy that compensated for
lot-to-lot variations in the rice and that would be suitable for on-line use in
a rice mill. This study utilized a slightly different approach from Kao's to
the problem of estimating rice degree of milling with NIR. Whole kernel as
well as ground rice samples were evaluated. The degree of milling values
used to calibrate the NIR instrument in this study were calculated from the
actual weight of bran removed during milling rather than subjective evalua-
tions. And, the NIR spectrum of the brown rice entering the rice milling
machine and the spectrum of the milled rice lea'ljpg the mill were both used
in developing the calibration.

A. Experimental
1. Rice Degree of Milling Samples
Five lots of long-grain rice (two lots of Labelle variety and one each of
Tebonnet, New bonnet, and Lemont varieties) and one lot of medium-grain
rice (Mars variety) were used. One of the Labelle lots was microwave-
vacuum dried and is referred to as Labelle-MV. The moisture contents of
the rice lots ranged from 11.4 to 15.9% (wet basis).
The six lots of rough rice were cleaned and shelled. The brown rice from
the sheller was hand-sorted to remove unshelled kernels. The pure brown
rice was then milled for periods ot time ranging from 0 seconds (brown rice
control) to 60 seconds to produce milled rice having various degree of
DEGREE OF MILUNG 153

milling treatments. The degree of milling (DOM) was calculated as 1.0

minus the ratio of WM (weight of the milled rice) divided by W8 (weight of
the brown rice). The DOMs ranged from 0 to 16%. The standard error for
degree of milling determination was 0.527. The milled rice samples were
also analyzed for total lipids content.
2. Near Infrared Spectrometer
The NIR spectra were measured with a grating-type monochromator scan-
ning from 1100 to 2500 nm in 2-nm intervals. The spectra were collected
from both ground rice and whole kernel rice in the reflectance mode. The
rice sample (either ground or whole) was spinning while the NIR spectrum
was collected.
3. NIR Software
Near infrared spectral analysis (NSAS) software version 3.14 was used to
analyze the NIR results (52). NIR data acquisition, calibration, prediction,
and file utilities are all contained in one package.
The utility of the NIR method depends on developing a good calibration.
NIR calibration equations are developed by regressing the NIR spectral data
on known values of a parameter (in this case DOM) for a set of samples that
covers the range of interest for that parameter. The calibration equation is
then validated with samples not included in the calibration set.
This study used a cross-validation procedure, recommended by K. H.
Norris (personal communication), based on the six rice lots. A calibration
equation based on the data from only five of the six rice lots was developed
and used to predict the DOM values for the sixth rice lot. This procedure
was repeated until each rice lot had been used as the prediction set. Then
the predicted DOM values and the laboratory measured DOM values of
the six prediction sets were combined into one validation set that was used
to estimate the standard error of prediction for the NIR method.
Two regression techniques were used in developing the calibration
equations-multiple linear regression (MLR) and partial least squares
(PLS) regression. MLR uses the values in the spectrum at each wavelength
to select the wavelengths that best correlate with the DOM. While only a
few pieces of information (usually less than six) from the 700 available in
each spectrum are used in the MLR calibration, this approach has the
advantage that the selected wavelengths can be compared with known
absorption bands that have a physical or chemical relationship with the
parameter being evaluated. Thus, for rice degree of milling, one would
expect bands associated with lipid, protein, fiber, and starch to be selected.
PLS regression first computes a principle component analysis on the
spectra. The principle components, which are vectors derived solely from
154 WADSWORTH

the spectral data, are independent variables in multidimensional space that

utilize information contained in the entire spectrum. PLS then calculates a
linear regression of the major principal components (usually less than 15)
on the known DOM values. This method has the advantage of using all of
the information contained in the spectra of the calibration sample set. For a
detailed discussion of the mathematics used in NIR calibration, see
Hruschka (53) and Martens and Naes (54).
The NSAS software can perform several mathematical transformations
on the individual spectra before developing the calibration equations. These
math transformations are performed to eliminate unwanted information
from the spectra such as noise or particle size effects and to enhance the
separation of absorbance peaks and shoulders. The transformations used on
the rice spectra included n-point smoothing and first and second derivatives.
4. Normalization
An additional mathematical transformation (normalization of the spectra
to compensate for variation among rice lots) was performed on the rice
spectra. The smoothed NIR spectrum of each milled rice sample was di-
vided by the smoothed spectrum of the brown rice control for that sample,
wavelength by corresponding wavelength. These normalized spectra were
then further modified by first and second derivative transformations and
used to develop calibration equations.

B. Results
1. Milling
The results of the milling experiments are shown in Figure 2. The micro-
waved dried rice lot (Labelle-MV) was milled with two weight settings on
the McGill. As expected for rice milled in a friction-type mill, the degree of
milling increased rapidly to the 8-10% level as the soft tissue was removed,
and then more slowly as the more resistant outer layers of the endosperm
were removed. It is obvious that the rice lots did not all mill the same,
which demonstrates one of the problems facing rice mill operators.
The differences in milling response can be related to several factors
including varietal differences, growing conditions, and moisture content of
the rice (9). In this case, the Tebonnet, which had the greatest DOM
values, was milled at an unusually high moisture content (15.5% w.b.). The
Mars lot had a higher moisture content than the Newbonnet lot but milled
to a lesser degree, demonstrating varietal differences. To maintain consis-
tent quality, the rice mill operator must adjust mill settings when the rice
entering the mill changes. Present practice is for the operator to make these
adjustments manually, based on the visual appearance of the rice.
DEGREE OF MILLING 155
20

RICE
• Labella MY wt • 20cm

~ + Labella MY wt • 10cm

- * Labella
~ 10
D Mara
0
X Tellonnet
CD
!a 0 Lamont
CD 6 /::;. Newbonnat
c

010 203040608070
Milling Time (sec)

Figure 2 Effects of milling time on degree of milling for six different rice lots.
The microwaved-vacuum dried Labelle lot was milled with the weight located at
two positions on the pressure arm of the McGill rice mill.

The DOM vs. milling time data in Figure 2 were not linear. Since the
NSAS software uses linear regression as one technique for developing an
NIR calibration equation, the DOM data were linearized using a log trans-
formation similar to that for a first order rate process: log(1-DOM/100). A
graph of the transformed DOM vs. milling time is shown in Figure 3. One
phenomenon immediately obvious by examining the data in this form is the
change in rate constant (slope) during the milling process. The data from 0
to 15 seconds can be fit to one straight line, while the data from 20 to 60
seconds fit another line with a lesser slope. This transition occurs at a DOM
level of approximately 8-9% and is probably related to the significant
change in hardness of the outer layers of the rice kernel as they are re-
moved. Figure 3 also indicates that the different rice lots have different
milling rates. The values of these slopes might be good parameters for
characterizing the milling performance of various rice lots.
2. Math Transformations
Figure 4 illustrates the three mathematical transformations that were ap-
plied to the reflectance data during the development of calibration equa-
tions. The upper curve is the optical density data, log(l/R), that have
been smoothed to eliminate noise. With NIR spectroscopy, the baseline
slopes upward as wavelength increases. Particle size effects can cause an
156 WADSWORTH

0.1

RICE:
• Laballa-MV wt • 20om

*
0
Labella
Mara
X lltllonnet
0 Lamont
/::,. Newbonnat

010 203040608070
Milling Time (sec)

Figure 3 Degree of milling data (linearized by a log transformation) vs. milling

time. The changes of slope in the linear regression lines demonstrate a change in the
rate of bran removal as milling time and degree of milling increase.

offset in the baseline (finer particles reflect and scatter more light). The
middle curve is a first derivative transformation of the upper curve. A
first derivative transformation eliminates the offset due to particle size.
With the first derivative transformation, the points of maximum ab-
sorbance in the upper curve are zero values in the middle curve (i.e.,
slope = 0). Thus, a calibration made with a first derivation transformation
will not utilize information contained at absorbance maxima wavelengths.
The maxima are restored and highlighted using a second derivative trans-
formation (lower curve). Not only are the maxima more sharply defined
(now as negative peaks), but shoulders and inflections not even visible in
the upper curve can be easily seen because they also appear as peaks in
the second derivative curve. The second derivative transformation also
eliminates particle size offset and any linear effects from a shift in baseline
slope. Thus, with the derivative transformations you lose information that
is hopefully of no value while you enhance resolution of the absorbance
information.
3. Sample Preparation Effects
The two log(l/R) spectra in the upper portion of Figure 5 were taken from
a ground rice sample and a whole kernel sample. As indicated by the
relative positions of the spectra on the graph, the whole kernel sample
 1.50

1.20

~
~"'
.60

.30

~
~

"'
...~ .100

.....
0

. . 050

...
<.
!

• 000

.080

~
"'
...!I .030

...
0

~ -.020
.!...
~
-.070

-. 120 -1;1~100:::----::::t::------:-:1 =----:19<4;:;o:O----;;::;I;:;""---;2500;w;'

lliiYE length lnl!l

Figure 4 Illustration of the three mathematical transformations used for develop-

ing NIR calibration equations. This spectrum was measured with a sample of
Tebonnet brown rice.

157
 1.20

s.oo

.80

@
= .60
!
.«t

.20

.00
uoo S300 1!500 1700 1 2100

.120
~
= .060
~
!5
!... .000
!
I
~-.060

-.120

-.sao
uoo

Figure 5 Comparison of rice spectra measured with a whole kernel rice sample
and a sample that had been ground into a powder. These spectra were measured
with a sample of Mars variety brown rice.

158
DEGREE OF MILLING 159

absorbed more light than the ground sample. This was expected, since it is
known that smaller particles scatter more light.
The second derivative transformations of these spectra show essentially
identical peaks in both spectra, indicating that they contain the same infor-
mation. The magnitudes of the peaks at the low wavelength end of the
spectrum are slightly higher for the whole kernel sample. At the high end
of the spectrum (>1800 nm), the ground sample has significantly greater
peak intensities.
Figure 6 shows spectra measured from a ground sample of brown rice
and a ground sample of very well-milled (DOM = 13.17%) rice. As
expected, since the chemical composition of the bran layers that are re-
moved during milling are very different from the composition of the
starchy endosperm, NIR can detect differences related to milling. Figure
7 shows similar results for whole kernel samples. The upper half of this
figure shows log(1/R) spectra measured with samples of whole kernel rice
that had been milled to different degrees. With the whole kernel samples
at wavelengths less than 1800 nm, the changes in the spectra with degree
of milling are relatively greater than they were with the ground samples.

1.20

:::
...
a:
- .72

•24

a:
......
~
.
.....
al
...I
.120

...>
Ill

1i: -.040
...>
'-
.!
~ _. 200 +-:-,=--=t--=-----:-::±-=--(D_D_M=:-::1±3,... 7_l_--:-:::t'='----:~=----=±=--=:::-'
1100 1500 1700 1900 2300 2500
Navelength (nml

Figure 6 Comparison of spectra taken from a ground sample of brown rice and
from a ground sample of well-milled rice (DOM = 13.17% ).
 160 WADSWORTH

1.65

1.35

...
'if
......
at
0
:::::! 1.05

........
>-

....
CD
c .75
....
u
....
a.
c
.45

.092

-....
a:
......
at
0

..... .024
0

........
CD
>
ftl
>
~ -.044
CD
c
'0
~ DOM • 0
-.112

Figure 7 Comparison of spectra taken from whole kernel samples of Labelle

variety rice that were milled to varying degrees of milling. The upper graph com-
pares smoothed optical density (log 1/R) spectra. The lower graph compares second
derivatives of these spectra over a reduced wavelength range.
DEGREE OF MILUNG 161

Calibration equations were developed from both ground rice and whole
kernel rice samples that had DOMs ranging from 0 to 13.2%. Multiple
linear regression was used to develop the calibration equations. These
results are summarized in Table 5. For both ground rice and whole kernel
rice, the derivation transformations gave significantly better calibration
equations than the log(l/R). There was not much difference between the
regression results for ground and whole kernel samples. All of the calibra-
tion equations were based on wavelengths known to be associated with
light absorption by oil. The wavelengths selected by the MLR procedure
for the ground sample calibrations came from the higher end of the NIR
wavelength range, while those for the whole kernel calibrations came from
the lower end. It appears that spectral measurements made with whole
kernel rice, which would be more practical for on-line control application,
are as useful for predicting degree of milling as those made with the tradi-
tional approach to reflectance-mode NIR, ground samples.
4. Normalized Spectra
Lipid content is one method for estimating degree of milling. One of the
problems is that there is considerable variation in the lipid content of
different lots of brown rice. This variation could be due to variety, cultural
practices, growing location, etc. Thus, with the same degree of milling, two
rice lots might have different lipid contents. The same problem exists with
other methods of measuing DOM such as rice color. A relative change in
the parameter being measured can be more important than the absolute
value of that parameter for estimating degree of milling.
For on-line control of degree of milling using NIR in a rice-processing
plant, it would be possible to measure the NIR spectra of the brown rice
entering the rice milling machine, as well as the milled rice leaving the
milling machine. The before and after spectra of the rice might be more

Table 5 Regression Statistics for Two-Wavelength Degree-of-Milling

Calibration Equations Developed from NIR Measurements Made with Whole
Kernel and Ground Rice Samples
Ground sample Whole kernel sample
Math treatment R• SECb R SEC
log(l/R) 0.814 2.08 0.850 1.88
1st derivative 0.987 0.67 0.972 0.84
2nd derivative 0.988 0.67 0.981 0.69
•Multiple correlation coefficient.
bStandard error of calibration.
 162 WADSWORTH

effective in estimating degree of milling than the after spectrum alone. To

test this hypothesis, the spectrum of each milled rice sample was divided by
the spectrum of the brown rice from which the milled rice was produced.
Several normalized spectra with various DOMs are shown in Figure 8. The
second derivative transformations of the normalized spectra from Figure 8
are shown in Figure 9. More peaks appeared in the normalized spectra than
in the original.
5. Multiple Linear Regression Analysis
Table 6 summarizes the calibration and validation results obtained by MLR
analyses of NIR spectra measured with whole kernel rice. Both the original
spectra and the normalized spectra were used to develop the calibration
equations.
All of the calibration equations adequately estimated degree of milling.
In general, the derivative transformations gave slightly better results than

1.80

1.64

ir
::::-
1. 48
0>
0
...J

.,
'D

.........
N

1.32
"''-e
0
z

1.16

1700 1900 2100 2300 2500

Wave length (nml

Figure 8 Normalized spectra for different degrees of milling. The normalized

spectra are smoothed optical density spectra of milled rice that have been divided
by the smoothed optical density spectrum of the brown rice from which the milled
rice samples were produced.
 DEGREE OF MILLING 163

.100

.060

.......
Q)
>

...>
Ill

~ .020
...
c

10
'CI
Q)

!:: -.020
....Ill
E
L.
0
z
-.060

-.100~~~~~--~~--~~----~~--~~--~~~~~
1100 1300 1500 1700 1900 2100 2300 2500
Wavelength lnml

Figure 9 Second derivative transformations of the normalized spectra shown in

Figure 8. The second derivative spectra were calculated from the normalized log(l/
R) spectra.

Table 6 Summary of Multiple Linear Regression Results Using Five

Wavelength Calibration Equations
Calibration Validation
Spectra Math
set treatment Ra SECb R SEPC Biasd Slope adj.e

Original lot(l/R) 0.959 0.898

Original 1st der. 0.981 0.779
Original 2nd der. 0.982 0.758 0.968 0.972 -0.596 1.11
Normalized log(l/R) 0.969 0.805
Normalized 1st der. 0.986 0.595
Normalized 2nd der. 0.988 0.569 0.985 0.744 -0.209 0.99

•Multiple correlation coefficient.

bStandard error of calibration.
'Standard error of prediction.
<!Correction for the intercept of the regression line.
'Correction for the slope of the regression line.
164 WADSWORTH

the log(l/R) spectra. The cross-validation results are shown for the second
derivative calibration equations. The standard error of prediction (SEP)
was slightly higher than the standard error of calibration (SEC). The valida-
tion regression lines had a bias that deviated less than one SEC from zero.
The slope adjustments were all close to one. Normalized spectra gave
better results than the original spectra.
The wavelengths and coefficients for the MLR calibration equations
derived from second derivative transformations of the original and normal-
ized spectra are given in Table 7. Two wavelengths (1200 and 1720 nm)
were selected by MLR for both calibration equations. Others researchers
have established that oil is a very strong absorber at 1722 nm and that
starch absorbs at 1198 nm (55). The other wavelengths in the original
spectra calibration equation are 1180 nm (a very strong protein band), 1760
nm (another strong oil band), and 1220 nm (a strong cellulose band). The
normalized spectra calibration equation contained a very strong protein
band (2060 nm) and another oil band (2120 nm). The other wavelength in
this calibration equation (1640 nm) is not close to any known strong absorb-
ers in rice.
Figures 10 and 11 show plots of the NIR-predicted DOM values vs. the
measured DOM values for the second derivative calibrations derived from
the original spectra and the normalized spectra, respectively. The calibra-
tions appear to fit the data quite well except for the very high DOM
samples (DOM > 14% ). These very high DOM samples all came from the
Tebonnet variety rice lot, which was milled at a high moisture content. It is
unlikely that a commercial rice processor would mill rice at greater than
14% moisture content. Also, the DOM for commercial rice processing
would usually not exceed 10-12%. The calibration equations fit the labora-

Table 7 Calibration Equations Developed with Multiple Linear

Regression and Second Derivative Spectra Transformations Using
Both Original and Normalized Spectra
Original spectra Normalized spectra

Wave- Wave-
length Coefficient F-value length Coefficient F-value

3.3 -1.8
1180 697.3 203 1220 -73.9 248
1200 -640.3 295 1640 -534.6 232
1720 1101.7 1041 1720 655.3 1276
1760 1224.2 317 2060 -1049.8 339
1220 305.6 11 2120 673.9 38
 20

Cl
c
§ 16

-
::E
0

!12
•
Q
Cl
RICE ~RIETY:

't:ll • Labelle MY
.! 8
+ Labelle
~

*c
't:ll Mara
!
D. Tebonnet
a: 4
z X
~
Lemont
N-bonnet

0 4 8 n ~ 20
Actual Degree of Milling (%)

Figure 10 Calibration data set. NIR-predicted DOM values vs. lab-measured

DOM values. Multiple liner regression calibration equation developed from second
derivative transformation of the original spectra.

20

01

=
.E

-•
::E 18
0

! 12
•
Cl
RICE ""RIETY:
Q
't:ll • Labelle MY
.! 8
+ Labelle
u
:0
!
D. 4
*c Mars
Tebonnet
a: X
z ~
Lemont
N-bonnet

0 4 8 12 18 20
Actual Degree of Milling (%)

Figure 11 Calibration data set. NIR-predicted DOM value vs. lab-measured

DOM values. Multiple linear regression calibration equation developed from sec-
ond derivative transformation of the normalized spectra.

165
 166 WADSWORTH

tory results linearly over an acceptable range of DOM (0-14% ). When the
high DOM Tebonnet samples were omitted from the calibration set, the
SEC decreased significantly.
Figures 12 and 13 show plots (predicted vs. measured) of the cross-
validation results for the second derivative calibrations derived from the
original and normalized spectra, respectively. The trends in the validation
plots were very similar to the calibration plots. The calibration equation
based on the original spectra underestimated the degree of milling at DOM
levels greater than 14%. The calibration equation from normalized spectra
predicted the high DOM samples better.
6. Partial Least Squares Analysis
Table 8 summarizes the calibration and cross-validation statistics obtained
by the PLS analyses of the original and normalized spectra. The calibra-
tion and validation data sets were the same ones used for the MLR
analyses discussed above. The PLS results were similar to those obtained
by MLR. All of the calibration equations yielded satisfactory estimates of
degree of milling. The SEC, SEP, bias, and slope adjustment were within
acceptable limits. The derivative transformations gave slightly better re-

20

Q
c

-
i 11
:::E
0

~ 12

,•
Q
RICE "'RIETY:
Ci
• Labelle MY
.I
0
I
+ Labelle
;;
!
A.
4
*c Mara
Tebonnet
a:
z X Le•ont
~ N-bonnet

0 4 8 12 18 20
Actual Degree of Milling (t.)
Figure 12 Cross-validation data set. NIR-predicted DOM values vs. lab-
measured DOM values. Multiple linear regression calibration equations for cross-
validation developed from second derivative transformation of the original spectra.
DEGREE OF MILUNG 167

20

-•
0

! 12
•
Q

c RICE ..aRIETY:
"a • Labelle MY
.! 8
+ Labelle
:a
0

!
IL 4
*
[]
Mara
Tebonnet
e:z X LeMont
~ N-bonnet

0 4 8 n • 20
Actual Degree of Milling (%)
Figure 13 Cross-validation data set. NIR-predicted DOM values vs. lab-
measured DOM values. Multiple linear regression calibration equations for cross-
validation developed from second derivative transformation of the normalized
spectra.

Table 8 Summary of Partial Least Squares Calibration and Validation Results

Validation
No. of Calibration
Spectra Math PLS Slope
set treatment factors Ra SECb R sEpc Biasd adj.e

Original log(l/R) 6 0.983 0.752

Original 2nd der. 7 0.986 0.648 0.962 1.08 0.647 0.943
Normalized log(1/R) 6 0.991 0.564
Normalized 2nd der. 8 0.992 0.511 0.965 1.07 -.102 1.04

•Multiple correlation coefficient.

bStandard error of calibration.
<Standard error of prediction.
dCorrection for the intercept of the regression line.
•Correction for the slope of the regression line.
 168 WADSWORTH

suits than the log(l!R) data. The calibrations derived from the normalized
spectra were slightly better than those derived from the original spectra.
The PLS analyses yielded somewhat better calibration (SEC) results than
the MLR approach.
Figures 14, 15, 16, and 17 show plots of NIR-predicted DOM values vs.
the measured DOM values for the calibration and cross-validation data sets
using the PLS calibration equations. These results correspond to the MLR
results shown in Figures 10 through 13. As with the MLR calibrations, the
PLS calibration's estimates of DOM were low for DOM levels greater than
14%. However, it appears that the PLS estimates at high DOMs were
slightly better than the MLR estimates.
The PLS technique produced NIR calibration equations that gave some-
what better estimates of degree of milling than the MLR analysis. How-
ever, for the application of NIR to on-line control of degree of milling
during rice processing, the PLS approach may not be as practical as MLR.
To use PLS, a scanning-type monochrometer would be required. The MLR
approach could use a filter-type NIR instrument. The scanning instruments
have a purchase cost about 10 times that of a filter instrument. A large rice
processor will have dozens of rice-milling machines, and the slightly better

20

= 18
Dl

-•
:i
0

!Dl 12
•
0 I RICE •RIETY:

s • Labella MY
"0
8 I +
•.
.2 Labelle
"0
I 31. Mara
Q.
4 I c Tabonnat
!!: I
X
z Lamont
0 N-bonnet

0 4 8 12 18 20
Actual Degree of Milling (%)
Figure 14 Calibration data set. NIR-predicted DOM values vs. lab-measured
DOM values. Partial least squares calibration equation developed from second
derivative transformation of the original spectra.
 20

Cl

=
.5

-
18
::IE
0
CD
! 12
Cl

,
CD RICE VARIETY:
0
• Labelle MV
.! 8
+
,_g ¥
Labelle
Mara
! c Tebonnet
Q. 4
!!: X Lemont
z ~ N-bonnet

0 4 8 12 18 20
Actual Degree of Milling (tt)
Figure 15 Calibration data set. NIR-predicted DOM values vs. lab-measured
DOM values. Partial least squares calibration equation developed from second
derivative transformation of the normalized spectra.

20

Cl
~
-
18
:i

.
0
CD
CD 12
Cl
CD
,
Q
• Labelle MV
.!
u 8
:0
+ Labelle

!
Q. *c Mara
Tebonnet
a: 4
z X Lemont
~ N-bonnet

0 4 8 12 18 20
Actual Degree of Milling (tt)
Figure 16 Cross-validation data set. NIR-predicted DOM values vs. lab-
measured DOM values. Partial least squares calibration equations for cross-
validation developed from second derivative transformation of the original spectra.

169
 170 WADSWORTH

20

Q
c

-•
§ 18
::::E
0

! 12

•
Q

Q
'a • Labella MY
.!
C)
8
+ Labella
:a
!
D.
*
[]
Mara
Tab on nat
4
a: X Lamont
z ~ N-bonnat

0 4 8 ~ • 20
Actual Degree of Milling (lift)
Figure 17 Cross-validation data set. NIR-predicted DOM values vs. lab-
measured DOM values. Partial least squares calibration equations for cross-
validation developed from second derivative transformation of the normalized
spectra.

precision obtained with a PLS calibration might not justify the extra cost of
the scanning-type NIR systems.
7. Lipid Analysis
Lipid analysis of milled rice is currently the objective method used most
frequently to estimate DOM. In order to properly evaluate the NIR proce-
dure, it must be compared with the current methodology.
Figure 18 shows the variation of the lipid content with milling time. The
initial lipid contents of the brown rice samples were different, and the lipid
contents for a given milling time were different for the different rice lots.
A plot of degree of milling vs. lipid content is shown in Figure 19. The
linear regression equation for the data in this figure was:
DOM = 14.44 - 5.85L
where Lis the lipid content of the rice. The standard error for this regression
was 1.62, which is much larger than the standard error of calibration for the
NIR procedure (see Tables 6 and 7 for regression statistics). This indicates
that the NIR procedure is better for estimating degree of milling in rice than
 DEGREE OF MILLING 171

3.0
RICE:

--...,.
~ Labelle MV wt • 20cm
• Labelle MV wt • 10cm

2.0 +
A
+ Labelle

c
.!
c
~
~
• *c Mara
Tebonnat

i
•*
0
0 1.5 c X Lemont

,Q.•
*•a ~
• •
~ N-bonnet

1.0 • • A •
:::i c ~
t

•
c ~ A A
~ ~ A
0.5
c
8~ i
I t
~ fj
0.0
0 10 20 30 40 50 80 70
Milling Time (sec)
Figure 18 Variation in rice total lipids content with time of milling for six rice lots.

20
RICE:
• Labelle MV

15 + Labelle
~
*c Mara

=
Q
Tebonnet
X

-.,
Lemont
:i 10 ~ New bonnet

*
0

.,a
! •
*
c 5 *
•
OL-----~L-----~L-------L---~~~~~~~~~~

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Lipids Content (Ill.)
Figure 19 Simple linear regression of degree of milling on rice lipids content.
Regression equation: DOM = 14.44 - 5.85 x LIPIDS.
172 WADSWORTH

20

Q
• Lipid prediction

~ * NIR prediction

-•
15
~
0
•
!Q 10 •
•
c
'U
.!
0
5 •
:a •
!
D.

-5~-----------L------------L-----------~----------~

0 5 10 15 20
Actual Degree of Milling
Figure 20 Predicted values for degree of milling using two methods (the lipid
regression method and the NIR analysis method) vs. the lab-measured degree of
milling values. Standard error of calibration for lipid regression method = 1.62.
Standard error of calibration for the NIR method = 0.57 (MLR using normalized
spectra).

the lipid regression procedure. This point is further illustrated in Figure 20,
which shows NIR-predicted DOM values and lipid regression-predicted
DOM values, both plotted against the measured DOM values. It is obvious
that the scatter about the 45° line is much greater in the lipid-derived values
than in the NIR-derived values.

C. Conclusions
NIR degree of milling calibration equations developed using spectra mea-
sured with whole kernel rice samples were as effective for estimating
DOM as calibration equations developed using rice samples that had
been ground into a powder.
NIR was a more effective method for estimating degree of milling in rice
than the lipid regression method. NIR is also much faster than the lipid
method.
Derivative transformations of the NIR spectra gave better calibration equa-
tions than the smoothed optical density (log 1/R) spectra.
DEGREE OF MILLING 173

Normalization of spectra from the milled rice, by dividing the milled rice
spectra by the spectrum of the brown rice from which the milled rice was
produced, led to improved calibration equations.
Partial least squares analysis produced better degree of milling calibration
equations than multiple linear regression analysis. However, the use of
PLS for on-line control of rice-milling machines is probably not practical
due to the higher cost of NIR scanning monochromators compared with
filter instruments.
Near infrared spectroscopy appears to be a suitable (fast and precise) tech-
nique to use in developing an on-line control system for rice-milling
machines. However, a more comprehensive approach would be the de-
velopment of a comprehensive expert system that uses both NIR and
color measurements.

REFERENCES
1. Lu, S., and Luh, B.S. (1991). Properties of the rice caryopsis, Rice, Vol. I,
Production (B.S. Luh, ed.), Van Nostrand Reinhold, New York, p. 389.
2. USDA (1976). United States standards for rough rice, brown rice for process-
ing, milled rice, U.S. Dept. Agri., Agricultural Marketing Service, Washing-
ton, DC.
3. Wadsworth, J. I. (1991). Milling, Rice, Vol. I, Production (B. S. Luh, ed.),
Van Nostrand Reinhold, New York, p. 347.
4. Autrey, H. S., Grigorief, W. W., Altschul, A. M., and Hogan, J. T. (1955).
Effects of milling conditions on the breakage of rice grains. J. Agric. Food
Chern., 3:593.
5. Rhind, D. (1962). The breakage of rice in milling. A review. Trop. Agric.,
39:19.
6. Hogan, J. T. (1969). Rice processing and products research Southern Divi-
sion. Rice J., 72(7):54,56,58.
7. Bajaj, B., and Sidhu, J. S. (1984). Extended milling of Indian rice. Effect on
physicochemical and milling characteristics. Rice J., 87(4):15.
8. Matthews, J., and Spadaro, J. J. (1974). Rice breakage during milling, Proceed-
ings 15th Rice Technical Working Group, March 12-14, Fayetteville, AR, p.
51.
9. Webb, B., and Calderwood, D. L. (1978). Laboratory studies on the relation
of moisture content to degree of milling and milling yield in rice, Proceedings
17th Rice Technical Working Group, Feb. 14-16, College Station, TX, p. 53.
10. Pomeranz, Y., and Ory, R. L. (1982). Rice processing and utilization, CRC
Handbook of Processing and Utilization in Agriculture, Vol II (1. A. Wolff,
ed.), CRC Press, West Palm Beach, FL, p. 139.
11. Resurreccion, A. P., Juliano, B. 0., and Tanaya, Y. (1979). Nutrient content
and distribution in milling fractions of rice grain. J. Sci. Food Agric., 30:475.
174 WADSWORTH

12. Villareal, C. P., Maranville, J. W., and Juliano, B. 0. (1991). Nutrient content
and retention during milling of brown rices from the International Rice Re-
search Institute. Cereal Chem., 68:437.
13. Bhattacharya, K. R. (1985). Parboiling of rice, Rice: Chemistry and Technol-
ogy (B. 0. Juliano, ed.), American Association of Cereal Chemists, St. Paul,
MN, p. 289.
14. Desikachar, H. S. R. (1967). Some aspects of the storage and processing of
rice in India. Int. Rice Comm. News/., Special Issue.
15. Piggott, J. R., Morrison, W. R., and Clyne, J. (1991). Changes in lipids and
sensory attributes on storage of rice milled to different degrees. Int. 1 Food
Sci. Techno/., 26:615.
16. Kwon, Y. W., and Jeon, W. B. (1991). Effect of period and store house grade
and degree of milling on the sensory taste of cooked rice. Korean J. Crop Sci.,
36:271.
17. Pehu, E., and Siddiq, E. A. (1986). Studies on variability for resistance to
milling and protein loss on milling and various grain characteristics influencing
them in rice (Oryza sativa L.). Cereal Res. Commun., /4(3):297.
18. Srinivas, T., and Desikachar, H. S. R. (1974). Removal of bran and aleurone
layers at different points on the surface of rice grain during progressive polish-
ing. J. Food Sci. Techno/., IJ(4):83.
19. McGaughey, W. H. (1970). Effect of degree of milling and rice variety on
insect development in milled rice. J. Econ. Entom., 63:1375.
20. Pingale, S. V., Kadkol, S. B., Rao, M. N., Swaminathan, M., and Subrah-
manyan, V. (1957). Effect of insect infestation on stored grain. II. studies on
husked, hand-pounded, and milled raw rice and parboiled milled rice. J. Sci.
Food Agric., 8:512.
21. Normand, F. L., Soignet, D. M., Hogan, J. T., and Deobald, H. J. (1966).
Content of certain nutrients and amino acids pattern in high protein rice flour.
Rice J., 69(8):13.
22. Kennedy, B. M., Schelstraete, M., and Del Rosario, A. R. (1974). Chemical,
physical, and nutritional properties of high protein flours and residual kernel
from overmilling of uncoated milled rice. I. Milling procedure and protein,
fat, ash, amylose and starch content. Cereal Chem., 51:435.
23. Kennedy, B. M., and Schelstraete, M. (1974). Chemical, physical, and nutri-
tional properties of high protein flours and residual kernel from overmilling of
uncoated milled rice. II. Amino acid composition and biological evaluation of
the protein. Ceremal Chem., 51:448.
24. Bajaj, B., and Sidhu, J. S. (1989). Extended milling of Indian rice. II. Effect
on cooking and sensory quality characteristics. Chem. Mikrobiol. Technsm.
Lebensm., 12:46.
25. Bajaj, B., Arora, C. L., Chhibba, I. M., and Sidhu, J. S. (1989). Extended
milling of Indian rice. II. Effect on Mineral Composition. Chem. Mikrobiol.
Technsm. Lebensm., 12:58.
26. Barber, S., and Benedito de Barber, C. (1976). An approach to the objective
measurement of the degree of milling. Rice Process Engr. Cntr., 2(2):1.
DEGREE OF MILLING 175

27. Desikachar, H. S. R. (1955). Determination of the degree of polishing in rice.

I. Some methods for comparison of the degree of milling. Cereal Chern.,
32:71.
28. Borasio, L. (1955). Macrocolorimetric method for evaluation of the final
product of rice processing. Ann. Stazione Spermentale di Risicolture e della
Colture Irrigue, 3:51.
29. FAO. (1972). Degree of milling of rice: The standard method adopted by the
food agency, government of Japan. Sub-Group on Rice Grading and Standard-
ization. Intergovernmental Group on Rice, Rome, May 10-12, 1972. FAO
Document CCP:RI/CS C.R.S. 1.
30. Bhattacharya, K. R., and Sowbhagya, C. M. (1972). A colorimetric bran
pigment method for determining the degree of milling of rice. J. Sci. Fd.
Agric., 23:171.
31. Stermer, R. A. (1971). Comments on objective methods of determination of
the degree of milling of rice, Annex D. Comments on the proposals for a
revised model system of grading rice in international trade, Sub-group on Rice
Grading and Standardization, Intergovernmental Group on Rice, Rome, De-
cember 1971. FAO Document CCP:RIIGS 72/2.
32. FAO. (1971). Proposed defintions of the terms used in rice trade, Comments
on the proposals for a revised model system of grading rice in international
trade, Sub-Group on Rice Grading and Standardization, Intergovernmental
Group on Rice, Rome, December 1971. FAO Document CCP:RI/GS 72/2.
33. Webb, B. D. (1976). Some potential uses of the Japanese NMG selective dye
staining technique for estimating degree of milling in rice breeding programs.
Proceedings 16th Rice Technical Working Group, March 2-4, 1976, Lake
Charles, LA, p. 77.
34. Hogan, J. T., and H. J. Deobald. (1961). Note on the method of determining
the degree of milling of whole rr.illed rice. Cereal Chern., 38:291.
35. Watson, C. A., Dikeman, E., and Stermer, R. A. (1975). A note on surface
lipid content and scanning electron microscopy of milled rice as related to
degree of milling. Cereal Chern., 52:742.
36. Matthews, J., and Spadaro, J. J. (1980). Milling degrees of Starbonnet brown
assayed. Rice J., 83(2):12,14.
37. Kik, M. C. (1951). Nutritive studies of rice. Arkansas Agric. Exp. Sta. Bull.
508.
38. Lyman, C. M., Ory, R., Trant, M., and Rich, G. (1952). Determination of
thiamine in rice and rice products. A rapid and simple method. Anal. Chern.,
24:1020.
39. Desikachar, H. S. R. (1955). Determination of the degree of polishing in rice.
II. Determination of thiamine and phosphorus for process control. Cereal
Chern., 32:78.
40. Kik, M. C. (1951). Determining the degree of milling by photoelectric means.
Rice J., 54(12):18.
41. Angladette, A. (1957). Note on the subject of photoelectric control of the
milling of rice. Riz Rizic, 3(4):139.
176 WADSWORTH

42. Stermer, R. A., Schroeder, H. W., Hartstack, A. W., and Kingsolver, C. H.

(1962). A rice photometer for measuring the degree of milling of rice. Rice J.,
65(5):22.
43. Johnson, R. M. (1965). Light reflectance meter measures degree of milling
and parboiling of parboiled rice. Cereal Chern., 42:167.
44. Stermer, R. A. (1968). An instrument for objective measurement of degree of
milling and color of milled rice. Cereal Chern., 45:358.
45. Pomeranz, Y., Stermer, R. A., and Dikeman, E. (1975). NMR-oil content as
an index of rice milling. Cereal Chern., 52:849.
46. Velupillai, L., andJ. P. Pandey. (1987). Color and bran removal in rice process-
ing, American Society of Agricultural Engineers, ASAE Paper 87-6042, St.
Joseph, MI.
47. Iwomoto, M., Suziki, T., Kongseree, N., Uozumi, J., and Inatsu, 0. (1986).
Analysis of protein and amino acid contents in rice flour by near-infrared
spectroscopy. Nippon Shokuhin Kogyo Gakkaishi, 33(12):848.
48. Hopkins, D. W., and Keneaster, K. K. (1980). Measurement of surface oil in
milled rice by near infrared reflectance. Cereal Foods World, 25:518.
49. Kao, C. (1986). Determination of rice milling degree with near infrared reflec-
tance method, Proceedings, 21st Rice Technical Working Group, Feb. 24-26,
Houston, TX, p. 101.
50. Kao, C. (1988). Determination oftotal rice oil content with NIR instrumenta-
tion, Proceedings, 22nd Rice Technical Working Group, June 26-29, Davis,
CA, p. 90.
51. Wadsworth, J. 1., Sequeira, D. J., Velupillai, L., and Verma, L. R. (1991).
Rice degree of milling measured by NIR, American Association of Agricul-
tural Engineers, ASAE Paper 91-6030, St. Joseph, MI.
52. NIRSystems. (1988). Near infrared spectral analysis software, version 3.14,
Silver Springs, MD.
53. Hruschka, W. R. (1987). Data analysis: wavelength selection methods, Near-
Infrared Technology in the Agriculture and Food Industries (P. C. Williams and
K. H. Norris, eds.), American Association Cereal Chemists, St. Paul, MN, p.
35.
54. Martens, H., and Naes, T. (1987). Multivariate calibration by data compres-
sion, Near-Infrared Technology in the Agriculture and Food Industries (P. C.
Williams and K. H. Norris, eds.), American Association Cereal Chemists, St.
Paul, MN, p. 57.
55. Williams, P. C., and K. H. Norris. (1987). Qualitative applications of near-
infrared reflectance spectroscopy, Near-Infrared Technology in the Agricul-
tural and Food Industries (P. C. Williams and K. H. Norris, eds.), American
Association of Cereal Chemists, St. Paul, MN, p. 241.
 8
The Influence of Rice Protein on Rice Quality
Bruce R. Hamaker
Purdue University, West Lafayette, Indiana

I. INTRODUCTION
Although cooking, eating, and processing qualities of rice are mainly attrib-
uted to the starch component of the grain, rice proteins appear also to play
some part in determining rice quality. Studies have shown that gross pro-
tein content in rices affects texture to some degree, particularly tenderness
and cohesiveness of cooked rice (1-3). Specific proteins have also been
suggested to influence quality attributes (4,5).
The chemical basis of cooked rice texture is still not completely under-
stood. It is generally thought that the amylose component, which leaches
out of the starch granule during gelatinization, has a principal affect on
cooked texture due to its tendency to associate through hydrogen bonding.
However, amylose, while usually good at predicting texture for a wide
range of contents, is not as accurate in discriminating among rices differing
slightly in content. Also, occasionally rices appear that do not fit the estab-
lished inverse relationship between amylose and cooked rice stickiness. In
studies related to these discrepancies, various other components of the rice
kernel have been suggested to influence quality, i.e., long-chain linear
portions of the amylopectin molecule (6,7), nonpolar and polar fatty acids
(8), as well as protein.
The influence of protein on rice quality has not been well studied, al-
though several papers have been published on the subject. This chapter

177
178 HAMAKER

reviews the literature and presents the author's recent research on the
relationship between rice endosperm protein and texture.

II. PROTEIN IN RICE

A. Protein Content
The protein content of rice varies somewhat among cultivars. Studies con-
ducted on rices grown in India (9), the Philippines (10), and the United
States (5) showed gross protein content of milled, debranned rice (dry
weight basis) ranging from 6.0 to 9.0 (n = 18), 6.1 to 11.4 (n = 16), and 7.7
to 10.0 (n = 32) with means of 7.6, 9.4, and 8.4, respectively (N x 5.95).
Protein content is affected by climate and agronomic conditions as evi-
denced in the higher protein contents of cultivars grown in the midsection
of the United States (Texas, Louisiana, Arkansas, and Mississippi) com-
pared to California (11). Textural properties of these rices also differ (the
former group are less sticky and tender), however, this has not been as-
cribed specifically to differences in protein content.
Protein is nonuniformly distributed in the rice grain. There are greater
concentrations in the bran and periphery of the endosperm and smaller
quantities towards the center of the grain (12). Accordingly, milled, pol-
ished rice has a lower protein content than brown rice; about 82% is
retained after milling (13).
The application of nitrogen fertilizer may increase the protein content of
rice depending on when the fertilizer is applied and what type is used
(15,16). Protein content increased up to 4% (from 6 to 10%) when ammo-
nium sulfate was top-dressed at full heading time (16). Perez et al. (15)
found only a slight increase in protein content when urea was applied.

B. Protein Classification
Rice proteins are in a broad sense classified by the Osborne scheme, which
categorizes them on the basis fo their solubilities in specific solvents. Meth-
ods of extracting proteins and their composition have been reported by
various investigators (17-19). Cagampang et al. (17), using three rice vari-
eties that were milled and polished (protein contents of 6.8-8.5% ), found
water-soluble proteins, or albumins, to constitute 3.8-8.8% of the total
protein, salt-soluble proteins, or globulins, 9.6-10.8%; alcohol-soluble pro-
teins, or prolamins, 2.6-3.3%; and alkali-soluble proteins, or glutelins,
66.1-78.0%. Glutelin is also commonly extracted using a detergent such as
sodium dodecyl sulfate (SDS) with or without the addition of a reducing
agent (18,20,21). Rice is unique among the cereals in that its storage pro-
tein is primarily glutelin, while in most other cereals it is prolamin. Because
RICE PROTEIN AND RICE QUALITY 179

glutelin has a more evenly balanced amino acid profile than prolamin,
which is quite deficient in lysine and tryptophan, rice, though somewhat
lower in protein content than most other cereals, has protein of better
quality than most other cereal protein.

C. Protein Location Within the Kernel

Glutelin proteins are thought to be found exclusively encapsulated in pro-
tein bodies distributed throughout the starchy endosperm of the grain (22-
24). Three types of protein bodies have been identified: crystalline, small
spherical, and large spherical. Bechtel and Pomeranz (23) reported that the
central region contains only large spherical protein bodies, however, over-
all the small spherical bodies were most numerous and large bodies least
numerous. The large spherical protein bodies, which show concentric rings
on transmission electron microscope observation, are a deposition site for
prolamin as well as glutelin protein, while the small spherical bodies are
primarily glutelin (25).
The protein bodies remain intact during the cooking process, and large
spherical protein bodies become more resistant to digestive enzymes, thus
decreasing protein digestibility (26). However, this has not been a problem
nutritionally since the biological value of rice protein increases with lower
protein digestibility and net protein utilization becomes comparable to
other cereal proteins due to the fact that the undigested protein is in part
prolamin.
Little or no matrix protein has been found in the rice endosperm (22).
This differentiates rice from other cereals, which contain large amounts of
protein existing as an intergranular matrix. It remains, however, a possibil-
ity that connecting protein fibrils between protein bodies may exist that
could form a network in the rice endosperm and would be difficult to detect
even with an electron microscope.
A starch granule-associated protein has been identified. The 60 kd pro-
tein, also termed the waxy gene product, was found to be embedded in the
starch granule (27 ,28). This protein is a starch synthetase and correlates
with amylose content (29).
A comprehensive review of rice proteins and their distribution and loca-
tions can be found in Juliano (13).

III. METHODS AND TERMINOLOGY INVOLVING RICE

QUALITY
For purposes of this paper, "quality" refers to the eating, cooking, and
processing qualities of rice. Of principal concern is the texture of cooked
180 HAMAKER

rice, which is considered the most important factor in eating quality (14).
Cooking quality, which is closely related to eating quality, refers to milled
rice behavior during cooking (i.e., water uptake, degree of cooking, and
cooking time). Processing quality refers to the "processability" of rice in
the manufacture of rice products such as parboiled, quick cooking, canned,
and frozen.
The texture of cooked rice contains many elements, and a rather large
nomenclature has developed to describe them. However, in a broad sense,
texture may be divided into two aspects: 1) stickiness, also termed adhesive-
ness (and related to cohesiveness), and 2) firmness, also referred to as
hardness or tenderness. These parameters are negatively correlated (30).
Cooked rice texture can be measured by sensory analysis (31), direct
instrumental methods (e.g., Instron Tester, General Foods-Zenken Textu-
rometer, Tensiometer, Haake Consistometer) (30,32-36), or indirect meth-
ods (e.g., amylose content, gelatinization temperature, alkali spread test,
gel consistency test, amylography, viscometers) (10,37). In screening for
quality in rice-breeding programs, amylose content, gelatinization tempera-
ture, and gel consistency are commonly used as predictive tests.

IV. EFFECT OF GROSS PROTEIN CONTENT ON TEXTURE

A report in 1960 (12) suggested that rice kernel components other than
starch may influence the cohesiveness of the cooked grain. The authors
noted that starch has the capacity to swell up to 64 times its volume, though
cooked rice does not swell nearly to this degree. Suppression of this swell-
ing in the starchy endosperm was postulated to be due to protein encasing
the starch granules and to cell walls.
Primo et al. (2) found that protein content, especially in the peripheral
part of the kernel, correlated with cooked rice texture. Cooked rice texture
of rices high in protein tended to be tougher and chewier than rices low in
protein; the two factors weakly correlated. Onate et al. (38) tested three
varieties, each taken from two different plots that produced grains mark-
edly different in protein content. Significant differences related to protein
content were observed in sensory ratings for flavor, tenderness, and cohe-
siveness (Table 1). Rices with lower protein content were more flavorful,
tender, and cohesive. Color, aroma, and amylose content were not affected
by protein content. Protein content did not, however, explain the texture
differences among varieties, which were attributed to amylose content.
Juliano et al. (1), in a more extensive study, similarly reported that low-
protein samples of the same variety had more flavor and were more tender
and cohesive than high-protein samples cooked in the same amount of
water. Amylose contents were essentially the same for samples within a
RICE PROTEIN AND RICE QUALITY 181

Table 1 Composition and Eating-Quality Characteristics of Cooked Rice

of Three Varieties of High and Low Protein Contents

Sensory ratings•
Variety and Protein Amylose
sample (%) (%) Flavor Tenderness Cohesiveness
Taichung 1
#1 12.54 31.2 3.6 2.5 1.2
#2 7.78 31.2 4.5 3.4 2.2
59-368
#1 11.12 32.2 4.8 3.6 2.4
#2 7.08 31.4 6.2 4.9 3.6
Chia-nan 8
#1 9.27 23.2 6.0 6.7 5.3
#2 6.15 20.0 6.3 6.9 6.3
Overall Mean
#1 Low 4.6 4.0 2.7
#2 High 5.4 4.8 3.7
LSD (5%) 0.16 0.23 0.18
•Quality rating scores: flavor: 9, very full, rich characteristics-1, very weak; tender-
ness: 9, very tender-1, very tough; cohesiveness: 9, pasty-1, well separated.
Source: Ref. 38.

variety. Differences in texture among cultivars again could not be ex-

plained by protein levels. However, these researchers suggested that one
variety, Taichung 65, with high protein content and intermediate amylose
content, may have had a low cohesiveness sensory score due to its protein
content. Because U.S. rice cultivars do not differ substantially in gross
protein content, it is unlikely that texture could be ascribed to this parame-
ter. On the other hand, cultivars grown in the south central United States
have higher protein content than the same cultivars grown in California
and differ in texture measurements (11). Thus, protein content may partly
explain this phenomenon.
Cooking time was positively correlated to protein content (1,9). High-
protein samples also showed lower water absorption than low-protein sam-
ples (1,3). The suggestion was made (9) that more protein formed a thicker
barrier around the starch granule, thus slowing water uptake by the granule.

A. Effect of Nitrogen Fertilizer

The application of nitrogen fertilizer at full heading time has been shown to
effect protein content and cooked rice texture. Tamaki et al. (16) found
water-insoluble protein, or storage protein, to increase by 3-4% (from
182 HAMAKER

about 6 to 10% protein) when ammonium sulfate (2 g N/pot) was top-

dressed at full heading time. Using the General Foods-Zenken Texturom-
eter they found that cooked rice chewiness increased linearly from 6 to
10% protein (Fig. 1). Stickiness and the textural palatability index (value
obtained by multiplying chewiness by stickiness) increased from 6% to
about 8.5% protein content and then decreased at higher levels up to 10%.
Thus, they found cooked rice with high protein content to be chewier and
less sticky, and rice with low protein content to be softer and less sticky
than intermediate-protein rice. They suggested that protein may play an
important role in regulation of starch swelling by restricting the expansion
of the gelatinizing granule.
Yanase et al. (3) did texture measurements on 16 rices fertilized with
nitrogen to different degrees, which resulted in protein contents from 6.0
to 10.3%. Rices with high protein contents were generally less adhesive;
correlation between brown rice protein contents and cooked rice adhesive-
ness was r = -0.561 (significant at 5% level).

B. Effect of Grain Development

Textural properties that change during development of the rice kernel may
be related to changes in protein content. Tamaki et al. (39) found that
cooked rice became more sticky and less chewy as rice reaches maturity.

2.0

1.6

Q)
::I

>
c 1.2
•
Q)
L.
.....::IX 0.8
Q)
1-

0.4

0.0
6 7 8 9 10

Protein Content (%)

Figure 1 Relationship between the texture of cooked rice and protein content <•
= chewiness, _. = stickiness, e = textural palatability index). (From Ref. 16.)
RICE PROTEIN AND RICE QUALITY 183

They found the percent protein in the grain to decrease approximately

0. 75% from 15 to 50 days after heading and attributed changes in texture to
differences in restriction of starch expansion during cooking due to water-
insoluble protein, amylose, and starch lipids.

V. STARCH-PROTEIN INTERACTION AND TEXTURE

Recent work on starch-protein interactions in rice have centered on either
molecular associations in an oryzenin (glutelin)-starch model system or the
relationship between starch granule-associated protein and starch gela-
tinization (4,5,40). These approaches are attempts to understand how pro-
tein may influence textural properties of whole cooked rice or flour pastes.

A. Starch-Oryzenin Model System

Chrastil (4) found stickiness of whole cooked rice to be positively correlated
to the binding ratios of oryzenin to starch, amylose, and/or amylopectin.
Molar binding ratios of oryzenin/starch were measured in a model system
using isolated starch, amylose, and starch-free oryzenin (Table 2). Binding
was measured by noting changes in the differential absorption spectra of
oryzenin in the presence of starch compared to oryzenin alone. Whole rice
(medium and long grain), as well as the separated components, were used
directly or stored at 4°C and 40°C for 12 months to produce cooked products
differing in stickiness. Stickiness, measured using a method in which clusters
of cooked rice are counted and the distribution noted (41), was highest in
postharvest medium-grain rice followed by rice stored at 4°C and then 40°C.
These findings support a definite role for rice protein in influencing sticki-
ness, and future endeavors in this area will help us understand the bio-
chemical basis of rice texture and give clues on how to control or modify it.

B. Rice Proteins and Starch Gelatinization

It is known that protein in the endosperm may form a barrier to expansion of
the gelatinizing starch granules (12). The barrier was at one time attributed
to a continuous protein matrix surrounding starch granules as is present in
most other cereal grains. However, subsequent studies on endosperm mi-
crostructure showed protein in rice to be almost entirely encapsulated in
protein bodies with little or no matrix protein present (22,23). Since the
glutelin-containing protein bodies remain intact after cooking (26), and
probably contribute little to texture, there has been little interest in pursuing
protein-texture relationships.
Another constraint to starch gelatinization during the cooking process
may be the presence of proteins closely associated with the granule. It is
184 HAMAKER

Table 2 Interaction of Oryzenin with Starch Components from

Rice•

Rice nb mb n:mb

J\mylose + oryzenin
medium-grain rice
control 0.265 0.107 2.48
4°C 0.258 0.106 2.43
40°C 0.196 0.101 1.94
long-grain rice
control 0.155 0.101 1.53
4°C 0.149 0.100 1.49
40°C 0.116 0.089 1.30
J\mylopectin + oryzenin
medium-grain rice
control 0.264 0.115 2.30
4°C 0.259 0.117 2.21
40°C 0.113 0.064 1.76
long-grain rice
control 0.150 0.103 1.45
4°C 0.144 0.102 1.41
40°C 0.090 0.072 1.25
Starch + oryzenin
medium-grain rice
control 0.268 0.099 2.70
4°C 0.261 0.100 2.61
40°C 0.187 0.101 1.85
long-grain rice
control 0.150 0.098 1.53
4°C 0.146 0.101 1.45
40°C 0.117 0.107 1.09
•Medium- and long-grain polished rice was stored at 4 and 40°C. Control
is freshly harvested rice.
bn is moles of oryzenin, m is moles of starch, n:m is molar binding ratio.
Source: Ref. 4.

conceivable that starch granule-associated protein could affect the manner

in which the starch in the granule gelatinizes and the manner in which the
granule maintains its shape even when swollen. This protein could be found
either in the form of a sparse matrix closely associated with the granule, a
protein associated with the periphery of the granule such as the 15 kd
protein identified in wheat by Greenwell and Schofield (42), or a protein
embedded inside the granule such as the 60 kd protein identified by Sano
(27) and others (28).
RICE PROTEIN AND RICE QUAUTY 185

Swelling potential of starch granules and dispersibility of the swollen,

gelatinized granule have been hypothesized as factors that may influence
cooked rice texture (7,40,43). Bhattacharya et al. (44) divided rices into
eight types based on differences in texture and found that certain paste
viscosity parameters, such as Brabender amylograph "relative breakdown"
viscosity, correlated with texture of the whole cooked grain. Other viscom-
eters were also used to differentiate the eight rice types (37,45). Sandhya
Rani and Bhattacharya (43) reported a peculiar phenomenon where at low
slurry concentration (5%) in the presence of shear, hot paste viscosity oflow-
amylose rice flour was higher than for high-amylose flour, whereas, at high
slurry concentration (12%) the situation was reversed, with high-amylose
flour showing higher viscosity (Fig. 2). They postulated that during the
gelatinization process, low-amylose granules swell to a greater extent than
the more rigid high-amylose granules and at low concentration levels pro-
duce a more viscous paste. In contrast, at high slurry concentration the
greater fragility of the low-amylose granules cause a breakdown of its swol-
len structure due to the shearing of the swollen granules as they move against
each other, thereby resulting in a less viscous paste. This relative fragility or
rigidity of the gelatinized rice starch granule could relate directly to the
textural attributes of stickiness and firmness of cooked rice.
Some studies have been done that might explain this relationship further.

-
a.._
Cl)
Ill 5000
"'9-
--- - - ..!_2X HA
·c;
a.
+='

-
t:
Cl)
0

~
·u;

5
0
0
Ill
4000

- '

'
12" LA'
'A
3000

0 25 50 75

Cooking Time (min)

Figure 2 Effect on apparent viscosity of rice flour slurries held at 95oc in a

Brabender Amylograph for three time periods (shear rate 5.4 s- 1) (HA = high
amylose, LA= low amylose). (From Ref. 43.)
 186 HAMAKER

Hamaker and Griffin (40) disrupted protein structure through cleaving

disulfide bonds by adding reducing agents to the cooking water. Following
this treatment, cooked rice stickiness, as measured using the Instron Univer-
sal Tester, significantly increased in 7 of 9 commonly grown Arkansas rices
(Table 3). For the lower-amylose, sticky rices, the increase in stickiness was
substantial. Brabender amylograph curves were lower when a reducing
agent was added to the ftour:water slurry (Fig. 3a) (peak viscosity was

1000 {a)

--
: :;)
Ill
800

600
WATER

~ /DTT
u; /

_-- ,."""
/
0 /
0 400 /
VI .......
>
200

1000 (b)

--
: :;)
Ill
800

600
~
u;
0
0 400
VI
>
200

0
55 85 92.5 92.5 62 32
TEMPERATURE {OC)

Figure 3 Amylograms of rice flour at 10% solids in water; (a) with and without
dithiothreitol added, (b) incubated 2 hours before analysis in water or a solution
containing chymotrypsin, pronase, or bovine serum albumin (BSA). BU =
Brabender units. (From Ref. 40.)
RICE PROTEIN AND RICE QUALITY 187

Table 3 Stickiness Values for Rice Varieties Cooked in

Water or Water with Dithiothreitiol (DIT)•
Stickiness (g-em)
Variety No treatment +DIT Increase Sig.b
Short
S201 68.0 ac 91.0 23.0 *
Nortai 49.0 ab 80.3 31.3 **
Medium
M201 65.0 a 92.5 27~5 *
Nato 52.8 a 69.0 16.2 NS
Mars 20.5 be 43.8 23.3 *
Long
Lemont 5.0 c 10.3 5.3 *
Newbonnet 4.0 c 5.5 1.5 NS
Lebonnet 16.0 c 20.8 4.8 *
Tebonnet 6.0c 7.3 1.3 **
•Data determined using an Instron universal testing machine.
bSignificance between untreated and treated rices is indicated by: NS
= not significantly different; • = p < 0.05; •• = p < 0.01.
<Values within the column that are followed by different letters are
significant at p < 0.05.
Source: Ref. 40.

significantly lower in 8 of 9 rices) and were likewise lower when treated with
proteinases (Fig. 3b). The observed lower viscosity after protein disruption
in the presence of high shear was thought to be due to an increase in the
fragility of the swollen granules. Starch granule-associated protein may nor-
mally confer strength to the gelatinized granule by reducing the leaching of
amylose molecules or by physically "holding together" the granule.
Hamaker and Griffin (46) showed that degree of gelatinization in-
creased when protein structure was disrupted with the reducing agent 2-
mercaptoethanol (Table 4). Flour from long-grain nonsticky rices had a
lower degree of gelatinization than flour from sticky rices; the former had
the greatest increase when disulfide bonds were cleaved. This may be
related to the 60 kd protein, which is embedded in the starch granule
(27 ,28) and is present in higher amounts in high-amylose rices. Hamaker et
al. (5) found the correlation between this protein and stickiness (r =
-0.85, p < 0.01) to be similar to that between amylose content and sticki-
ness (r = -0.87, p < 0.01).
Studies relating viscosity measurements of unreduced and reduced rice
flour pastes to breakdown of the swollen granule have been done that indi-
188 HAMAKER

Table 4 Degree of Gelatinization in Rice Flours Before

and After Treatment with 2-Mercaptoethanol (2-ME)•·b

Variety N'fC +2-ME Change

% % %
Short
S201 92.9 a,b 96.8 a 3.9
Nortai 91.9 a,b,c 93.5 a,b,c 1.6
Medium
M201 93.7 a 96.9 a 3.2
Nato 90.4 a,b,c 87.2d -3.2
Mars 86.9 a,b,c,d 91.0 c,d 4.1
Long
Lemont 90.4 a,b,c 96.0 a,b 5.6*
Tebonnet 85.9 b,c,d 88.9 c,d 3.0
Newbonnet 85.1 c,d 91.6 b,c,d 6.5**
Lebonnet 81.6 d 88.5 d 6.9**

•Means within a column with the same letter are not significantly
different (p < 0.05).
bTreated group significantly different from untreated (* = p <
0.05, ** = p < 0.01).
cNT = not treated.
Source: Ref. 46.

cate that protein does give the gelatinized granule a degree of strength or
rigidity (46). As mentioned above, Brabender viscosity measurements de-
creased after protein structure was disrupted. However, when previously
gelatinized rice flour pastes were measured using the lower shear Brookfield
Viscometer or Bostwick Consistometer, paste viscosity increased in the pres-
ence of a reducing agent (Table 5). This suggested that, in the absence of the
normal disulfide-bound protein network associated with the granule, the
granule swelled to a greater degree and produced a more viscous paste.
However, with the high amount of shear present in Brabender amylography,
the swollen, gelatinized granules broke down more easily, resulting in a less
viscous paste. An experiment using a Carri-Med Controlled Stress Rheome-
ter showed breakdown of structure of the gelatinized paste at a much lower
degree of stress (dynes/cm2) when the reducing agent was present (Fig. 4).
These studies suggest that specific proteins may influence the gela-
tinization properties of rice starch granules and that this may be directly
related to textural properties of whole grain cooked rice. Identification of
the exact protein-starch relationships needs to be obtained. If specific pro-
teins are responsible for the observed changes in texture, rice texture might
be manipulated through breeding or processing techniques.
 RICE PROTEIN AND RICE QUALITY 189

Table 5 Apparent Viscosities of Short-, Medium-, and Long-Grain Rice Flour

Treated with or Without Dithiothreitol (DTT) Using the Brabender Visco/
Amylograph, Brookfield Viscometer, and Bostwick Consistometer
Brabender (BU)• Brookfield (cps
(peak viscosity) X 103)• Bostwick (mm)b
Variety NT• +DTT NT +DTT NT +DTT
Nortai 770 570*c 7.6 13.0** 23.5 16.3*
Mars 800 670* 14.7 16.5** 15.3 10.0*
Lemont 840 770* 14.2 16.4* 11.0 10.3

•BU = brabender units, NT = not treated, cps = centipoise.

bLower values represent increased viscosity.
<Treated group significantly different from untreated (* = p < 0.05, ** = p < 0.01).
Source: Ref. 46.

100000

- 10000
-------- .. --

I
Q)
rn
\
·sa. \
.._, 1000 \
\
~
·u;
0
0 100 '
rn
>
- +dithlothreltol
10 - - water

1
0 1000 2000 3000 4000 5000 6000 7000
2
Shear Stress (dynes/em )
Figure 4 Viscosity of an 8% gelatinized rice flour paste with and without
dithiothreitol subjected to incremental increases of shear stress. (From Ref. 46.)

VI. RICE AGING, PROTEIN, AND TEXTURE

Rice textural properties change significantly in the months following har-
vest (14). Freshly harvested rice is sticky and pasty when it is cooked.
However, after a few months of storage, it becomes less sticky and more
flaky. These changes have been attributed to a number of physico-chemical
190 HAMAKER

changes that occur in the grain, most of them being empirical relationships.
The biochemical basis for aging is still not well established. A few research-
ers have explored the role of proteins in aging and specifically have exam-
ined changes in sulfhydryl content and disulfide bonding (4,46).
Moritaka and Yasumatsu (47) found that rice flour consistency, using a
Brabender farinograph, increasing when an oxidizing agent was added.
Also, hydrogen sulfide, a constituent of cooked rice aroma, was found in
larger amounts in rice stored at soc than at 40°C. Lower temperatures
retard the aging of rice. Disulfides form in an oxidizing environment and
hydrogen sulfide is released only from free sulfhydryls. They concluded
that a decrease in free sulfhydryls (increase in disulfide bonds) during
storage may contribute to rice becoming less sticky and firmer by inhibiting
the swelling of the starch granule.
Chrastil (4) also showed a decrease in free sulfhydryls and an increase in
disulfide groups during aging. About 40% of total cysteine was in the
oxidized (disulfide) form after harvest compared to 60% after storage for
12 months at 40°C.
The formation of disulfide-bound protein complexes either through oxi-
dation of two cysteine groups or through sulfhydryl-disulfide interchange
reactions could conceivably restrict the expansion of the starch granules
during gelatinization and, thus, affect texture. As mentioned above,
Hamaker and Griffin (40) found a decrease in Brabender peak viscosity
when a reducing agent was added to cleave disulfide bonds (see Fig. 3a).
During the first months after harvest, Brabender viscosity measurements
increase (14,48). This could be due in part to disulfide bond formation.

Vll. CONCLUSIONS
The evidence presented in this chapter indicates that rice protein plays a
role in influencing rice quality, especially texture. The mechanism by which
protein can influence texture is currently not clear but may involve regula-
tion of water diffusion to the starch granule and control of granule swelling
during cooking. Chemical interactions between protein and starch may also
influence rice quality. However, since protein bodies remain intact upon
cooking, interaction between bulk (storage) protein and starch, while
shown to interact as isolated components, may be minimal in situ. This
interaction could be of considerable importance in defining the viscoelastic
properties of rice flour pastes. This does not, however, diminish the possi-
ble interaction of granule-associated protein with starch during cooking of
whole grain rice. At this point, no definitive conclusions can be made
regarding the control of texture by proteins. If proteins do in fact regulate
RICE PROTEIN AND RICE QUAliTY 191

texture to a significant degree, then texture may be regulated through

breeding for specific proteins or modifying them during aging or process-
ing. In this regard, future work in this area should stress the identification
of the protein or proteins responsible for influencing rice texture.

REFERENCES
1. Juliano, B. 0., Oiiate, L. U., and del Mundo, A. M. (1965). Relation of
starch composition, protein content, and gelatinization temperature to cook-
ing and eating qualities of milled rice. Food Techno/., 19:1006.
2. Primo, E., Casas, A., Barber, S., and Barber, C. B. (1962). Factores de
calidad del arroz. VI. Influencia de las proteinas sobre la calidad de cocci6n.
Proteinas en la capa externa. Rev. Agroquim. Teena/. Alimentos, 2:135.
3. Yanase, H., Ohtsubo, K., Hashimoto, K., Sato, H., and Teranishi, T. (1984).
Correlation between protein contents of brown rice and textural parameters
of cooked rice and cooking quality of rice. Rept. Nat/. Food Res. Inst., 45:118.
4. Chrastil, J. (1990). Protein-starch interactions in rice grains. Influence of stor-
age on oryzenin and starch, J. Agric. Food Chern., 38:1804.
5. Hamaker, B. R., Griffin, V. K., and Moldenhauer, K. A. K. (1991). Potential
influence of a starch granule-associated protein on cooked rice stickiness. 1
Food Sci., 56:1327.
6. Juliano, B. 0., Villareal, R. M., Perez, C. M., Villareal, C. P., Takeda, Y.,
and Hizukuri, S. (1987). Varietal differences in properties among high
amylose rice starches. Starch/Starke, 39:390.
7. Tester, R. F., and Morrison, W. R. (1990). Swelling and gelatinization of
cereal starches. II. Waxy rice starches. Cereal Chern., 67:558.
8. Maniiigat, C. C., and Juliano, B. 0. (1980). Starch lipids and their effect on
rice starch properties. Starch/Starke, 32:76.
9. Chakrabarthy, T. K., Dwarakanath, K. R., and Prabhakar Bhat, B. (1972).
Studies on physico-chemical properties of some varieties of rice. 1 Food Sci.
Techno[., 9:140.
10. Juliano, B. 0., Bautista, G. M., Lugay,J. C., and Reyes, A. C. (1964). Studies
on the physicochemical properties of rice. J. Agric. Food Chern., 12:131.
11. Juliano, B. 0. (1990). Rice grain quality: Problems and challenges. Cereal
Foods World, 35:245.
12. Little, R. R., and Dawson, E. H. (1960). Histology and histochemistry of raw
and cooked rice kernels. Food Res., 25:611.
13. Juliano, B. 0. (1985). Polysaccharides, proteins, and lipids in rice, Rice:
Chemistry and Technology (B. 0. Juliano, ed.), The American Association of
Cereal Chemists, Inc., St. Paul, MN, p. 59.
14. Juliano, B. 0. (1985). Criteria and tests for rice grain qualities, Rice: Chemis-
try and Technology (B. 0. Juliano, ed.), The American Association of Cereal
Chemists, Inc., St. Paul, MN, p. 443.
192 HAMAKER

15. Perez, C. M., Juliano, B. 0., De Datta, S. K., and Amarante, S. T. (1990).
Effects of nitrogen fertilizer treatment and source and season on grain quality
of IR64 rice. Plant Foods Human Nutr., 40:123.
16. Tamaki, M., Ebata, M., Tashiro, T., and Ishikawa, M. (1989). Physico-
chemical studies on quality formation of rice kernel. I. Effects of nitrogen top-
dressed at full heading time and air temperature during ripening period on
quality of rice kernel. lap. 1 Crop Sci., 58:653.
17. Cagampang, G. B., Cruz, L. J., Espiritu, S. G., Santiago, R. G., and Juliano,
B. 0. (1966). Studies on the extraction and composition of rice proteins.
Cereal Chern., 43:145.
18. Tecson, E. M.S., Esmama, B. V., Lontok, L. P., and Juliano, B. 0. (1971).
Studies on the extraction and composition of rice endosperm glutelin and
prolamin. Cereal Chern., 48:168.
19. Padhye, V. W., and Salunkhe, D. K. (1979). Extraction and characterization
of rice proteins. Cereal Chern., 56:389.
20. Juliano, B. 0., and Boulter, D. (1976). Extraction and composition of rice
endosperm glutelin. Phytochem., 15:1601.
21. Snow, S. D., and Brooks, J. R. (1989). Fractionation of rice glutelin poly-
peptides using gel filtration chromatography.l Food Sci., 54:730.
22. Harris, N., and Juliano, B. 0. (1977). Ultrastructure of endosperm protein
bodies in developing rice grains differing in protein content. Ann. Bot., 41:1.
23. Bechtel, D. B., and Pomeranz, Y. (1978). Ultrastructure of the mature
ungerminated rice (Oryza sativa) caryopsis. The starchy endosperm. Am. J.
Bot., 65:684.
24. Bechtel, D. B., and Juliano, B. 0. (1980). Formation of protein bodies in the
starchy endosperm of rice (Oryza sativa L.): A re-investigation. Ann. Bot.,
45:503.
25. Tanaka, K., Sugimoto, T., Ogawa, M., and Kasai, Z. (1980). Isolation and
characterization of two types of protein bodies in the rice endosperm. Agric.
Bioi. Chern., 44:1633.
26. Tanaka, Y., Resurreccion, A. P., Juliano, B. 0., Bechtel, D. B. (1978). Prop-
erties of whole and undigested fraction of protein bodies of milled rice. Agric.
Bioi. Chern., 42:2015.
27. Sano, Y. (1984). Differential regulation of waxy gene expression in rice
endosperm. Theor. Appl. Genet., 68:467.
28. Villareal, C. P., and Juliano, B. 0. (1986). Waxy gene factor and residual
protein of rice starch granules. Starch/Starke, 38:118.
29. Villareal, C. P., and Juliano, B. 0. (1989). Comparative levels of waxy gene
product of endosperm starch granules of different rice ecotypes. Starch/
Starke, 41:369.
30. Perez, C. M., and Juliano, B. 0. (1979). Indicators of eating quality for non-
waxy rices. Food Chern., 4:185.
31. del Mundo, A. M. (1979). Sensory assessment of cooked milled rice, Proc.
Workshop on Chemical Aspects of Rice Grain Quality, International Rice
Research Institute, Los Banos, Laguna, Philippines, p. 313.
RICE PROTEIN AND RICE QUAUTY 193

32. Okabe, M. (1979). Texture measurement of cooked rice and its relationship to
the eating quality. J. Texture Stud., 10:131.
33. Lee, S., and Peleg, M. (1988). Direct measurement of the attractive force
between individual cooked rice grains of sticky and flaky cultivars. J. Food
Sci., 53:1113.
34. Despande, S. S., and Bhattacharya, K. R. (1982). The texture of cooked rice.
J. Texture Stud., 13:31.
35. Juliano, B. 0., and Perez, C. M. (1983). Major factors affecting cooked
milled rice hardness and cooking time. J. Texture Stud., 14:235.
36. Sowbhagya, C. M., Ramesh, B. S., and Bhattacharya, K. R. (1987). The
relationship between cooked-rice texture and the physicochemical characteris-
tics of rice. J. Cereal Sci., 5:287.
37. Sandhya Rani, M. R., and Bhattacharya, K. R. (1989). Slurry viscosity as a
possible indicator of rice quality. J. Texture Stud., 20:139.
38. Onate, L. U., del Mundo, A. M., and Juliano, B. 0. (1964). Relationship
between protein content and eating quality of milled rice. Philipp. Agric.,
47:441.
39. Tamaki, M., Ebata, M., Tashiro, T., and Ishikawa, M. (1989). Physico-
chemical studies on quality formation of rice kernel. II. Changes in quality of
rice kernel during grain development. lap. Jr. Crop Sci., 58:659.
40. Hamaker, B. R., and Griffin, V. K. (1990). Changing the viscoelastic proper-
ties of cooked rice through protein disruption. Cereal Chern., 67:261.
41. Chrastil, J. (1990). Chemical and physicochemical changes of rice during stor-
age at different temperatures. J. Cereal Sci., 11:71.
42. Greenwell, P., and Schofield, J.D. (1986). A starch granule protein associated
with endosperm softness in wheat. Cereal Chern., 63:379.
43. Sandhya Rani, M. R., and Bhattacharya, K. R. (1989). Rheology of rice-flour
pastes: Effect of variety, concentration, and temperature and time of cooking.
J. Texture Stud., 20:127.
44. Bhattacharya, K. R., Sowbhagya, C. M., and Indudhara Swamy, Y. M.
(1982). Quality profiles of rice: A tentative scheme for classification. J. Food
Sci., 47:564.
45. Sandhya Rani, M. R., and Bhattacharya, K. R. (1985). Rheological proper-
ties of rice flour slurries and pastes. J. Food Sci. Techno/., 22:322.
46. Hamaker, B. R., and Griffin, V. K. (1993). Effect of disulfide bond-
containing protein on rice starch gelatinization and pasting. Cereal Chern., (in
press).
47. Moritaka, S., and Yasumatsu, K. (1972). The effect of sulfhydryl groups on
storage deterioration of milled rice studies on cereals (part 10). J. lap. Soc.
Food Nutr., 25:59.
48. Indudhara Swamy, Y. M., Sowbhagya, C. M., and Bhattacharya, K .R.
(1978). Changes in the physicochemical properties of rice with aging. J. Sci.
Food Agric., 29:627.
 9
Enrichment of Rice
Diane W. Hoffpauer
Gourmet Technologies, Inc., Crowley, Louisiana
Salmon L. Wright III
Wright Enrichment Inc., Crowley, Louisiana

I. INTRODUCTION

Although in 1941 the Food and Drug Administration passed the first
Enrichment Act of the Federal Food Code for bread and flour (1), few
people other than those in processing or research understood its im-
plications. The statute was originally developed by government, aca-
demic, and private concerns so that the basic food staples could provide
thiamine, riboflavin, and niacin lost in the bran layer of milled grains.
During World War II when food rations were required, fresh meat, poul-
try, and eggs were also scarce, so the addition of iron was included in
the Standard of Identity for Enrichment. During this era every year
brought more amazing biotechnical discoveries in the vitamin and min-
eral field. Food processors and scientists began to realize the positive
impact of the fortification of various foods with the essential vitamins
and minerals (1).
Enrichment generally refers to the restoration of vitamins and minerals
lost during processing. Fortification generally means adding vitamins and
minerals to foods in higher amounts than were present before processing.
Before the standards of identity were adopted to define enriched foods, the
term fortification was used when any vitamins and minerals were added to
foods.

195
196 HOFFPAUER AND WRIGHT

II. FOOD ENRICHMENT AND FORTIFICATION

A. General
Niacin deficiency or "black tongue" was first reported in the United States
in 1902, but the vitamin niacin was not determined to be the cure until
1937. Many dreaded diseases such as goiter, rickets, and blindness in chil-
dren were shown to be results of the lack of the intrinsic dietary factors
iodine, vitamin D, and vitamin A, respectively (1). As the research gained
momentum, more vitamins and minerals were added to stored foods. Even
today, dried milk and margarine are fortified with vitamin A and breakfast
cereals have a variety of vitamins and minerals added to give a more
complete meal. Although it has been disputed that the current diet of the
American people provides adequate supplies of the 12 vitamins and 7
minerals in the National Academy of Sciences Recommended Dietary Al-
lowances, we must wonder what effect low-fat/low-calorie diets, two-
income "fast food" families, microwavable foods, and unfortified foods are
having on the nutrition of the general population. Groups with special
dietary needs such as diabetics, heart disease patients, the elderly, and
infants all pose challenges to the traditional "basic four food groups" and to
the present-day food scientist.

B. Health Implications
In the world today concern for population health is low because our food
supply is more than adequate. Nutrition requirements in the animal and
poultry industry are taken very seriously, however, because of the commer-
cial value of the products. If animals and poultry do not gain weight cor-
rectly and quickly or the animal becomes ill, serious financial losses can
result. When the above conditions exist it is readily detected. In the popula-
tion, if there are marginal deficiencies, it is very hard to detect the symp-
toms, and physicians do not routinely run the tests required to detect the
preclinical problems associated with such deficiencies. In countries where it
is known that food supplies are low, it is expected that these deficiencies of
calories, protein, vitamins, and minerals will cause illness, and supplemen-
tal doses of all the required nutrients are administered.
Preclinical symptoms of vitamin A deficiency include increased suscepti-
bility to infection. Vitamin D- and vitamin C-deficient infants are restless
and sleep poorly. Preclinical symptoms of deficiencies of vitamin E, ribofla-
vin (vitamin B2), and niacin include some type of mild skin disorder such as
flaky dermititis and skin lesions. Deficiencies of several vitamins and miner-
als cause symptoms common to many illnesses such as fatigue, headaches,
irritation, sleep disturbances, poor appetite, and abdominal discomfort.
The requirement for the vitamins thiamine, riboflavin, niacin, and folic
ENRICHMENT OF RICE 197

acid is increased in the case of alcoholism. Vitamin B 12 should be increased

in tobacco smokers due to the presence of cyanide in the smoke (2).
In Japan there were many thousands of deaths from beriberi or thiamine
deficiency in the 1940s. After a major government effort to enrich rice with
thiamine, deaths from beriberi were eliminated by 1965 (3). Enrichment of
rice by consumers and housewives is encouraged to this day in Japan and is
readily available in the grocery stores.
Before World War II many people in Mississippi died from niacin defi-
ciency from eating a diet consisting mainly of corn meal. During World War
II, as a defense effort, enrichment laws or fortification laws, as they were
known then, were enacted. After World War II in 1948, an experiment on
the island of Bataan in the Philippines was conducted to dramatize the cause
and cure for beriberi. Highly milled or polished rice was being consumed as
the main diet of the people. Since most of the thiamine was being removed
with the bran layer, the disease was causing a high mortality rate on the
island. One half of the island was fed rice that had been enriched with
thiamine, niacin, and iron, resulting in no deaths and reduced symptoms of
the disease. The other half of the island had no decrease in the death and
disease rate (4). Armed with this information, vitamin manufacturers helped
get a fortification regulation enacted in Puerto Rico and initiated a "golden
grain" campaign in Cuba. In the United States, some states passed enrich-
ment laws for rice as well as for other grains and grain products. As of 1991,
the United States provides enrichment laws and standards of identity for
flour, bread, cornmeal, corn grits, macaroni and noodle products, and rice.

ill. ENRICHMENT STANDARDS

The original standard of identity and definition for enriched rice was
adopted by the Commissioner of Food and Drugs in the Code of the Federal
Register (CFR) on February 28, 1958. The current regulations covered un-
der Title 21 CFR 137.350 Enriched Rice (Aprill, 1990 edition) gives the
definitions and standard of identity for forms of milled rice to which nutri-
ents have been added for each pound of rice. The levels and nutrients are:
1. Not less than 2.0 mg and not more than 4.0 mg of thiamine; not less
than 1.2 mg and not more than 2.4 mg of riboflavin; not less than 16 mg
and not more than 32 mg of niacine or niacinamide; and not less than
13 mg and not more than 26 mg of iron (Fe).
2. Each pound of rice may contain not less than 250 U.S.P. units and not
more than 1000 U.S.P. units of vitamin D.
3. Each pound of rice may contain not less than 500 mg and not more
than 1000 mg of calcium (Ca).
4. In the case of enriched parboiled rice, butylated hydroxytoluene
198 HOFFPAUER AND WRIGHT

Table 1 Rice Enrichment Requirements and Standards of Identity•

Enrichment Standard
State or territory mandatory of identity
Alabama No Federal
Alaska No Federal
Arizonab No Federal
Arkansas No Federal
Californiac
Colorado No Federal
Connecticut Yes Federal
Delawareb No Federal
Florida Yes Federal
Georgia No Federal
Hawaii No Federal
Idaho No Federal
Illinois No Federal
Indiana No Federal
Iowa No Federal
Kansas No Federal
Kentucky No Federal
Louisiana No Federal
Mainec
Maryland No Federal
Massachusettsc
Michigan No Federal
Minnesotab No Federal
Mississippi No Federal
Missouri No Federal
Montana No Federal
Nebraska No Federal
Nevada No Federal
New Hampshirec
New Jerseyb No Federal
New Mexico No Federal
New Yorkd
North Carolina No Federal
North Dakota No Federal
Ohio No Federal
Oklahoma No Federal
Oregon No Federal
Pennsylvania No Federal
Rhode Island No Federal
South Carolina Yes State
South Dakota No Federal
ENRICHMENT OF RICE 199

Table 1 Continued
Enrichment Standard
State or territory mandatory of identity
Tennessee No Federal
Texas No Federal
Utah No Federal
Vermont No Federal
Virginia No Federal
Washington No Federal
West Virginia No Federal
Wisconsinb No Federal
Wyoming No Federal
American Samoad No
Guamd No
Puerto Ricoc
Virgin Islandsc
•Table 1 is a result of compilation of responses to inquiries directed to the
state or territory governments. It is provided for illustrative purposes and
should not be relied on as legal authority.
blnformation based on 1986 response from state or territory, no response to
1991 inquiry.
cNo information provided, state or territory did not respond to 1986 or 1991
inquiry.
dPartial reponse received, addressing information provided only.

(BHT), an antioxidant, may be added as an optional ingredient in an

amount not to exceed 0.0033% by weight of the finished food.
As of the April1, 1990, edition of the CFR law, the regulations in 137.350
(formerly 15.525) are stayed insofar as they require each pound of the rice
to contain not less than 1.2 mg and not more than 2.4 mg of riboflavin. This
stay shall continue until final action is taken disposing of the objections,
after public hearing thereon (5).
The levels of enrichment in the standard of identity for rice are designed
according to the levels for flour, cornmeal and pasta enrichment. Some
states require enrichment for flour, cornmeal, pasta and rice. Some states
require the enrichment of one or a combination of two or more of the
products. Ifthe enrichment is optional a state is not required to enrich, but
if the term "Enriched" is added to the label, vitamins and minerals must be
added according to the Federal Standard of Identity as outlined above [5].
At least one state, South Carolina, adopted a state wide standard of iden-
tity for rice enrichment for sales within the state and adopts the federal
standard of identity for interstate shipment of the product (Table 1). In
200 HOFFPAUER AND WRIGHT

Table 1, the 50 states and 4 United States territories are listed along with
their enrichment requirement if any, and the standard of identity used for
conformity.

IY. RICE-ENRICHMENT PROCESSES

A. Powder Enrichment
Currently two forms of enrichment are used by packagers and processors of
rice. The first form is a preblended powder mixture (Table 2) of U.S.P.
thiamine, riboflavin (if requested), niacin or niacinamide, and ferric ortho
phosphate (white iron), ferrous sulfate (yellow iron), or reduced iron.
Riboflavin and ferrous sulfate give the powder and the rice a slight off-
white to yellow color, which is undesirable to some consumers. Reduced
iron has the potential to turn the rice gray to black. Ferric ortho phosphate
is the most requested form of iron used in the industry due to its white color
and blending ability with the white rice. When ferric ortho phosphate is
oxidized or contains excessive moisture, it can turn tan, yellow, purple,
and/or black. When powder enrichment is used in the packaged product,
the statement -"To retain vitamins do not rinse before or drain after cook-
ing" is required according to 21 CFR 137.350 4(c) (5). One ofthe disadvan-
tages of powder enrichment in rice is that 20-100% of the enrichment will
wash off of the rice depending on the amount of water used in rinsing and
the application time. Other disadvantages of powder enrichment are that
the vitamins and minerals are less stable, they are difficult to detect and/or
assay in the final product, and they can easily react with food components.

Table 2 Forms of Enrichment Showing Preblended Levels

Vitamin/ Level
Form of enrichment Mineral (glib) Addition ratea

Vitamin/Mineral powder Thiamine 3.2 1 oz/100 lb rice

Niacin 25.6
Iron 33.6
Filler 391.2
Vitamin/Mineral-coated kernel Thiamine 0.4 1lb/199 lb rice
Niacin 3.2
Iron 2.6
Filler 447.4
•When vitamin/mineral enrichment is added at this rate, the final product
will contain the Federal Standard of Identity for thiamine, niacin, and iron.
ENRICHMENT OF RICE 201

The advantages of powder enrichment are that the blend is less expensive
than other forms of enrichment and it is easy to prepare.

B. Coated Kernel Enrichment

The second form of enrichment available in the industry is a premixed and
treated kernel-type enrichment. This "premix" as it is known in the rice
industry, is a powder-blended enrichment, which is applied to the milled
rice grains and coated with a water-insoluble surface (see Table 2) so as to
retain the vitamins and minerals during rinsing. The enrichment of the rice
grains is concentrated so that when added to milled rice at 0.5%, all of the
enrichment required per pound of rice is provided according to the stan-
dard of identity and can be rinsed without losing 85% of the vitamins and
minerals. The grains are not noticeable in the rice as they are the same
color as the rice. Parboiled, precooked, and regular rice can all be enriched
in this manner. Enrichment-coated grain is especially useful in instant and
precooked rice, which is cooked in an excessive amount of water. The
insoluble food-grade grain coating is broken down with the release of the
vitamins and minerals when it reaches the acid environment of the stom-
ach. 21 CFR 137.350 4(c) states in part that
unless the label of the food bears the statement, "To retain vitamins do
not rinse before or drain after cooking" immediately preceding or follow-
ing the name of the food and in letters not less than one-fourth the point
size of type used for printing the name of the food (but in no case less
than 8 point type) and the label bears no cooking directions calling for
washing and draining or unless the food is precooked and it is packaged
in consumer packages which are conspicuolously and prominently la-
beled with directions for preparation which, if followed, will avoid wash-
ing away or draining off enriching ingredients, the vitamins and miner-
als, (optional and otherwise) named in paragraphs (a) (1), (2), and (3)
of this section shall be present in such quantity or in such form that when
the enriched rice is washed as prescribed in paragraph (e) of this section,
the washed rice contains not less than 85% of the minimum quantity of
the substances.
One of the advantages of the "premix" or kernel-type enrichment method
is that washing instructions may be placed on the package and the product
will conform with the 85% clause in the above law. Other advantages are
that the vitamins and minerals are stable and will not react with other food
components and they are easy to detect and assay in the final product. The
disadvantage is that the product is slightly more expensive than powder
enrichment.
202 HOFFPAUER AND WRIGHT

C. Other Enrichment Processes

A number of kernel-type premix processes have been attempted, with or

without success. They can be described as a) coating grains with active and
inactive ingredients, b) infusing active ingredients into the grains and cross-
linking the starches, c) making simulated grains containing the ingredients,
and d) combinations of the above processes. The coating of the active and
inactive ingredients onto the surface of the grains has been the most success-
ful and inexpensive as a whole. When infusing ingredients into the grains, a
soaking and drying phase is usually employed, which is complicated and
expensive because of the extra drying step. Much hope has been given to
the simulated grain experiments, but drawbacks exist as to blending with
the natural product and consistency after cooking.

V. APPLICATION OF ENRICHMENT MATERIAL TO RICE

Powdered vitamins and minerals are applied to rice at various points during
the milling and packaging process. Powder enrichment mixtures are avail-
able for addition rates of 1, Vz, V4 ounce per 100 pounds of milled rice. The
most effective application of powder enrichment is soon after milling with
white and parboiled rice. The powder adheres to the grain well at this point
due to the heat and moisture on the grain surface. Kernel-type enrichment
does not require adhesion consideration because the grain is sealed and
does not have to stick to the milled rice. This enrichment simply blends
with the final product and is often added immediately prior to packaging. It
has been found in the industry that aspiration of rice will remove much of
the powder enrichment if applied before this step in packaging. Powder
does not adhere well to precooked rice. In this case it is easier to apply 1
ounce to 100 pounds of rice (see Table 2) rather than 1/4 ounce of enrich-
ment. Because brown rice has more oil on its surface, vitamins and miner-
als adhere readily. Antioxidants can be added at the same time to prevent
the common rancidity problem associated with brown rice. Brown rice is
not included in the Standard of Identity for Enriched Rice (5) and there-
fore can be enriched or fortified at any level desired.
Enriched kernel-type permix is added to rice at the rate of 1 pound of
premix to 199 pounds of rice, or 0.5% (see Table 2). Products that are on
the market with instructions to rinse the rice on the package or with instruc-
tions to boil in excess water (such as boil-in-the-bag) after which the water
is discarded are usually enriched with a cook and rinse-resistant kernel-
type enrichment.
ENRICHMENT OF RICE 203

VI. RINSE, COOK, AND WASH RESISTANCE OF

ENRICHED RICE
It is possible to attain rinse and cook resistance using any of the above
methods with the selection of the best ingredients for the purpose. One
consideration is that the desired ingredients can be secured so completely
within the grain that the vitamins and minerals will not be released for
absorption in the digestive system and thus go through the system without
benefit to the individual. Incorporating ingredients that are digestible be-
comes very important when assimilation is considered.
Since humans discovered the undesirability of eating dirty food, they
have been washing rice and other foods to remove actual or imagined dirt.
Most food processors and manufacturers take great efforts to ensure that
their products are as clean and as sanitary as possible. In some parts of the
world this is not the case. Some people will always wash their rice whether
it is necessary or not. A product can be processed and be very clean, only to
be poorly packaged, handled, and stored and thus need washing or even
sterilizing. In the early days of rice packaging, much attention was given to
the appearance of rice, and a clear, translucent grain was very desirable in
the market. The effect was accomplished by using glucose to make talc
adhere to the grain. Talc, also known as "soapstone" or "steatite," consists
of finely powdered native hydrous magnesium silicate (6). Since talc was
considered inedible, instructions were placed on the package to remove it
by washing. Because of universal washing of rice, early fortification (enrich-
ment) efforts were made to utilize wash-resistant treated grains that were
added to regular milled rice at the rate of 1 pound of enrichment to 199
pounds of rice (0.5% ). One grain in every 200 was coated with enough
thiamine, niacin, and iron to fortify to the desired levels. In the 1960s, a
movement developed to stop the practice of adding talc to rice because its
addition was considered cosmetic only and did not contribute any food
value to the rice. There has been no clinical evidence or research produced
to substantiate the claim that talc is harmful to humans, however.

VII. CONCLUSIONS
Much debate has been carried out as to the benefits of adding vitamins and
minerals to foods (7). Other than the enrichment of the traditional cereal
grains wheat, corn, and rice, fortification has been added to a number of
processed foods. Recently the powdered drink mixes used in weight reduc-
tion plans have enjoyed a positive response from consumers. The drink
mixes are fortified with a variety of vitamins and minerals. The long-term
204 HOFFPAUER AND WRIGHT

effect of such products has yet to be determined. Enrichment of rice has

endured and is continuing to reach new populations worldwide as a basic
way to combat starvation and dietary deficiency. New research is being
carried out to fortify grains with the missing factors of a particular popula-
tion and not change their normal diet. An example is the fortification of
wheat for Africa with vitamin A. New challenges include stabilization and
storage techniques for areas without facilities enjoyed in the United States.
The Food and Drug Administration has traditionally discouraged fortifica-
tion of foods with extra ingredients for fear of "overvitaminization" of the
population (8). This fear may be justified in the case of the fat-soluble
vitamins and some minerals, but the effects and need for other ingredients
including water-soluble vitamins will continue to be explored.

REFERENCES
1. Erdman Jr., J. W. (1989). Nutrition: Past, present and future. Food Techno/.,
43:220.
2. Berkow, R. (ed.) (1977). The Merck Manual, 13th ed. Merck & Co., Inc.,
Rahway, NJ, p. 1160.
3. Mitsuda, H. (1969). New approaches to amino acid and vitamin enrichment in
Japan, Protein-Enriched Cereal Foods For World Needs (Dr. Max Milner, ed. ),
American Association of Cereal Chemists, St. Paul, MN, p. 215.
4. Salcedo, J., et al. (1950). Artificial enrichment of white rice as a solution to
endemic beriberi. Report of field trials in Bataan. J. Nutr., 42:501.
5. Title 21 Code of the Federal Register, Chapter 1, Part 137.350-Enriched
Rice, April 1, 1990 edition and as amended in 1977, 1982, 1984, and 1989.
6. Windholz, M. (ed.). (1983). The Merck Index, lOth ed. Merck & Co., Inc.,
Rahway, NJ, p. 1299.
7. Austin, J. E. (1978). Cereal fortification reconsidered. Cereal Foods World,
23:233.
8. "Minimum Daily Requirement." Chemical Business (Jan. 1990), p. 38.
 10
Starch Gelatinization in Brown and Milled
Rice: A Study Using Differential Scanning
Calorimetry
Wayne E. Marshall
Agricultural Research Service, U.S. Department of Agriculture,
New Orleans, Louisiana

I. INTRODUCTION
The literature contains numerous references to gelatinization of starch
from plant material from such diverse sources as cereal grains, tubers, and
legumes. Starch gelatinization has been most extensively examined in ce-
real grains, and there is a growing body of literature pertaining to the
gelatinization of rice starch and, to a lesser extent, rice flour. These studies
have added considerable information about the gelatinization properties of
starch from different rice varieties. They have examined changes in gela-
tinization parameters when rice starch or flour has been exposed to other
food components, such as different moisture levels (1-4), protein (5), lipid
(6), salt (3), and both salt and sugar (4).
Rice is unique among cereal grains in that it is primarily consumed in
the cooked, whole grain form (7). Since milled rice (the most commonly
eaten form) is 85% starch, starch gelatinization during rice cooking is the
most important determinant of texture in the cooked product. Starch
must be gelatinized during cooking, otherwise rice is often too firm to be
eaten comfortably. The temperature at which starch gelatinizes is also
important because this parameter is related to the cooking time (8). Cook-
ing time is an important variable in both household and commercial appli-
cations. Finally, the heat energy required to completely gelatinize starch
in rice is critical to the rice processor, who must optimize heat input,

205
206 MARSHALL

cooking time, and temperature and, at the same time, minimize the cost
of the entire cooking process. Therefore, for such cooking applications,
rice starch gelatinization should be studied in situ, that is, directly in the
whole grain, and gelatinization should be followed during the cooking
process. Due to the lack of suitable instrumentation, there has been little
information generated about the thermal properties of rice starch during
rice cooking.
An analytical instrument used extensively to determine gelatinization
temperature, the heat energy input required for gelatinization, and the
degree of starch gelatinization is the differential scanning calorimeter
(DSC). Much of our quantification of rice starch gelatinization has been
attributed to measurements made with this instrument. Since the DSC
allows programmed heating of the sample at a constant rate and precise
control of temperature, this instrument can simulate the rice-cooking pro-
cess and yield valuable information to the rice processor. Until recently,
most calorimeters could not analyze whole rice grains because of sample
size limitations. The introduction of newer calorimeters, with slow scan
rates and large sample compartments, has made possible the gathering of
information on starch gelatinization in situ during simulated rice cooking.
An advantage of measuring the thermal characteristics of whole grain rice
is a more realistic representation of the gelatinization process related to
whole grain cooking.
The purpose of this chapter is to present a comprehensive review of
recent research done in the author's laboratory on the in situ determination
of starch gelatinization in brown and milled rice using differential scanning
calorimetry. The chapter will examine the effect of soaking and milling of
rice and the effect of grain particle size on starch gelatinization. The chap-
ter will then discuss the results in terms of the role played by the mainte-
nance of grain structure in influencing starch gelatinization and conclude by
describing future research directions where in situ DSC analysis can be
extended to other foods.

II. DIFFERENTIAL SCANNING CALORIMETRY

This section is devoted to a general description of the calorimetric method
used to analyze both unmilled and milled rice along with rice flour. The
first in situ description of rice starch gelatinization was reported by Nor-
mand and Marshall (9). Subsequent studies have been reported (10-13). In
all of these studies, a Hart Scientific model 7708 Differential Scanning
Calorimeter (Hart Scientific, Pleasant Grove, UT) with three 1-ml capacity
sample ampules and a reference ampule was used. The ampules normally
STARCH GELATINIZATION 207

contained 13-14 brown or milled rice grains (260-270 mg) or about 260 mg
of rice flour. The water:rice or water:flour ratio was 2.3:1, which produced
a moisture content of 70% (wet basis) in the rice or flour after heating in
the calorimeter. Seventy percent is the usual moisture content of fully
cooked rice (14). At this final moisture level, the starch in the rice or flour
was completely gelatinized. Inspection of the ampule contents at the end of
the DSC run revealed that all of the water had been absorbed by the
samples. For rice, there was no evidence of hard centers, indicating the
grains had been completely cooked in the calorimeter.
Milled rice was subjected to a presoak period of 45 minutes in order to
reach an equilibrium moisture content of 38-40% (wet basis) at room
temperature. Brown rice was presoaked for 5 hours in order to equilibrate
the moisture content. After the equilibration period, the samples were
analyzed in the calorimeter.
Two successive heating and cooling cycles were run on all samples.
Heating and cooling rates were normally 1.0°C/min but were altered de-
pending upon the study undertaken. The first cycle scanned the tempera-
ture range of 20-ll0°C and, after a brief 5- to 8-minute hold time at
ll0°C, scanned back to 20°C. The second cycle followed almost immedi-
ately and was identical to the first. The second heating established a
baseline for each run since, in all cases, no thermal transitions attributed
to starch gelatinization appeared during the seond heating due to the well-
known irreversibility of the gelatinization process. Baseline subtractions
were made on all thermal curves. By using the gelatinized starch to repre-
sent the baseline (reference) for the ungelatinized samples, differences in
heat capacities between sample and reference could be eliminated,
thereby creating a horizontal baseline for ease and greater accuracy of
measurement.
Thermal curves generated by the calorimeter depicting rice starch gela-
tinization were characterized by three temperatures-onset (T0 ) , peak
(Tp), and conclusion (Tc)-and by the gelatinization enthalpy (~H), which
was expressed as J/g of rice or flour at 11% moisture. Values for these
parameters were obtained from the thermal curves employing a series of
intersecting lines, as shown in Figure 1. Once the points of intersection
were fixed, perpendicular lines were dropped from the points and values
for T0 , TP, Tc were read on the temperature axis. ~H was obtained by
integrating the area under the curve using a baseline as shown in Figure 1.
Placement of the lines in Figure 1 was based on a procedure by Lund (15).
The calorimeter was periodically calibrated using internal heaters or by
determining the heat of fusion ( ~Hru.) of water. The experimental value for
~Hru. normally was within ±5% of the literature value.
208 MARSHALL

3:
0
...J
u...
~
w
J:
u
:IE
a:::
w
J:
t-
o
0
z
w

I 60 70 80 90
TEMPERATURE (•C)

Figure 1 Determination of the thermal parameters T 0 , TP, Tc, and AH from a

thermal curve depicting starch gelatinization in whole grain rice. Thermal curves for
rice flour are treated in a similar manner, except there is no low temperature
shoulder.

III. STARCH GELATINIZATION IN MILLED RICE AND

MILLED RICE FLOUR
A. Thermal Curves for Milled Rice and Milled Rice Flour
Thermal curves depicting starch gelatinization for four different varieties of
milled rice are presented in Figure 2. The four varieties represent three
different grain types: Lemont (long), Mars (medium), and S-201 and
Calmochi (short). In addition, high (Lemont), medium (S-201, Mars), and
low (Calmochi) amylose varieties are also represented (Table 1). All four va-
rieties exhibited a bimodal starch gelatinization endotherm characterized by
a low temperature shoulder followed by the main endotherm. These thermal
curves are in contrast to the thermal curves developed for flours from the
same four varieties (Fig. 3). The flour samples yielded a unimodal or single
starch gelatinization endotherm followed by a small, higher temperature en-
dotherm. For Calmochi, the small endotherm was absent. These endotherms
were attributed to melting of an amylose-lipid complex formed during heat-
ing in the calorimeter or already present in the flour (6). Calmochi appar-
 Table 1 Amylose Content8 and Thermal Curve Data8 of Whole Grain Milled Rice and Milled Rice Flom·b
Whole grain milled rice Milled rice flour
Amylose Gelatinization temperatures eq Gelatinization temperatures eq
Rice content Enthalpy Enthalpy
variety (%) To TP Tc (i\H) (J/g) To TP Tc (i\H) (J/g)
Lemont 25.0 ± 0.1 72.7 ± 0.2 93.5 ± 0.5 NDC ND 70.3 ± 0.1 75.9 ± 0.0 83.8 ± 0.1 10.3 ± 0.1
Mars 15.8 ± 0.1 65.4 ± 0.4 86.4 ± 0.7 98.6 ± 0.2 15.8 ± 0.4 65.3 ± 0.4 72.9 ± 0.4 82.2 ± 0.2 10.5 ± 0.1
S-201 20.6 ± 0.8 55.4 ± 0.7 82.6 ± 0.4 97.6 ± 0.8 13.3 ± 0.1 59.2 ± 0.2 66.6 ± 0.1 77.1 ± 0.1 9.5 ± 0.1
Calmochi 6.6 ± 0.0 60.5 ± 0.5 81.7 ± 0.3 93.5 ± 0.5 14.9 ± 0.4 61.0 ± 0.1 68.8 ± 0.1 78.8 ± 0.01 11.2 ± 0.4
•Values given are means ± SEM of triplicate determinations.
hMoisture content of all samples was 70%.
<Not determined due to absence of concluding baseline.
Source: Ref. 9.

g
210 MARSHALL

Lemont

40 50 60 70 80 90
TEMPERATURE (°C)

Figure 2 DSC thermal curves of different milled rice varieties. The final mois-
ture content of all samples was 70% (w/w) and the heating rate of the calorimeter
was 1.0°C/min. (From Ref. 9.)

ently does not have sufficient lipid-associated amylose to produce a detect-

able endotherm. Melting of these complexes likely occurred in whole grain
rice as well as in the flour. The presence of these thermal curves was probably
obscured by the much larger gelatinization endotherms or occurred at tem-
peratures higher than could be recorded by the calorimeter (12).
Table 1 gives the thermal parameters taken from the thermal curves in
Figures 2 and 3. Milled rice for all varieties had TP and Tc values about 13-
200C higher than comparable flour samples, but T 0 values were similar. ~H
values were 4-5 Jig higher in the milled rice compared to the flour. The
four varieties were characterized by distinctly different gelatinization tem-
peratures, which were variety specific.

B. Models to Explain Rice Starch Gelatinization In Situ

The observed differences in thermal curves and differences in the thermal
parameters for milled rice and milled rice flour could be attributed to one
or more of the following models: a) the existence of two populations of
starch granules with their own unique gelatinization characteristics, b) the
STARCH GELATINIZATION 211

;t
0
....J
1&.
....
~
w
J:
0
:i
a:
w
....J:
0
Q
z
w

l
40 50 60 70 80 90 100 110
TEMPERATURE (°C)

Figure 3 DSC thermal curves of different milled rice flour varieties. The final
moisture content of all samples was 70% (w/w) and the heating rate of the calorime-
ter was 1.0°C/min. (From Ref. 9.)

existence of temperature gradients in the grain (heat transfer effects) which

would broaden the thermal curves thereby increasing the gelatinization
time, or c) the existence of water diffusion gradients in the grains, which
are caused by structural barriers such as the interface between different cell
types, e.g., cell walls between the subalurone and central endosperm layers
in milled rice (16).
1. Two Different Starch Granule Populations
Rice has compound granules within each individual grain, which are rela-
tively small compared to other cereal grains. If two different populations of
starch granules exist in rice, then they should exist in the flour samples.
However, only one endotherm attributed to starch gelatinization was ob-
served in the flour samples (see Fig. 3). Therefore, the presence of granules
that contain starch of distinctly different gelatinization profiles is unlikely.
2. Heat Transfer Effects
Heat transfer effects can be studied by determining the effect of heating
rate on the gelatinization temperatures and gelatinization enthalpy of the
 212 MARSHALL

G 90
0

.....
0: T,IFI
----
::::l

<
.....
0:
D..
----· • • -eTPIFI
:
--- :
~
.....
---- : I T01FI
T01Wtl

600 0.2 0.4 0.6 0.8 1.0 .4

HEATING RATE (°C/minl

Figure 4 Changes in onset (T0 ), peak (Tp), and conclusion (Tc) temperature of
Lemont milled rice (WG) and flour (F) as a function of heating rate. (From Ref. 9.)

starch. The resulting curves are then extrapolated to zero heating rate. At
this value, heat transfer effects are eliminated. The effect of heating rate on
starch gelatinization temperatures was determined for Lemont long-rain
rice (Fig. 4). Onset temperatures were essentially unaffected by heating
rate. There was a small (2°C) difference in To between milled rice and flour
at zero heating rate, with milled rice having the higher value. Decreases in
TP and Tc occurred with decreasing heating rate for both milled rice and
flour. This relationship was linear. At zero heating rate, the TP for milled
rice was still almost 9°C higher than for flour. The Tc for milled rice was
most affected by heating rate, as it showed the greatest rate of change
(steepest slope). At zero heating rate, the Tc for milled rice was 4°C higher
than for flour. Gelatinization enthalpies showed a small but consistent
increase with decreased heating rate (Fig. 5). At zero heating rate, enthal-
pies for milled rice were more than 3 J/g higher than for flour samples.
Extrapolated values for T 0 , T P' Tc, and L\H were larger in milled rice than
flour at zero heating rate (see Figs. 4 and 5). If each of these extrapolated
values had been the same for both milled rice and flour, the the heat transfer
model could completely account for the differences in gelatinization parame-
ters. In our example, the heat transfer model can, at most, only partially
explain the differences in thermal curves between milled rice and flour.
3. Barriers to Water Diffusion
Thus far, the first two models put forth to account for differences in starch
gelatinization between milled rice and milled rice flour have not been
STARCH GELATINIZATION 213

-.,
16.0

----·•~-,·~--------·
~ 1-4.0 WHOLE GRAIN
0
en
m
......... 12.0

;; ----IL-----••r---~::~·----------~
<I 10.0 FLOUR ....

8.0 +-~--.--~---r-~---T-----,.-----.
0.0 0.2 0.-4 0.6 0.8 1.0
HEATING RATE (°C/min)

Figure 5 Changes in enthalpy (AH) values as a function of heating rate for

Lemont milled rice (whole grain) and flour. (From Ref. 9.)

satisfactory. The third model, the existence of barriers to water diffusion,

would appear to be the most likely explanation simply by the process of
elimination. However, evidence needs to be presented to prove that this is
the correct model. In this scenario, the rate of water diffusion within the
grain dictates when starch gelatinization occurs. If this model is correct,
then disruption of diffusion barriers within the grain, such as removal of
successive layers of rice tissue by milling or destruction of the structural
integrity of the rice grains by grinding or pulverization, should cause an
observable reduction in gelatinization temperatures and gelatinization
enthalpies. The next section will describe how starch gelatinization is influ-
enced by alterations to grain structure.

rv. GRAIN STRUCTURE AND STARCH GELATINIZATION

A. The Effect of Cracks and Fissures
The most obvious difference between rice and rice flour is that the struc-
tural integrity of the rice grain has been destroyed when the grains are
reduced to small particle sizes to form a flour. This situation represents an
extreme example of rice grain modification. Actually, one of the most
common modifications of intact grain structure is the formation of surface
cracks and the deeper fissures that can farm in milled rice through im-
proper drying, milling, and exposure to aqueous and nonaqueous media.
Marshall et al. (10) examined starch gelatinization after soaking rice in
214 MARSHALL

water or treating rice with hexane and chloroform-methanol (C-M). Rice

grains were soaked in an excess of distilled water for 30 minutes and then
air dried for 24 hours in order to bring the moisture content of the grains
back to their original value (11-12%). This simple method is known to
cause transverse fissures to form in the grains (17). Milled rice was also
exposed to hexane or C-M, two solvents that remove rice surface lipids
(18). Treatment with both aqueous and nonaqueous media caused exten-
sive cracking and fissuring in representative Lemont rice as seen in scan-
ning electron micrographs (Fig. 6B-D). Note the virtual absence of surface
imperfections in the untreated (control) grain (Fig. 6A).
Thermal curves showing the starch gelatinization profiles of untreated
and treated Lemont rice are shown in Figure 7. The thermal curve for the
control (curve 1) was considerably different than the thermal curves for
soaked (curve 2) and C-M-treated (curve 3) milled rice. Curves 2 and 3 were
similar to each other. In separate studies (10), the thermal profiles for un-
treated Mars, S-201, and Calmochi varieties were found to be different from
the corresponding treated rice within the same variety. As with Lemont, the
thermal profiles for soaked, hexane-, and C-M-treated samples were similar
when compared to each other within the same variety. Curves 2 and 3 are
characterized by a shift in TP and Tc to much lower values and an increase in
the size of the low temperature shoulder compared to the main gelatinization
endotherm. Table 2 gives the thermal parameters for these curves. The data
show T 0 values were unaffected by either soaking or C-M treatment, but a 6-
70C decrease was observed in the TP values of the treated grains when com-
pared to the control. Comparisons between untreated and treated sam-

Table 2 Thermal Parametersa of Lemont Whole Grain Milled Riceb

Soaked in Water or Treated with Chloroform-Methanol (C-M)

Gelatinization temperaturesc eq
Enthalpy, .:\H
Treatment To TP Tc (Jig)
None 71.2 ± 0.2 93.0 ± 0.5 NDct ND
Soaked 71.4 ± 0.1 86.8 ± 0.2 95.2 ± 0.4 13.6 ± 0.8
C-M 70.0 ± 0.6 85.7 ± 0.8 95.9 ± 1.3 15.1 ± 0.7

•Values are the means ± standard errors of the means of duplicate determinations.
hMoisture content of calorimeter samples was 70%.
'To = onset temperature, TP = peak gelatinization temperature, T, = conclusion
temperature.
dNot determined due to absence of concluding baseline.
Source: Ref. 10.
STARCH GELATINIZATION 215

Figure 6 Scanning electron micrographs of Lemont milled rice. (A) Untreated,

(B) hexane-treated, (C) chloroform-methanol treated, (D) soaked in water. (From
Ref. 10.)
216 MARSHALL

~
0
..J
IL.
1-
<
w
J:
0
:::E
a:
w
J:
t-
o
c
z
w

l
TEMPERATURE (•C)

Figure 7 DSC thermal curves of Lemont milled rice: untreated (curve 1), soaked
in water (curve 2), and chloroform-methanol-treated (curve 3). The final moisture
content of all samples was 70% (w/w) and the heating rate of the calorimeter was
1.0°C/min. (From Ref. 10.)

pies could not be made for Tc or dH due to temperature limitations of the

calorimeter. Experiments with Mars, S-201, and Calmochi showed that Tc
and dH exhibited a 9-13°C and up to a 0-3 J/g decrease, respectively, when
treated rice was compared to its untreated control (10).
All measured thermal parameters were similar when water-soaked and
C-M-treated grains were compared to each other (see Table 2). This simi-
larity was also found in Mars, S-201 and Calmochi [10]. Less than 5% ofthe
nonstarch lipid was removed during soaking. Almost 80% of the nonstarch
lipid was removed by C-M treatment (10). Thus, the results displayed in
Figure 7 suggest that the major determinant affecting starch gelatinization
is the formation of cracks and fissures in the rice. The removal of nonstach
lipid appears to play only a minor role in influencing starch gelatinization.

B. Effect of Milling
Milling of brown rice is another method of modifying grain integrity by re-
moving successive layers of grain tissue. Milling does not destroy grain integ-
STARCH GELATINIZATION 217

rity as would grinding or pulverizing the rice. Milling could modify starch
gelatinization by removing potential barriers, such as the bran layer, to water
transport phenomena that are essential to the gelatinization process.
Champagne et al. (11) investigated the effect of bran removal on starch
gelatinization in Lemont brown rice. They used abrasion milling to remove
successive layers of bran and polish in 1-2% degree of milling increments
in order to obtain rice milled at 9.1% degree of milling. Table 3 presents
the results of the milling study. T0 and TP tended toward lower values as the
degree of milling increased. The greatest decline in gelatinization tempera-
tures occurred after the outer 1.3% of the grain was removed. At this
degree of milling, T 0 and TP had decreased 1.6 and 4.8°C, respectively.
Measurements of Tc and AH in most cases could not be obtained due to the
inability of the calorimeter to measure conclusion temperatures beyond
110°C.
Marshall (13) expanded the milling study of Champagne et al. (11) to
include rice milled to 66% degree of milling. The effect of deep milling on
starch gelatinization temperatures is given in Figure 8. T 0 decreased until
about 10% degree of milling was reached. This degree of milling removed
the bran and polish layers from the grain. Subsequent removal of layers of
starchy endosperm had no further effect on this parameter. T 0 decreased by
3°C over the entire milling range. TP and Tc showed the same general trend
as T 0 • A Tc value could not be obtained for brown rice due to the absence of

Table 3 The Effect of Degree of Milling

on Gelatinization Temperatures of Whole
Grain Ricea
Gelatinization
temperatures eq
%Milled To TP
Unmilled 76.4 ± 0.1 101.5 ± 0.1
1.3 74.8 ± 0.2 96.7 ± 0.2
1.9 74.1 ± 0.3 96.4 ± 0.1
3.1 73.6 ± 0.1 95.1 ± 0.1
5.0 73.6 ± 0.3 94.2 ± 0.0
7.2 72.8 ± 0.3 94.0 ± 0.2
9.1 72.5 ± 0.2 94.1 ± 0.1

•Moisture content of samples was 70%. Values

given are means ± standard error of the mean
of duplicate determinations.
Source: adapted from Ref. 11.
 218 MARSHALL

110

........
u 105
........,
0

w
a:: 1 00
::::>
~
f5
D..
95
:::!:
w
I- 90
z
0
~ 85
N
z
80
~w
" 75

70 L-~--~~--L-~--~~--~~--~~~~~~~ •
0 10 20 30 40 50 60 70
DEGREE OF MILLING (~)

Figure 8 The effect of degree of milling on the starch gelatinization temperatures

for Lemont unmilled (brown) and milled rice. Onset (T peak (Tp), and conclu-
0 ),

sion (Tc) temperatures were obtained from thermal curves at the degree of milling
shown. (From Ref. 13.)

a concluding baseline on the thermal curve. Therefore, the Tc for brown

rice was >l10°C. There was a steady decline in TP and Tc until about 20%
degree of milling was reached. Further milling had little effect on either
value. At 20% degree of milling, TP had decreased about 10-l1°C com-
pared to brown rice.
The absence of a concluding baseline precluded measurement of LlH for
brown rice. However, the results in Figure 9 show that for milled rice, LlH
increased until the degree of milling reached 20-25%. The degree of mill-
ing appeared to have no apparent further effect on LlH, although the
variability among data points was high. LlH was presented as J/g starch
because corrections were made for the starch content of the residual grain
as the outer layers were milled away.

C. Effect of Particle Size

We have shown that the introduction of cracks and fissures compromises
rice grain integrity and that both cracks/fissures and milling affect starch
STARCH GELATINIZATION 219

19

•
• • •
18

.........

- • •
.s::.

•
0
.....
c 17
rn
0'1
....,
'-
....._,
16
I
<I

15

14L-~~--~~--~-L--~~~--~~--L-~~

0 10 20 30 40 50 60 70
DEGREE OF MILLING (~)

Figure 9 The effect of degree of milling on gelatinization enthalpy (~H) for

Lemont milled rice. ~H was calculated from endotherms at the degree of milling
shown. (From Ref. 13.)

gelatinization. However, in both cases grain structure remained largely

intact. In this section, we examine the effect of incrementally disrupting the
grain structure of milled rice by sectioning or pulverizing to smaller particle
size.
Figure 10 depicts the thermal curves generated by sectioning Lemont
milled rice with a scalpel into pieces of roughly equivalent size. As the
sections became shorter, TP and Tc decreased and the low temperature
endotherm became more prominent. As particle size was further reduced
by pulverizing milled rice in a mortar with a pestle and the particles were
partitioned into size ranges, the gelatinization endotherms underwent con-
siderable change (Fig. 11). The most obvious change was endotherm
shape. Curves A through F show a progressive increase in the size of the
low temperature shoulder and a progressive decrease and eventual disap-
pearance (Curve E) of the major starch gelatinization endotherm. A quali-
tative examination of Figures 10 and 11 reveal an inverse relationship
between particle size and the contribution of the low temperature shoulder
to the entire gelatinization event. That is, a decrease in particle size results
220 MARSHALL

3':
_.
0
.....
~J:
(.)
::::1!:
0:::
w
J:
1-
0
0
z
w

50 60 70 80 90 100 110
TEMPERATURE {°C)

Figure 10 DSC thermal curves for Lemont milled rice and milled rice sections.
(A) Unsectioned (whole) grains, (B) one-half grains, (C) one-quarter grains, (D)
one-eighth grains. The heating rate of the calorimeter was 1.0°C/min, and the final
moisture content of all samples was 70% (w/w). (From Ref. 13.)

in an increase in the percentage of the gelatinization endotherm, which was

originally the low temperature shoulder.
The effect of particle size on starch gelatinization parameters can be
seen in Tables 4 and 5. For sectioned grains (Table 4), TP and Tc progres-
sively decreased as the number of sections increased and particle size de-
creased. When intact grains were compared to kernels sectioned into eight
parts, a decrease of 4-5°C and 5-6°C was observed for TP and T0 , respec-
tively. There was no consistent trend for T0 and 6-H values. Reducing the
STARCH GELATINIZATION 221

31:
0
...J
LL

!<L.J
::t: c
~
::IE
0:::
L.J
::t:
1-
0 D
0
z
L.J

so so 70 ao go 1 oo 11 o
TEMPERATURE (°C)

Figure 11 DSC thermal curves for Lemont rice grains pulverized to particle sizes
of710-1400 #Lm (A), 500-710 ~£m (B), 355-500 ~£m (C), 250-355 #Lm (D), 180-250
l£m (E) and 125-180 #£m (F). The heating rate of the calorimeter was 1.0°C/min,
and the final moisture content of all samples was 70% (w/w). (From Ref. 13.)

particle size further (see Table 5) caused an additional decrease in TP and Tc

and also a decrease in ~H. TP remained stable at particle sizes above 250
/LID before decreasing 1oc at particle sizes below 250 IJ-ID. At the smallest
particle size range evaluted (53-64/Lm), decreases of 16°C and 20°C for TP
and Tc, respectively, and a decrease of 7.7 Jig for ~H were seen when
compared to gelatinization temperatures for intact milled rice (see Table
4). Particle sizes smaller than 250 /LID yielded no additional changes in TP.
The particle size had to be reduced to less than 125 IJ-ID before no further
222 MARSHALL

Table 4 Thermal Parameters• of Milled Lemont Riceb Sectioned to Pieces of

Equivalent Size

Number Gelatinization temperatures (oqct

of sections Enthalpy, ~H
(length in ~tmc) To TP Tc (J/g)

0(6,770 ± 37) 73.3 ± 0.2 93.4 ± 0.7 103.7 ± 0.4 14.1 ± 1.4
2(3,680 ± 36) 72.8 ± 0.2 92.1 ± 0.0 103.2 ± 0.5 14.8 ± 0.4
4(1840 ± 29) 73.4 ± 0.5 89.9 ± 0.1 102.0 ± 0.0 14.4 ± 0.4
8(1300 ± 25) 72.8 ± 0.4 88.8 ± 0.2 98.2 ± 0.1 15.1 ± 0.1
•Values are means ± standard errors of the means for duplicate determinations.
bMoisture content of calorimeter samples was 70%.
'Mean length ± standard errors of the means of 30 sectioned kernels selected at random.
Section lengths determined by image analysis.
ctT0 , TP, and T, = Onset, peak, and conclusion temperatures, respectively.
Source: Ref. 13.

change in Tc was observed. ~H continued to decline until a particle size

range of 64-90 ~-tm was reached.
Starch gelatinization parameters of samples prepared with a mortar and
pestle were compared with a flour sample prepared in a Udy cyclone mill
(see Table 5). The Udy cyclone mill was used in our laboratory to prepare
the rice flour used to develop the data seen in Figure 3. The flour yielded a
broad range of particle sizes, but the single particle size range that had the
highest percentage of flour particles was 180-250 ~-tm. When the samples,
prepared by two different methods, were compared for this size range,
starch gelatinization parameters were similar. This result indicates that the
method of sample preparation is less important than particle size, at least
for this particle size range and only if starch granule damage is minimal.
At larger particle sizes, the above statement may not be true. Grains
sectioned into eight equivalent parts were about 1300 ~-tm thick (see Table
4) and are within the particle size range (710-1400 ~-tm) of the largest
particles prepared with mortar and pestle (see Table 5). However, their
thermal parameters, except for T0 , were considerably different. There was
a 6-7°C differences in TP and Tc and a 3.3 Jig difference in ~H. In this
example, the amount of exposed, internal grain surface may be important
in defining starch gelatinization. For the sectioned grain, the top and
bottom of a disk-shaped section are internal surfaces and contain clusters
of exposed starch granules. The perimeter of the disk is the milled or
polished surface and shows no morphologically distinct granules. For the
pulverized grain, most of the surface is internal surface with exposed
starch granules.
STARCH GELATINIZATION 223

Table 5 Thermal Parameters• of Milled Lemont Riceb Reduced to Different

Particle Size Ranges
Particle size Gelatinization temperatures eqc
range Enthalpy, AH
(p.m) To TP T, (J/g)
710-1400 73.1 ± 0.1 82.0 ± 1.1 92.0 ± 0.1 11.8 ± 0.5
500-710 73.5 ± 0.1 80.8 ±0.2 90.2 ± 0.1 11.8 ± 0.3
355-500 73.4 ± 0.1 77.8 ± 0.1 89.8 ± 0.1 11.4 ± 0.2
250-355 73.3 ± 0.1 77.7 ± 0.1 89.4 ± 0.1 10.6 ± 0.2
180-250 72.4 ± 0.1 77.2 ± 0.0 85.8 ± 0.0 9.5 ± 0.3
125-180 72.4 ± 0.1 77.2 ± 0.2 85.1 ± 0.1 9.3 ± 0.3
90-125 72.3 ± 0.3 77.0 ± 0.0 83.9 ± 0.1 8.2 ± 0.6
64-90 72.5 ± 0.3 77.2 ± 0.1 84.2 ± 0.1 6.4 ± 0.5
53-64 72.4 ± 0.1 77.2 ± 0.1 83.4 ± 0.0 6.4 ± 0.3
Sample prepared in Udy cyclone mill
53-710d 72.8 ± 0.0 77.4 ± 0.1 86.0 ± 0.2 10.0 ± 0.2
•Values are means ± standard errors of the means for duplicate determinations.
bMoisture content of calorimeter samples was 70%.
c'f0 , TP, and T, = Onset, peak, and conclusion temperatures, respectively.
dSeventy-nine percent of the sample was collected in the particle size range 125-355 p.m,
with 43% in the 180-250 p.m range.
Source: Ref. 13.

D. Discussion
This section has described experimental evidence showing that modifica-
tion of rice grain structure, either by milling, introduction of cracks/
fissures, or pulverization, considerably reduced TP, T,, and AH. The only
gelatinization parameter relatively unaffected by structural modification
was T 0 • T 0 can be slightly altered by successive removal of bran layers (see
Table 3). T 0 values are variety dependent (see Table 1). For a specific
variety they are probably influenced more by the arrangement of amylose
and amylopectin (molecular architecture) within the starch granule than by
modification of the grain. TP, T, and AH were also variety dependent.
However, they were at least equally dependent on grain integrity. These
observations have practical applications. If a rice processor specifies a set
of starch gelatinization parameters, then they can either find the rice vari-
ety that has those parameters or manipulate the physical structure of the
grain for the variety at hand.
The effect of milling on rice starch gelatinization can be explained by
considering the milling process as removing layers of different types of
grain tissue. This eliminates structural barriers that hinder the movement
224 MARSHALL

of water through the grain and create water diffusion gradients. Milling
studies by Champagne et al. (11) and Marshall (13) serve as examples. The
steepest decline in T0 and TP occurred after the inital 1.3% of brown rice
surface was removed (see Table 3). In this outer layer is the pericarp and
seed coat with its associated high wax content (19). Removal of the
pericarp and seed coat had the greatest impact on water penetration to the
starch granules, most likely due to the presence of wax in these layers. A
smaller rate of decline in TP and Tc was observed until approximately 20%
degree of milling was reached (see Fig. 8). In this case, subalurone and
starchy endosperm layers may provide additional resistance to water pene-
tration because of the heterogeneity of the tissue in these layers. The
starchy endosperm remaining after 20% degree of milling apparently pro-
vides a constant resistance to water diffusion as no changes were observed
in starch gelatinization temperatures. This part of the grain is probably
homogeneous in its cellular makeup compared to outer layers (20).
The decrease in LlH at degree of milling values below 20% cannot be
readily explained (see Fig. 9). The amount of heat input was expected to
decline with improved water penetration though the kernel. Milling ini-
tially reduced TP. It also increased the size of the low temperature shoulder
in relation to the main endotherm (13). For milled rice, LlH may be more
closely related to changes in the low temperature shoulder than TP.
The size of the low temperature shoulder does not always appear depen-
dent on the degree of milling. A low temperature shoulder was not ob-
served in intact brown rice but was prevalent in milled rice after the bran
layer was removed (13). Cracked, fissured, or pulverized brown rice will
display a prominent low temperature shoulder during gelatinization (W. E.
Marshall, unpublished observations). In milled rice, further milling pro-
vided a noticeable but small increase in the size of this shoulder (13). Thus,
the appearance and size of the shoulder can be assisted by milling, but large
changes appear to depend on the degree to which the internal surfaces of
the grain are exposed by cracks/fissures or pulverization. Exposure of the
internal surfaces to water could eliminate water diffusion gradients to the
granules closest to the granule/water interface. These granules would then
gelatinize at the lowest possible temperature, creating a small endotherm.
Lund (21) has noted that water diffusion into isolated starch granules is not
a rate-limiting step for gelatinization. The population of granules physically
removed from the interface between water and granule would still be under
the control of diffusion gradients. This population of granules would gelati-
nize at higher temperatures. Thus, an important factor in determining the
size of the low temperature shoulder may be the ratio of exposed internal
surface to total surface. Exposure of increasing amounts of internal surface
relative to total surface, as exemplified in the particle size study described
STARCH GELATINIZATION 225

in this section, would increase the population of starch granules not sub-
jected to diffusion gradients. At a sufficiently small particle size, diffusion
is no longer rate limiting and gelatinization depends only on the gela-
tinization characteristics of the individual granules. Brown rice represents
the other extreme, where starch gelatinization is totally dependent on wa-
ter diffusion gradients controlled by the unique structure of the grain.

V. CONCLUSIONS
We have demonstrated that starch gelatinization in brown or milled rice is
dependent upon grain structural integrity. Gelatinization can be varied
over a wide range of temperatures and a range of enthalpy values by
modifying grain structure. Modification affects water diffusion gradients in
the grain that allow the diffusion process to be the rate-limiting step in
gelatinization. Previous characterizations of rice starch gelatinization that
have appeared in the literature have dealt with rice flour and isolated starch
granules. Under these conditions, diffusion is no longer the rate-limiting
step in gelatinization. Results from the literature could not be used to
accurately model starch gelatinization in the intact rice grain because of the
complex and critical role structure plays in the gelatinization process.
Therefore, our results have opened a new perspective on starch geatlini-
zation in rice that could not have been quantified without in situ analysis.
What kind of practical value do our results have? Since use of a differen-
tial scanning calorimeter simulates the rice cooking process, our quantifica-
tion of gelatinization temperatures and gelatinization enthalpies can have
practical application to the processing of rice. As noted earlier, starch
gelatinization temperatures and cooking times appear to be directly related
(8). Lower gelatinization tempeatures and lower gelatinization enthalpies
would be desirable to processors. Cooking time and the amount of heat
energy required to cook the rice would be reduced. Processors could use
rice milled to different degrees to develop faster cooking products. To
create a rice product with substantially shorter cooking times, the internal
structure of the grains must be exposed. Alexander (22) and Bardet and
Giesse (23) described commercial processes where brown rice was exposed
to hot, dry air to produce fissures in the grain surface. The extensive
fissuring reduced cooking time by allowing rapid water penetration to the
starch granules. The processes described in these patents are based on the
principles discussed in this chapter. One product derived from these pat-
ents is quick cooking, brown rice seen on U.S. grocery shelves. Preceeding
further along these lines, if retention of grain structure is not a require-
ment, cooking times can be substantially reduced by controlling the particle
size of the broken grain.
226 MARSHALL

VI. FUTURE RESEARCH

Future research efforts need to be directed toward a quantitative descrip-
tion of the kinetics of water movement in whole grain rice. The kinetic
studies would supply time, temperature, and variety data needed by proces-
sors to minimize production costs yet obtain the desired end result. In
addition, studies are needed to compare the relationship between starch
gelatinization and cooking times using the whole grain model versus a flour
or starch model. After these questions have been answered, rice processors
will have a better understanding of the interrelationships among rice grain
structure, starch gelatinization, and cooking.
Future research should also be directed toward using the whole grain
model of rice starch gelatinization. It might be applied to quantify starch
gelatinization in other food products that are cooked before being con-
sumed. These products would include legumes, such as beans and peas,
and tuber crops, all of which have starch as their major component.

REFERENCES
1. Maurice, T. J., Slade L., Sirett, R. R., and Page, C. M. (1985). Polysaccharide-
water interactions-thermal behavior of rice starch, Properties ofWater in Foods
in Relation to Quality and Stability (D. Simatos and J. L. Multon, eds.), M.
Nijhoff Publishers, Dordrecht, The Netherlands, p. 211.
2. Biliaderis, C. G., Page, C. M., Maurice, T. J., and Juliano, B. 0. (1986).
Thermal characterization of rice starches: A polymeric approach to phase
transitions of granular starch. J. Agric. Food Chern., 34:6.
3. Chungcharoen, A., and Lund, D. B. (1987). Influence of solutes and water on
rice starch gelatinization. Cereal Chern., 64:240.
4. Chang, S.-M., and Liu, L.-C. (1988). Investigations on the gelatinization of
rice starches with various amylose contents by differential scanning calorim-
etry. Bull. Inst. Chern. Acad. Sin., 35:97.
5. Hamaker, B. R., and Griffin, V. K. (1993). Effect of protein on rice starch
gelatinization. Starch/Starke, in press.
6. Biliaderis, C. G., Page, C. M., and Maurice, T. J. (1986). On the multiple
melting transitions of starch/monoglyceride systems. Food Chern., 22:279.
7. James, C., and McCaskill, D. (1983). Rice in the American diet. Cereal Foods
World, 28:667.
8. Juliano, B. 0., and Perez, C. M. (1983). Major factors affecting cooked
milled rice hardness and cooking time. J. Texture Stud., 14:235.
9. Normand, F. L., and Marshall, W. E. (1989). Differential scanning calorim-
etry of whole grain milled rice and milled rice flour. Cereal Chern., 66:317.
10. Marshall, W. E., Normand, F. L., and Goynes, W. R. (1990). Effects oflipid
and protein removal on starch gelatinization in whole grain milled rice. Cereal
Chern., 67:458.
STARCH GELATINIZATION 227

11. Champagne, E. T., Marshall, W. E., and Goynes, W. R. (1990). Effects of

degree of milling and lipid removal on starch gelatinization in the brown rice
kernel. Cereal Chern., 67:570.
12. Marshall, W. E., and Normand, F. L. (1991). Exothermic transitions in whole
grain milled rice and milled rice flour studied by differential scanning
calorimetry. Cereal Chern., 68:606.
13. Marshall, W. E. (1992). Effect of degree of milling of brown rice and particle
size of milled rice on starch gelatinization. Cereal Chern., 69:632.
14. Juliano, B. 0., Perez, C. M., Barber, S., Blakeney, A. B., Iwasaki, T.,
Shibuya, N., Keneaster, K. K., Tatsumi, K., and Webb, B. D. (1981). Interna-
tional cooperative comparison of instrument methods for cooked rice texture,
J. Texture Stud., 12:17. ·
15. Lund, D. B. (1983). Applications of differential scanning calorimetry in
foods, Physical Properties of Foods (M. Peleg and E. B. Bagley, eds.), AVI
Publishing Co., Inc., Westport, CT, p. 125.
16. Bechtel, D. B., and Pomeranz, Y. (1978). Ultrastructure of the mature
ungerminated rice (Oryza sativa) caryopsis. The starchy endosperm. Am. J.
Bot., 65:684.
17. Desikachar, H. S. R., and Subrahmanyan, V. (1961). The formation of cracks
in rice during wetting and its effect on the cooking characteristics of the cereal.
Cereal Chern., 38:356.
18. Hogan, J. T., and Deobald, H. J. (1961). Note on a method of determining the
degree of milling of whole milled rice. Cereal Chern., 38:291.
19. Bechtel, D. B., and Pomeranz, Y. (1977). Ultrastructure of the mature
ungerminated rice (Oryza sativa) caryopsis. The caryopsis coat and the
aluerone cells. Am J. Bot., 64:966.
20. He, G.-C., and Suzuki, H. (1989). Properties of starches from the outer layer
and central core of the rice endosperm. Nippon Nogeikagaku Kaishi, 63:981.
21. Lund, D. (1984). Influence of time, temperature, moisture, ingredients, and
processing conditions on starch gelatinization. CRC Crit. Rev. Food Sci. Nutr.,
20:249.
22. Alexander, W. P. (1954). Process of preparing quick-cooking rice, U.S. patent
507,242, November 9.
23. Bardet, G. V., and Giesse, R. C. (1961). Processing of brown rice, U.S. patent
2,992,921, July 18.
 11
New Methods and Equipment for Processing
Rice
Robert S. Satake
Satake Corporation, Hiroshima, Japan

I. INTRODUCTION
Since rice first became a staple food product, there has been a steady and
consistent effort to increase and enhance the yield, food value, and quality
aspects of harvested rice as a food for human consumption. These efforts
have led over the years to significant gains in all these areas. For example,
improved breeding has led to sizable harvested yield increases through
more productive strains, improved resistance to plant diseases, optimum
plant stature, and so on. Improvements in growing and harvesting technol-
ogy have resulted in better yields and quality by better insect control,
efficient use of fertilizers, better harvesting techniques, and other new and
improved crop-management practices. Safe storage of the harvested grain
has also significantly added to the availability of an abundant and high
quality rice product. In a similar fashion, the striving for higher yields and
better quality has not gone unrecognized by rice processors.
This chapter will discuss recent and ongoing developments in methods and
equipment for processing rice to improve yield, efficiency, and quality. This
chapter will not seek to provide basic information regarding rice processing
itself, except where deemed necessary for better understanding, since the
subject has been adequately covered in several other publications {1-6).
Instead, the focus will be on those new processing technologies that have
only recently been developed and introduced into the world's rice industries.

229
230 SATAKE
Topics will include new equipment for the actual milling (i.e., de-
branning) of rice, machinery for improving quality factors, and processes for
enhancing yields and efficiencies. Since many of these new machines and
processes have overlapping impacts on yields, processing efficiencies, and
quality enhancement, the discussion will be arranged by machine/process
rather than by benefit area. With this in mind, rice-milling and -polishing
equipment such as the new high-yield/high-quality Vertical Whitening Sys-
tem and the Water-Mist Polisher will be discussed first. Next, the emerging
technique of adjusting the moisture levels of rice for quality and economic
advantage through the use of "rice conditioners" will be addressed. Finally, a
newly developed process for "prewashed rice" will be examined.

IT. RICE-WHITENING EQUIPMENT

A. Introduction
One of the primary steps in the preparation of rice grains for human
consumption is the removal of the pericarp, or bran layer. The term "mill-
ing" is applied to this process, although it may be more accurate to call this
process "whitening" in order to distinguish it from the totality of processing
steps (i.e., cleaning, hulling, whitening, grading, etc.), which are also com-
monly termed "rice milling."
Rice whitening can be conceptualized as the partial destruction of the
grain surface and its separation from the internal portion of the grain body.
The rice grain structure consists of a hard core of starch (endosperm) and
soft bran layers covering the hard core, so that milling action is largely a
matter of material strength. The objective is to exert sufficient work against
the outer bran layer to overcome its resistance to removal and yet avoid
excessive work and pressure, which may crush or otherwise damage the
integrity of the harder endosperm. Several physical mechanisms are in-
volved in mechanical whitening action, as illustrated in Figure 1. These
include friction (tearing), cutting, grinding, and impact, all of which may
be present in varying degrees in an actual whitening system. Modern rice-
whitening theory recognizes that this rice-whitening action can be classified
into two categories when discussing whitening equipment, namely at high
pressure with low peripheral speed (tearing and cutting action) and at low
pressure with high peripheral speed (grinding and impact action).

B. Milling Action at High Pressure with Low Peripheral Speed

This action removes the bran layer of the grain by effective friction force
(tearing) and cutting force of grains being rubbed vigorously against other
grains and against the metal surfaces of the milling unit. The rice grains are
NEW PROCESSING METHODS AND EQUIPMENT 231

RN

Model of Tearing Action Model of Grinding Action

Figure 1 Milling action on rice grains. Mechanical whitening action involves

several physical mechanisms including friction (tearing action), and cutting and
grinding (grinding action). (Courtesy Satake Corporation.)

enclosed in a perforated metal enclosure or cylinder into which is inserted a

rotating eccentric milling rotor that agitates the grains and provides the
pressure and movement to bring the friction forces into play.
Since friction is seen to be the major contributor to the whitening action,
this type of milling is termed "friction type." A typical friction-type whiten-
ing machine is shown in Figure 2. To be effective, the friction-type whiten-
ing unit requires a peripheral speed of the milling rotor of less than 600 m/
min and an average milling pressure of over 100 gf/cm2 (10,000 N/m2).
Figure 3 shows that average milling pressure falls within 200-300 gf/cm2
for two typical friction-type whitening machines (models CP4A and CP6A)
operating at various horsepower loadings. Peripheral speed is between 200
and 300 m/min in the zone of average milling pressure.

C. Milling Action at Low Pressure with High Peripheral Speed

This action removes the bran layer of the rice grains by a grinding action and
an impact force when the grains come into contact with a rotating abrasive
medium mounted on a central shaft turning in a perforated metal enclosure
or cylinder. Because it is abrasion that in this case provides the medium for
milling, this type of milling is termed "abrasive type." A typical high-speed,
abrasive-type whitening machine is shown in Figure 4. Abrasive milling
requires a peripheral speed of over 600 m/min and an average milling pres-
sure of less than 50 gf/cm2 ( 5000 N/m2). Figure 5 shows that for typical models
of abrasive-type whitening machines (models CP2A, CP4A, and CP6A), the
abrasive rotor must operate above 600 rpm to achieve the minimum periph-
eral speed requirement of 600 m/min. Average milling pressure can be seen
to be inversely proportional to the RPM and for two models (CP4A and
232 SATAKE

Feed hopper

Perforated steel plate

Figure 2 Typical low-speed, friction-type rice-whitening machine. Whitening ac-

tion is a result of friction forces applied to the rice grains in the friction-type
whitener. (Courtesy Satake Corporation.)

400
'
'' CP6A

Peripheral
speed
(m/min.) 300

~O~~~~LLLL£L£L~~~~~~~~~~~LL
'', ' • , , (CP6A 50ff'
',, ' CP4A 25ff'
'•,,, ····-(CP6A 40ff'
Average milling ' , '• CP4A 20ff'
pressure (gf I cm2 ) 100 '··---cCP6A 30ff'
C P4A ISIP

~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Revolution <R P Ml

Figure 3 Peripheral speed and average milling pressure in friction-type whiten-

ers. Average milling pressure falls with 200-300 gf/cm2 for two typical friction-type
whitening machines (CP4A and CP6A) at various horsepower loadings. (Courtesy
Satake Corporation.)
NEW PROCESSING METHODS AND EQUIPMENT 233

Feed hopper

Figure 4 High-speed, abrasive-type rice-whitening machine. Grinding action

against an abrasive medium provides the whitening action in abrasive-type rice
whiteners. (Courtesy Satake Corporation.)

BOO 80
CP4A
CP6A
700 70

600

500

400 40

300 30
Peripheral
speed
200 '' <: ··., Average milling pressure
Average milling
20 pressure (gf I cm2)
(rnlmin.)
',<:::(
100 10
' ... ,>'CP4A 20W
',c P6A 301-P

~ a ~ ~ - ~ ~ - - - ~
Revolution {RPM)

Figure 5 Peripheral speed and average milling pressure in abrasive-type whiten-

ers. For typical abrasive-type whiteners (models CP2A, CP4A, and CP6A) at vari-
ous horsepower ratings, peripheral speed must exceed 600 m/min. Average milling
pressure will be less than 50 gf/cm2 . (Courtesy Satake Corporation.)
234 SATAKE

CP6A) at different horsepower ratings. And the average milling pressure

remains well below the maximum average milling pressure of 50 gf/cm2 •

D. Historical Development of Whitening Machines

In order to better understand later developments, it may be useful to
briefly review the origins and developments of modern rice-whitening ma-
chines. Although there had been some early crude mechanical devices for
rice whitening, the first modern rice-milling machine can be said to have
been an abrasive-type rice-whitening device (Fig. 6a) invented by the Doug-
las and Grant Company of Kirkcaldy, Scotland, about 1860. Specifically,
the Douglas and Grant machine was first developed for indica rice grown in
the British colony of Burma (now Myanmar). It became the most widely
used whitening machine in the tropical rice zones prior to World War II. It
incorporated an inverted "cone-shaped" abrasive roll constructed of natu-
ral emery, rotating inside a perforated steel plate of a similar shape. This
was the first of the vertical abrasive rice-whitening machines and represents
the origin of the high-speed type rice whitener where the milling action is
provided primarily by abrasive (grinding) action.

Abrasive roll
Material feeding hopper
(a) (b)

Milling chamber
Perforated steel plate
""""'~,- Conveyor

Outlet
Perforated steel plate

Figure 6 Rice-milling machine by Douglas & Grant (a) and Engelberg-type rice-
milling machine (b). The Douglas & Grant machine utilized a cone of natural
emery abrasive rotating in a perforated steel plate screen to remove bran, and is
considered to be the first modern high-speed abrasive-type rice whitener. The
Engelberg machine removed bran through friction provided by a horizontally ar-
ranged iron roll rotating within a chamber fitted with a perforated steel plate. It is
considered to be the first modern low-speed friction-type rice whitener. (Courtesy
Satake Corporation.)
NEW PROCESSING METHODS AND EQUIPMENT 235

Some years later in 1888, the first machine representative of the low-
speed type of rice whitener (Fig. 6b) was introduced by the Engelberg
company of Syracuse, New York. This was a horizontal unit incorporating
an iron roll with a screw and stirring bar portion revolving within a milling
cylinder, part of which was a perforated steel plate. The Engelberg "huller-
mills" were utilized to not only whiten the rice but also to remove the hull,
with each step usually being performed in a separate pass through the
machine. The milling action in the Engelberg came from friction and cut-
ting action, and therefore this machine can be said to be the origin of the
low-speed rice whiteners.
During the intervening years, different rice-whitening machines have
appeared, varying in form and function but all falling into either the high-
speed abrasive-type or low-speed friction-type classifications. While all
these machines could succeed to a greater or lesser extent, it could be seen
that each type had its own strengths and weaknesses. Specifically, the low-
speed friction-type whiteners had the advantages of a better polished,
smoother grain surface than could be provided by the high-speed abrasive
types. On the other hand, the tremendous pressures needed in the low-
speed type to overcome the resistance and natural waxiness of the brown
rice bran layer create significant stresses on the grains and lead to the
creation of broken rice. The high-speed abrasive type was much better for
preserving milling yields, but there were drawbacks in that the surface of
the milled rice was very rough and unattractive. It seemed that neither type
could achieve perfect results in and of itself.
It was only with the development of the combined high- and low-speed
milling system in the 1960s that the goals of modern rice-whitening technol-
ogy, namely high milling efficiency together with high whole kernel yield,
could be practically attained. This was accomplished by placing together in
one system initial high-speed abrasive-type whiteners with subsequent low-
speed friction-type whiteners. Tests performed in Japan confirmed that the
initial abrasive action of a combined system abrades the rice grain surface
and thereby effectively raises the coefficient of friction of the brown rice
surface. With an increased friction coefficient, less milling pressure is
needed in the subsequent friction action, and less milling pressure results in
a reduction in the generation of broken rice. As well as a whole grain yield
increase, there is also a benefit in finishing the whitening process with low-
speed friction action so that the rice grain surface is smooth and polished.
Moreover, because milling with friction almost ceases when the surface on
the hard endosperm (aleurone layer) is reached, overmilling is thereby
prevented.
Figure 7 shows the characteristics of a combined high- and low-speed
whitening system. Note that at 0% milling by abrasion (bottom scale),
236 SATAKE

Milling
yield
('l.)

910

905

10 20 30 40

Milling Ratio in Abrasive Rice Milling Machine ('l.)

Figure 7 The effect of combined high- and low-speed types. For this test on
japonica rice, milling yield is maximum when high-speed abrasive milling accounts
for 20% of total milling. (Courtesy Satake Corporation.)

which of course indicates 100% milling by friction, generation of broken

grains is highest. In this particular data set, which was derived from tests on
japonica rice in Japan, the optimum combination of abrasive and friction
milling is found when the abrasive portion is at about 20% of the total
milling load. At this point, milling yield is highest, as shown on the graph.
In general, the high-speed abrasive portion of milling should be limited so
as to avoid milling into the endosperm yet be sufficient to substantially
increase the coefficient of friction on the surface of the rice grains.
The data in Table 1 and the corresponding figure (Fig. 8) also show the
results of milling tests comparing the effect of milling first with abrasion
versus milling only with friction at low, medium, and high load level (repre-
sented by 0.315, 0.333, and 0.366 hp load). Again this shows that in each
case efficiency has increased and that broken rice generation and grain
temperature have decreased due to the initial treatment of the rice grains
with the high-speed abrasive action.

E. New Vertical Whitening System

Truly modern rice-whitening systems are combined high-speed abrasive
and low-speed friction-type configurations usually arranged in horizontal-
NEW PROCESSING METHODS AND EQUIPMENT 237

Table 1 The Effects of Milling Pressure in Combined Milling Systems

Milling Power Temp. Broken White. Moist.
Test yield req. Flow Efficiency rise rice deg. loss
no. (%) (hp) (kg/hr) (kg/hr hp) eq (%) (%) (%)
1 92.4 0.315 38.3 121.6 2.6 3.8 32.2 0.2
2 93.2 0.315 36.6 116.2 5.1 10.3 30.4 0.3
3 91.2 0.333 37.6 112.9 4.0 11.2 33.4 0.2
4 91.6 0.333 36.4 109.3 8.1 16.6 33.5 0.2
5 89.5 0.366 36.5 99.7 3.2 14.7 35.2 0.2
6 90.7 0.366 34.2 93.4 8.2 16.2 34.7 0.3
Source: Courtesy Satake Corporation.

IZZI Combined £2Zl Combined 172'.3 Combined

20 20
CJ Low-speed CJ Low-speed D Low-speed
150
0.315 hp load 0.333 hp load 0.366 hp load

Figure 8 Milling pressure effects in combined milling systems. Test results show
that efficiency increases and broken rice generation and grain temperature decrease
when milling with a combined high-speed abrasive and low-speed friction milling sys-
tem compared with systems using low-speed friction only. (Courtesy Satake 1990.)
238 SATAKE

type whitening machines. "Horizontal" here designates the juxtaposition

of the machine's main shaft as well as the direction of the movement or
flow of rice. Both are in a horizontal orientation. Figures 2 and 4 are
examples of so-called horizontally aligned whitening machines. As efficient
as these horizontal whitening machines have become after many years of
refinement, there are nonetheless several technical drawbacks that prevent
the horizontal milling machines from reaching optimum performance. One
factor has been that density and pressure within the milling chamber
tended to vary, especially from the low or bottom side of the milling cham-
ber to the upper or top side. The main cause of this was simply that gravity,
and in most units the suction air used to pull the loosened bran away from
the milling chamber, pulled the rice to the bottom of the milling chamber.
Thus, it was subjected to the weight of the rice above, causing higher
pressure and density at the milling chamber's bottom portion. This dispar-
ity of pressure and density created conditions where efficiency was less than
the theoretical optimum due to the fact that the upper portion of the
milling chamber was underutilized. Thus, uneven pressure created condi-
tions for increases in broken rice generation. Significant improvements to
the horizontal whiteners over the years have minimized these shortcom-
ings, but it was only recently that most of these drawbacks were overcome.
This was accomplished with the development of the vertical-type whiten-
ing machines such as those shown in Figures 9 and 10. These units and their
predecessors are of two types, namely high-speed abrasive and low-speed
friction, and are meant to be utilized in a combined abrasive and friction
system. The hallmark feature of these new machines is that the shaft and
milling chamber are vertically oriented and the flow of rice is in a vertical
direction, from the bottom of the unit toward the top. A spiral shaft at the
bottom of the chamber forces the rice grains to flow upward against the
weight of the rice above. The interaction of pressure and weight through all
layers of rice within the chamber achieves maximum utilization of the milling
surfaces, thus enhancing efficiency as well as preventing uneven pressures
and the excess pressure that causes broken rice. In this way, the process
uniformly generates a safer maximum pressure on each rice grain, which
results in improved milling efficiency and rice-whitening yield with reduced
generation of broken rice. This then corrects the problem of density and
pressure differences found in earlier horizontal whitening machines.

F. Construction of Vertical Whiteners

The abrasive-type vertical unit utilizes a two-stage whitening chamber that
provides a central spiral screw, in addition to the one on the bottom of the
chamber, to propel the rice upward for a second whitening process. This
NEW PROCESSING METHODS AND EQUIPMENT 239

Resistance ring Abrasive roller

Feeding screw

Outlet

Feed roller
Clean-out outlet

Figure 9 Cross-section of bottom-up abrasive-type vertical rice-milling machine

(Sa take). Rice entering the inlet is augured horizontally by the feeding screw where
it comes into contact with the feed roller and is propelled upward into a milling
chamber consisting of abrasive rollers rotating within a perforated metal screen
fitted with a resistance ring to regulate internal flow. Milled rice is forced out the
top of the machine and is discharged from the outlet. A clean-out outlet is fitted to
remove any residual grain remaining in the milling chamber after milling. (Courtesy
Sa take Corporation.)

enables high flow rates through the unit, nearly double that of conventional
horizontal units, without the uneven pressure and density problems associ-
ated with horizontal types. This ability to achieve high flow rates and still
maintain proper pressure and density is one of the main reasons the vertical
type machines were developed.
The vertical orientation allows for another significant improvement for
the vertical low-speed friction whitener. Previously, almost all friction-type
whitening machines had utilized milling rotors that were almost the same
diameter as the milling cylinder and which utilized ribbed protrusions run-
ning approximately parallel to the main shaft (Fig. 11). Unfortunately, this
arrangement created high rice density at the front of the ribbed protrusions
 / Milling roller

Screen

Feeding screw

Outlet

Clean-out outlet--- Feed roller

Figure 10 Cross-section of bottom-up friction-type vertical rice-milling machine

(Satake). Rice entering the inlet is augured horizontally by the feeding screw where
it comes into contact with the feed roller and is propelled upward into a milling
chamber consisting of a steel milling roller rotating within a perforated metal
screen. Milled rice is forced out of the top of the machine and is discharged from
the outlet. A clean-out outlet is fitted to remove any residual grain remaining in the
milling chamber after milling. (Courtesy Satake Corporation.)

--
Horizontal Horizontal Horizontal Vertical
Engelberg type friction type friction type friction type

Figure 11 Low-speed milling rotor cross sections. The diameter of the new
vertical-type milling rotor is larger at the rear of the protrusion than in front of the
protrusion, which is roughly opposite the configuration of earlier horizontal milling
rotors. This shape greatly increases efficient action around the milling rotor and
provides more consistent rice grain agitation. (Courtesy Satake 1990.)

240
NEW PROCESSING METHODS AND EQUIPMENT 241

and also brought about lower and uneven densities at the rear of the
protrusions. In the new vertical-type milling rotor, the milling rotor is
configured differently so that the diameter at the rear of the protrusion is
larger than that in front of the protrusion, which is roughly opposite the
configuration of earlier horizontal milling rotors. This shape greatly in-
creases efficient action around the milling rotor and provides more consis-
tent rice grain agitation. This new configuration allows, for the first time,
high rice germ removal together with high polishing, high yield recovery,
and a sharp decrease in the generation of broken tipped rice. Before the
development of this new type of vertical friction whitener, such accomplish-
ments, particularly high germ removal and high yields, were thought to be
mutually exclusive goals that could not be obtained simultaneously. Inter-
estingly, this benefit is unique to the vertical-type low-speed machines. The
same type of milling rotor and chamber, if applied to a horizontal type
friction unit, would have the opposite result. Geometry and forces in the
vertical-type whitener allow this unique action to be realized.

G. Vertical Milling Systems

By combining the vertical high-speed abrasive together with the vertical,

low-speed friction units, an improved system can be formed for almost any
type of rice. For the japonica short- and medium-grain rices, high germ
removal and high polishing can be achieved without sacrificing whole grain
yield. For indica long grains, this vertical approach can be of significant
value for reducing the amount of broken kernels particularly when milling
brittle rices or rices of low quality.
Another advantage of the combined vertical system is that a water-
polishing feature can be added to the final pass through the vertical
friction-type whitener. This greatly enhances the final bran removal process
and helps clean the milled rice grains so that very little or no residual bran
adheres to the discharging grains. The mist application also serves to help
cool the rice. A more detailed discussion of the benefits of water mist
polishing is undertaken in the following section.
These vertical milling units allow for two additional benefits related to
installation. First, because the units are vertically oriented, the floor space
required for installation is minimized and is usually less than the floor space
required by conventional horizontal whiteners. This makes them useful
where mills wish to upgrade to higher efficiences and capacities within
limited floor space. The second installation benefit of the vertical whiteners
involves a reduction of conveying equipment. Because the units are
bottom-up flow, units placed adjacent to one another can flow automati-
cally from one machine to the next without any intermediary conveyors or
242 SATAKE

II II II

Figure 12 Typical four-break vertical rice-milling machines consisting of two units

of abrasive type and two units of friction type (Satake). Rice entering the first
abrasive unit is milled and discharged out of the top of the unit where it is conveyed by
a diagonal chute into the inlet of the adjoining machine. This rice continues in the
same manner to the subsequent two friction-type units. Rice is finally discharged
from the system from the fourth-break friction unit. (Courtesy Satake Corporation.)

elevating equipment. In a four-break system, for example, there is only one

input and only one discharge, thereby decreasing the complexity and ex-
pense of installation. Figure 12 shows how the units can be joined in a
multibreak installation.
In the context of processing equipment for rice, this new type of vertical
milling technology can be seen as a generational transition to a new plateau
of more efficient, higher yielding, higher quality rice-whitening equipment.

ill. WATER MIST POLISHING

A. Cleaning and Polishing Effect
As mentioned in the introduction, one of the ongoing goals in rice process-
ing is quality enhancements of the finished rice, which means a rice that not
only is well milled and free of contaminants but has an attractive appear-
ance or "eye appeal." Although an inferior appearance may not detract
from the nutritional and functional value of a rice, a rice consumer will be
more pleased and satisfied if in some way the raw product "looks deli-
cious." Commercially, this can be of particular value in certain consumer
markets where "shelf appeal" is a very real competitive factor.
The water mist polisher was developed as a processing step to enhance
the finish or appearance of milled rice. Although this process was invented
NEW PROCESSING METHODS AND EQUIPMENT 243

in 1975, it is just now emerging as a processing alternative in the rice

industry in the United States. From that standpoint, it is a relatively new
technology. As the name suggests, this device utilizes a fine-mist spray of
water to remove final traces of dust, bran, and rice polish from the surfaces
of milled rice grains. Moreover, bran and polish hidden deep inside the
exterior grooves of rice kernels can be stripped away. As well as serving
this cleaning function, the water mist and the subsequent rubbing action of
the machine impart a high polish or luster to the milled rice, which substan-
tially alters its appearance. From a dull, rather flat and chalky appearance,
the rice grains can be transformed into shiny, almost translucent kernels,
which have good positive visual impact. This shiny polished appearance is
even visible in photographs (Fig. 13).

B. Construction of Water Polishing Machines

The actual apparatus for water polishing is very similar to a horizontal, low-
speed friction-type rice whitener. The main differences are that the milling
action is much less severe, as suited to its polishing function rather than a
whitening one, and that the actual milling chamber is elongated to almost
twice the length of the friction units used for milling. In this way, the
relatively lower millling pressure of the water mist polisher is compensated
for by a longer retention time of product inside the milling chamber. This
arrangement is ideal for producing a polishing effect at low pressure, much
in the same way as a rock polisher, without the generation of broken grains.
Water and compressed air tubing are carried to the back of the unit
through the hollow main shaft. The tubing is connected to an atomizing
device just inside the first portion of the milling chamber. A plug in the
hollow shaft prevents the water mist from being applied except for the first
few inches of the milling chamber. This brief exposure, however, is suffi-
cient to wet the grains in preparation for cleaning and polishing. After
wetting, the moistened grains are agitated and scrubbed by the milling
action of the central friction-type rotor. During this phase, residual bran
and polish interact with the moisture in such a way as to be more easily
removed, particularly from the grooves in the rice as well as from the
kernel surfaces. A bran vacuum air line attached to the bottom of the
milling chamber pulls air through the chamber and removes any loosened
material and prevents it from being redeposited on the product. The final
stage of the process takes place at the discharge end of the unit where
ambient air is drawn into the unit by the force of the bran vacuum system
and serves to dry the rice thoroughly before discharge. Thus, the polished
rice product becomes easy to handle and safe to store without any apprecia-
ble increase in moisture content.
244 SATAKE

A
Figure 13 Ordinary milled rice (A) and water-polished milled rice (B). Water
polishing cleans the final traces of dust, bran, and rice polish from the surfaces of
rice grains and at the same time imparts a high polish to the milled rice by rubbing
the grains together vigorously. The result is shiny, almost translucent rice grains
whose improved performance is evident even in photographs. (Courtesy Satake
Corporation.)
NEW PROCESSING METHODS AND EQUIPMENT 245

The degree of cleaning and polishing depends mainly on product flow

rate and amount of water introduced. For truly brilliant, highly polished
rice, multiple passes through the machine are necessary. The water addi-
tion rate, which will be discussed in more detail in a later section, is typi-
cally about 0.3-0.4% by weight. Insufficient water addition rates will
246 SATAKE

lessen the polishing effect. Exceeding the optimum will not lead to further
polishing since the rice surfaces become overmoistened and this interferes
with polishing.

C. Other Benefits
Although the water mist polisher was developed for precisely the process
its name implies, its utilization has yielded other interesting benefits. Fore-
most among these is the discovery that moisture addition softens the sur-
face of partially milled rice kernels so that it is easier to remove the bran
layer for the production of milled rice. Because water addition facilitates
bran removal, lower milling chamber pressures can be utilized in the whit-
ening process, and lower pressure leads to less broken rice and higher
yields. Furthermore, the water addition cools the rice during the whitening
process and also helps reduce grain moisture loss.
Figure 14 shows a comparison of milling results from a 4-break (abrasion-
friction-friction-friction) system with a 5-break (abrasion-friction-friction-
friction-water polisher) system. The degree of milling was equivalent in both
systems. The water polisher system had better bran removal, slightly higher
total milling yield, lower rice temperature increases, lower power consump-
tion, and less than half the amount of broken rice.
Figure 15 shows the effect of water addition rate on the performance of
the water whitening system. As can be seen in Figure 15, machine motor
load, broken rice generation, grain temperature, and moisture loss are at
their lowest at the 0.3-0.4% addition level.
Because of these significant benefits, the water mist polisher is starting
to be used as an integral part of the rice-whitening system rather than
simply a polishing step after completion of the whitening process. As
mentioned in the above section on the vertical milling units, a similar
water mist function is incorporated into the final unit of the vertical whiten-
ing system unit precisely to take advantage of these subsidiary benefits as
well as to provide a product with a better appearance. Hopefully, these
milling benefits will come to be recognized by the rice industries at large.
Rice mills with conventional equipment will begin to use water mist polish-
ers in their whitening process in order to take advantage of these milling
benefits.

D. Enhancement of Storage Stability

Evidence suggests that treatment of rice by the water mist polisher enhances
the storage stability of whitened rice. Residual bran on the surface of milled
rice kernels can quickly oxidize and create free fatty acids, which impart a
NEW PROCESSING METHODS AND EQUIPMENT 247

100% 92% 90%

Ventilating, humidifying
Abr---+ Fri---+ Fri---+ Fri---+ Hu·Fri
friction action
Ventilating friction Abr---+ Fri ---+ Fri ---+ Fri
Brown rice Milled rice

Bran White. Milling Temp. Broken Moist. Effi-

rem. de g. yield nse nee loss ciency
(%) (%) (OC) (%) (%) (kg/h ·hp)
Vent. hum.
friction
good 40.0 91.1 14.5 2.0 0.2 60.0
Ventilat.
friction
fair 40.0 90.5 16.5 4.2 0.4 62.0

100 ~Ventilating humidifying friction typ e

10 20 1.0 100 100
c:::J Ventilating friction type

5 15 0.5 50 50 50

E
v
5:
·;:
~
.
(i
7
"'"'liD
~
:J
~ .......
L;
~
~ "",., Oi
~
c. "'
"' "'
"0 ~
.,.,
"0

.g"' E
.2
"' ~
"'"'c:
u
~ ~ c:
c:
t; ;g"' ~""
""'"'
c: ~
·;;; ·c; :E
w ~ v:
0
cD <; ::;: ;:
,.,
"'"' "'
"' u
c:
"'"'
"'
c: "'
.2
"'<;!
-oo
~ 00
"' w
~~
c:
~0
5

Figure 14 Characteristics of water polishing machines. Rice-whitening systems

utilizing water polishing (ventilating, humidifying friction action) can achieve better
bran removal, slightly higher total milling yield, lower rice temperature increase,
lower power consumption, and less than half the amount of broken rice of systems
not utilizing water polishing. (Courtesy Satake 1990.)
248 SATAKE

/x
X
/Grain
temperature
110 26 900
/X
)(

Load

~
::I
100 25 8 3 tl
.... 'iS
::::;:
E ~
.:::::_
-0.1
Moisture loss

~ "'
E
2 ~
lil
~ ~
Q)
~ ~
Q)
c. c. Q) Q)

E E E <.)
·;:::
~
::I Weight-to-volume
~ 0
c >
-o
.s
I
'iii
"'
0
_J (5 I
EbD
'(ij
3:: -0.3
0 0.1 0.2 0.3 0.4 0.5 0.6 ):!. 7 0.8
Water addition rate (%}

Figure 15 Effects of added water quantity for water polishing. Water addition
rates of about 0.3-0.4% produce the best results in terms of motor load, broken rice
generation, grain temperature, and moisture loss. (Courtesy Satake Corporation.)

rancid smell and taste to rice. Removal of the residual bran with the water
polisher can therefore extend the safe storage period for milled rice. In some
instances, the water mist polisher has been used to reclean and polish previ-
ously whitened rice that had deteriorated during long storage or transport pe-
riods. The water mist polishers can restore a fresh taste and smell to the rice.
There is also data, shown in Figure 16, that suggests that water polished
rice ("clean rice") is more resistant to mold growth. Tests comparing water
polished and regular rices show that the water polished rice resisted mold
growth 7-10 days longer than regular rice. This phenomenon may be due
to the presence of residual bran remaining on the surface and in the
grooves of ordinary rice where it presents an ideal environment for the
promotion of mold growth.
 :z
tl1

-- -- ....
~
I
g
loci
Mold Mold 1/4 Mold 1/4
Clean rice

- --
13% M.C. tl1
Mold Mold 2/3 M11d r:IJ
r:IJ
Ordinary rice
:z
~

--
13%M.C.
n
Mold
--
Mold 1/2
...
ML 3/4 a:tl1

.
I
Clean rice

-- -
I
;!
---
14% M.C.
Mold Mold (15%) Mold 3/4 M11d
Ordinary rice
I
0
t:l
14%M.C. r:IJ

lean rice - -
Mold Mold (10%)

-
.Mold 3/5
...
M14/5 :z>
t:l

-- - -
tl1

--
15% M.C.
Mold Mold (15%) Mold (40%) Mold Mlld. ,0
Ordinary rice e
15% M.C.
-- -- -- - ----
- j __ ~ =a
a:tl1
0 10 20 30 40 50 60 Day
:z
,..,
Humidity: 98%
Akitsuho rice grown in Hiroshima Prefecture, 1979
Temperature: 28"C

Figure 16 Mold growth comparison for ordinary and water-polished rice. Tests indicate that water-polished rice
("clean rice") resists mold growth 7-10 days longer than ordinary rice at a given moisture content. (Courtesy Satak:e
1990.)

~
CC)
250 SATAKE

Iv. RICE MOISTURE CONDITIONING

A. Problems with Low Moisture Rices

In commercial rice milling, moisture content can have a significant impact

on the processing characteristics of the rice and the economic return to the
miller. Usually moisture content concerns at the rice mill level are not
about high moisture, since rices cannot be stored safely with excessive
moisture. Thus, high moisture problems have to be corrected before the
rice reaches the mill. Instead, the moisture content problems a rice miller is
likely to encounter have to do with low moisture content. These rices tend
to be more rigid and inelastic than higher moisture rices and are more apt
to break under the strain of rice milling, thereby reducing whole kernel rice
yield. Cooking and taste quality can also be seriously affected by kernel
cracking and starch leaching to the cooking water observed during cooking
of low moisture rices.
The low moisture problem usually originates in the field during hot dry
weather in late summer or at the rice dryer when inattention and tight
scheduling create conditions for overdrying of some rice lots. When the
rice reaches the mill, the miller can do little but try to adjust the equipment
so as to make the most possible out of the inferior raw material. A method
is needed by which moisture could be safely added to rice kernels so that
the optimum moisture content could be restored. Water addition also in-
creases total yield of milled rice.

B. Difficulties of Moisturizing Rice

However, there are some practical difficulties that make moisture addition
difficult. Primarily, moisture cannot be added to rice kernels, either brown
or milled, by simply soaking in water. Too rapid rehydration creates large
moisture concentration gradients that cause the overly dry rice to develop
serious cracks, which later become broken kernels. Instead, moisture must
be added slowly with periods of resting or tempering time to allow the
moisture gradients inside the rice kernels to reach equilibrium. In order to
provide this resting time, additional storage bins or tanks are required.

C. Early Rice Conditioners

In the past, this phenomenon of kernel cracking due to rapid rehydration

has limited moisture addition to less than 0.2% per hour. Ifhigher moisture
levels were desired, subsequent additions of moisture could only be per-
formed after at least one hour of resting time in an abundance of interim
NEW PROCESSING METHODS AND EQUIPMENT 251

LSU moisture adding area Elevator

Volumetric discharger
Blower

"""
Humidifier

Figure 17 First moisture conditioner for milled rice. This moisture conditioner
utilized humidified blowing air through a unit resembling an LSU-type dryer to add
moisture to rice. Rice passes from the top of the unit, receiving moisture-laden air
through the inverted troughs, and is discharged out the bottom by the volumetric
discharger. An elevator is provided for recirculating the rice through the unit for
multiple passes. (Courtesy Satake 1990.)

storage tanks. Rice with 15% moisture content could be produced in this
fashion. The first units of milled rice conditioners (Fig. 17) resembled a rice
dryer of the LSU type that utilized blowing air, humidified almost to satura-
tion with a water mist to slowly moisturize the rice. Moisture could be
added safely and not crack the rice but only if the rate of moisture addition
was held below 0.2% per hour.

D. Pressurized Rice Conditioners

For some time, a rate of moisture addition of 0.2% per hour represented
the maximum rate practical. However, later research discovered that im-
provements could be made if the water mist was applied in a pressurized
vessel. Previous rice conditioners had always added moisture under nor-
mal atmospheric pressure. It was discovered that pressurizing a chamber
to about 1.2 x 105 N/m2 could increase the safe penetration of moisture
into the grain by a factor of 5 or 10, thus accelerating the moisture absorp-
tion rate from 0.2% per hour up to 1.0% or even 2.0% per hour without
252 SATAKE

Water supply
t---:s._device

Pressurized chamber

Air cutoff
valve

Figure 18 Pressure-type moisture conditioner. This rice conditioner utilizes a

pressurized vessel to apply moisture to rice grains under pressure, thus accelerating
moisture addition. Rice enters the top of the chamber through an air cutoff valve
and passes through the chamber absorbing moisture. The moisturized rice is dis-
charged through the bottom air cutoff valve and can be recirculated for multiple
passes by an external elevator. (Courtesy Satake 1990.)

damaging the rice grains. Figure 18 shows the device constructed to accom-
plish this. Basically this is a continuous flow pressure vessel with air lock
cutoff valves at the inlet and outlets. Pressurized moist air is injected into
the vessel in a regulated fashion, and the pressure inside the chamber
helps promote the penetration of the moisture into the kernels. Retention
time inside the chamber is about 20 minutes. A moisture sensor on the
interior of the chamber regulates the amount of moisture injected into the
chamber.
The use of a pressure vessel has led to improvement in the ability to
rehydrate overly dry rices. Table 2 shows typical moisture addition results
for the atmospheric pressure, white rice moisture conditioner. Notice that a
1% moisture gain is achieved at 300 minutes, or 5 hours.
Table 3 shows the results of moisture addition under pressure. In this
case, 1% moisture addition can be accomplished typically in about 20
minutes without any significant cracking or breaking of the rice.
Table 2 Moisture Conditioning of White Rice Using
an Atmospheric Pressure Moisture Conditioner
Elapsed Grain Moisture Broken
time temperature content Cracks rice
(min) eq (%) (%) (%)
0 34.0 13.4 1.0 2.0
30 34.0 13.4 1.0 1.9
60 33.8 13.6 1.0 2.0
90 33.6 13.7 1.0 1.8
120 33.5 13.8 1.0 2.0
150 32.5 13.9 1.0 2.1
180 31.0 14.0 1.0 2.2
210 30.0 14.1 1.2 2.0
240 29.3 14.2 1.2 2.0
270 28.0 14.3 1.2 2.1
300 27.0 14.4 1.2 2.0
Source: Courtesy Satake Corporation.

Table 3 Moisture Conditioning of White Rice Using a Low Pressure Moisture

Conditioner
Grain Moisture
temperature content Cracks Broken rice
Elapsed
time Before After Before After Before After Before After
(min) eq eq (%) (%) (%) (%) (%) (%)
0 32.7 31.4 13.7 14.1 1.5 1.5 2.0 2.0
0.5 32.8 32.6 13.7 14.2 1.6 1.5 2.0 1.9
1 34.7 32.8 13.6 14.2 1.5 1.5 1.9 1.9
2 35.2 33.1 13.6 14.3 1.8 1.9 2.1 2.0
3 35.0 33.5 13.7 14.6 1.9 2.0 2.0 1.9
4 35.4 34.1 13.6 14.6 1.0 1.1 1.8 2.1
5 35.5 34.5 13.6 14.6 0.8 0.9 2.1 2.0
8 36.0 35.3 13.6 14.7 0.0 0.0 2.0 2.1
11 35.9 35.8 13.6 14.6 0.5 0.9 1.9 1.7
14 37.3 36.3 13.5 14.7 0.9 0.9 2.1 2.0
17 36.5 36.3 13.6 14.7 2.0 1.5 1.8 1.7
20 36.0 36.3 13.5 14.6 1.8 1.5 1.9 1.9
23 36.4 36.1 13.6 14.7 1.3 1.8 1.9 1.8
26 36.5 36.5 13.7 14.6 1.0 1.5 2.0 2.0
29 37.0 36.8 13.6 14.7 1.4 1.8 1.8 1.6
32 37.1 37.1 13.7 14.7 2.0 1.2 1.7 1.6
35 37.5 37.2 13.7 14.6 0.8 0.4 2.2 2.1
38 38.1 37.7 13.8 14.7 1.0 1.0 2.0 1.3
Source: Courtesy Satake Corporation.

253
254 SATAKE

Figure 19 Rice cracking during soaking (in zooc water for 40 minutes) moistur-
ized to (A) 13%, (B) 14%, and (C) 15% moisture content. Rice moisturized to 15%
moisture content shows the least number of cracks compared to rices of lesser
moisture. (Courtesy Satake Corporation.)

E. Benefits of Rice Moisture Conditioning

Figure 19 shows that moisturizing rice can have a significant positive bene-
fit on the cooking and eating quality of milled rice. Samples of rice at 13.0,
14.0, and 15.0% moisture content were soaked in zooc water for 40 min-
utes. Under these conditions, moisturized rice of 15% moisture content
develops the least number of cracks. During actual tests, a distinctive snap-
crackle-pop noise can be heard coming from the low moisture rices, indicat-
ing the severity of the stresses being built up within the rice grains as
moisture begins to enter the kernel. These stresses build up until released
NEW PROCESSING METHODS AND EQUIPMENT 255

c
256 SATAKE

in a cracking of the rice grain, which can become so violent that the crack-
ing sound is audible.
This cracking phenomenon during soaking is of great importance, par-
ticularly to those countries where cultural practices entail prolonged wash-
ing and soaking of the rice prior to cooking. Still, the same phenomenon
also takes place during the initial phase of cooking rice so that this becomes
a universal problem. Since this cracking allows inner starch to flow out into
the cooking water and breaks down the integrity of the rice kernels, the
result is a cooked rice that is starchy and mushy and generally unattractive
and flavorless.
The potential benefit of moisture addition to rice is one of quality. Not
only are the weight and volume of the rice increased, but the quality in terms
of better-tasting, better-cooking rice is improved. It is conceivable that
future milling technology will be- expanded to include moisture condition-
ing as a standard processing stop. And as modem, efficient milling equip-
ment maximizes milling yields in the rice industry, the competitive edge in
the future will lie in rice quality and taste. Certainly rice moisture condition-
ing will play a major role in this area.

V. RICE-WASHING TECHNOLOGY
A. Origins of Traditional Rice Washing
In some cultures, rice cooking is as simple as putting rice in a pot of water
and boiling until done. In other cultures, however, especially those where
rice is the staple food and the main ingredient of the meal, rice preparation
is a more serious affair. This can probably be best illustrated by the rice-
preparation practices in Japan. In Japan, much time and attention are
devoted to the preparation of rice before cooking. The principal step in this
preparation is washing the rice. Ordinarily it may seem that washing rice is
not really necessary, and, in the United States, it is recommended that rice
not be washed in order to retain the required enrichment. But in rice-
centered cultures such as Japan, washing is seen as indispensable for the
proper preparation of rice.
In Japan specifically, washing is done because washed rice simply tastes
better. Even in well-milled rice, some bran residue usually adheres to the
kernel surface. Depending on the interval between milling and consump-
tion, the residual bran can oxidize and produce free fatty acids that can be
detected as an off-taste or rancid flavor. Discriminating palates can readily
detect these off-flavors, which make the cooked rice unpalatable. Washing
the rice removes the residual bran and thereby eliminates the off-flavors.
Washing is usually more than just a simple rinsing; rather, it can include a
NEW PROCESSING METHODS AND EQUIPMENT 257

WATER I
I
FILTER I
~·· · · · · · ·l· · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·i
i ALEURONE
MILLED
RICE r+ LAYER
SCRUBBER
MOISTURE
-CONDITIONER - POLISHER ~
j

L·············f·····························wA8i-iiN<iii:ii:iviNG".Lir·ilr··························j
WASTE WATER _I

................................................................................................................ .
: :
COLOR ROTARY METAL
1--
SORTER SIFTER DETECTOR

................................................................................................................ .
SEPARATING PROCESS

~ • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 00 • • • 0 • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • 0 0 • • • • 0 . ~

+
VACUUM
TANK WEIGHING/ f-- PACK
PACKAGING MACHINE

'······································fi-'Ci<ACiiNCi.Pf:iocess·····················
................. :

JIFF
PREWASH ED
RICE

Figure 20 "Jiff" rice process. The "Jiff" rice process consists of three steps:
washing/drying, separating, and packaging. (Courtesy Satake Corporation.)

lengthy process of adding water, scrubbing the rice by hand, pouring off the
water, and adding more water again and again. Usually this is repeated
until the water finally runs clear, at which point heat is applied and actual
cooking is begun.

B. Problems with Rice Washing in a Modern World

Traditionally, in most households where washing is practiced, this time-
consuming chore is performed by the housewife as part of her regular
cooking duties and is not seen as a problem. However, in the case of
258 SATAKE

A
Figure 21 "Jiff" rice washer (A) and treatment plant (B). The washing unit
washes and dries the rice. The waste water treatment plant treats the waste water
from the rice washer and reduces the level of contaminants before discharge of the
water to municipal sewage systems. (Courtesy Satake Corporation.)

institutional preparation (restaurants, hospitals, schools, etc.), rice wash-

ing can be overly time consuming and costly, not to mention rather cumber-
some, due to the large volumes of rice involved. Moreover, even in individ-
ual households, rice washing is becoming bothersome as more and more
housewives enter the labor market and have less time to devote to pro-
longed cooking tasks.

C. Development of Rice-Washing Machines

To respond to this problem, rice-washing machines have been developed.
Washing rice should be a simple matter, however, large-scale rice washing
poses special problems. First, water addition must be handled carefully,
NEW PROCESSING METHODS AND EQUIPMENT 259

otherwise extensive cracking of the grains can occur, and, as discussed

earlier, this can seriously degrade rice, completely negating any benefits
from washing. Attention must be paid also to the drying step applied after
washing. Too little drying will leave the rice moist, inviting mold and han-
dling problems. Too much drying, however, could partially gelatinize the
grains, leading to uneven cooking and poor flavor. Finally, there is the
problem of waste water. Any washing process produces waste water. A
rinsing process, such as that usually used for washing rice, intrinsically
utilizes copious amounts of water, many times the volume of the rice being
washed. This can create substantial handling and disposal problems.

D. The "Jiff" Rice Process

To overcome the problems associated with commercial, large-scale wash-
ing of rice, a new washing system has been developed. The emphasis is
on the whole system, since the technique not only controls the amount of
water and the drying rate but also handles waste water treatment of the
effluent. In this sense, the rice-washing system becomes a complete, self-
contained process. The system consists of two separate parts: the washer/
dryer and the waste water treatment device. The entire system has been
termed the "Jiff" rice process in Japan to denote the fact that its washed
260 SATAKE

rice product can be prepared quickly and easily or, in American slang,
"in a jiffy." The flow sheet for the complete "Jiff'' process is shown in
Figure 20.
The washer/dryer unit (Fig. 21A) adds water to the rice, agitates the
moistened rice, and dries the rice, all in a continuous process flow. The
moisture content of the rice can be adjusted to remain constant with no
moisture absorption or loss of moisture.
The second part of the process (Fig. 21B) is the treatment of the waste
water. During the washing step organic solids become dissolved in the wash
water, and these solids pose a significant disposal problem. Usually this
type of waste water is discharged into the municipal sewage system; how-
ever, due to the high loading of suspended and dissolved organic solids, the
costs charged by municipal sewage plants can be substantial. Therefore, a
method has been developed to reduce this contaminant loading before it
reaches the municipal system.
This system (Fig. 22) consists of integrated waste water treatment,
which, by precipitating and settling action, removes almost all of the
suspended and dissolved solids from the water. The solids and sediment
are discharged from the system and can be discarded or dried and added
to the bran fraction since the nutritional content of the dissolved solids is
suitable for mixing with rice bran for animal feed. The treated water
discharged from the system is acceptable for discharge into municipal
sewage systems at no penalty or at greatly reduced surcharges. The BOD
and suspended solids values before and after purification are shown in
Table 4.
Although this process is one that has maximum advantages in cultures
where rice washing is the norm, the contribution of the process to the
quality of milled rice may make it desirable even in those markets where
rice washing is not customary.

Table 4 Efficiency of Waste Water Treatment of

Rice Washing Using the "Jiff" Rice Washing Process
Before After
Treatment
type BOD• BOD ss
Biological treatment 15,000: 15,000 50 20
Activated sludge 7,500 7,500 600 600
• BOD = biological oxygen demand.
b SS = suspended solids.
c Values given in parts per million.
Source: Courtesy Satake Corporation.
 ztrl
~
"'::I
~
r:l
trl
rJJ
.....
rJJ

li
z
n

r-
BIOLOGICAL SLUDGE PURIFIED ~
WASTE AGGUJTINATIONJ I-- SEPARATING trl
SEDIMENTATION I--- TREATMENT

=
WATER TANK WATER ""l
TANK TANK
0
I t::l

4-
rJJ

HIGH MOISTURE I--- DRIED >

z
WRINGER I-- DRYER SOLIDS
ORGANIC SCUDS t::l
--
trl
.0

Figure 22 "Jiff" rice washer water treatment plant. This system removes almost all of the suspended and dissolved ~
~
solids from the waster water of the Jiff process. (Courtesy Satake Corporation.) trl
z""l

w
....C"l
262 SATAKE

VI. CONCLUSION
Although rice processing is already a fully mature technology, there are
areas where new and improved techniques can be incorporated to enhance
the yield and quality of milled rice. Without doubt, progress to be made in
rice processing in the 1990s will focus on these areas. The benefit will be
better tasting and higher quality rice products to supply sophisticated and
discerning consumers around the world.

REFERENCES
1. Modern Rice Milling Technology, (1991}. Satake Corporation, Hiroshima,
Japan.
2. Luh, B. S. (1980}. Rice: Production and Utilization, AVI Publishing Com-
pany, Westport, CT.
3. Satake, T. (1942}. Efficiency of grain milling machines, Shin-tairiku, Japan.
4. Satake, T. (1951}. Efficiency and heat generation of Grain Milling Machines,
Farming Mechanization, Shin-Norin-sha Press.
5. Satake, T. (1964}. Centralized Rice Milling, Satake Corporation, Hiroshima,
Japan.
6. Satake, T. (1990}. Modern Rice-Milling Technology, University of Tokyo
Press, Tokyo, Japan.
 12
Parboiling Rice with Microwave Energy
Lakshman Velupillai
Louisiana State University Agricultural Center, Baton Rouge, Louisiana

I. INTRODUCTION
A. The Conventional Parboiling Process
Parboiling, a precooking process applied to rough rice, has been practiced
for centuries in parts of South Asia. Presently about 20% of the rice pro-
cessed in the world is parboiled. The subject of parboiling and its uses are
described extensively in the literature. Notable among these is a fascinating
account of the story of rice parboiling by Bhattacharya (1) (see also Refs. 2
and 3). The process, now practiced in Asia, Africa, Europe, and the Ameri-
cas, involves soaking the rough rice until the grains absorb moisture to
approximately 30% (wet basis), draining the excess water, and heat treat-
ing to gelatinize the starch in the kernels. The grain is then dried to facili-
tate processing or storage. Although the process, according to Garibaldi
(2), was probably invented to ease the removal of the husk, economic,
nutritional, and other advantages were later recognized. With the advent
of mechanical processing methods, the process of parboiling not only sur-
vived over the years but also became widely practiced in many countries. In
the United States, the process is fast gaining acceptance not only for export
purposes but also for widespread domestic use. Of the rice exported from
the United States, nearly 50% is in the parboiled form. In the domestic
market, a rapidly growing use for parboiled rice is in hotel and restaurant

263
264 VELUPILLAI

chains, which have recognized its many desirable cooking and keeping
qualities.

B. Research Emphasis
As the process of parboiling evolved and its use spread, the research empha-
sis changed based upon the state of the technology available and the cur-
rent consumer interest. The early studies date back to the turn of the
century, when it was recognized that consuming parboiled rice prevented
the occurrence of beriberi, a disease later associated with the lack of thia-
mine in the diet. Thus, most of the early research was devoted to the
nutritive aspects of parboiled rice (2, 4-9). The next stage of development
of the process was the advent of the steam boiler, which spurred the larger-
scale commercialization of the process. Research activity was thus concen-
trated on the processing conditions and techniques to improve the quality
of the parboiled product. These improvements were geared to the reduc-
tion or elimination of the undesirable characteristics of the end product.
Some of these improvements included:
1. Cleaning and size grading of rough rice prior to parboiling
2. Improvements to the soak process including the use of hot water to
reduce the undesirable odor associated with "cold" soaking
3. The application of heat (primarily saturated steam) to facilitate uni-
form gelatinization of the kernels
4. Mechanical drying
Additionally, much research effort was devoted during this phase to the
scientific evaluation of the quality characteristics of parboiled rice.
During this "commercialization" phase of development of the parboiling
process, both the researcher and the processor realized the importance of
energy (process heat) that was required to complete the process. This
realization became particularly relevant when the commercial-scale opera-
tors began to resort to mechanical systems to dry the parboiled rice. On a
per ton basis, a complete mechanical system that parboils rice utilizes
almost 80% of the total process heat for the drying stage. This explains the
widespread use even today of solar drying in many countries that parboil
rice. However, the industrialized nations used mechanical drying systems
almost exclusively due to the labor intensive nature of solar drying systems.
Research was therefore directed at systems utilizing nonfossil fuel sources
to keep costs down. A prime target was rice hulls, a byproduct of the
milling process and available at the mill site. Since steam boilers were
available at the time, modification of existing boilers and even new designs
of "husk-firing" systems for steam boilers came into use. As a result, husk-
PARBOILING RICE WITH MICROWAVE ENERGY 265

fired water tube boilers of a wide range of capacities have been in use at
parboil rice mills for decades. Typically, the supply of rice hulls is sufficient
for the production of process steam for the entire parboiling process includ-
ing drying (via steam heat exchangers or direct/indirect fired drying sys-
tems). However, due to efficiency and cost considerations, the quantity of
hulls is insufficient for the production of process heat as well as motive
power for the rice mill. These developments are embodied in what is
termed "cogeneration," a concept receiving increased attention today.
Using the energy available in rice hulls seems a logical field of endeavor
for "parboil" rice mills as well as "white" rice mills (i.e., those that produce
only "white" nonparboiled rice). The use of rice hulls as a source of "wet"
heat is the most appropriate form for the small-capacity (1-2 t/hr) parboil
mill. The larger parboil mills could improve efficiencies if the investment
could be made to produce motive/electrical power and then utilize the low
pressure steam for the process wet heat in the parboiling process. On the
other hand, large capacity "white" mills in the United States have shown
that production of electrical power that could be sold to utility companies
through husk-burning technologies is feasible. Electrical power in these
instances is produced through conventional steam turbine-generator tech-
nologies. Capitalizing on these technologies and their applications offers
much promise in utilizing electrical power for both parboiling and milling
phases. Electrical power could be utilized through the application of
electrotechnologies such as microwave parboiling-the subject of discus-
sion in this chapter.

IT. THE MICROWAVE PARBOILING PROCESS

A. Background
The desire among processors to seek alternative sources of energy, pa;ticu-
larly for grain drying, was a result of the energy crisis of the late 1970s.
Utilizing electrical power generated from coal, nuclear energy, or sources
such as rice hulls becomes an important alternative. These choices are not
fully exploited primarily due to economic factors (availability of low-cost
natural gas for grain drying in the United States, for example). Serious
consideration is merited, however, when the investment can be made to
expand the use of rice hulls, a renewable energy source, for the production
of electrical energy for processing. There have been efforts, particularly in
the United States where certain processes have been developed for the
application of electrotechnologies to the processing of grains. The develop-
ment of the MIVAC (microwave-vacuum) drying process (10) was one such
effort primarily for the drying/dehydration of agricultural crops. This pro-
266 VELUPILLAI

cess involved the use of microwave heating of the product in an evacuated

atmosphere. The application of microwave heating technology as applied
to rice was first recorded by Wadsworth and Koltun (11) when they evalu-
ated the drying of harvested rice in a chamber under the combined effect of
microwave energy input with an evacuated atmosphere. Subsequently,
Wadsworth et al. (12) and Velupillai et al. (13) have reported the applica-
tion of a similar process to the drying of conventionally parboiled rice.
Based upon a 5-year effort begun in 1984, Velupillai et al. (14) were named
as inventors in a U.S. patent awarded to the Louisiana State University
(LSU) Agricultural Center for a process for parboiling rice termed the
"microgel" process. Since this development, researchers at the LSU Agri-
cultural Center and the Southern Regional Research Center, U.S. Depart-
ment of Agriculture, are collaborating in the development of a complete
process for parboiling rice using microwave energy.

B. Basic Features of Microwave Parboiling

Regardless of the source of energy, the basic steps in parboiling rice include
hydration, heat treatment, and drying. There are, however, certain distinct
differences in the application of electromagnetic radiation in the micro-
wave frequency range to effect the soaking, cooking, and drying steps.
Some of these differences include the mode of transfer of heat to and
within the rice kernels and the removal of moisture during the drying stage.
A unique feature of this approach, as the present status of research indi-
cates, is that all three steps of soaking, cooking, and drying can be per-
formed as a series of sequential steps in a single piece of equipment. The
microwave energy application and vacuum pressure cycles are programmed
in sequence to produce a finished product at the end of the process.
Provided in the following paragraphs is a discussion of the three basic
steps in an experimental microwave energy application to parboil rice.
Published research data are provided where applicable. However, as the
process is still in a state of development, the author has made projections in
areas where information is lacking.
1. The Soaking Process
The most important objective of the soak process is the increase of the
moisture content to a level and distribution (within the kernel) whereby the
subsequent cook stage will ensure the complete and uniform gelatinization
of the starch in the rice kernels. Conventional processes have determined
that when the grain (total) moisture is approximately 30% (wet basis), then
the moisture at the innermost parts of the kernel is sufficient to fully
gelatinize the entire kernel during the subsequent steam heat treatment
PARBOILING RICE WITH MICROWAVE ENERGY 267

stage. The soak temperature and its uniformity throughout the grain mass
during the soak period are additional factors that determine grain quality
and cost of parboiling. Research (15) has shown that a total grain moisture
content of about 23% (wet basis) after the soak process is sufficient to
produce parboiled rice of acceptable commercial quality if the grains are
allowed to partially equilibrate in terms of moisture. A lower total mois-
ture at the end of the soak stage not only means a lower cost of soaking but
also reduces the amount of moisture that must be removed later (adding to
the cost) at the drying stage. This advantage must be considered in light of
the cost of "holding" the grain during the partial equilibration as well as the
increase in process time.
A key to successful soaking or hydration of the rice grains in the micro-
wave parboiling process is to start the soak stage with the minimum amount
of excess water and end the stage at a near optimum grain moisture. At the
end ofthe cook stage, which is the start of the drying stage, there ought to
be little or no excess water. This is very important from an energy and
process time standpoint. On the other hand, excess moisture must be pres-
ent throughout the cook stage to ensure complete and uniform gela-
tinization of the rice kernels. As a result of these requirements, present
studies at the LSU Agricultural Center assume a target moisture at the end
of the soak process of 30% (wet basis), a grain pore space volume of 50%
and approximately 20% by weight of excess water to make up for evapora-
tion losses. When information on the effect of soaking rice using the appli-
cation of microwave energy on partial equilibration and target moisture
level to ensure commercially acceptable end grain quality become known,
the above assumptions could be revised. At this writing, information is
available on target moisture levels achieved under vacuum at 300 Torr,
indicating that 30% moisture (wet basis) could be achieved in 40-60 min-
utes under the application of between 0.6 and 1.2 kW of microwave power
in 700-g rough rice samples.
Soak temperature is another factor that must be given serious consider-
ation. In conventional parboiling processes, this factor is controlled exter-
nally in that the hot water is circulated via an external tank or small con-
tainer where makeup heat is added to compensate for losses from the soak
tank. In the microwave process, one of the controlling mechanisms is the
pressure in the chamber. Additionally the pressure in the chamber could
also be used to advantage in increasing hydration rates. A high vacuum
(approximately 50 Torr) at the beginning of the soak stage enables the air
trapped between the kernels and the hulls to be bled out of the system. The
pressure in the chamber could then be increased gradually to a level corre-
sponding to the boiling point of water, which is also the target soak tem-
perature for the soak step (e.g., 300 Torr corresponding to about 75°C).
268 VELUPILLAI

The soak process can then proceed at a constant pressure with the time and
microwave power adjusted to achieve the desired results. Again, the crite-
ria governing variables of microwave power and soak time should include
the grain moisture at the end of the soak stage and the quantity of water
remaining. The suitability of these parameters could only be judged by the
final outcome of parboiled rice quality as seen by the present completely
translucent kernels, grain color, and yields. Present research indicates that
this approach yields approximately 30% (wet basis) final moisture in a time
period of 40-60 minutes with suitable microwave power levels.
In the conventional process, partial equilibration has been found to be
beneficial to a certain extent as discussed previously. However, the suitabil-
ity of partial equilibration prior to the cook stage in a microwave process is
yet to be determined. From a process time and operational point of view,
time devoted to just "holding" a product is not desirable. Thus any partial
equilibration stage must be carefully investigated with respect to potential
benefits. In addition, the electric field of the microwave radiation interacts
with the dipoles in the material under treatment, generating heat. Heat is
normally generated within the material due to the reversal of the polarity of
the field when oscillation at the frequencies approaching that of the micro-
waves occurs. Since water possesses a large number of dipoles, it reacts
strongly to the application of microwave energy. These factors and mecha-
nisms may or may not aid in producing a uniform distribution of moisture
in the kernels at the end of the soak stage.
2. The Cooking Process
The next stage of the parboiling treatment involves heating the soaked rice
kernels. During this phase, the starch granules in the endosperm, which
have absorbed water during the soak stage, are changed in structure from a
crystalline to an amorphous form. This is an irreversible process referred to
as gelatinization. It is essential that heat and moisture are present through-
out the kernel to completely parboil each grain. The key criteria are the
presence of sufficient moisture and the transfer of heat at or above the
gelatinization temperature for the variety of rice under treatment. Assum-
ing that the grains are sufficiently hydrated, a drawback of the conven-
tional steam heating process is heat transfer from the steam source to and
within the kernels. The present technology, which utilizes "autoclaving" or
steam in tanks under pressure, provides the best form of effecting gela-
tinization of the soaked rice grains. By comparison, heat treating the
soaked grains of rice in the presence of excess moisture by the application
of microwave power possesses an inherent advantage over conventional
heating in that the water molecules between and within the grains selec-
PARBOILING RICE WITH MICROWAVE ENERGY 269

tively respond, generating heat. With appropriately adjusted microwave

power input, the rice kernels are subjected to heat, both outside (due to the
presence of the excess water) and inside as the kernels are already satu-
rated. This mechanism therefore makes the heat-treatment phase indepen-
dent of conduction heating. An additional advantage is that, unlike conven-
tional soaking and steam heat treating, the vessel or chamber that contains
the rice is not heated as much. This results in reduced conduction and
convection losses, translating to energy savings.
Current research at the LSU Agricultural Center involves the evaluation
of the effect of pressure in the chamber, the input microwave power, and
the time of the cook stage. The required input power levels, the process
times, and the pressure in the chamber will be functions of the equipment
configuration and the quantity of rice under treatment. One would assume
that a properly designed equipment configuration would ensure adequate
penetration by the microwaves at the selected frequency (2450 or 915
MHz) as well as ensure uniform heating of the rice water mixture. It is
critical, however, that one track the increase of grain temperature via a
suitably calibrated infrared sensor throughout the period of the heat treat-
ment to ensure that (a) the grain temperature reaches or slightly exceeds
the gelatinization temperature for the variety of rice under treatment, and
(b) the grain temperature does not increase to a point at which scorching
may take place. Once the rate of application of microwave power is deter-
mined for the particular equipment-rice/water configuration, future re-
search should concentrate on determining the optimum point at which the
power should be reduced or cut off. That is important due to the fact that
one does not want the drying phase to begin sooner than necessary, which
results in partly gelatinized grains or "white bellies." White bellies are an
undesirable component of the processed parboiled rice. Another factor
that merits consideration is the pressure in the microwave chamber. A
pressure level at or higher than atmospheric pressure would prevent prema-
ture drying effects. Future research must address these issues, and the heat
treatment parameters must be refined based primarily on the quality charac-
teristics of the end product.
3. The Drying Process
In conventional commercial processes, the steam-heat treated rice is di-
rectly transported into a series of dryers to complete the drying step. This
step reduces the grain moisture from a level of approximately 36% (wet
basis) to a level suitable for storage and/or processing, i.e., 12-14% (wet
basis). In the case of microwave parboiling, the heat-treated rice could at
once be subjected to a new set of appropriate conditions of pressure and
270 VELUPILLAI

power input to begin the drying step. Chapter 14 discusses in detail the
implications of applying microwave technology to the drying of parboiled
rice. However, a few salient points are discussed here.
Previous studies by Wadsworth et al. (12) and Velupillai et al. (13) have
reported on both the drying aspects and the grain end quality parameters.
The primary finding as reported by the above studies was that the applica-
tion of microwave energy to dry parboiled rice in an evacuated atmosphere
produced rice of a quality comparable to parboiled rice produced by con-
ventional methods. Analysis of the conditions of drying, such as the micro-
wave power input and the pressure in the cavity, revealed that the drying
mechanisms were inherently different from conventional methods of drying
parboiled rice. For instance, the "falling rate" trend of the drying rates was
absent; instead, the drying rates (observed to be directly proportional to
the microwave power input) essentially remained constant until the final
moisture content of the parboiled rice was 18% or lower. This is a signifi-
cant difference when the drying mode is compared with conventional hot
air drying and indicates that the diffusion of water within the kernels of rice
is not the limiting factor. Another important finding in this study was that
rice kernel density did not increase as in the case of conventional drying
when the kernels tend to shrink as drying progresses. These two important
trends under microwave energy application to drying of parboiled rice
indicate that with the proper adjustment of the microwave input power, the
pressure in the chamber and the drying time, parboiled rice could be dried
in one single pass to safe storage levels. This has a profound bearing on
potential equipment cost and process time savings when compared with
conventional commercial drying installations. Although the proper equip-
ment configuration is not yet determined at this stage of development, the
impact on drying equipment could be seen when one reviews the present
equipment layout in a typical large capacity (20 t/hr) parboil rice drying
section. Normally, the rice leaving the autoclave is conveyed directly to a
rotary dryer in which hot air at temperatures ranging from 150 to 260°C is
utilized. The first-pass drying reduces the grain moisture to between 18 and
22% (wet basis). Typically, the rice is then conveyed to a series of column
dryers to gradually reduce the moisture (at lower air temperatures of less
than 93°C) to between 12 and 14% (wet basis).
The microwave-vacuum method of drying parboiled rice has been found
to produce rice yields (total as well as whole kernel yield) comparable to
yields from conventional methods. The yield variation was primarily influ-
enced by the drying rate, which, in turn, was influenced by the power input
and the pressure in the chamber. A similar trend was seen for the percent
variation of the gelatinized kernels, i.e., when the input power and vacuum
were increased, there was a reduction of the percent gelatinized kernels.
PARBOILING RICE WITH MICROWAVE ENERGY 271

III. FUTURE OUTLOOK

The foregoing discussions lead one to ask the question, could a complete
process for microwave parboiling be developed? What would be some
distinct advantages over conventional processes? Although a complete pro-
cess has yet to be formulated, it is worth while to explore certain inherent
differences and some advantages to using microwave technologies in rice
parboiling.
First, there has been a greater awareness among the food processors,
particularly in the United States, who have in the last few years installed
commercial scale microwave based processing equipment. This, according
to Decareau (16) is due in part to the growth in the consumer use of
microwave ovens in the United States as well as other parts of the world.
Second, reviewing the general features of microwave processing systems,
one finds certain inherent differences when a microwave-based system re-
places a conventional system:
1. Equipment configuration
2. Space savings
3. Process time
4. Energy use
5. Quality of product
6. Maintenance

On the other hand, microwave-based systems are generally technologically

more advanced than conventional systems and require a different set of
operational and safety standards.
Examining the above inherent differences in light of a potential rice
parboiling system utilizing microwave energy, several comments could be
made. The equipment configuration and the resulting space savings are a
necessary consideration when investments are to be made. At a minimum,
soak tanks, a hot water supply tank, a boiler, an autoclave, a rotary dryer,
one or more column dryers, ancillary equipment including valves and
steam pipes, and a conveying system of elevators and belts are required in a
conventional parboiling plant. In addition to a large investment for the
equipment, the layout necessitates large, tall buildings to house the parboil-
ing equipment. By comparison, a single unit utilizing microwave energy to
parboil the rice is expected to save considerable plant floor space and not
require tall buildings. A related point for discussion is the potential process
time savings. As discussed earlier in this chapter, the soak times have been
shown to be as much as a third less than for the conventional process.
Process time savings generally mean that a smaller amount of rice is pro-
cessed at any one time, translating into a smaller amount of the product
272 VELUPILLAI

being lost (or lowered in quality) in the event of equipment failure. An-
other aspect of this configuration is the possibility that parallel systems
could be set up to make up the desired throughput. As a consequence, only
partial reduction of throughput will result if an equipment component in a
single stream fails. This alternative is not financially feasible in the present-
day systems of rice parboiling.
The energy use in microwave parboiling is an important consideration.
However, research data are not available at this time to illustrate the en-
ergy use on a commercial scale. One needs to consider both the types of
energy application and the source of the energy when energy use compari-
sons are made. In conventional systems, natural gas, kerosene, and furnace
oil are the primary sources of energy. Rice hulls have also been used for
decades, particularly in the smaller mills worldwide. This source has the
advantage that rice hulls are available at the mill site and are a renewable
source. Ifcommercial-scale parboiling systems utilizing microwave technol-
ogy become a reality, two factors merit consideration. First, the environ-
mental impact (i.e., pollutants from combustion of either fossil fuels or rice
hulls) could be better controlled at a single location of the electric utility
company (which usually must conform to the environmental standards).
Second, the use of a renewable source of energy such as rice hulls could
reduce the use of fossil fuels. As stated earlier, research data are not
available to illustrate a commercial-scale comparison. However, data are
available on an experimental scale for comparison of the energy use in the
drying step of the parboiling process. The drying step is the most energy
intensive step (requiring nearly 80%) of the total energy to soak, heat-
treat, and dry rice. Concentrating on the drying step will therefore provide
a good indicator in making energy use comparisons between conventional
and microwave parboiling systems. Data available as a result of the LSU
Agricultural Center studies (17) indicate a savings of 46% in the energy
available at site or at the magnetron to dry the soaked and steamed rice.
This comparison, however, does not take into consideration the conversion
efficiency of fossil fuels to electrical energy.
Product quality in the general sense when microwave systems are used
generally tends to be better primarily due to the fact that the heating is
throughout the product rather than on the surface first and then by conduc-
tion heat transfer into the product. Previous discussions in this chapter
have addressed this issue illustrating that uniform heating throughout the
kernels of rice is essential. In the marketplace today, particularly in the
United States, parboiled rice must meet certain "performance" criteria
required by consumers. Cooking quality has long been the hallmark of
parboiled rice in that it is easier to cook and the cooked rice is dry and
fluffy in comparison with unparboiled rice. Another criterion is the steam-
PARBOiliNG RICE WITH MICROWAVE ENERGY 273

table keeping quality. Restaurant chains and hotels desire good steam-table
keeping quality. Previo.us research (14) has shown that microwave technol-
ogy produces rice of comparable quality to that produced by conventional
technology. Additionally, the microwave configuration lends itself to fine-
tuning process variables resulting in a uniform product. The point to keep
in mind here is that a smaller amount of rice (in comparison with conven-
tional parboiling) is being treated at any one time. The microwave system
also has a great potential for "postparboiling" treatments that are now
utilized to enhance, among other things, the cooking aspects (rapid water
uptake resulting in reduced cooking time).
The foregoing discussion leads one to the question raised in the begin-
ning of this section regarding the formulation of a complete process for
microwave parboiling. The response is a positive one in that a commercial-
scale system is feasible. The state of technology development is at a point
today where a commercial-scale prototype needs to be developed. This
prototype must address equipment (microwave chamber, wave guides,
etc.), configuration, controls, instrumentation, the conveying system, and
ancillary equipment including vacuum locks.

REFERENCES
1. Bhattacharya, K. R. (1985). Parboiling of rice, Rice: Chemistry and Technol-
ogy (B. 0. Juliano, ed.), American Association of Cereal Chemists, Inc., St.
Paul, MN, p. 289.
2. Gariboldi, F. (1974). Rice parboiling. FAO Agricultural Development Paper
No: 97, Food and Agriculture Organization of the United Nations, Rome,
Italy.
3. Kik, M. C. (1946). Rice conversion. Rice 1, 49(1):4,12.
4. Kik, M. C. (1956). Polished rice and beriberi. Ricel, 49(4):608.
5. Kik, M. C. (1955). Influence of processing on nutritive value of milled rice. J.
Agric. Food Chern., 3(7):600.
6. Hunter, I. R., Ferrel, R. E., and Houston, D. F. (1956). Free amino acids of
fresh and aged parboiled rice.l Agric. Food Chern., 4:874.
7. Kik, M. C. (1943). Thiamine in products of commercial rice. Cereal Chern.,
20:103.
8. Simpson, I. A. (1951). The distribution of thiamine and riboflavin in rice
grains; and of thiamine in parboiled rice. Cereal Chern., 28:259.
9. Subba Rao, P. V., and Bhattacharya, K. R. (1966). Effect of parboiling on
thiamine content of rice.l Agric. Food Chern., 14:479.
10. Aeroglide, and McDonnell Douglas, MIVAC microwave vacuum drier. Prod-
uct Brochure.
11. Wadsworth, J. 1., and Koltun, S. P. (1986). Physicochemical properties and
cooking quality of microwave-dried rice. Cereal Chern., 63(4):346.
274 VELUPILLAI

12. Wadsworth, J.l., Velupillai, L., and Verma, L. R. (1990). Microwave-vacuum

drying of parboiled rice. Trans. ASAE, 33(1):199.
13. Velupillai, L., Verma, L. R., and Wadsworth, J. I. (1989). Quality aspects of
microwave-vacuum-dried parboiled rice. Trans. ASAE, 32(5):1759.
14. Velupillai, L., Verma, L. R., and Tsangmuichung, M. (1989). Process for
parboiling rice. U.S. Patent No. 4,810,511, March 7.
15. Velupillai, L., and Verma, L. R. (1982). Parboiled rice quality as affected by
the level and distribution of moisture after the soaking process. Trans. ASAE,
25(5):1450.
16. Decareau, R. V. (1988) Microwave processing. Food Industry Trends in FPE
News, The Food and Process Engineering Institute of ASAE, 22(1):1.
17. Verma, L. R., Velupillai, L., Wells, J. H., Ransibrahmanakul, V., and Wads-
worth, J.l. (1991). Microwave-vacuum process in rice parboiling. 1991 Winter
Meeting of ASAE, Paper No. 913531, December 1991.
 13
New Methods for On- The- Farm Rice Drying:
Solar and Biomass
Lalit R. Verma
Louisiana State University Agricultural Center, Baton Rouge, Louisiana

I. INTRODUCTION
The agriculture industry is crucially dependent on fossil fuels for efficient
production of food and fiber. Rising fuel costs and dependence of the
United States on foreign oil are critical factors for U.S. agriculture and
other segments of our economy and make energy independence an urgent
but difficult goal. Significant amounts of energy are used in the postharvest
processing of crops. It is estimated that about 3.3 x 109 MJ of fossil fuel
energy is used in the United States for this phase of crop agriculture (1).
Grain drying in the United States requires the equivalent of 15 million
barrels of crude oil per year, most of which is in the form of propane and
natural gas. Crop drying is a very energy-intensive farm operation. The on-
the-farm energy use for crop drying in the United States is estimated to be
7.3 x 1010 MJ (2). As world demand for efficient food production and
processing increases due to growing population, limited resources, and
emphasis on keeping our environment clean, alternative sources of energy
for on-the-farm use need to be developed and substituted for the more
expensive and increasingly scarce fossil fuels.

Approved for publication by the Director of the Louisiana Agricultural Experiment Station as
manuscript number 91-07-5377.

275
276 VERMA

Solar and biomass energies have unique possibilities in the agricultural

system. On-the-farm energy requirements for applications such as drying
and space heating can be easily met by these renewable energy sources. In-
storage low-temperature drying is estimated to use less than 25% of the
energy of other, faster-drying alternatives. Drying and space heating on the
farm can efficiently utilize low quality heat, which can be generated with
simple and inexpensive solar or biomass systems.
The technology necessary to design/modify, construct, operate, and
maintain such systems is available on the farm, which make them attractive
alternatives. Alternative renewable energy systems are ideally suited for
on-the-farm applications because drying and space heating are required at
different times of the year, which increases the efficiency of solar energy
systems by using them for longer periods during the year. Another factor in
favor of these supplemental energy systems is that they can be used with
the existing air-handling and control systems on the farm. These factors
reduce complexity and cost. Conventional back-up energy can be left in
place to meet increased demand or emergency situations.

II. ON-THE-FARM RICE DRYING

Rice, like many other grain crops, often requires more energy for drying
than for preplanting, planting, and harvesting combined. The amount of
energy used in on-the-farm rice drying systems and methods to reduce this
energy input are of great concern to the rice farmers as this can determine
the profitability of the crop.
Drying offreshly harvested rice, using air as the transfer agent, is essential
for safe storage. To maintain quality, drying should be started as soon as
possible after harvest. And to produce high-quality milled white rice with a
minimum amount of broken kernels, the drying operation must be carefully
executed. During the rice harvest season, the farm drying and storage facili-
ties may become a bottleneck to the immediate and complete drying of the
rough rice. Thus, rice drying on-the-farm is a critical operation requiring
careful management of the available harvesting and drying resources.
On-the-farm rice-drying equipment is essentially the same as that for
other grain crops, however, the operation of the drying systems is different.
Rough rice is a unique type of grain in that its bran and husk layers are
intact while it is being dried. The weather conditions in rice-growing areas
are also different from those in the midwestern states. The high humidity
and ambient temperatures in the South make the rice very susceptible to
quality deterioration and spoilage if not dried quickly, and yet to preserve
milling quality, it cannot be overheated. If rice is heated above a critical
temperature and/or dried too fast below a certain moisture content level,
rough rice will fissure. It is critically important to avoid fissuring in the rice
NEW DRYING METHODS 277

during drying as fissured grains tend to break during subsequent handling

and milling, which drastically reduces the market value of the rice.

A. Equipment
The most common system used for on-the-farm rice drying is the in-bin,
deep-bed drying system. A deep-bed system consists of a metal bin with
perforated floor and plenum chamber below it, a fossil fuel burner, air fan,
duct system for air distribution in the bin, and a dryer control system. Such
a system of rice drying can be easily used for either solar- or biomass-
assisted drying.
The most commonly used fossil fuels for rice drying are natural gas,
propane gas, LP gas, and oil. Burners utilize either a direct or indirect
mode of heat transfer to the air. In the direct mode, the products of
combustion are forced through the grain with the drying air. Such a system
is usually less expensive and utilizes the thermal energy more efficiently,
but the grain may be adversely affected by the smoke and combustion
products. In the indirect mode, a heat exchanger surface is heated and the
drying air is circulated around it and then through the grain to be dried,
keeping the combustion products away from the grain.
Both axial fans and centrifugal fans are used for air movement in the in-
bin rice-drying systems. Maintaining an adequate airflow throughout the
drying period is vital for keeping deterioration in rice quality to a mini-
mum. Electric motors are the most common sources of energy to power the
fans. Airflow rate is the single most important factor in determining the
progress of the drying zone through the rice mass in the bin.
Uniform and level distribution of the rice in the bin can be achieved with
the use of mechanical spreaders provided they do not discharge the grain at
excessive velocities and damage the rice. Gentle mixing of the rice during
drying promotes uniform quality and reduces the likelihood of heat and
moisture accumulation due to respiration, which can promote mold damage.

B. Drying Conditions
Unheated air is recommended for on-the-farm rice drying in a storage bin
except during prolonged periods of high humidity (3). Even then it is
recommended that the drying air be heated by no more than 7°C. Adding
small amounts of heat is desirable to lower the humidity of the drying air
but does not greatly influence the rate of movement of the drying zone and
the time required to complete the drying of the rice. Proper control of the
most common on-the-farm drying system requires monitoring of the ambi-
ent conditions, the temperature rise of the drying air by the burner, and the
temperature and moisture content of the material being dried. Thermostats
and/or humidistats are commonly used to shut off the fuel flow when the
278 VERMA

Table 1 Average Energy Used During On-the-Farm

Drying of Long- and Medium-Grain Rough Rice
Grain No. of kJ/kgH20 Coef. of Heating
Year type tests removed variation (%)

1980 long 3 2513 0.158 92.0

medium 1 3123 94.8
1981 long 1 3364 91.4
medium 8 2370 0.393 89.0
1982 long 4 4120 0.134 92.6
medium 9 3996 0.161 91.7
1983 long 1 3672 92.5
medium 10 3956 0.159 90.6
1984 long 7 3385 0.224 88.3
medium 5 2734 0.371 86.5
1985 long 6 4947 0.532 88.9
medium 5 4178 0.528 87.1
Avg. long 22 3780 90.2
medium 38 3481 89.6

Source: Ref. 4.

preset temperature is exceeded in the plenum and turn it on again when the
temperature drops below a certain point or the humidity exceeds a certain
value for the drying air. A newer control strategy used with the conven-
tional on-the-farm rice-drying systems, to reduce drying energy consump-
tion, is a modulating control on the gas line to the burner of the dryer. Such
a control system has been shown to save up to about a third of the drying
energy using a conventional burner system (4).
Drying only with ambient air, although possible, takes much longer time
than heat-assisted drying (18-24 days versus 7-10 days) and results in
milling quality loss. This loss may be due to the extended storage of high-
moisture rice before it dries, resulting in excessive respiration heat, mois-
ture, and deterioration, and to desorption and adsorption of moisture in
the rice due to changing humidity of the ambient air resulting in fissuring.
Moisture adsorption in dried rice also causes fissuring, similar to that
caused by excessively rapid drying or desorption.

C. Energy Use
A theoretical minimum of 2.5-2.7 MJ/kg of water removed, depending on
the temperature at which water is evaporated, is required in crop drying
(5). The actual energy requirements for evaporating water from grain have
been found to range from 3.0 to 7.0 MJ/kg of water removed (6). The
NEW DRYING METHODS 279

Table 2 Comparison of Energy Consumption Using

Gas-Modulating Control with Conventional Gas
Burners for On-the-Farm Rough Rice Drying
Modulating Conventional burners,
control, kJ/kg H 2 0 removed
kJ/kg H2 0
Year removed Bin 1 Bin 2 Bin 3
1984 1980 3808 3120 3830
2914 4069 3976 2575
1866 2291 2529
4510
Average 2253 3389 3534 3202
1985 3495 3912 9850 7976
2369 4411 2989
3056 5644 4242
2629
Average 2887 4656 5694 7976

energy requirements in fixed bed and cross flow rice driers are reported as
3.96 and 5.31 MJ/kg of water removed, respectively (7). High air tempera-
ture and low air flow increase the energy efficiency of a concurrent flow
rice dryer, whereas the grain flow rate has only a marginal effect (8). An
LSU-type commercial dryer uses about 7.9 MJ/kg of water removed (9),
whereas the energy used in an LSU-type model lab dryer for rice was 16.5
MJ/kg of water removed (10).
The results for a long-term on-the-farm rice-drying study conducted in
southwest Louisiana are shown in Table 1 (4). The 7-year average energy
requirement for on-the-farm drying of long- and medium-grain rice was
3.46 MJ/kg of water removed. However, large variations were observed
from year to year. A low average energy input of 1.9 MJ/kg water removed
and a high average of 4.9 MJ/kg water removed for long-grain rice were
observed. About 90% of the total energy was used in heating the ambient
air to 38°C during drying. The drying potential for the system (defined as
the actual water removed during drying compared with the theoretical
maximum possible as calculated from psychometric data) was generally in
the range of 40-80% depending on the time of day and weather conditions.
Higher drying potential was observed for lower ambient temperatures,
typically in the morning, when the air was higher in humidity and required
heating. The average potential was 63%.
A burner control to regulate the fuel flow (modulator) as well turn it on
or off to maintain a stable drying air temperature in the plenum was evalu-
ated (4). The average energy usage (Table 2) with such a modulating con-
280 VERMA

trol (2.40 MJ/kg water removed) was about 28% lower than the conven-
tional burners (3.32 MJ/kg water removed) during 1984 to 1986. Energy
savings of as high as 48% were reported in 1985 when prolonged periods of
rainy weather occurred during the drying tests. No adverse effect on rice
milling quality was observed during these drying tests.

III. SOLAR ENERGY USE FOR RICE DRYING

Solar heat can be harnessed to heat ambient air for low-temperature drying
with a relatively simple system. The purpose of solar collectors used in
drying is to intercept solar radiation, convert it to thermal energy, and
transfer this energy to air used for drying or heat storage. Farms have a
practical advantage for the use of solar energy for both drying and space
heating because they have ample area for solar collectors.
One of several possible uses of solar energy in agricultural production
systems is for drying rough rice. The slow drying required for on-the-farm
drying of rough rice has the advantage of using a long period for collection
of solar energy. Low to moderate temperature rise solar energy systems,
which are more efficient than high temperature rise systems, can provide a
large percentage of the energy requirements normally met by using high-
quality fossil fuels. Simple and inexpensive solar collectors can add enough
heat to lower the relative humidity of the drying air sufficiently to dry the
rice crop to a moisture level safe for storage and marketing.

A. Equipment
Various physical configurations for solar collectors are shown in Figure 1.
The solar energy collector can be either a flat plate type or a concentrating
type. A flat plate collector with reflectors works as both of the above.
Typically, in a flat covered plate collector, an insulated black surface acts as
the solar energy absorber and a transparent cover or glazing above it, which
reduces heat losses from the absorber, is used to form the heated space
through which air is circulated. The bare plate collector does not have a
glazing above it, which causes a loss of absorbed energy to the wind, but has
a back plate below it to form the channel for airflow. An ideal solar collec-
tor absorber should: (a) absorb a high percentage of incident solar radiation
(emissivity value close to 1), (b) have minimum losses to its surroundings,
and (c) efficiently transfer the absorbed energy to the circulating air stream.
The glazing layer reduces convection heat losses, is transparent to short-
wave solar radiation, and reduces long-wave radiation heat loss from the
absorber. A covered plate collector is more efficient than a bare plate
collector. A suspended plate collector is more efficient than the covered
plate collector. It has the absorber in the middle with the transparent layer
NEW DRYING METHODS 281

[ Sola.r Ra.dia."tion
~Absorber

~Air-Flow

<...-----J ~ Ba.ck Plo.te

A. Bare Plate Collector.

Tro.nspo.rent Cover

B. Covered Plate Collector.

F
~Air-Flow
...-----:=
Sola.r Ra.dia."tion
Tro.nspa.rent
~Absorber
Cover

~Ba.ck Pla.te
C. Covered, Suspended Plate Collector,
Airflow tmder absorber.
~ Sola.r Ra.dia.tion
~ Tra.nspa.rent Cover

~ ~Aiosorloer
~Air-Flow

~Ba.ck Pla.te

D. Covered, Suspended Plate Collector,

Airfl~v on both sides of absorber.
Figure I Different physical arrangements for air-type flat plate solar collectors.
(From Ref. 12.)

above and back plate below it. It can have airflow either under the absorber
or on both sides of the absorber (11). Concentrating collectors are not
necessary for low-temperature rice drying. Plans are available from local
agricultural extension services or the Midwest Plan Service for wrap-around
solar collectors (Figs. 2, 3) on a grain bin or wall of a farm building or for a
portable collector (12). Any south-facing wall or roof can be used, or the
collector can be placed on the ground. Such a simple collector can capture
as much as 70% of the solar energy incident on it to heat the drying air.
282 VERMA

Sliding Vent Door

Figure 2 Schematic diagram of a wrap-around solar collector attached to a dry-

ing bin. (From Ref. 12.)

B. Drying Conditions
Most of the solar grain drying research has been conducted with corn in the
midwestern United States. The major differences between rough rice and
corn for solar drying are the ambient weather conditions during the drying
season and the vulnerability of rough rice to fissuring and subsequent
breakage, resulting in lower head rice yields and reduced market value.
Results from the midwestern states cannot be directly transferred for use in
NEW DRYING METHODS 283

Flo.shlng

Po.nel Mo. terlo.l

Spo.cer Block

F"oo.M filler

F"RP Cover. corrugo. tea

Bin 1./o.ll, po.1ntea blo.ck -----<1

t• 1./ooa Screws. l'o.c.

3/8' Bolt

Outer rmg of two lx2s

continuous o.rouna collector
Fo.sten spo.cers with
7a rlng-sho.nk no.lls.

Block 2'x (width required)

X 8' long-------..,..___

Vertlco.l Section Horlzonto.l Section

Figure 3 Vertical and horizontal cross sections of the wrap-around solar collector
shown in Figure 2. (From Ref. 12.)

the humid South. Although the midwestern systems can be used under the
more humid and warm conditions, the airflow rates must be adjusted to the
local weather conditions.
A solar collector can provide enough supplemental heat for low-
temperature drying (13). The basic principles of solar energy in drying and
curing, which are applied in different types of practical solar drying sys-
tems, are discussed by Huang (14). The technical feasibility of solar energy
284 VERMA

Collector

Drytng Bin Rock Becl

A
Figure 4 Two configurations for drying rice using a rock bed as a heat sink for
storing solar energy. (A) A single air flow path for both day and night operation.
(B) A dual air flow path. For daytime operation, dampers 1 and 3 are open. For
nighttime operation, damper 2 is open and dampers 1 and 3 are closed. (From Ref.
15.)

for providing a significant percentage of the heat energy for drying has
been demonstrated for various agricultural commodities. However, eco-
nomic considerations and reliability data on specific systems and their com-
ponents have limited the widespread adaptation of solar energy for drying.
Calderwood (15) evaluated three methods using three types of solar
collectors for applying solar heat to dry rough rice in metal bins. These
methods were (a) solar heat applied directly to a stationary bed of rice, (b)
solar heat applied directly to rice displaced by a stirring auger, and (c)
stored solar heat from a rock bed applied to rice displaced by a stirring
auger. Figure 4 shows schematics of two rice-drying systems using a heat-
sink (rock bed) to store solar energy. System A uses only one airflow path.
Ambient air is heated in a solar collector and passed through a rock bed
and rice storage bin connected in series. When the air exiting the solar
collector is hotter than the rock bed (daytime conditions), the air heats the
NEW DRYING METHODS 285

Do.ytiMe Air Po.ths - -

NighttiMe Air Po.th -----

Dry;ng Bin Rock Bea

B

rock bed before it passes through the rice bin. When the air exiting the
solar collector is cooler than the rock bed (nighttime conditions), the air
absorbs energy from the rock bed. Schematic B shows two airflow paths. In
the daytime, ambient air is used unheated to dry the rice and the solar-
heated air is used to store energy in the rock bed. At night, ambient air is
heated by the rock bed before it passes through the rice bin, and the solar
collector is shut down.
The above solar methods were compared with rice dried with unheated
air. It was reported that drying time and fan energy for drying were lower
for solar-heated air than the unheated air. The application of solar heat
directly to the bed of rice was the simplest of the techniques evaluated for
drying rough rice in depths of up to 2.4 m with an average daily solar heat
input of up to 2.2 kWh/m3 of rice under southeast Texas conditions. The
additional investment necessary for storing solar heat in a rock bed and
associated control system was not justified.
Rice drying using solar energy was conducted using systems with (a)
flatbed-type forced air dryer in a greenhouse, (b) a circulation-type dryer in
286 VERMA

a dryer roofhouse, and (c) a ground-level stirrer in a semi-circular solar

energy collection house (16). The second system performed the best out of
the three evaluated in this study. It had a higher equipment cost than the
other two but required less floor area. This cost can be further reduced by
modifying an existing on the farm circulation-type dryer and using a farm
building, such as a shop, for the solar energy collection house.
A solar rice dryer designed for natural convective heat transfer, which
utilized a layer of burned rice husk on the ground as an absorber and
insulator (Fig. 5), resulted in enhanced thermal energy collection (17). In
this dryer, the rough rice to be dried does not remain in contact with the
ground. This small prototype flatbed dryer with a simple air heater made
from burned rice husk and clear plastic sheet was successfully used to dry
rice during the wet harvest season in Thailand. Rice was dried in depths of
up to 150 mm in 2-3 days with no adverse effect on milling quality. The
suggested dryer capacity is 80 kg/m2 of bed.
The satisfactory drying of rice with heat produced by solar energy has
also been reported by Mayeux et al. (18). The recommendation is to use
about 16 cm 2 of bare plate rooftop collector area for every m3 of rice to be
dried. The air velocity used for drying in the bin should be limited to 10 m/
sec by adjusting the depth and width of the collector plenum, and an
airflow rate of 2 m 3/min per m3 of rough rice to be dried should be main-
tained. A supplemental heat source for prolonged inclement weather is,
however, recommended.
Supplemental heat is normally used in on-the-farm drying during peri-
ods of high relative humidity or cool temperatures to accelerate the drying
process. The ambient air temperature in the southern rice-growing states of
Texas, Louisiana, Arkansas, and Mississippi is typically above 27°C, and
the relative humidity is above 65% during the rice-drying months. Supple-
mental heat to reduce the relative humidity should not increase the drying
air temperature above 38°C to keep from overdrying and fissuring the rice
and yet provide an acceptable drying rate. The challenge with a solar
energy-supplemented rice dryer is to limit the temperature rise of the
ambient air in order to prevent rice overdrying resulting in reduced head
yields. This is especially true on clear days, because, when the solar radia-
tion is the greatest, the ambient relative humidity is usually less than 65%.
Solar heating of the air would further reduce this humidity and may result
in overdrying, especially in the lower rice layers in the bin, if the tempera-
ture rise is above about 38°C. Temperatures well above this may cause
faster than desirable drying and large moisture gradients in the grain,
resulting in fissuring and subsequent breakage. A more complete discus-
sion of the rice moisture gradients and their effect on kernel breakage is
presented in Chapter 15. This confirms the need for gradual and slow
 zt!j
~
~Mney t:::l

~
~
BlQCk plQstic
sheets ~
t!j

~
0
t:::l
r;r.,

Door

Collector 'Woven
BQMioOO

Air Burnt rice husks

Inlet

Figure 5 Solar-heated, natural convection rice dryer that uses burnt rice husks as a radiant Nl
00
energy absorber. (From Ref. 17.) ---1
288 VERMA

moisture removal from rough rice, especially at moisture levels below

about 16% (w.b).
A controller using a thermostat and a humidistat is commonly used in
on-the-farm drying systems to accomplish the drying management pro-
gram described above. A properly designed and operating control to limit
this temperature rise but maintain the airflow can virtually eliminate· this
problem and maintain rice head yields. Such a drying system in combina-
tion with a conventional natural gas or LP-gas system for night or pro-
longed cloudy spells can provide drying as fast as a conventional gas
system without adverse quality effects and keep the drying energy costs to
a minimum.
The above research has shown that grain drying in humid conditions
using solar energy is possible, preferably with a humidistat controller for
the blower to enable 24-hour operation (19). Drying in depths not to
exceed 1 m is recommended with adequate airflow for satisfactory drying,
and a backup system is recommended for periods of bad weather.

Iv. BIOMASS AND RESIDUES FOR DRYING

Agricultural crop residues are produced as by-products in normal crop
production systems and are not used beneficially at present. Crop residues
and biomass are among the cheapest, most readily available energy sources
on the farm. These renewable fuels provide an option for partially meeting
future energy needs.
Biomass from agriculture as an alternative energy source is predicted to
supply up to 5% of U.S. energy needs by the year 2000. An estimated 365
x 109 kg of residue is produced each year from 10 major crops in the United
States. The potential of biomass and residues to supplement energy needs
in agriculture and other areas is well documented (12, 20). The residues
recovered from one hectare of corn should provide enough heat energy to
dry corn from 10 or more hectares (21). A similar situation also exists for
crops such as soybeans, wheat, and rice. Baling and storage of rice straw at
different moisture levels has been reported (22, 23).

A. Equipment
Direct combustion of crop residues and biomass provides the greatest utili-
zation of stored energy in biomass to heat air for grain drying and is
presently the most widely used technique. Successful systems have primar-
ily used wood or forest products as fuel (24). Other biomass fuels could be
viable alternatives if appropriate technology and equipment were available
to utilize the biomass energy. Equipment suitable for one biomass may not
NEW DRYING METHODS 289

be as suitable for other materials. Round balers or stackers are available on

most farms and can be used to harvest various types of crop residues and
straw. These packages of crop residue are easy and efficient to handle and
store. Dry residue in a typical bale of 500 kg can have 9000 MJ of thermal
energy, equivalent to over 1.45 barrel of crude oil. The usable energy from
a bale depends on the moisture content, heat of combustion of the residue,
and the furnace efficiency.
Furnaces utilizing biomass and crop residue for grain drying applications
are either direct fired or indirect fired. The former type usually uses gasifi-
cation, whereas the latter uses a heat exchanger. All furnaces utilize either
updraft, downdraft, or crossdraft gasifiers. The biomass fuel typically un-
dergoes drying, pyrolysis, combustion, and reduction in conversion to ther-
mal energy (25). Several biomass and residue-fired burners for drying
grains, direct or indirect fired, have been evaluated in the recent past to
reduce fossil energy consumption in crop drying (26-35). At least one
furnace of each type is commercially available in the United States and has
been used to dry com and a few other grains.
An indirect-fired biomass furnace was evaluated with crop residues for
furnace efficiency (36). Large round bales of cotton stalks, com stover, and
soybean residues burned satisfactorily in the tests. Heat energy available
for grain drying ranged from 45 to 64% of the biomass heat energy deter-
mined from bomb calorimetry tests. The same furnace was utilized for
forage dehydration using cotton stalks, small grain straw, and Bermuda
grass in large round bales (37). Cotton stalks were the most efficient.
Bermuda grass was found to be a good source of heat if the bales were kept
dry or burned with another biomass source.
One of the major concerns in grain drying with heat from direct-fired
furnaces using biomass is the contaminant residue on the grain from the
combustion gases. Several studies have been conducted that quantify the
contaminants and their effects on the grain (primarily com) and air pollu-
tion (38-43). In most of these studies, preliminary results showed that the
grain was of acceptable quality.

B. Rough Rice Drying

Rough rice has been successfully dried in a bin from 21 to 12.7% (w.b.)
moisture using an indirect fired furnace with a crossdraft gasifier (44). The
furnace used a heat exchanger to heat the air blown through the rice. This
heat exchanger was located around a batch-type combustion chamber in
which a fixed supply of biomass was allowed to bum completely before
another supply of biomass was placed in the furnace. The heated air was
delivered to the floor plenum of a batch-type bin dryer. The air tempera-
290 VERMA

ture entering the bin dryer was controlled by an adjustable thermostat that
regulated the on-off mode of the conbustion air fan. This fan supplied
oxygen to the combustion chamber and controlled the burning rate of the
biomass to maintain the desired temperature for drying. The furnace was
rated to raise the drying air temperature an average of 25°C above ambient
conditions but was used in the rice-drying tests with plenum air below 40°C
to maintain the milling quality (Table 3). Average furnace efficiency was
41% with bales of rice straw and Bermuda grass as fuel. Drying efficiency
of over 60% was achieved with an air temperature rise of at least 5°C. More
energy was required in these tests to power the fan as the heat output by
the biomass was irregular, forcing longer drying times. Rice quality, as
determined by the mill and head yields of samples from the three tests, was
not adversely affected by this biomass drying system as compared with the
rice dried in a conventional in-bin drying system.
A rice husk-fired rotary dryer (Fig. 6) was designed and evaluated for
rapid drying using conduction heating of rough rice at heating surface
temperatures of 100 to 180°C [45). This system was also evaluated for
insect disinfestation of the high moisture rough rice. Drying was termi-
nated at about 16% (w.b.) moisture, and the reduction in milling yields
was non-significant. Such a system may not be easily adaptable for on-the-
farm drying unless the rice husk is readily available in ample quantities.
A rice dryer (Fig. 7) with energy derived from a rice husk gasifier was
developed and tested at the International Rice Research Institute (IRRI).
Unlike the rotary dryer system described above, this system uses combustion

Table 3 Drying Rough Rice with Energy Generated by Burning Rice Straw and
Dallis Bermudgrass in a Stormor Furnace

Test 1 Test 2 Test 3

Dates of drying tests 7/29-7/31 9/30-10/4 10/29-11/3
Wet weight ofrice (kg) 17,2S9 16,S1S 19,232
Initial rice moisture(% w.b.) 19.8 20.1 22.8
Final rice moisture(% w.b.) 13.8 12.1 12.2
Total heat in biomass (MJ) 4,648 24,130 32,S40
Total plenum heat gain (MJ) S,424 18,96S 27,919
Water evaporated in drying (kg) 1,202 1,S03 2,322
Furnace efficiency (%) 2S.8 47.6 S0.2
Total energy used (MJ/kg of H 20) S.41 1S.84 13.S1
Energy from furnace(% of total) 83.4 79.7 89.0
Drying efficiency with tl T <:: S°C (%) 61.0 S9.6 60.2
Drying efficiency with tl T < soc (%) 4S.6 48.7 S3.0
Source: Ref. 44.
 ~ ~
[J
6 Roto.ry dryPr

[J
l Rlct1 husk hopp•r
2 Fur-nace 7 Perforated roto.ry cool<'r
2o illdJ. Inclined gro.te 8 Electric I'IOtor
2b illdJ. osh trop 9 Chain drlv•
2c illsh collector tO Boll brorlng
2d Brick wo.ll II Po.ddy outl•t gUidP
2• Husk flow t1quo.llz1Pr 12 Anglo• bar frnMOP
3 Po.ddy Inlet hopper 13 Hrllco.l bo.ffln
4 Electro-I'IQgn et vlbro.tor 14 Ch1Mnt1y
5 Dry<'r brick housing 15 rlong" JOint
14
14

:5

II

Figure 6 A schematic diagram of a rotary dryer/cooler system that uses rice husks for fuel. Rice is heated
by conduction at surface temperatur es between 100°C and 180°C for short times. (From Ref. 45.)

~
~
292 VERMA

Flue ga.s

Pa.ddy
t
Hopper

Air Hull

t t
Rea.ctor Dryer

Ga.sifier
Rota.ry

Figure 7 Schematic diagram of the DA-IRRI reactor dryer with rice hull gasifier.
(From Ref. 46.)

gases from the burner to directly heat the rice by forced convection. Short-
time, single-pass exposures of the rice to temperatures in the range of 80-
950C dried the rice to 16-19% (w.b.) moisture without affecting the milling
quality (46). This dryer was developed for Asian regions for farmer coopera-
tives and rice millers when sun drying is not possible due to adverse weather
conditions. This system is also not entirely suitable for on-the-farm drying.
A warehouse dryer with a center-tube furnace fueled with combustible
agricultural residues and by-products such as rice and coconut husks, and a
vortex wind machine as the power source for air movement, was developed
and tested for drying rice at IRRI ( 47). The system was found suitable for
drying and storage, and the airflow rate was found to be the critical factor
in drying of the rice.

V. COMBINED SOLAR AND BIOMASS

A grain dryer using both biomass and solar energy sources has also been
proposed (48). This system for on-the-farm rice drying, to prevent quality
loss, enabled the drying to be spread over a longer time period with com-
bined solar/thermal energy as compared with solar energy alone. Biomass,
crop residues, and husk can be utilized for periods of no solar energy
availability. The use of solar and/or husk/biomass briquette dryers for
NEW DRYING METHODS 293

drying rice at the farm level has also been suggested for the rural situations
in the Far East, where dependence on sun drying at roadside areas and
other floor space is predominant (49). Both these options provided low-
temperature drying and can be adapted to in-bin drying for maintaining
milling quality of the rice. The briquette was able to sustain burning for
over 2 hours.

VI. OTHER NONCONVENTIONAL METHODS

The energy utilization and drying characteristics of low-temperature, solar-
assisted, heat pump, and combination higMow-temperature drying for
shelled corn have been discussed (50). The energy savings are estimated at
about 17, 48, and 54% for in-bin, low-temperature drying of a ton of corn
at an initial moisture content of 26% (w.b.), with conventional, solar, and
heat pump drying, respectively, over in-bin drying with a 20°C air tempera-
ture rise. A simulation of aeration of rice using cooled air was reported to
maintain moisture below 13% (w.b.), to lower rice temperature below
15.6°C, and to be economically feasible in southwest Texas (51). Cooling of
rice in hot and humid conditions can result in savings due to lower drying
costs, less grain shrinkage, and lower discount due to quality deterioration
than conventional drying.

VII. CONCLUSIONS
Under humid conditions, solar heat can be harnessed to assist low-
temperature rice drying with a relatively simple system. Low to moderate
temperature rise solar energy systems, which are more efficient than high
temperature rise systems, are well suited for rice drying. The slower drying
that must be used for on-the-farm drying of rough rice provides a longer
period for collecting the solar energy. A 24-hour humidistat controller for
the blower is more efficient than completely shutting down at night. Drying
in depths not to exceed 1 m is recommended with adequate airflow for
satisfactory drying. A backup heating system is recommended for periods
of bad weather.
Crop residues and biomass are among the cheapest and most readily
available energy sources on the farm. Direct combustion of crop residues
and biomass provides the greatest utilization of stored energy in biomass to
heat air that can be used for grain drying. Rough rice can be dried in a bin
using a biomass furnace with bales of rice straw and low-quality grasses as
fuel, without sacrificing milling yields. Several other drying systems using
combustible agricultural residues and by-products such as rice and coconut
husks are also possible for rough rice drying. Several of these systems
294 VERMA

provide low-temperature drying and can be adapted to in-bin drying for

maintaining milling quality of the rice. In on-the farm bin drying systems,
energy usage with a modulating control can be significantly reduced com-
pared with conventional burners.
Drying using both biomass and solar energy sources is also possible, with
biomass, crop residues, and husk being utilized for periods of no or low
solar energy availability. This can provide low-temperature drying and can
be adapted to in-bin drying for maintaining milling quality of the rice.

REFERENCES
1. Federal Energy Administration. (1976). Energy and U.S. agriculture: 1974
data base. FEA/D-76/459. Federal Energy Administration, USDA, Washing-
ton, D.C.
2. U.S. Department of Agriculture. (1980). Energy and U.S. agriculture: 1974
and 1978. USDA, Economics, Statistics, and Cooperative Services. Statistical
Bulletin No. 632, Washington, D.C.
3. Kunze, 0. R., and Calderwood, D. L. (1980). Systems for drying of rice,
Drying and Storage of Agricultural Crops (C. W. Hall, ed.), AVI Publishing
Co., Westport, CT, p. 225.
4. Verma, L. R., and Jacobsen, L. A. (1987). On-farm rice drying energy use.
App. Engr. Ag., 3(1):79.
5. Fluck, R. C., and Baird, C. D. (1980). Agricultural Energetics. AVI Publish-
ing Co., Westport, CT, p. 113.
6. Maddex, R. L., and Bakker-Arkema, F. W. (1978). Reducing energy require-
ments for harvesting, drying, and storing grain. Energy Fact No. 18, Extension
Bulletin E-1168, Michigan State University, E. Lansing, MI.
7. Singh, R. P., Chandra, P. K., and Zuritz, C. A. (1981). Energy accounting of a
rice dryer. ASAE Paper No. 81-6017. American Society of Agricultural Engi-
neers, St. Joseph, MI.
8. Walker, L. P., and Bakker-Arkema, F. W. (1978). Energy efficiency in concur-
rent flow rice drying. ASAE Paper No. 78-3540. American Society of Agricul-
tural Engineers, St. Joseph, MI.
9. Velupillai, L. (1977). Report on the rice mill complex at Morkeytiya. RPDC
No. R1/77. Rice Processing Development Center, Anuradhapura, Sri Lanka.
10. Verma, L. R., and Thomas, M.D. (1981). Laboratory measurement of energy
use in rice processing. ASAE Paper No. SWR-81-402. American Society of
Agricultural Engineers, St. Joseph, MI.
11. Midwest Plan Service. (1980). Low Temperature and Solar Grain Drying
Handbook. Publication MWPS-22. Midwest Plan Service, Iowa State Univer-
sity, Ames, lA, p. 31.
12. Midwest Plan Service. (1980). Low Temperature and Solar Grain Drying
Handbook. Publication MWPS-22. Midwest Plan Service, Iowa State Univer-
sity, Ames, lA, p. 63.
NEW DRYING METHODS 295

13. Ritchie, J.D. (1983}. Sourcebook for Farm Energy Alternatives. McGraw-Hill
Book Co., New York, p. 11-1.
14. Huang, B. K. (1986}. Use of solar energy for drying and curing, Alternative
Energy in Agriculture, Vol. I (D. Y. Goswami, ed.), CRC Press, Inc., Boca
Raton, FL, p. 109.
15. Calderwood, D. L. (1979}. Rice drying with solar heat. Trans. ASAE,
22(6}:1435.
16. Nakagawa, K., Horibe, K., and Oshita, S. (1987}. Utilization of solar heat for
grain dryer, International Symposiumon Agricultural Mechanization and Inter-
national Cooperation in High Technology Era. Japanese Society of Agricul-
tural Machinery, Tokyo, Japan, p. 63.
17. Exell, R. H. B. (1980}. Basic design theory for a simple solar rice dryer. Ren.
En. Rvw. J., I(2):1.
18. Mayeux, M. M., Wratten, F. T., and Pringle, L. (1979). Conditioning grain
with solar energy. Louisiana Ag., 22(4):4.
19. Chau, K. V., and Baird, C. D. (1981}. Solar grain drying under hot and humid
conditions, Agricultural Energy, Vol. 1. ASAE Publication 3-81. American
Society of Agricultural Engineers, St. Joseph, MI, p. 31.
20. Stout, B. A. (1984). Energy Use and Management in Agriculture. Breton
Publishers, North Scituate, MA, p. 172.
21. Morey, R. V., and Thimsen, D.P. (1981}. Combustion of crop residues to dry
corn, Agricultural Energy, Vol. 1. ASAE Publication 3-81. American Society
of Agricultural Engineers, St. Joseph, MI, p. 142.
22. Dobie, J. B., Miller, G. E., and Parsons, P. S. (1977). Management of rice
straw for utilization. Tran. ASAE, 20(6):1022.
23. Dobie, J. B., and Haq, A. (1980}. Outside storage of baled rice straw. Tran.
ASAE, 23(4}:990.
24. Hiler, E. A., and Stout, B. A. (1985). Biomass Energy A Monograph. TEES
Monograph Series. Texas A&M University Press, College Station, TX.
25. Rajvanshi, A. K. (1986). Biomass gasification, Alternative Energy in Agri-
culture, Vol. II (D. Y. Goswami, ed.), CRC Press, Inc., Boca Raton, FL,
p. 83.
26. Payne, F. A., and Ross, I. J. (1979). Crop residue as an energy source. ASAE
Paper No. SER 79-217. American Society of Agricultural Engineers, St. Jo-
seph, MI.
27. Claar, P. W., Buchele, W. F., and Marley, S. J. (1981}. Development of a
concentric-vortex agricultural residue furnace, Agricultural Energy, Vol. 2.
ASAE Publication 3-81. American Society of Agricultural Engineers, St. Jo-
seph, MI, p. 349.
28. Kajewski, A., Marley, S. J., and Buchele, W. F. (1977). Drying corn with a
crop residue fired furnace. ASAE Paper No. 77-3525. American Society of
Agricultural Engineers, St. Joseph, MI.
29. Sukup, C. E., Bern, C. J., and Buchele, W. F. (1982}. Performance of A
biomass furnace for grain drying. ASAE Paper No. 82-3524. American Soci-
ety of Agricultural Engineers, St. Joseph, MI.
296 VERMA

30. LePori, W. A., Anthony, R. G., Lalk, T. R., and Craig, J. D. (1981).
Fluidized-bed combustion and gasification of biomass, Agricultural Energy,
Vol. 2. ASAE Publication 3-81. American Society of Agricultural Engineers,
St. Joseph, MI, p. 330.
31. Raman, K. P., Walawender, W. P., Shimizu, Y., andFan,L. T. (1981). Gasifica-
tion of corn stover in a fluidized bed, Agricultural Energy, vol. 2. ASAE Publica-
tion 3-81. American Society of Agricultural Engineers, St. Joseph, MI, p. 335.
32. Peart, R. M., Zink, H. R., and Doering, 0. C. (1981). Corn cob gasification
for corn drying, Agricultural Energy, Vol. 2. ASAE Publication 3-81. Ameri-
can Society of Agricultural Engineers, St. Joseph, MI, p. 338.
33. Barrett, J. R., and Jacko, R. B. (1984). Corncob channel gasifier design,
performance and safety, Proceedings of the Fourth Annual Solar & Biomass
Workshop, Southern Agricultural Energy Center, Tifton, Georgia, p. 108.
34. Rice, C. E., and Brown, R. H. (1979). Solar heat for broiler production from
a biomass burner system. ASAE Paper No. SER 79-039. American Society of
Agricultural Engineers, St. Joseph, MI.
35. Hill, R. C. (1979). Design, construction and performance of stick-wood fired
furnace for residential and small commercial application. U.S. Department of
Energy, EC 77-S-02-45, Washington, D.C.
36. Summer, H. R., Sumner, P. E., Hammond, C. W., and Monroe, G. E. (1981).
Energy available from biomass for grain drying. ASAE Paper No. 81-3014.
American Society of Agricultural Engineers, St. Joseph, MI.
37. Hellwig, R. E., Sumner, H. R., and Monroe, G. E. (1983). Systems for drying
forages and grain with biomass products, Proceedings of the Third Annual
Solar & Biomass Workshop, Southern Agricultural Energy Center, Tifton,
Georgia, p. 250.
38. Jacko, R. B., Foster, G. H., and Barrett, J. R. (1982). Contaminant resi-
dues on corn dried by combustion gases from biomass furnaces. ASAE
Paper No. 82-3523. American Society of Agricultural Engineers, St. Jo-
seph, MI.
39. Sizemore, S. W., Loewer, 0. J., Taraba, J. L., Ross, I. J., and White, G. M.
(1983). Retention of gasification-combustion products by corn. ASAE Paper
No. 83-3005. American Society of Agricultural Engineers, St. Joseph, MI.
40. Jacko, R. B. (1984). Grain contamination and air pollution from residue
furnaces, Proceedings ofthe Fourth Annual Solar & Biomass Workshop, South-
ern Agricultural Energy Center, Tifton, Georgia, p. 104.
41. Mwaura, E. N., Bakker-Arkema, F. W., Van Ee, G. R., and Braselton, W. E.
(1983). Grain contamination in drying by direct biomass heating. ASAE Pa-
per No. 83-3512. American Society of Agricultural Engineers, St. Joseph, MI.
42. Anderson, M. E., Bern, C. J., and Baker, J. L. (1983). Corn drying with
biomass combustion products. ASAE Paper No. 83-3005. American Society
of Agricultural Engineers, St. Joseph, MI.
43. Williams, R. S., Barrett, J. R., and Field, W. E. (1983). Emissions from
biomass furnaces. ASAE Paper No. 83-3547. American Society of Agricul-
tural Engineers, St. Joseph, MI.
NEW DRYING METHODS 297

44. Verma, L. R. (1990). Biomass energy for rice drying. Biomass, 23(1):13.
45. Jindal, V. K., and Obaldo, L. G. (1986). Rice-husk fired rotary dryer for
disinfestation and rapid drying of paddy, Grain Protection in Postharvest Sys-
tems. Proceedings of the Ninth ASEAN Technical Seminar on Grain Post-
harvest Technology (B. M. de Mesa, ed.), ASEAN Crops Postharvest Pro-
gram, Manila, Philippines, p. 42.
46. Stickney, R. E., Piamonte, V. N., and Belionio, A. T. (1988). DA-IRRI
rotary paddy dryer with rice hull gasifier, Advances in Grain Postharvest Tech-
nology Generation and Utilization. Proceedings of the Eleventh ASEAN Tech-
nical Seminar on Grain Postharvest Technology (J. 0. Naewbanij, ed.),
ASEAN Grain Postharvest Program, Bangkok, Thailand, p. 46.
47. Jeon, Y. W., Bockhop, C. W., and Halos, L. S. (1984). A warehouse dryer
using non-conventional energy sources, Health and Ecology in Grain Post-
harvest Technology. Proceedings of the Seventh ASEAN Technical Seminar
on Grain Postharvest Technology (R. L. Semple and A. S. Frio, eds.),
ASEAN Crops Postharvest Program, Manila, Philippines, p. 241.
48. Kamaruddin, A., Tahir, R., and Sutrisno, R. (1987). Performance of a dual
energy source grain dryer, International Symposium on Agricultural Mechani-
zation and International Cooperation in High Technology Era. Japanese Soci-
ety of Agricultural Machinery, Tokyo, p. 71.
49. Soedjatmiko, and Madjid, R. (1987). Grain drying at the small scale farm-
the Indonesia case, International Symposium on Agricultural Mechanization
and International Cooperation in High Technology Era. Japanese Society of
Agricultural Machinery, Tokyo, p. 272.
50. Peart, R. M., Brook, R., and Okos, M. R. (1979). Energy requirements for
various methods of crop drying, Handbook of Energy Utilization in Agricul-
ture (D. Pimental ed.), CRC Press, Inc., Boca Raton, FL, p. 49.
51. Maier, D. E., Weidmayer, K. D., Bakker-Arkema, F. W., and Moreira, R. G.
(1991). Regular and chilled aeration of grains in Texas. ASAE Paper No. 91-
6062. American Society of Agricultural Engineers, St. Joseph, MI.
 14
Microwave-Vacuum Drying
James I. Wadsworth
Agricultural Research Service, U.S. Department of Agriculture,
New Orleans, Louisiana

I. INTRODUCTION
One of the most important economic considerations in rice processing is
the prevention of breakage; broken rice is worth about one-half the value
of head (whole grain) rice. Farmers must harvest their rice before it
completely matures to prevent fissuring of the rice in the field (see Chap-
ter 15). The average moisture content of freshly harvested (green) rice is
usually between 18 and 24% (wet basis). Rice at this high moisture con-
tent is susceptible to the growth of microorganisms, and it must be dried
to less than 14% moisture content to prevent spoilage during storage. To
prevent deterioration, drying should be started within 24 hours of harvest.
Rough rice is the normal form in which rice is dried because it stores
better than brown rice and it is easier to remove the hulls after the rice
has been dried.
Drying can have a profound effect on rice breakage. As a rice kernel
dries, the outer layers shrink, establishing stress within the kernel (1). The
magnitude of this stress depends on the moisture gradient and temperature
gradient within the rice kernel, while the tensile strength of the kernel is a
function of its moisture content and temperature. Ifthe rice is dried improp-
erly, the internal stress in some kernels might exceed their tensile strength,
resulting in cracked kernels that tend to break during milling. Thus, drying
methods are extremely important for the rice industry.

299
300 WADSWORTH

This chapter will first briefly review conventional rice-drying technology

currently being used by the rice industry. Then it will describe in detail a new
method for drying rice using microwave-vacuum technology. The micro-
wave method for drying both freshly harvested green rice and parboiled rice
will be discussed. Finally, the cooking and processing quality of microwave-
vacuum dried rice will be compared with conventionally dried rice.

ll. CONVENTIONAL DRYING

Conventional rice drying has been treated in detail by Steffe et al. (2),
Kunze and Calderwood (3), and Wang and Luh (4). Mossman (5) reviewed
critical areas of rice-drying research. Basically, in conventional rice drying,
air is forced through the rice by fans. Forced-air dryers currently being
used include two types: the deep-bed dryer and the continuous-flow,
heated-air dryer.

A. Deep-Bed Dryers
Deep-bed dryers (Fig. 1) serve both to dry the rice and to store it after
drying without need for transferring the grain during these operations.
These structures are generally used on-farm. The air-distribution systems
commonly used are a perforated floor (Fig. lA) or different arrangements
of ducts and laterals (Fig. lB). Airflow direction is usually upward. A zone
of dried rice begins at the bottom and moves upward through the bin. The
difficulty with this type of drying is that bad weather may unduly prolong
the drying of the rice, particularly at the top of the bin, until the rice is
damaged. During prolonged wet or humid weather, supplemental heat can

Figure 1 Bin-type deep-bed dryers. (A) False perforated floor; (B) main duct
and laterals.
MICROWAVE-VACUUM DRYING 301

improve the effectiveness of deep-bed drying. However, supplemental heat

may result in overdrying the bottom layers, with consequent loss of quality.
The drawback of supplemental heat in deep-bed dryers has been largely
obviated by the development of an auguring system, extending from the
top to the bottom of the bin, programmed to move around the bin to
ensure complete mixing and uniform moisture content of the rice. With
adequate mixing, a greater amount of supplemental heat can be safely
applied to speed drying. Both natural gas and LPG are used as fuels.
Combustion products are mixed with the drying air. For a detailed discus-
sion of on-the-farm drying, see Chapter 13.

B. Continuous-Flow, Heated-Air Dryers

The majority of the U.S. rice crop is dried at commercial installations using
continuous-flow, heated-air dryers (Fig. 2). Natural gas and heating oil are
the preferred fuels; the combustion products are mixed with the drying air.
Commercial dryers are of the columnar type, in which rice flows by gravity
from the top to the bottom of the dryer, where it is conveyed to receiving

Adjustable Speed
DischarQe Rail

Screw Conveyor
Discharge

A B

Figure 2 Continuous flow, heated-air columnar dryers. (A) Nonmixing type; (B)
mixing type.
302 WADSWORTH

bins. Feed rolls or other devices at the bottom of the dryer are used to
regulate the rice flow. Columnar dryers are classified as nonmixing or
mixing.
Figure 2A shows a cross section of a nonmixing dryer. Rice flows be-
tween two parallel screens 15-23 em apart. Heated air flows horizontally
through the screens and rice, and the rice kernels flow straight downward.
The rice closest to the screen where the air enters is always exposed to
hotter, drier air than the rice on the opposite side. Hence, the rice on one
side tends to be dried more than that on the other side. Figure 2B shows a
mixing columnar dryer. Grain initially on the inside surface may move to
the outside surface in falling from one baffle to the next, producing uniform
drying. In nonmixing dryers air temperatures up to 54°C are commonly
used. Airflow rates vary from 112 to 262 m3/min/t. The mixing dryer uses
lower air velocities and higher temperatures than the nonmixing type.
Typical are air temperatures up to 66°C and airflow rates of 44-97 m3/min/t.
Rice is usually not dried from harvest moisture content to storage mois-
ture (14%) in a single pass through a continuous dryer. To minimize break-
age due to the development of excessive internal moisture gradient stress in
individual grains, only 2-3% moisture is removed in each pass through the
dryer. Between passes the rice is conveyed to bins for tempering. Temper-
ing allows moisture migration from the interior of a rice kernel to the
exterior, thereby relieving the internal stress. It is customary to aerate the
tempering bin with low airflow rates of ambient air to cool the rice and also
remove additional moisture.

C. Mechanism of Conventional Drying

The mechanism for conventional hot-air grain drying is shown in Figure 3.
An energy source, such as natural gas, is used to heat air. The heated air is
then used to transfer energy to the surface of the grain kernels by convec-
tion. The transfer of energy within the kernel is by conduction. Water is
desorbed into the drying gases at the surface of the kernel. A moisture
gradient is established within the kernel, and liquid water molecules diffuse
to the surface. The temperature and moisture profiles within the kernel
determine the rate of liquid water diffusion to the surface. The air tempera-
ture, humidity, and water concentration at the grain surface determine the
rate of water desorption from the grain. As long as the rate of water
evaporation from the surface does not exceed the rate of liquid water
diffusion from within the kernel to its surface, the desorption step is rate
limiting and the overall drying rate is constant. However, when the rate of
diffusion falls below the rate of desorption, the drying process becomes
diffusion limited and enters the falling rate period. A constant rate period
MICROWAVE-VACUUM DRYING 303

ENERGY TRANSFER
BY CONVECTION

EVAPORATION
FROM SURFACE

RICE KERNEL

Figure 3 Mechanism for conventional hot-air drying of rice.

followed by a falling rate period is typical for most conventional grain-

drying operations.
An increase in the drying gas temperature increases the rate of desorption
of water from the surface of the kernel, which results in a greater moisture
gradient inside the kernel. Since a high internal moisture gradient can result
in cracked or broken kernels, there is a practical upper limit to the drying
temperature that can be used. Commercial dryers generally operate at tem-
peratures around 60°C and use multiple passes through the dryer.

III. MICROWAVE-VACUUM DRYING

A new technique for drying rice in a single pass using microwaves as an
ennergy source was reported by Wadsworth (6), Wadsworth and Koltun
(7), and Wadsworth et al. (8). The energy crisis of the late 1970s demon-
strated the need for multiple energy sources. With the continuing political
turmoil in the Middle East, there still exists the threat that supplies of
natural gas and oil might become restricted and more expensive. If such a
circumstance were prolonged, processors would be required to seek alterna-
tive sources of energy for grain drying. It seems likely that electrical power,
generated from coal or nuclear energy, would be a prime alternative.
Microwave-vacuum (MY) drying is one technique that utilizes electrical
power for grain drying.
304 WADSWORTH

A. Mechanism of Microwave Drying

Microwave-vacuum drying (Fig. 4) is fundamentally different from conven-
tional hot-air grain drying. The energy source must be electrical. The elec-
trical energy is converted via a magnetron into electromagnetic radiation in
the microwave frequency range (915 and 2450 mHz are the FCC allowed
frequencies for use in food processing). The microwave energy is transmit-
ted by wave guides into the chamber that contains the rice. Microwaves
penetrate the rice kernel, where they are preferentially absorbed by highly
polar water molecules. In some products during microwave heating, the
internal product temperature can exceed the surface temperature (9). The
high-energy water molecules rapidly diffuse through the rice kernel to the
surface where they desorp into the reduced-pressure atmosphere. In some
cases the water molecules attain a sufficiently high energy level to enable
them to diffuse through and out of the product as vapor.
The reduced pressure in the vacuum chamber lowers the concentration
of water in the air surrounding the grain, thereby increasing the rate at
which the water molecules desorb from the surface of the kernel. For a
given rate of drying, a reduced pressure enables the grain to be dried at a
lower grain temperature than would be required at atmospheric pressure.
Moisture is removed from the chamber by a vacuum pump and condenser.
The temperature of the grain during MV drying can be regulated by the
rate of input of microwave energy and by the vacuum level maintained in
the chamber. During MV drying there will still be a moisture gradient estab-

Figure 4 Mechanism for microwave-vacuum drying of rice.

MICROWAVE-VACUUM DRYING 305

lished within the rice kernel, but the magnitude of the gradient will be less
than that for conventional drying.

B. MV Drying Freshly Harvested Rice

1. Experimental
a. Rice
Tests to obtain definitive information on the suitability of MV technology for
drying freshly harvested green rice were run with six lots of rice. Three Nato
medium-grain lots, two Labelle long-grain lots, and one lot of Starbonnet
long-grain grown in different crop years were evaluated. The rice moisture
contents at time of harvest ranged from 17.5 to 22.0%, wet basis (w.b.).
b. Drying
A batch-operated microwave-vacuum dryer (Fig. 5) was used to assess the
drying characteristics of green rice. A schematic diagram of the MV system

Figure 5 Laboratory model batch-operated microwave-vacuum dryer.

306 WADSWORTH

VACUUM CHAMBER

AIR

BLEED
ROTATING DRUM

ALTERNATE SAMPLE HOLDER

EXHAUST

Figure 6 Schematic diagram of batch operated microwave-vacuum dryer system.

is shown in Figure 6. The dryer operated at a microwave frequency of 2450

MHz. The power was continuously adjustable up to a maximum of 2.5 kW.
Reflected power was measured with a microwave power meter installed in
the wave guide. The vacuum pump/condenser system (condenser tempera-
ture approximately 1-4°C) could maintain pressures as low as 1.3 kPa.
During drying, the experimental rice sample was either continuously tum-
bled in a rotating microwave-transparent drum located in the drying cham-
ber (Fig. 7) or spread on a revolving turntable that was suspended from a
strain gauge (Fig. 8). The turntable had the advantage that continuous
sample weight measurement and infrared temperature measurement (the
infrared sensor was calibrated for rough rice) could be made while the rice
was drying. The rotating drum produced more uniform drying of the rice,
which was important for quality evaluations.
The dryer was operated at three microwave power levels (600, 1200, and
2000 watts) and three vacuum levels for each power level. Initial rice mois-
ture content varied with the particular lot of rice being dried. Drying was
accomplished in a single pass, with the drying times adjusted to achieve the
desired final moisture content. Control samples were shade dried at ap-
proximately 22°C and 60% relative humidity.
MICROWAVE-VACUUM DRYING 307

Figure 7 Rotating drum sample holder. Rice sample is continuously tumbled

during microwave-vacuum drying.

c. Milling
The MV-dried samples were allowed to equilibrate with the milling room
environment (22°C and 60% relative humidity) for two weeks before mill-
ing. Milling moisture content was approximately 12%. Samples were
shelled and milled in accordance with directions in the USDA rice inspec-
tion handbook (10). A well-milled rice was obtained for all samples. The
head rice yield was determined with a sizing device using the plates and
conditions described in the USDA handbook.
308 WADSWORTH

Figure 8 Revolving turntable suspended from a strain gauge inside MV dryer

that enables continuous rice sample weight and temperature measurements to be
made during drying.

2. Results
Figure 9 shows typical weight and temperature changes that occur in 4.54-
kg rice samples dried at three power levels. One interesting feature of the
weight traces was that, after a brief initial lag, the weight loss was essen-
tially linear during the entire drying period. This was not, however, the
traditional constant-rate drying period usually observed with hot-air drying
because the temperature of the sample increased as it dried.
 MICROWAVE-VACUUM DRYING 309

_
- WEIGHT CURVES
100
•
(.)
- - TEMP. CURVES

I&J
....... 1200 a:
80
.....Cit ....
;::)

/
""
---
a:

-- --
/

- -
I&J
....::z:: 5000
600 a..
60 ~
(!)
I&J
....I&J
I&J
~ -J
4500 40 a..
~ ~
a..
~ ""en
""en 4000
0 20 40 60 80 100 120
20

DRYING TIME (min)

Figure 9 TYPical weight and temperature curves for 4.54-kg rice samples dried at
microwave power inputs of 600, 1200, and 2000 watts.

a. Drying Rate
Figure 10 shows the effects of the microwave dryer power level on the
drying rate. The response of drying rate to power level change was linear
over the range investigated. The maximum drying rate shown was equiva-
lent to drying a rice sample from 20% moisture to 12% moisture in 20
minutes (24 percentage points decrease per hour). The fact that the drying
rate was linearly proportional to the power input, even at the very high
rates, indicates that microwave drying was not controlled by the rate of
water diffusion within the grain.
Table 1 presents data for three different rice lots dried to 13% final mois-
ture content using three power levels and three vacuum levels. The initial
moisture contents for Nato 82, Nato 83, and Labelle 83 were 20.1, 18.3, and
19.5%, respectively. Analysis of variance (ANOVA) indicated that there
were significant differences (p = 0.01) in drying rates obtained for the three
different rice lots for all conditions investigated. This may be related to the
initial moisture content of the rice lot rather than the variety or grain type.
As seen in Table 1, the medium-grain Nato 82 lot (20.1% initial moisture)
had the highest drying rates, while Nato 83 lot (18.3% initial moisture) had
the lowest and the long-grain Labelle 83 (19.5% initial moisture) was inter-
mediate. Data from other rice lots have also indicated that rice with higher
initial moisture had a higher initial drying rate and maintained the higher
rate throughout the drying process even as the moisture content decreased.
 w
1-'
Q

Table 1 Drying Rate, Maximum Drying Temperature, and Drying Efficiency for Three Lots of Green Rice Microwave-
Vacuum Dried at Three Power Levels and Three Vacuum Levels for Each Power Level

NAT082 NATO 83 LABELLE 83

Power Pressure Rate Temp. Efficiency Rate Temp Efficiency Rate Temp Efficiency
(kW) (kPa) (g/min/kg) (OC) (g/kJ) (g/min/kg) (OC) (g/kJ) (g/min/kg) ("C) (g/kJ)
0.6 2.7 1.23 52 0.156 1.13 52 0.145 1.19 50 0.149
0.6 13.3 1.10 57 0.149 0.91 58 0.117 0.99 60 0.128
0.6 53.2 0.95 64 0.124 0.80 64 0.105 0.74 67 0.095
1.2 6.7 2.46 70 0.152 2.21 67 0.138 2.23 65 0.139
1.2 13.3 2.44 70 0.152 2.10 66 0.129 2.23 68 0.137
1.2 53.2 2.17 77 0.136 1.90 75 0.119 1.89 74 0.118
2.0 8.0 4.65 71 0.160 3.55 73 0.140 3.88 67 0.149
2.0 13.3 4.52 77 0.162 3.47 77 0.139 3.72 77 0.150
2.0 53.2 4.41 88 0.156 3.15 80 0.129 3.46 83 0.141

~
t::)
\1)

:E
0
1:':1

=
~
MICROWAVE-VACUUM DRYING 311

24

E
~ 16
w
I-
<(
a:
(!)
z

.~
>- 8
a:
0

~:

400 800 1200 1600 2000

POWER LEVEL (watts)

Figure 10 The effect of microwave power level on the rate of water removal
from 4.54-kg rice samples.

The effect of MV drying on the relationship between moisture and the

volume of a rice kernel might explain this phenomenon. It is generally
understood that kernel volume is proportional to moisture content (11).
During conventional hot-air drying, the kernels shrink as moisture is re-
moved, resulting in a higher density endosperm. However, with MV drying
the vacuum and the internal mechanism of heating enable the kernels
retain a more open, less dense structure as they dry, thereby allowing the
moisture to escape at a constant rate throughout the drying process.
The effect of vacuum level on drying rate can be seen from the data in
Table 1. At all power levels and for all rice lots, the drying rate was
inversely related to dryer pressure. ANOVA indicated that the interaction
effect of power and vacuum levels on rate was significant (p = 0.05)-
changing pressure had more effect at low power than it did at high power.
This indicates that the rate of adsorption/desorption of water on the surface
of the rice was a significant factor in establishing the overall drying rate.
b. Drying Temperature
Also shown in Table 1 are the maximum temperatures reached by the rice
samples during drying. Higher maximum temperatures were observed
when the rice was dried at the higher power levels. Also, the rate of
temperature increase was greater with higher power. The operating pres-
312 WADSWORTH

sure in the dryer had a significant effect on maximum rice temperature. As

the dryer pressure decreased, the maximum temperature reached by the
rice also decreased.

c. Drying Efficiency
The drying efficiency data shown in Table 1 are based on the total amount
of water removed relative to the total amount of energy input to the dryer.
ANOVA indicated significant differences (p = 0.01) in efficiency related to
power level, vacuum level, and rice lot. The lower dryer operating pres-
sures yielded higher efficiencies. This was expected since the lower dryer
pressure would give a greater water vapor pressure differential between the
rice and its environment, thereby increasing the rate of drying. Thus, more
of the microwave energy went toward evaporating water and less toward
heating the rice. This is reflected in the data on maximum rice tempera-
tures, which show lower temperatures associated with lower pressures. The
efficiency and temperature data also support the theory that the ad-
sorption/desorption rate was a significant factor in overall drying rate.
An unexpected result was that both higher efficiencies and higher tem-
peratures were obtained at higher microwave power levels. This indicates
that more of the microwave energy was being coupled with the rice for both
drying and heating. One possible explanation is that, with the batch-
operated system, less energy was lost from the dryer during the higher
power runs because the drying times were shorter. Two sources of direct
microwave energy loss are energy reflected back into the magnetron and
energy absorbed by the dryer itself.
Figure 11 shows the reflected power values measured at various micro-
wave power settings and dryer pressure settings. Reflected power re-
mained essentially constant during the course of a drying experiment. Re-
flected power loss was higher at higher dryer operating pressure over the
entire power range. Thus, at the lower operating pressure, more micro-
wave energy was available to be absorbed by the rice sample.
The relationship between reflected power and input power was more
complex than the relationship of reflected power with pressure. The re-
flected power increased with increasing power level to a maximum value at
an input power setting of approximately 1.1 kW, and then decreased. The
percentage of input power that was reflected decreased continuously with
increasing input power. These results demonstrate the importance of prop-
erly matching the cavity, the load, and the input power level to attain
optimal efficiency.
The relationship between rice lot and efficiency was similar to that be-
tween rice lot and drying rate. The rice lots with higher initial moisture
contents dried more efficiently.
MICROWAVE-VACUUM DRYING 313

0.14 14

0.12 !::::::::!::::::::!::·· 12 ~
,
i
.
6

I 0.08
0.10

·. <; .; : "' 10 !
..!
i
i
,
a.
0.08
.5
....0
.!
•
0

•
;:: •.. .-··...
·······Percentage ritf~~~CI*··
····.. .
J
0.04 r . - - - - - - - - - - - - - - - - - . · ·..... t:::.:··.. .
•e
~ &:

*
Pressure
0.02 +- 13.3 kPa 53.3 kPa 2 :.

u 0
u ~ u u u ~ u u u u u
Microwave Input Power (kW)

Figure ll Effects of microwave-vacuum dryer operating conditions on the rate

of microwave energy reflected back into the wave guides.

The efficiencies shown here are somewhat lower than those obtained
with conventional drying because a batch drying operation is inherently
less efficient than a continuous one. A well designed continuous microwave
system could expect to achieve a microwave-rice coupling efficiency of
approximatley 80%. The MV system is suitable for latent heat energy
recovery through waste-heat evaporators (12). Thus, one could expect an
overall drying efficiency significantly greater than conventional drying.
d. Head Yields
Figure 12 shows the final rice moisture contents obtained for different
MV drying intervals. These runs were made to measure changes in head
yields at various stages of the drying process. Samples that were not MV
dried to storage moisture levels were shade dried before milling. The
head yields obtained at the various final moisture contents from Figure 12
are shown in Figure 13 for three different lots of rice. A moisture content
of approximately 15% appeared to be critical. When the microwave dry-
ing was stopped before the rice reached 15% moisture, there was very
little change in head yield relative to the shade-dried controls. However,
below 15% moisture the head yields began to decrease. The lower the
final moisture attained, the lower the resulting head yield. The responses
of the three rice lots were slightly different with respect to the exact
314 WADSWORTH

25

x- 600 walls
• -1200 "
• -2000 "
;p.
~ 15
w
0:::
:::l
1-
(/) 10
0
~

0 o~---L--~2~o~--~---4~0~--~--~s~o----L---~a~o----L---7.1o~o~los
DRYING TIME (min)

Figure 12 Relationship between rice moisture content and drying time for the
1982 Nato rice dried with three power levels.

moisture content at which the head yield began to decrease and the rate
of decrease.
Preliminary reports on MV drying of rice (13,14) indicated that drying
could be accomplished in a single pass at greatly improved drying rates and
without any increase in breakage as compared with conventional drying.
The results of this study were disappointing with respect to anticipated
head yields at high drying rates. Only when the rice was dried slowly (less
than 2-4 percentage points decrease in moisture content per hour) could it
be dried by MV procedures to safe storage moisture level in a single pass
without any reduction in head yield.
3. Conclusions
For freshly harvested green rice, the MV drying rate was linearly propor-
tional to the rate of microwave energy put into the system up to the maxi-
mum rate of water removal obtained (from 20 to 12% moisture content in 20
minutes). Drying rate over the duration of a run was essentially constant for
a given power input and initial moisture content. Lower operating pressures
in the dryer increased the drying rate, increased drying efficiency, and de-
creased the rice temperature. This indicates that the rate of water
adsorption/desorption on the rice kernel surface is an important factor in
determining overall drying rate. Drying rate varied significantly for different
lots of rice. There is an indication that this is related to the initial moisture
content of the rice rather than variety or type. Rice with higher initial mois-
ture had a higher drying rate. The maximum drying rate that could be used to
MICROWAVE-VACUUM DRYING 315

LABELLE 1982

1981
NATO 1982

0
...J
w
>-
0
<
w
:I:

FINAL MOISTURE (%)

Figure 13 Relationships between final moisture content and head yield for three
rice lots MV dried in a single pass. Microwave power = 600 watts. Chamber
pressure = 13.3 kPa.

dry the rice to safe storage moisture (14%) in a single pass, without increas-
ing breakage, was approximately 2-4 percentage points per hour.

C. MV Drying of Parboiled Rice

Parboiling, a hydrothermic process that invovles soaking and steaming of
rough rice, increases the rice moisture content to approximately 35%
(w.b.). For safe storage and subsequent milling, it is essential that the rice
moisture content be reduced to less than 14% (w.b.). If the drying is not
performed properly, one of the important economic advantages of rice
parboiling (the reduction of kernel breakage) will not be realized. Thus,
drying is a very important step in the overall process of parboiling.
316 WADSWORTH

The use of microwave energy in a reduced-pressure environment for

drying parboiled rice in a single pass was studied by Wadsworth et al. (8).
The effects of parboil soaking conditions, steaming conditions, microwave
power level, drying pressure, drying time, and final moisture content were
evaluated.
1. Experimental
a. Rice
Two lots of Saturn variety rice grown near Crowley, Louisiana, were used
in these studies. The rice had been commercially dried using conventional
methods. The moisture content was 12% (w.b.).
b. Parboiling
Rice samples weighing 1500 g were first soaked in water at 70°C. While
soaking, nitrogen was used to pressurize the parboiling apparatus to 205
kPa (29.7 psia). Soaking time was either 60 or 120 minutes. After soaking,
the excess soak water was drained and the rice sample was equilibrated in
the closed apparatus for 90 minutes at room temperature and atmospheric
pressure. The apparatus was then pressurized with saturated steam at 205
kPa (29.7 psia). Steaming time was either 20 or 25 minutes. These time-
temperature conditions were selected because they had previously been
determined sufficient to increase the rice moisture content into a range
considered optimum for parboiling and to gelatinize the starch without
excessive discoloration of the milled parboiled rice product (15).
c. Drying
The equipment described above was used to study MV drying of parboiled
rice. The samples were transferred without cooling from the parboiling
apparatus to the MV dryer. The moisture content of a parboiled rice sam-
ple at the start of drying (initial moisture) was a function of the four sets of
parboiling conditions. Three microwave power levels (0.6, 1.2, and 1.8
kW) and two dryer pressure levels (6.7 and 53.2 kPa absolute pressure)
were used. Drying was accomplished in a single pass with the drying time
adjusted to achieve a final moisture content of 14, 16, or 18% (w.b.).
2. Results
All tests for statistical significance were performed at the p = 0.01 level.
a. Parboiling Conditions
Table 2 shows the effects of the parboiling soak and steam times on the
change in moisture content of the rice. The effects of parboiling conditions
on the microwave-vacuum drying of the parboiled rice are shown in Table 3.
Parboiling soak time and steam time both significantly affected the moisture
MICROWAVE-VACUUM DRYING 317

Table 2 Moisture Content of Rice After Various

Parboiling Treatmentsa
Soakingb Steaming" time Moistured content
time (min) (min) (%, w.b.)
60 0 32.3a
60 20 35.4c
60 25 36.0d
120 0 34.7b
120 20 36.7e
120 25 37.8f

•Initial rice moisture content was 12.0% (w.b.).

bSoaking in water at 70°C temperature and 205 kPa absolute pres-
sure.
csteaming with saturated steam at 205 kPa absolute pressure.
dThe same letter within a column indicates no significant differ-
ences between means at p = 0.01 (Duncan's multiple range test).

content at the start of drying and, thus, the required drying time. ANOVA
also indicated that steam time, but not soak time, significantly affected total
yield and head yield. Final moisture content, drying rate, drying efficiency,
rice temperature, and brown rice density were not significantly affected by
parboiling conditions. Also, none of the interaction effects involving soak
time and steam time were significant.
b. Drying conditions
The values presented in Table 4 are results averaged over all of the parboil-
ing conditions and show the effects of microwave power level and dryer
pressure level on drying results. A general linear model (GLM) for main
effects, with drying time treated as a continuous independent variable,
indicated that microwave power level significantly affected all of the drying
results shown in Table 4 except the initial and final moisture contents. The
only factor significantly affecting final moisture content was the drying
time. The dryer pressure level significantly affected final rice temperature,
drying rate, drying efficiency, and head yield. Rice temperature, drying
rate, and drying efficiency were also significantly related to the interaction
of power with pressure.
c. Drying Curves
Araullo et al. (16) describe conventional hot-air drying curves for parboiled
rice as being divided into three periods or zones: (a) a zone of constant rate
drying that occurs between 35% and 26% moisture content; (b) a zone of
first falling rate drying that occurs between 26 and 21%; and (c) a zone of
 Wj
~
00

Table 3 Effects of Parboiling Conditions (Soaking Time and Steaming Time) on Microwave-Vacuum Drying Results•
Moisture Moisture Final Brown Total Head
Soak Steam at start Drying at end rice Dryingb Drying' rice rice rice
time time of drying time of drying temp. rate efficiency density yield yield
(min) (min) (%, w.b.) (min) (%, w.b.) eq (g/min/kg) (g/kJ) (g/ml) (%) (%)
60 20 35.4a 55.8a 15.6a 86a 9.1a 0.161a 1.4574a 74.1a 70.4a
60 25 36.0b 57.4b 15.5a 86a 9.1a 0.162a 1.4537a 73.4b 67.1c
120 20 36.7c 60.8c 15.2a 87a 9.2a 0.163a 1.4540a 74.1a 70.2a
120 25 37.8d 64.7d 15.4a 86a 9.2a 0.163a 1.4572a 73.4b 68.1b
•Values averaged over all drying conditions for each parboiling condition. The same letter within a column indicates no significant difference
between means at p = 0.01 (Duncan's multiple range test).
hRate basis: g water evaporated per minute per kg of rice (dry weight basis).
'Efficiency basis: g water evaporated per kJ of energy input into system.

~
~
r:r.J

~
~
 ....~
n
~
0
~<!
~

n~
Table 4 Effects of Microwave Dryer Pressure and Power Settings on Drying Results•
Dryerb Dryer Final Brown c
pressure power Initial Final rice Dryingc Dryingd rice Total Head c
~
level level moisture moisture temp. rate efficiency density yield yield
(kPa) (kW) (%, w.b.) (%, w.b.) eq (g/min) (g/kJ) (g/ml) (%) (%) t::='

....~
6.7 0.6 36.6a 15.1b 68a 4.93b 0.181d 1.4599c 74.0bc 71.4d z
6.7 1.2 36.6a 15.7bc 88c 10.42d 0.191e 1.4602c 73.8b 69.5c C"l
6.7 1.8 36.3a 16.1c 98d 16.41f 0.201f 1.4447b 72.6a 60.0a
53.2 0.6 36.2a 14.6a 72b 2.98a 0.109a 1.4627d 73.9bc 73.0e
53.2 1.2 36.7a 15.6bc 89c 7.48c 0.137b 1.4638d 74.3c 72.7e
53.2 1.8 36.4a 15.2b 102e 12.76e 0.156c 1.4422a 73.8b 67.0b

"Values averaged over all drying conditions for each parboiling condition. The same letter within a column indicates no significant difference
between means at p = 0.01 (Duncan's multiple range test).
hAbsolute pressure values.
'Rate basis: g water evaporated per minute per kg of rice (dry weight basis).
ctEfficiency basis: g water evaporated per kJ of energy input into system.

w
......
~
320 WADSWORTH

second falling rate of drying occurring between 21 and 14%. This probably
indicates that, for conventional drying of parboiled rice, moisture desorp-
tion from the surface of the rice and heat transfer to the rice are rate limiting
from 35 to 26% moisture content and that diffusion of liquid water within the
kernel is rate limiting below 26%. During conventional drying the rice tem-
perature is constant and near the drying air temperature.
Figure 14 shows four typical MV drying curves for parboiled rice dried
with 53.2 and 6.7 kPa dryer operating pressures and two microwave power
levels (0.6 and 1.8 kW) for each pressure. Note that the sample weight scale
is linear but the sample moisture content (w.b.) scale is not. The corre-
sponding rice temperature curves for the same sets of drying conditions are
shown in Figure 15.
The parboiled rice was still hot when placed in the MV dryer. While the
dryer pressure was decreased from atmospheric to the desired operating
pressure, a small amount of water would flash off and the rice temperature
would decrease. When the operating pressure was attained and the micro-
wave power turned on, the rice temperature would initially increase rapidly
and then continue to increase, though more slowly, throughout the drying
experiment. Rice temperature was lower at lower operating pressure. The
sample weight would decrease slowly at first. After this brief initial lag, the

2.2 aouum 40.0

turned/ Mlorowaw turned on
on= attlme•O POWER, PRE88URE
···\·····... ......

---
2.1

--
0.1 kW, 118.2 kPa - 37.1
'\ ······.... --
-
0.1 kW, 1.7 kPa

-I
#
~
2.0
'\ ··.. ····... 1.1 kW, 58.2 kPa - 34.0

.c '\ - 1.8 kW, 1.7 kPa

c
.!
c
Q 1.9 '\ ··············... - 30.5 0

-•
0
II 1.8 '\ ·············... !::::J
- 28.7
a.
·········...
•
E '\ 0
(I) 1.7 '\ ··. - 22.4 :::1
\. ··········...
-
''
1.8 17.5
·············....
1.5 L-...---L----L--1---....L..----L---'-----'-----L-~ 12.0
~ o ~ ~ ~ ~ ~ m ~ ~
Drying Time (min)

Figure 14 MV-drying curves for parboiled rice at two dryer operating pressures
and two microwave power levels for each pressure. Note-moisture scale on right
side vertical axis is not linear.
MICROWAVE-VACUUM DRYING 321
120

--
100 vacu.-
turned
00 on~

!:I 80 \:
.
ii ·············::.:::···:::.:,:.···::.::.:································
...... ;;:..·····;.:...-
II
Q,
E
80
/.
~

~ POWER,PREUURE
40
\
II
...... 0.1 kW, 18.2 kPa
a:
()

- - 0.1 kW, 1.7 kPa

Mlcrowav. turned on
20
at time • o - 1.8 kW, 58.2 kPa
- 1.1 kW, 1.7 kPa
OL_~--L-~--L-~~~~~~~~
-20 0 20 40 80 80 100 120 140 180 180
Drying Time (min)
Figure 15 Rice temperature curves during microwave-vacuum drying of par-
boiled rice at two dryer operating pressures and two microwave power levels for
each pressure. Initial sample weight = 2130 g. Initial moisture content = 38.0%
(w.b.).

weight loss was essentially linear with time until a moisture content level of
16% (w.b.) was attained. At that point, the slope of the drying curve
decreased slightly indicating a possible change in drying mechanism or
water activity.
These results verify that the mechanisms for conventional hot-air drying
and MY drying of parboiled rice are different. Even the period of MY
drying where weight loss is linear with time is not equivalent to the
constant-rate drying period observed with hot-air drying because, with MY
drying, the rice temperature is continuously increasing.
d. Drying Rate
Figure 16 summarizes the drying rate results for the various microwave
power and dryer pressure conditions. There were large changes in drying
rate related to power and pressure settings and a more subtle decrease
in drying rate, as explained below, that occurs as the rice is dried below
16% moisture content (w.b.). Above 16% rice moisture, the drying rate
was essentially constant for each microwave power and dryer pressure
combination.
322 WADSWORTH

POWER = 0.8 kW POWER • 1.2 kW POWER • 1.8 kW

20 20 20
•
D.
"" i....
C\1 i.... i.... a._ I am a
~ 1: 1: 1:
10
i1o i.... 10 i.... 10
I

w ....
••
0: ~ ~ a•D Da~D ~
;:)
It)
It)
!0: Ill •• ,_aD
!0: 0:
w 0 0 0
0:
D. 10 15 20 10 15 20 10 16 20
Final Moisture (") Final Moisture (") Final Moisture (")

20 20 20
•
D.
""
:i i....
i.... i.... • D D a aDD 8D

,. .......
1: 1:

... .
1:
u
i.... 10 i.... 10
w i.... 10

••
0:
;:) .!! ~ .!!
.!
It)
It)
w !0: • 0: •
0:
0: 0 0
D. 0
10 15 20 10 15 2~ 10 16 20
Final Molature (,.) Final Moisture (") Final Moisture (")

Figure 16 Effects of operating conditions for the microwave-vacuum dryer sys-

tem on the rate of drying of parboiled rice.

The trends for parboiled rice were similar to those seen with freshly
harvested rice. The drying rate was linearly proportional to microwave
power. Also, the drying rate was greater at the lower dryer pressure set-
ting. These two pieces of information indicate that drying rate was not
limited by liquid water diffusion and that the adsorption/desorption of
water on the surface of the rice was a significant factor in the overall drying
rate. GLM and regression analyses of the data, with final moisture content
treated as a continuous independent variable, indicated that the rate of
increase in drying rate with increasing power was greater at a pressure of
6.7 kPa (regression line slope= 12.63) than it was at a pressure of 53.2 kPa
(slope = 10.75). The percentage change in drying rate with change in
pressure was greater at lower power levels than it was at higher power
levels. This indicates that adsorption/desorption rate was a more important
factor in determining overall drying rate at lower power levels. And at the
higher power levels, it may indicate that water molecules inside the rice
kernel have attained an energy level high enough to diffuse through and
out of the kernel in a gaseous state.
MICROWAVE-VACUUM DRYING 323

GLM and regression analyses indicated that, below 16% moisture con-
tent, the drying rate decreased as the rice sample was dried to lower mois-
ture contents. This reduction in drying rate could be due to a diffusion
effect that was manifesting itself at the lower moisture levels. Or it might
be due to water activity changes (i.e., the physical binding of water mole-
cules to the starch at these lower moisture contents that increases the
energy required to vaporize the water molecules). The interaction effect of
pressure and final moisture content on drying rate was also significant. The
decrease in drying rate with decreasing moisture was greater at 6. 7 kPa
than at 53.2 kPa. At the 53.2 kPa pressure, the slopes of the regression lines
for drying rate on final moisture content at all power levels were not
statistically greater than zero. This interaction effect indicates that the
contribution of adsorption/desorption rate to the overall drying rate is
greater at 6. 7 kPa than at 53.2 kPa.

e. Drying Efficiency
The coupling efficiencies (percentage of microwave energy entering the
drying cavity that is absorbed by the load) attained in this work ranged
from 31 to 62%. A well-designed continuous microwave system, where the
cavity, load, and power level have been matched, should have coupling
efficiencies in excess of 80% (17). However, an analysis of the experimen-
tal data herein will show the relative effects of the MV operating variables
on drying efficiency for a batch-operated system.
Microwave energy that enters the drying cavity is either absorbed by the
rice sample, reflected back into the wave guide, or lost through dryer
inefficiencies. Energy absorbed by the sample can be calculated from the
sample temperature and weight changes. Reflected power was experimen-
tally measured. System losses were estimated by differences. Drying effi-
ciency was calculated as the amount of water evaporated from the sample
divided by the amount of microwave energy entering the drying cavity.
Table 5 shows the total amount of energy absorbed by the rice sample
during drying at different microwave power and dryer operating pressure
settings. A specific heat value of 1.8 kJ/kgfDC (11) and a heat of vaporiza-
tion value of 2722 kJ/kg (2) were used in the calculations. Since approxi-
mately equal amounts of water were evaporated (latent heat) from the
parboiled rice samples for all of the power/pressure combinations, the
differences in total amount of energy absorbed were due to the differences
in sensible heat changes in the rice. The rice samples dried at higher power
levels and higher pressure levels attained higher final temperatures and,
thus, had greater increases in sensible heat.
Table 5 also shows the average rate of energy absorption by the rice at
each of the power and pressure settings. At a fixed power level, the aver-
324 WADSWORTH

Table 5 Energy Absorbed by Rice Sample at Different Microwave Power

Levels and Dryer Operating Pressures.
Microwave Dryer Sensible Latent Total Mean rate Percentage
power operating heat heat energy of energy of input
level pressure effect effect absorbed absorption power
(kW) (kPa) (kJ) (kJ) (kJ) (kW) (%)
0.6 6.7 75.6 1429 1505 0.3097 51.6
0.6 53.2 88.2 1418 1505 0.1887 31.5
1.2 6.7 142.2 1407 1548 0.6793 56.6
1.2 53.2 145.8 1388 1534 0.4914 41.0
1.8 6.7 176.4 1425 1602 1.1084 61.6
1.8 53.2 187.6 1412 1598 0.8601 47.8

age rate of energy absorption by the rice was greater at lower pressures.
The differences in energy absorption rate at different pressures could not
be totally accounted for by the differences in reflected power at the differ-
ent pressures. Thus, energy losses from the dryer system, as explained
below, must have been greater at higher operating pressures.
At a fixed operating pressure level, energy absorption by the sample
increased as microwave power level increased. This increase was at a rate
greater than the rate of input power increase (i.e., the second derivative of
the curve was positive). At the lower operating pressure (6.7 kPa), this
additional increase in energy absorbed by the sample (the amount over and
above that accounted for by the input power increase) was approximately
equal to the decrease in reflected power with increasing input power. How-
ever, at the higher pressure (53.2 kPa), the additional increase in energy
absorbed exceeded the decreases in reflected power, which indicates that
the system losses decreased at the higher input power levels.
Also shown in Table 5 are the percentages of input power absorbed by
the rice (coupling factors).
Table 6 shows the estimated energy losses from the dryer system at the
different dryer operating conditions. These losses, which were due to vari-
ous causes described below, were estimated to be the difference between
the microwave energy input to the system and the sum of the energies
absorbed by the rice and reflected.
Since instrumentation was not available to directly measure the system
energy losses, one can only speculate about the sources and magnitudes of
these losses. The rice sample holders, which were constructed of po-
lypropylene, had masses that were 2.5 and 3.4 times the sample weight.
Since the rice is in direct physical contact with the sample holder, it is
MICROWAVE-VACUUM DRYING 325

Table 6 Estimated System Energy Losses at Different Microwave

Power Levels and Dryer Operating Pressures
Dryer
Microwave operating Total• system Average Percentage of
power level pressure losses loss rate input power
(kW) (kPa) (kJ) (kW) (%)
0.6 6.7 1069 0.2203 36.7
0.6 53.2 2927 0.3367 56.1
1.2 6.7 947 0.4148 34.6
1.2 53.2 1829 0.5866 48.9
1.8 6.7 871 0.6316 35.1
1.8 53.2 1588 0.8528 47.4

•Energy losses from dryer system estimated by difference (losses = input

energy - absorbed energy - reflected energy).

probably safe to assume that the sample holder is heated from room tem-
perature to approximately the final rice temperature. The specific heat of
polypropylene is 1.9 kJ/kg/°C {18). Thus, the energy loss to the sample
holder would be more than three times greater than the sensible heat
change of the rice. Another source of energy loss was due to the room air
bled into the system to help carry the evaporated moisture away and pre-
vent it from condensing inside the rotating drum sample holder and
rewetting the rice. The amount of energy lost in this manner would depend
on the rates of heat transfer from the rice and sample holder to the air.
Since convective heat transfer is highly pressure dependent, this explains
the much greater energy losses at higher dryer operating pressure. Another
energy loss was due to water condensing on the inside dryer walls and then
dripping onto the sample holder where it was revaporized. It is not possible
to estimate the magnitude of this energy loss, but undoubtedly it was
significant. This loss was also greater at the higher dryer operating pres-
sure, because this condensation-revaporization process would continue un-
til the interior surfaces of the dryer attained a temperature equal to the
flash temperature of water at the dryer operating pressure. The boiling
points of water at 6.7 and 53.2 kPa are 38 and 83°C, respectively. Our
operating experience for this dryer was that, with the 6. 7 kPa pressure runs,
the interior surfaces of the dryer were free of moisture at the end of the
run. This indicates that at some point during the run the interior surface
temperature reached 38°C and the condensation-revaporization process
ceased. With the 53.2 kPa pressure runs, the interior surfaces were still wet
at the end of the run indicating that an interior surface temperature of 83°C
had never been attained. Thus, the condensation-revaporization process
326 WADSWORTH

Table 7 Summary Energy Balance on Batch-Operated

Microwave-Vacuum Dryer System at Different Microwave
Power Levels and Dryer Operating Pressures•
Percentage of input powerb
Dryer
operating Microwave Absorbed System
pressure power level Reflected by rice losses
(kPa) (kW) (%) (%) (%)
6.7 0.6 11.7 51.6 36.7
6.7 1.2 8.8 56.6 34.6
6.7 1.8 3.3 61.6 35.1
53.2 0.6 12.4 31.5 56.1
53.2 1.2 10.1 41.0 48.9
53.2 1.8 4.8 47.8 47.4
•Energy rates averaged over complete drying cycle for drying par-
boiled rice from approximately 36 to 14% moisture content
(w.b.).
bEnergy rates normalized as a percentage of the rate of micro-
wave energy input to the drying chamber.

had continued, probably at a diminishing rate, throughout the 53.2 kPa

runs. Another source of energy loss in a microwave system is called wall
loss. No material reflects 100% of the microwave energy striking its sur-
face. Of the various energy loss mechanisms described above, only the wall
loss should be a factor in a well-designed continuous system.
Table 7 summarizes the energy balance on the dryer system in terms of
the average rate of energy transfer, normalized as a percentage of input
rate, while drying parboiled rice from approximately 36 to 14% moisture
using various combinations of microwave power levels and dryer operating
pressures. Reflected power decreased as input power increased and was
lower at low operating pressure. The power absorbed by the rice increased
as input power increased and was considerably higher at low operating
pressure than at high operating pressure. The average power lost from the
system was much greater at higher operating pressure than at low pressure.
The change in system losses with changing input power was negligible at
low operating pressure but decreased with increased power input at high
operating pressure.
Figure 17 summarizes the drying efficiency data for the various micro-
wave power and operating pressure combinations. Each subplot shows
the effects of final dried moisture content. The ANOVA and GLM statis-
tics indicated that microwave power level, dryer operating pressure, and
MICROWAVE-VACUUM DRYING 327

POWER • 0.8 kW POWER • 1.2 kW POWER • 1.8 kW

:. -.., .30
.30 .30

"' ~... ... ...

.X ~ ~
.X .X .X

"'
10
I g-.16
~
g-.16 a• a ,,Jiaa
~
g-.15
,_,,.,
w
• ..... rll.• • •
- - -
a: 0 ~~~
0 0
::» ;: ;: ;:
CD
CD w w w
w .00 .00 .00
a:
II. 10 u; 20 10 15 20 10 15 20
Final Molature (") Final Molature (") Final Molature (")

• .30 .30 .30

.
II.
...ci ...
.X ~
...
~
...
~

. .t'"
.X .X .X

~ ~ , . . OlD . . ~
.,,.,,,
•
w
aa

g-.16
a: g.15
g-.16
• • •
-
::»

--
0
CD
CD ~ ~ ;:
w
a: w .00 w
.00
w
II. .00
10 16 20 16 10 16 20
Final Molature (.,) ~l~al (~~ Final Molature (")

Figure 17 Effects of operating conditions for the microwave-vacuum dryer sys-

tem on the efficiency of drying of parboiled rice.

final moisture content significantly affected drying efficiency. Also, the

interactions of power with pressure and final moisture with pressure were
significant.
Drying efficiencies were considerably higher at lower operating pres-
sure. Drying efficiency increased with increasing power level, and the rate
of increase was greater at higher pressure. The higher drying efficiencies at
lower pressures can probably be explained by differences in the energy
losses related to the condensation and revaporization of water that were
different for the different pressures. The differences in drying efficiencies
at different power levels can also be explained by differences in energy
losses. Reflected power losses were lower at higher power levels. Also, the
rate of energy loss by convective heat transfer to the air bled into the dryer
and by condensationlrevaporization of water are probably independent of
microwave power level. Since the high power drying runs required less
drying time, there was relatively more energy available to heat the rice and
evaporate water. And because these losses (convection and condensation!
revaporization) were greater at higher pressures, the effect of the shorter
328 WADSWORTH

drying times for higher power levels was relatively greater at higher pres-
sures, which explains the interaction effect.
The drying efficiency decreased slightly as the rice moisture decreased
below the 16% level. This is the same region where a slight decrease in
drying rate was observed. As with drying rate, the effect was greater at the
lower operating pressure.
f Brown Rice Density
The effects of microwave power level, dryer operating pressure, and final
moisture content on the density of the dried brown parboiled rice are
shown in Figure 18. ANOVA and GLM indicated that the significant fac-
tors affecting the density of MY-dried parboiled rice were microwave
power level and the interaction of power level with final moisture content.
It has been reported by Bhattacharya (19) and many others that as
parboiled brown rice moisture content decreases, the density of the rice
increases. However, the rice density changes observed with MV drying of
parboiled rice, and with MV drying of green rice, do not conform with

POWER • 0.8 kW POWER • 1.2 kW POWER • 1.8 kW

-
1.5 1.5 1.5
•
a.
e :a,. aa -e.... ... -e.... ..... .
.
.
~ rP DB at aDD-

"' -;.... 1.4

-• -• .
'lo

-•
~ al al al
10 -; 1.4 -; 1.4
•
w
a: iic:
;::)
ii
c:
ii
c:
10 iCI iCI iCI
10
w 1.3 1.3 1.3
a:
a. 10 15 20 10 15 20 10 15 20
Final Molature (,.) Final Molature (,.) Final Molature (,.)

•
. .,.. ..
1.5 1.5 1.5
a.
- -i.... :c
...
.... i.... i....
~
DID lib II B D a 11 ~

-•
ci

-• -•
al
• Cl 1:11
w -; 1.4 -; 1.4 -; 1.4
a:
;::) ii ii ii
10 c: c: c:
10
w iCI iCI iCI
a:
a. 1.,0 1.3 1.3
15 20 10 15 20
Final Molatura (,.) ~l~al Mo~~ure (~f Final Molature (,.)

Figure 18 Effects of operating conditions for the microwave-vacuum dryer sys-

tem on the density of MY-dried parboiled rice.
MICROWAVE-VACUUM DRYING 329

conventional drying results. At microwave power levels of 0.6 and 1.2 kW

the brown rice density remained constant as the rice was dried from 18 to
12% moisture content. The densities obtained at 1.2 kW were approxi-
mately equal to the densities obtained at 0.6 kW (means of 1.4620 g/ml and
1.4613 g/ml, respectively). For the 1.8 kW drying runs, the densities at 18%
moisture content were not significantly different from the densities ob-
tained with the 0.6 and 1.2 kW power levels. However, at 1.8 kW, as the
rice was dried below 18% moisture content, the densities actually de-
creased. This indicates that a slight puffing effect was taking place at the
higher power level. The fact that brown rice density did not increase as the
rice was MV dried to lower moisture contents is evidence that vaporization
of moisture within the rice kernel was occurring during microwave drying.

g. Milling Yields
Total milled rice yield was significantly affected by microwave power level,
dryer operating pressure, and the moisture content to which the rice was
dried. At a final moisture content of 18% (w.b.) there were no significant
differences in total yield related to power or pressure levels. Regression
analyses of total yield on final moisture content below 18% indicated no
significant reduction in total yield at power levels of 0.6 and 1.2 kW for
both pressure levels. However, at the power level of 1.8 kW, the total yield
significantly decreased with decreasing final moisture content. These re-
sults were very similar to the relationships found for brown rice density
with microwave power, operating pressure, and final moisture content.
Since milling conditions were identifcal for all samples, the samples with
lower densities were probably softer and lost more endosperm material
during milling, which resulted in reduced total milling yields.
Bhattacharya and Indudhara Swamy (20) reported on hot-air drying of
parboiled rice in a single pass at various temperatures from 40 to 80°C and
sun drying. At all conditions, rice head yields decreased dramatically when
the parboiled rice was conventionally dried in a single pass to less than 16%
final moisture content.
Head yields for MV drying (Fig. 19) were affected by microwave power
level, operating pressure level, and final rice moisture content. Regression
analysis of the results obtained for the MV drying of parboiled rice (Table
8) shows that, at the lower power level (0.6 kW), it was possible to dry the
rice in a single pass to less than 12% moisture content without any signifi-
cant reduction in head yield. The head yields at a pressure level of 6. 7 kPa
(final rice temperature = 68°C) were only slightly less than those at 53.2
kPa (final rice temperature = 72°C) even though the average drying rate
was 1.63 times greater. At the high power level (1.8 kW), regression
analysis indicated significant reductions in head yield as the moisture con-
330 WADSWORTH

POWER " 0.8 kW POWER • 1.2 kW POWER • 1.8 kW

• 80 80 80

"-
a.

"-
.¥ 'Ia s 0 a0 o11 aa
~ a• a oaollbo a

.
N a aa a
a
C') l! '0 '0 a
10
• ~85 ~86 :! 86
>-
a

w
a: 1 1 1 a

:z:• :z:• :z:•

a
::;)
tn
tn
w 60 60 50 a
a:
a. 10 16 20 10 16 20 10 16 20
Final Molature (~) Final Molature (~) Final Molature (")

:.
.¥
80 80 80

.... ~
.; a
dloaa
•'lo a
a ~ 0 D'b
a
'L
~ a
aa
'0 '0 aa '0
•
w ~81! ~85 a
a
a
~85 a a
a

a: a
1 1 1
• :z:• :z:•
::;) a
a
tn
w :z:
tn
a: 5, 50 50 a
a.
10
Final Molature (fl)
15 20
J.~ •. 15
~~r
10
Final Molature (~)
15 20

Figure 19 Effects of operating conditions for the microwave-vacuum dryer sys-

tem on the milled head rice yield for MY-dried parboiled rice.

tent decreased below 16%. The reduction was greater at an operating

pressure of 6.7 kPa (final rice temperature= 98°C) than it was at 53.2 kPa
(final rice temperature = 102°C). The results at a power level of 1.2 kW
were intermediate to those at the lower and higher power levels. At an
operating pressure of 53.2 kPa, regression analysis of the 1.2 kW data
indicated no significant reduction in head yield with decreasing final mois-
ture content, whereas, at 6.7 kPa operating pressure, the decrease in head
yield was significant. Thus, at water removal rates of less than 7.4 g of
water/min/kg of dry rice, there was no significant decrease in head yield as
the moisture content decreased from 18 to 12% (w.b.). At water removal
rates of 10.5 glmin/kg or greater, head yield decreased as the final rice
moisture content fell below 18%. As the drying rate increased, the reduc-
tion in head yield also increased. Figure 20 shows the relationship between
drying rate and the decrease in head yield (percentage points) for each
percentage point decrease in final moisture content below a moisture level
of 18% (w.b.).
MICROWAVE-VACUUM DRYING 331

v
.. 4
;::
:
J:
.5 3
•
i
-e
~ 2
Cll
c:
II

:. 1

oL-~==~~~~--~~--~--L--L~
0 2 4 8 8 10 12 14 18 18 20
Drying Rate (g H10/mln/kg of dry rice)

Figure 20 Effect of drying rate on the rate of decrease in head yield as parboiled
rice is microwave-vacuum dried from 18 to 12% moisture content (w.b.). The head
yield decrease is the change in head yield per percentage point decrease in moisture
content below 18%.

Table 8 Effects of Water Removal Rate on Change in Head Yield with

Decreasing Moisture Content Below 16% (w.b.)
Operating Regression statistics
pressure Microwave Water•
level power level removal rate Slopeb Probability
(kPa) (kW) (g/min/kg) (%/%) F-Value level

6.7 0.6 4.9 0.18 0.4 0.55

6.7 1.2 10.4 1.25 8.8 0.01
6.7 1.8 16.4 4.05 9.8 0.01
53.2 0.6 3.0 0.12 1.3 0.28
53.2 1.2 7.5 0.15 0.7 0.41
53.2 1.8 12.8 2.50 12.5 0.01

•Grams of water removed/min/kg of rice (dry basis).

hDecrease in head yield (in percentage points) for each percentage point decrease
in final moisture content.
332 WADSWORTH

3. Conclusions
Parboiling conditions had no direct effect on the rate and efficiency of MV
drying of parboiled rice. Longer soaking and steaming times increased the
moisture content of the rice, which then required longer drying times.
Drying rate and drying efficiency were significantly affected by micro-
wave power level, dryer operating pressure, and the final moisture level to
which the rice was dried. Energy losses in the batch-operated system were
highly dependent on the dryer operating pressure (lower losses at lower
pressure). Drying runs at higher power levels also had lower energy losses
due to shorter drying times and lower reflected power losses. Drying rates
were directly proportional to microwave power level, were higher at lower
dryer operating pressures, and were constant during a run until the rice
moisture content fell below 16%(w.b.). For equivalent MV operating condi-
tions, parboiled rice dried more rapidly than freshly harvested rice.
The brown rice density did not increase with decreasing moisture as it
normally does during conventional rice drying. At the higher microwave
power levels, the brown rice density actually decreased, showing that a
slight puffing effect was occurring during drying.
The results indicate that liquid diffusion of water within the rice kernels
was not rate limiting, that the rate of adsorption/desorption of water played a
significant role in determining overall mass transfer rate, and that vaporiza-
tion of water within the rice kernel was probably occurring during drying.
Milling yields were significantly related to the same variables that af-
fected the drying rate (power, pressure, and final moisture). Parboiled rice
could be rapidly dried to 16% moisture content without any significant
affect on milling yields. However, below 16% moisture, head rice yield was
highly dependent on drying rate. At water removal rates of less than 7.4 g
of water/min/kg of dry rice, the parboiled rice could be dried in a single
pass to safe storage moisture levels ( <14%, w.b.) without any significant
decrease in head yield. At water removal rates equal to or greater than 10.5
g/min/kg, rice head yields decreased dramatically as the rice moisture con-
tent fell below 16% (w.b.).
Future research on the MV drying of parboiled rice should investigate
the effects of varying the microwave power level during the course of a
drying run in order to develop an optimum power/moisture content rela-
tionship so that the rice could be dried in a minimal amount of time without
adversely affecting milling yields. Another research area that might be
fruitful is a study of postdrying rice-handling procedures in order to mini-
mize breakage due to stress development after drying (21). The possibility
of modifying the parboiled rice density to produce products with unique
rehydration characteristics should also be investigated.
MICROWAVE-VACUUM DRYING 333

D. Quality Characteristics of MV-Dried Rice

Hogan and Planck (22) reported that drying conditions can influence the
hydration characteristics of rice. Goebel et al. (23) reported the effects of
microwave energy on wheat starch granule transformations. It is known
that microwaves can have subthermal effects (9). Anyone who has heated
bakery products in a home microwave is aware of the textural changes that
can result. While it has been demonstrated that MY technology can suc-
cessfully dry rice in a single pass to safe storage moisture levels without
adversely affecting milling yields, the process will not be acceptable to the
rice industry unless rice dried by this method has the cooking and process-
ing characteristics of conventionally dried rice. Wadsworth and Koltun
(24) reported on the physicochemical properties and cooking quality of
microwave-dried rice.
1. Rice
The rice samples evaluated for quality characteristics were taken from the
six rice lots used in the study described above on MY drying of freshly
harvested green rice. The samples were selected so that their MY-drying
conditions spanned a range of mild to harsh treatments. The MY-drying
conditions for each sample evaluated are shown in Table 9. Drying rates
ranged from 2-24 percentage point decreases in moisture content per hour.
The head rice yields for the milder treatments were comparable to the
controls, whereas the harsh treatments gave head yields as low as 13%.

Table 9 Microwave-Vacuum Drying Conditions• and Drying Codes

Drying time (min)
Microwave Dryer Labelle Nato
Drying power pressure Starbonnet
code (kW/kg) (torr) 1981 1982 1983 1981 1982 1983

MVll 0.132 20 54 49 50 42 48 44
MV12 0.132 100 65 68 58 67 61
MV13 0.132 400 98 99 88 100 92
MV21 0.264 50 34 33 33 28 30 29
MV22 0.264 100 37 37 32 37 34
MV23 0.264 400 47 48 40 46 42
MV31 0.441 60 18 18 18 14 16 15
MV32 0.441 100 21 21 18 21 19
MV33 0.441 400 30 28 28 25 29 26

•Control samples were shade dried at zzoc and 60% relative humidity.
334 WADSWORTH

Table 10 Alkali Spreading Values for MY-Dried Rice

Labelle Nato
Drying Starbonnet
code• 1981 1982 1983 1981 1982 1983
MVll 4.6 2.5 3.2 6.6 6.2 6.6
MV12 2.9 3.2 6.2 6.2 6.2
MV13 3.2b 3.3 6.2 6.4 6.2
MV21 4.3 2.8 3.1 6.4 6.1 6.3
MV22 2.5 3.5 6.5 6.3 6.5
MV23 3.3b 3.4 6.5 6.4 6.6
MV31 4.5 2.4 3.4 6.4 6.3 6.4
MV32 2.5 3.3 6.6 6.7b 6.6
MV33 4.9b 3.4b 3.8b 6.8 6.7b 6.7
Control 4.3 2.5 3.0 6.3 6.1 6.2
•See Table 9.
bSignificantly different from the control at p = 0.05 (LSD = 0.6).

2. Physicochemical Properties
The quality evaluation tests were selected from tests used at the USDA
Rice Quality Laboratory (25). These were water uptake capacity at 77°C
(26), alkali spreading reaction (27), and amylographic gelatinization and
pasting characteristics (26).
The results of the alkali reaction tests are shown in Table 10. ANOVA
indicated significant differences among the rice lots (p = 0.01) and among
the treatments (p = 0.05). For each variety, when the control was com-
pared with each treatment using Fisher's least significance difference
(LSD) test, the alkali spreading values for some of the samples receiving
the most severe MV treatments were significantly higher (p = 0.05) than
the controls. However, these differences (less than 1.0 units) are not great
enough to be of any practical consequence in terms of rice quality.
Similar results were obtained with the water uptake capacity tests (Table
11). ANOVA indicated significant differences among rice lots (p = 0.01)
and among treatments (p = 0.05). Some of the more severely treated
samples had significantly higher water uptake when compared with their
controls (Fisher's LSD,p = 0.05). However, the differences were too small
to substantially affect rice cooking and processing quality.
A possible explanation for the higher alkali spreading values and water
uptake capacities might be that the more severely treated samples had
either small fractures in the kernels and/or lower density, more porous
kernels. These would allow alkali solution or water to infuse the kernels at
MICROWAVE-VACUUM DRYING 335

Table 11 Water Uptake Values for MY-Dried Rice

Water uptake (mUlOO g)b

Labelle Nato
Drying Starbonnet
code• 1981 1982 1983 1981 1982 1983

MVll 114 62 132 214 234 310

MV12 84 130 210 219 332
MV13 74 125 263C 229 297
MV21 106 73 128 177 255 335
MV22 84 126 225 262 338
MV23 66 133 227 246 351C
MV31 138 ll4C 124 237 253 340
MV32 75 136 215 248 350
MV33 115 lOSe 144 235 268C 354c
Control 123 72 124 225 236 316

•See Table 9.
bWater uptake measured at 77°C.
cSignificantly higher than the control at p = 0.05 (LSD = 38).

a slightly greater rate than for the uncracked or denser kernels, resulting in
slightly higher test values.
Table 12 shows the peak viscosities and setback viscosities as measured
with a Brabender viscoamylograph. The amylographic viscosity curves for
the control and the MV-treated samples within each rice lot were essen-
tially identical. ANOVA indicated significant differences among the rice
lots but not among the treatments for both peak and setback viscosities. It
seems safe to conclude that the MV treatments did not materially affect the
cooking and processing quality of the milled rice.
3. Taste Panel Evaluation
A taste panel evaluated the flavor and texture of standard cooked rice
samples. A triangular test was used to compare the MV-dried rice with its
shade dried control. At a panel session, each member was presented three
samples, two of which were identical. The panelists were asked to identify
the sample that was different by either flavor or texture. Thus, the probabil-
ity of a correct identification being made by random chance was one third.
The hypothesis was that there was no difference between the control and
the MY-treated sample. The results of the taste panel evaluations for two
of the rice lots are given in Tables 13 and 14. Similar results were obtained
with all lots.
336 WADSWORTH

Table 12 Brabender Amylographic Peak and Setback

Viscosities for MY-Dried Rice
Viscosity (BU)b
Labelle Nato
Drying Starbonnet
codea 1981 1982 1983 1981 1982 1983
Peak viscosity
MVll 359 423 387 442 532 512
MV12 424 392 443 620 503
MV13 443 395 453 595 504
MV21 374 435 406 518 552 505
MV22 380C 371 455 575 511
MV23 398 381 470 588 507
MV31 356 418 388 460 525C 520
MV32 408 365 458 575 501
MV33 354 393 369 426 533 497
Control 367 434 396 467 582 505
Setback viscosity
MVll 480 641 537 619 604 523
MV12 647 570 658 650 522
MV13 650 570 644 656 517
MV21 502 640 553 655 661 522
MV22 615 597 620 692 528
MV23 614 589 672 647 526
MV31 494 610 546 608 615 519
MV32 627 581 639 680 525
MV33 495 660 576 653 644 507
Control 480 652 551 660 655 526
•See Table 9.
bBrabender units, measured with a 1000 cm-g cartridge.
cSignificantly different than the control at p = 0.05 (LSD = 51).

Table 13 shows the number of times a given MY-dried sample was cor-
rectly identified by a panelist as different from its control. The chi-square
test, which is a measure of deviation from the expected probability of one-
third correct judgments, indicated that the panel as a whole could not distin-
guish between the flavor or texture of the control and those of the MY-
treated sample. Table 14 shows the performance of the individual panelists in
correctly distinguishing between MY-dried samples and the control. The
"random chance probability" is the probability that a panelist would, by
chance alone, get a number of correct responses equal to or greater than the
number actually attained. The best performance of any panelist was eight
correct out of 18 sessions. The probability of getting eight or more correct
MICROWAVE-VACUUM DRYING 337

Table 13 Taste Panel Comparison of MV-Dried Rice with Control Samples

Using the Triangular Test

Taste test evaluations

Correctly Incorrectly
identified identified Chi-Squarea
Microwave Dryer
power pressure Labelle Nato Labelle Nato Labelle Nato
(kW/kg) (torr) 1982 1982 1982 1982 1982 1982

0.132 20 3 4 10 11 0.61 0.30

0.264 50 44 10 10 0.14 0.14
0.441 60 9 3 12 10 0.86 0.61
0.441 400 3 69 12 0.37 0.00
All samples 19 17 41 43 0.08 0.68

•Chi-square for a binomial population with a probability of success equal to one third.

Table 14 Ability of Individual Members of the Taste

Panel to Distinguish MV-Dried Rice Samples from
Control Samples

Number of evaluations
Random chance
Panelist Correct Incorrect probabilitya

A 6 12 0.59
B 2 12 0.97
c 6 12 0.59
D 4 10 0.74
E 4 8 0.61
F 4 12 0.83
G 4 6 0.44
H 8 10 0.22

•The probability that the panelist would, by random chance only,

get a number of correct responses equal to or greater than that
shown in the "Correct" column.

out of 18 tries by chance is 0.22. This indicates that no individual panelist was
able to consistently detect a difference in flavor or texture resulting from MV
drying. Thus, the null hypothesis cannot be rejected.
4. Conclusions
The variabilities in water-uptake and alkali-spreading values normally
found among rice lots with similar cooking and processing characteristics
are much greater than the observed changes caused by the MY-drying
338 WADSWORTH

treatments. There was no evidence to indicate that MV treatments suffi-

cient to dry green rough rice in a single pass to safe storage moisture levels
had any adverse effects on the physicochemical properties or eating quality
of the resulting milled rice.

REFERENCES
1. Choudhary, M. S. (1970). The effects of moisture adsorption on the tensile
strength of rice. PhD dissertation. Texas A&M University, College Station,
TX.
2. Steffe, J. F., Singh, R. P., and Miller, Jr., G. E. (1980). Harvest, drying and
storage of rough rice, Rice: Production and Utilization (B. S. Luh, ed). AVI
Publishing Co., Westport, CT, p. 311.
3. Kunze, 0. R., and Calderwood, D. L. (1985). Rough rice drying, Rice Chemis-
try and Technology (B. 0. Juliano, ed), American Association of Cereal
Chemists, St. Paul, MN, p. 233.
4. Wang, C. Y., and Luh, B. S. (1991). Harvest, drying and storage of rough
rice, Rice Production, Vol. I (B. S. Luh, ed.), Van Nostrand Reinhold, New
York, p. 311.
5. Mossman, A. P. (1986). A review of basic concepts in rice-drying research.
Crit. Rev. Food Sci. Nutr. 25(1):49.
6. Wadsworth, J. I. (1984). Microwave-vacuum drying of rice, Proceedings of
20th Rice Tech. Working Group, Lafayette, LA, p. 85.
7. Wadsworth, J. 1., and Koltun, S. P. (1990). Rice drying by microwave-
vacuum, Rice J., 93(6):20.
8. Wadsworth, J.l., Velupillai, L., and Verma, L. R. (1990). Microwave-vacuum
drying of parboiled rice, Trans ASAE, 33:199.
9. Copson, D. A. (1975). Microwave Heating, 2nd ed., AVI Publishing Co.,
Westport, CT.
10. USDA. (1974). Inspection Handbook for the Sampling, Inspection, Grading
and Certification ofRice, Agricultural Marketing Service, Grain Division, HB
918-11.
11. Wratten, F. T, Poole, W. D., Chesness, J. L., Bal, S., and Ramarao, V. (1969}.
Physical and thermal properties of rough rice, Trans. ASAE,12(6}:801.
12. Bryan, W. L. (1977). Recovery of waste heat from drying citrus by-products,
Water Removal Processes: Drying and Concentration ofFoods and Other Mate-
rials, (C. J. King and J.P. Clark, eds.), AIChE Symposium Series, p. 25.
13. Wear, F. C. (1979}. Microwave-vacuum drying system. Progress Report No.5
(Final}. C00-2918-Tl. U.S. Department of Energy, Washington, DC.
14. Gardner, D. R., and Butler, J. L. (1980). Preparing crops for storage with a
microwave vacuum drying system, Proceedings Second International Sympo-
sium on Drying, Montreal, Canada.
15. Velupillai, L., and Verma, L. R. (1982}. Parboiled rice quality as affected by
the level and distribution of moisture after the soaking process. Trans ASAE,
25(5):1450.
MICROWAVE-VACUUM DRYING 339

16. Araullo, E. V., dePadua, D. B., and Graham, M. (1976). Parboiling, Rice
Postharvest Technology, International Development Research Centre, Ot-
tawa, Canada, p. 161.
17. Schiffman, R. F. (1977). Microwave processing for the food industry, Interna-
tional Microwave Power Institute, Edmonton, AB, Canada.
18. Perry, R. H., Chilton, C. H., and Kirkpatrick, S.D. (1963). Chemical Engi-
neer's Handbook, 4th ed., McGraw-Hill, New York.
19. Bhattacharya, K. R. (1985). Parboiling of rice, Rice Chemistry and Technol-
ogy, 2nd ed. (B. 0. Juliano, ed). American Association of Cereal Chemists,
St. Paul, MN, p. 289.
20. Bhattacharya, K. R., and lndudhara Swamy, Y. M. (1967). Conditions of
drying parboiled paddy for optimum milling quality. Cereal Chern., 44:592.
21. Kunze, 0. R. (1984). Physical properties of rice related to drying the grain.
Drying Technol., 2(3):369.
22. Hogan, J. T., and Planck, R. W. (1958). Hydration characteristics of rice as
influenced by variety and drying method. Cereal Chern., 35:469.
23. Goebel, N. K., Grider, J., Davis, E. A., and Gordon, J. (1984). The effects of
microwave energy and convection heating on wheat starch granule transforma-
tions. Food Microstruct., 3(1):73.
24. Wadsworth, J. 1., and Koltun, S. P. (1986). Physicochemical properties and
cooking quality of microwave-dried rice. Cereal Chern., 63(4):346.
25. Webb, B. D. (1980). Rice quality and grades, Rice: Production and Utilization
(B.S. Luh, ed.), AVI, Westport, CT, p. 543.
26. Halick, J. V., and Kelly, V. J. (1959). Gelatinization and pasting characteristics
of rice varieties as related to cooking behavior. Cereal Chern., 36:91.
27. Little, R. R., Hilder, G. B., and Dawson, E. H. (1958). Differential effect of
dilute alkali on 25 varieties of milled white rice. Cereal Chern., 35:111.
 15
Role of Moisture Content in Affecting Head
Rice Yield
Terry J. Siebenmorgen
University of Arkansas, Fayetteville, Arkansas

I. INTRODUCTION
Rice quality is judged by criteria different from those used for other grains.
While several grading factors are used in determining rice quality, one of
the primary indices is head rice yield (HRY). The term "head rice" denotes
milled rice comprised of kernels three-fourths or more the original kernel
length (1). "Brokens" are milled rice fragments other than head rice. It is
critical to maintain optimum conditions during drying, storage, and milling
operations to obtain high HRY (2,3). Fryar et al. (4) quantified the eco-
nomic value of HRY as $0.075 per cwt per percentage point of HRY
reduction. While this value is subject to domestic and international market
trends, it does indicate the relative value of head rice. A rule of thumb
often followed is that head rice is worth twice that of brokens.
It is difficult to ascribe reduction in HRY to a single cause or to assign
relative weights to factors that result in a lower HRY. A major contributing
factor is the environment surrounding the rice kernels, both in the field
after reaching maturity and during postharvest processing operations.
Moisture content (MC) is used perhaps more than any other property in
managing rice from harvest to milling. Many studies have used MC as a
benchmark in quantifying the effects of various harvest, drying, storage, and
milling practices on HRY. The MC of a bulk rice sample is dependent on the
MC of individual rice kernels forming the bulk sample. Many investigators

341
342 SIEBENMORGEN

(5-8) have indicated that large differences in individual kernel MCs exist
throughout maturation and at harvest. Nelson and Lawrence (9) have shown
that in equilibrated (no moisture exchange between the kernel and surround-
ing air) samples of corn, large differences in kernel-to-kernel MCcan exist.
Siebenmorgen et al. (10) have shown similar differences in rice. Since there
is kernel-to-kernel MC variation, different MC distributions could presum-
ably result in the same bulk MC. Differences in HRY have been observed in
samples at the same bulk MC; therefore, some of the variability in HRY
could be due to different individual kernel MC distributions in samples.
This chapter summarizes recent research that has addressed the interrela-
tionship between rice MC, at various stages from preharvest to milling, and
resultant HRY. A summary of past research pertinent to the subject area
being scrutinized is presented as a prelude to detailed discussion on current
studies. Specifically, preharvest studies leading up to and directly involving
the correlation of HRY to individual kernel MC distributions within sam-
ples are first described. A second area is research addressing the rewetting
of rice and resultant effects on HRY. Finally, recent work on the effects of
rice properties, including MC, and settings of the McGill #2 rice miller on
HRY and degree of milling (DOM) is presented.

II. PREHARVEST AND HARVEST MOISTURE CONTENT

EFFECTS
A. Previous Studies
The rice kernel starts developing after pollination and fertilization. Matura-
tion continues over a period of time that is longer in temperate climates
than in tropical climates and is characterized by a general decline in MC.
When the kernel is fully mature, its MC is no longer controlled by moisture
transfer from within the plant, but varies in response to the surrounding
environment. This dependence on environmental conditions is a result of
the hygroscopic nature of the rice kernel, in that it loses moisture (desorp-
tion) when the kernel MC is greater than the equilibrium MC of rice
associated with given temperature and relative humidity (RH) conditions
and gains moisture (adsorption) if the conditions are reversed.
Milling quality is directly affected by the condition of rough rice prior to
harvesting. Since MC is often used as an indicator of kernel maturity, it is
one of the most important factors affecting the field yield of rough rice and
the HRY. Rice should be harvested at a proper stage of maturity, or at an
optimal MC, to maximize both the yield of rough rice and the HRY.
Rice kernels in a field, and even on the same panicle, do not reach a
given maturity at the same time. Thus, at harvest, some kernels may al-
MOISTURE CONTENT AND HEAD RICE YIELD 343

ready be well past maturity, while others may still be immature. During
normal weather conditions at harvest, rice will lose moisture during the day
but gain moisture at night because of high relative humidity (RH) and dew.
Rain will cause dramatic MC change. Rapid moisture adsorption by ker-
nels with low moisture content can have detrimental effects on HRY. How-
ever, harvesting rice at a high MC can result in a large proportion of
immature kernels, which often break during milling. Therefore, harvesting
rice at too high or too low a MC may result in significantly reduced HRY.
Many studies have quantified the effect of MC at harvest on HRY and
have reported different optimal harvest MCs. Smith and Jones (11) sug-
gested that rice be harvested at a MC of about 23-28% in order to obtain
the maximum HRY and field yield. McNeal (12) reported that the highest
HRY of four varieties tested was obtained when the harvest MC was in the
16-23% range.
Kester et al. (13) reported that the optimal MC for harvesting short-
grain rice in California to obtain maximum HRY occurred between 25 and
32%. Steffe et al. (14) reported a parabolic relationship between HRY and
harvest MC for a short-grain rice variety. They recommended that most
early-maturing rice grown in California should be harvested at an average
MC of 24-26% in order to obtain the maximum HRY and total milling
yield.
In a study on the effect of harvest MC on the HRYs of two medium-
grain varieties, M-101 and M- 103, K. E. McKenzie (personal communica-
tion) reported that the HRYs for both varieties began to decrease as har-
vest MC declined from approximately 23%. The two varieties showed
different responses to changes in MC.
Determination of the optimal harvest MC is complicated by the fact that
the relationship between HRY and preharvest MC can be variety depen-
dent. Cultural practices, climatic conditions, and soil types are also consid-
ered to be factors, although Kunze et al. (15) state that management prac-
tices affect the number of rice kernel fissures only to the extent that they
affect grain MC. In addition to the quality aspect, farmers must also con-
sider other factors such as harvest losses and drying costs since these factors
are also a function of MC. Thus, the optimal harvest MC for maximizing
HRY may not necessarily be the same as that for maximizing economic
return, which would include considerations of field losses, drying costs, and
harvest capacity.

B. Recent Research
Work by Counce et al. (16) has also shown that the MC at which rice is
harvested is critical in determining HRY. A study was conducted over a 2-
MOISTURE CONTENT AND HEAD RICE llELD 345

mately the same throughout the harvest season. The average MC trend for
the 0-week time of draining in 1988 is attributed to the frequent occurrence
of rain. Thus, the trends in the MC and HRY data of this study indicate
that there were interactive effects among soil type and other variables,
most notably weather conditions.
The trend of increasing HRYs as MC decreased at the early stages of
harvest can be attributed to the decline in the number of immature kernels.
If MC is an indicator of maturity, then as average MC decreases, kernels
reach maturity and attain sufficient structural strength to withstand the
stresses of milling without cracking. Draining at heading apparently did not
allow this maturation process to occur in several situations.
The above research indicates that HRY is affected by the MC of rice at
harvest. Harvesting rice at too high or too low a MC results in reduced
HRY, but for different reasons. At the high end of the harvest MC spec-
trum, the presence of immature kernels appears to reduce HRY. Con-
versely, at the low end of the MC spectrum, HRY reduction results from
rapid adsorption of moisture and concomitant development of fissures that
increases the propensity of rice kernels to break during milling.
1. Variation in Individual Kernel Moisture Contents
a. Preharvest Moisture Content Variation
The work of several investigators (17-19) has indicated that when rice
below a critical MC rapidly adsorbs moisture, fissuring will occur with
resultant reductions in HRY. This knowledge has prompted researchers to
investigate the possibility that preharvest HRY reduction may be corre-
lated to the number of low-MC kernels adsorbing moisture.
Chau and Kunze (7) showed that average MC differences ranged from
21 to 29 percentage points for groups of grains harvested from the top of 10
of the most mature panicles and from the bottom of 10 of the least mature
panicles on a given day during the normal harvest season of medium-grain
Brazos rice. This work identified a need for accurately quantifying the
kernel MC variation of rice throughout harvest.
The means for addressing this need were provided by the introduction of
the Shizuoka Seiki CTR-800A single kernel moisture meter (20). This
meter is capable of measuring the MC of individual kernels of rice and will
display the MC of each kernel tested, the average MC of a sample, and the
standard deviation of kernel MCs. Calibration by Siebenmorgen et al. (21)
showed the meter to be in close agreement with an oven-drying method in
measuring both individual kernel and sample average MCs.
Kocher et al. (5) used this moisture meter to measure kernel-to-kernel
MC variation in Katy, a long-grain rice variety. Kernel MC frequency
distributions as measured throughout the 1988 harvest season are shown in
 346 SIEBENMORGEN
...., ....,
Ill
,... HARVEST DATE 238
Ill
.... HAAYEST DATE UO

......Jz

-...
....z
..J

....0:1200 ....a:,..,.
..."
0
..."
0 ...
....a:
...
n:
....ID
,. ,.::>
ID

.
:! z

0
• 10
• 5 ,
MOISTURE CONTENT (S) MOIS~UAE CONTENT (S)

....,
. 1600
KAAYEST OAT£ 243

Ill
,... HARVEST DATE 250

~ ....z
..J

ffi tZOO ....a: ,..,.

"...
0 ... ..."
...
0 ...
a:
w
,.::> ....a:CD
ID
,.
z ~ 400

0
5

MOISTURE CONTENT (I) MOISTURE CONTENT II)

Ill
....z..J
1600
HARVEST OA TE 253

Ill
..J
.... H4RVEST OA TE 2!16

a:,..,.
UJ
ffi 1200
z

..."
0 ...
""'
...
0
a: a:
UJ
CD
%
~ 400
"':>:
CD
::>
z

MOISTURE CONTENT (I) MOISTURE CONTENT i')

.
....z
..J
1600
HARVEST DATE z,e

....
([ 1~0

...O" BOO
a:
....CD
~ 400

j _,... -~1'
C'O
...... ••·······
25 30 35
-.-......,
•o •s
MOISTURE CONTENT Ill
MOISTURE CONTENT AND HEAD RICE YIELD 347

Figure 2. The early harvest dates showed a trimodal kernel MC distribution.

As time progressed and the rice matured, the frequency distribution peaks
for kernels with high MCs decreased and the peaks for kernels with low MCs
increased. This process continued until the last harvest date, which generally
showed a single mode. The cause of these multimodal distributions was
speculated to be either kernel location effect on panicles or kernel maturity
differences resulting from different tiller emergence times.
b. Postharvest Moisture Content Variation
Based on the fact that a large variation in kernel MCs was present at
harvest, especially if the rice was harvested at high average MC, Sie-
benmorgen et al. (10) conducted a study to determine the kernel MC
variation in apparently equilibrated samples. Rice that had been harvested
and stored at 18% MC for one year was conditioned to average MCs
ranging from 9 to 37%. Individual kernel MC distributions were deter-
mined at 3 weeks and 3 months after conditioning using a Shizuoka Seiki
CTR-800A moisture meter. There was no difference (p < 0.05) in the
average MCs of the samples measured after either equilibration time pe-
riod. However, a characteristic response in the standard deviation of the
individual kernel MCs was observed for both storage periods (Fig. 3). The
standard deviation increased exponentially from approximately 0.5% at
10% average MC to 4.5% at 26% MC. The coefficient of variation also
increased with average MC. The dramatic increase in standard deviation
with average MC and the increasing coefficient of variation suggested that
the hygroscopic response of rice kernels within a sample to MC change was
different. The cause of this varying hygroscopic behavior was hypothesized
to be due in part to varying kernel maturities when the rice was harvested.
Because of the variability in kernel MCs at harvest, an associated variabil-
ity in kernel hygroscopic behavior would also be expected.
2. Correlation of Kernel Moisture Content Variation to Head
Rice Yield
The progression of research findings described above, along with the capa-
bilities afforded by the individual kernel moisture meter, prompted further
studies investigating the relationship between individual kernel MC distri-
bution at harvest and HRY. Specifically, a hypothesis was formulated that
the number of kernels with MC less than a critically low level prior to
moisture adsorption could be correlated to HRY reduction. Kocher et al.
(5) give a summary of research that supports such a hypothesis.

Figure 2 Frequency distributions of individual rice kernel moisture contents

throughout a harvest season for a May 3, 1988, seeding date. (From Ref. 5.)
348 SIEBENMORGEN

5.0~-----------------------------------------,

.0

N
~ 4.5

5 4.0
·-
-<J

-~ 3.5
>
m
1J
3.0
1J
-oJ
01 2.5

0CTR-800A. 3wks of aqulllbratla-1

AcTR-SOOA. 3nns of equilibration

0 Individual kernel oven
test. 3nns of equlllbratla-1
0.010~-----L------J-------~-----L------~----~40
15 20 25 30 35
Bulk sample oven moisture content. Zw.b.

Figure 3 Average indiviudal kernel moisture contents standard deviations deter-

mined from MC data taken after 3 weeks and 3 months of equilibration using a
Shizuoka Seiki CTR-800A moisture meter. (From Ref. 10.)

Tests were conducted by Siebenmorgen et al. that directly addressed this

hypothesis. Time-of-harvest experiments were conducted using the long-
grain variety Newbonnet in 1989 and two long-grain varieties, Lemont and
Tebonnet, in 1990. Rice was harvested on eight dates for the Newbonnet in
1989 and on 13 dates for the Lemont and 15 dates for the Tebonnet in 1990.
Immediately following each harvest, individual kernel MCs were deter-
mined using a Shizuoka Seiki, CTR-800A moisture meter. Samples for
HRY determination were refrigerated at approximately 1°C, subsequently
dried to 12.5% MC and milled.
The trends in MC frequency distribution patterns were in agreement
with those reported by Kocher et al. (5). Tables 1 and 2 summarize the
average MCs, total milling yields (TMYs), and HRYs for the Newbonnet
and Lemont rice, respectively. As harvest progressed, environmental condi-
tions had an increased influence on kernel MCs. Frequently, rice harvested
at a later date had a higher MC than earlier dates because of rewetting by
rain. Table 1 shows that TMY for the Newbonnet rice increased as the
average MC decreased to about 22%. Below this MC level, TMYs did not
MOISTURE CONTENT AND HEAD RICE YIELD 349

Table 1 Average Moisture Contents at Harvest, Total Milling

Yields and Head Rice Yields for Newbonnet Rice, 1989

Harvest date
Average
Calendar DOY• MCb (%,w.b.) TMYC(%) HRYC(%)

Sept. 11 254 25.7 62.3c 39.7d

13 256 22.9 65.7b 49.7c
18 261 21.4 67.6a 54.0a
20 263 19.8 67.7a 54.7a
22 265 18.1 67.9a 55.3a
25 268 17.7 68.1a 55.5a
27 270 15.0 67.7a 54.2a
Oct. 2 275 17.8 68.3a 52.2b

•Day of year.
bMean of 8000 kernel MC measurements.
<Mean of 8 milling determinations. Means within a column followed by
the same letter are not significantly different at the 0.05 level as deter-
mined by Duncan's multiple range test.
Source: Ref. 21.

Table 2 Percentages of Kernels with Moisture

Contents Below Given Levels for Lemont Rice, 1990

Harvest date Percentage of kernels with MCs

Calendar DOY• <12.5% <13.5% <14.5%

Sept. 25 268 0.0 0.0 0.2

27 270 0.1 0.1 0.1
Oct. 2 275 0.1 0.4 8.7
11 284 0.1 0.1 0.3
16 289 0.4 1.3 20.0
18 291 0.6 1.9 28.8
24 297 0.3 1.0 11.4
26 299 0.8 9.1 45.1
29 302 71.5 95.5 99.0
31 304 30.7 77.3 96.3
Nov. 7 311 0.2 1.8 20.6
12 316 1.0 13.7 61.4
14 318 10.6 52.5 90.5

•Day of year.
Source: Ref. 21.
350 SIEBENMORGEN

show a significant change throughout the remaining harvest season. The

TMYs for the Lemont or Tebonnet rice also did not show any appreciable
change after the average MC reached about 22%.
The results in Table 1 show that HRY for the Newbonnet variety in-
creased until the average MC was about 22%. Thereafter, HRY did not
show a significant change until the average MC was about 15%. The HRY
from the last harvest date was lower than those from earlier harvest dates
(Table 1). It appears that HRY began to decrease after MC reached 15%.
However, no dramatic reduction in HRY was observed for the Newbonnet
rice. It is speculated that this was because the harvest was ended above the
critical MC level for most of the kernels.
a. Environmental Effects on Head Rice Yield
Daily temperature, RH, and rainfall data were analyzed for both harvest
seasons to investigate the influence of environmental conditions on HRY.
Figure 4 shows HRYs and average MCs for the Newbonnet rice and the
incidence of rain during the 1989 harvest season. Figures 5 and 6 show
similar data for the Lemont and Tebonnet rice, respectively, plus the calcu-
lated equilibrium MC based on the daily average temperature and RH
during the 1990 harvest season. The rain data shown in these figures only
reflect those dates on which there was rain and do not represent rain
amount. As seen in Figures 5 and 6, the rice MCs were increasingly influ-

100·~------------------------------------------rSO

··•·· HAY ·+·· MC ~ RAIN

80 40

!':.
:2 60
~
CD
.!:!
a:
"tl 40
Cll
CD
I

20

0
250
Harvest Time (Day of Year)

Figure 4 Head rice yield (HRY) and average moisture content (MC) for
Newbonnet rice on each harvest date and the incidence of rain during the 1989
harvest season. (From Ref. 21.)
MOISTURE CONTENT AND HEAD RICE YIELD 351

100.--------------------------------------------.50

• HAY + MC - EMC ~ RAIN

80

'#.
"tl 60
~
>-Q)
a:
(J

"tl 40

......
Cll
Q)
J:

20

0
260
Harvest Time (Day of Year)

Figure 5 Head rice yield (HRY) and average moisture content (MC) on each
harvest date for Lemont rice, calculated equilibrium moisture content (EMC)
based on the daily average ambient air conditions, and the incidence of rain during
the 1990 harvest season. (From Ref. 21.)

100,---------------------------------------------~50

80
··•·· HAY ··+· MC - EMC ~ RAIN
40 ~

.ri
'#. ~
"tl
a; 60 30 iE
> ·•················•······•··································•·········. Q)

·. E
Q)

a:
(J
0
20 ~
,._ .,'~\".:- ~·.: :·. : ...
"tl 40
Cll :;
Q)
u;
J:
·a
~
20 10

o~+-~~--.-u.~~~-.--~~~-.,--.~~~.---+0
260 270 280 290 300 310 320
Harvest Time (Day of Year)

Figure 6 Head rice yield (HRY) and average moisture content (MC) on each
harvest date for Tebonnet rice, calculated equilibrium moisture content (EMC)
based on the daily average ambient air conditions, and the incidence of rain during
the 1990 harvest season. (From Ref. 21.)
352 SIEBENMORGEN

enced by the ambient air conditions after the average MC reached about
17%. In the 1989 experiment, there was about 1.0 em of rain (September
23, DOY 266) when the rice MC was about 18% (Fig. 4). At this high MC,
HRY was not affected when the rice was harvested 2 days after the rain
(Table 1). However, rain did show some effect on HRY (2 percentage
points reduction) when the Newbonnet rice was harvested after a rain was
received at a time when the average MC was about 15%.
The 1990 experimental results for the Lemont and Tebonnet rice (Figs. 5
and 6, respectively) show that rain had an adverse effect on HRY only after
the average MC reached 15%. Dramatic HRY reductions occurred for both
Lemont and Tebonnet as a result of rewetting of low-MC rice by rain.
Lemont HRY dropped 37.5 and Tebonnet 23.1 percentage points over a
time interval in which rain began on DOY 308. Based on these reductions
and the fact that both varieties had reached 12% MC prior to the rains, the
Lemont rice appeared to be more susceptible to fissuring than Tebonnet.
The results in Figures 5 and 6 show that HRYs for both Lemont and
Tebonnet varieties had not been reduced when rice was harvested at 12%
average MC, which was believed to be below the critical MC level. How-
ever, it is to be noted that it had not rained for several days prior to the
dates when these MCs were measured and thus conditions for rapid mois-
ture adsorption did not exist. Further, when the Lemont rice was harvested
at 12.9% MC (October 31, DOY 304), 2 days after the rice had reached
12.1% MC, a HRY reduction of only 1.6 percentage points was measured.
Figure 5 indicates that there was no rain over this 2-day period and thus the
increase in MC was attributed to dew. Therefore, it appears that dew or
frost did not fissure a large percentage of the rice kernels, even at this low
MC level.
In summary, HRY reduction for the three rice varieties appears to have
been mainly caused by rewetting of rice kernels by rain. Rain had no
discernible effect on HRY when the average MC was above 15%. Severe
HRY reductions occurred as a reuslt of rewetting by rain after kernel MCs
reached about 12%. These results agree with the work of other investiga-
tors (17 ,19).

b. Head Rice Yield vs. Kernel Moisture Content Distribution

The percentages of rice kernels with MCs below 12.5, 13.5, and 14.5% for
the Lemont and Tebonnet rice are shown in Tables 2 and 3, respectively.
Head rice yield on a given date was first correlated to the percentage of
kernels with MCs below given levels for that particular harvest date. The
correlation coefficient was found to be very low because the percentage of
kernels below an assumed critical MC level could be very low, but unless
rain or some other means of rewetting occurred, a corresponding reduction
MOISTURE CONTENT AND HEAD RICE YIELD 353

Table 3 Percentages of Kernels with Moisture

Contents Below Given Levels for Tebonnet Rice, 1990
Harvest date Percentage of kernels with MCs
Calendar DOY• <12.5% <13.5% <14.5%
Sept. 17 260 0.0 0.1 0.2
19 262 0.0 0.0 0.1
24 267 0.0 0.1 0.2
26 269 0.1 0.1 0.1
28 271 0.0 0.0 0.0
Oct. 1 274 Data missing due to data collec-
tion error
5 278 0.3 0.6 13.1
12 285 0.2 0.4 0.7
15 288 0.7 1.4 25.3
17 290 0.6 1.5 40.7
19 292 3.5 44.5 87.5
25 298 3.6 36.7 80.7
30 303 73.7 97.2 99.7
6 310 0.2 1.4 22.9
13 317 4.7 46.1 92.3
•Day of year.
Source: Ref. 21.

in HRY did not appear. Since the HRY reduction usually occurred as a
result of rewetting by rain, it was deemed more logical to correlate HRYs
from those harvest dates after a rain to the percentages of kernels with MCs
above (or below) a given MC level measured on the harvest date immedi-
ately before a rain.
It was found that HRY on a given harvest date after a rain was most
favorably correlated to the percentage of kernels with MCs above 10.5%
for Newbonnet, 13.5% for Lemont, and 12.5% for Tebonnet rice as mea-
sured immediately prior to the rain. Figures 7 and 8 illustrate this correla-
tion for Lemont (r = 0.992) and Tebonnet rice (r = 0.991), respectively. It
appears from the above discussion that the critical MC level for individual
kernels was about 13% for the Lemont and Tebonnet rice and about 11%
for the Newbonnet rice.
In summary, the study by Siebenmorgen et al. (21) showed that TMY and
HRY were a function of MC at harvest. Further, there was a relationship
between individual kernel MCs and HRY reduction. However, this relation-
ship must be viewed as a correlation of the number of kernels below (or
354 SIEBENMORGEN

100~------------------------------------------r120

+··························+-·-··-··········+···········-··+
.. ~ ....... ~'"!: 100
80 (..)
~

~ \\ 80 ~

...
10
"0 cr)
a; 60
>Gl •··········································-····-·····-·-······\\.
.8
Gl
~--~. 60

.,' , ~\
0
a: c(
"0 40 U)
a;
as --- HRY -+- % Ko-
Gl
:I:
40 E
,
Gl
.......... \\\ ~

··....
~
20 ·.. + 20

0 0
270 280 290 300 310
Harvest Time (Day of Year)

Figure 7 Comparison of head rice yields (HRYs) of Lemont rice harvested after
raining to the percentages of kernels with MCs above 13.5% measured before
raining. (From Ref. 21.)

100 120

+-----~---------r-----+------~---+ 100
80 (..)
~
~ 80 ~
10
"0 60 N
a;
>Gl .
•················•·····························•················•··············•··········•
60
Gl
~
0 .J:J
a: 40
.......... c(
"0 .......... .!1!
as
Gl
:I: -+- % Kernels
·.·. 40
Gl
E
··•·· HAY ~
Gl

20 + ~
20

0 0
265 275 285 295 305 315
Harvest Time (Day of Year)

Figure 8 Comparison of head rice yields (HRYs) of Tebonnet rice harvested

after raining to the percentages of kernels with MCs above 12.5% measured before
raining. (From Ref. 21.)
MOISTURE CONTENT AND HEAD RICE YIELD 355

above) a critical level on a given date to the HRY measured on a subsequent

date after a severe rewetting condition, such as rain, has prevailed.

III. EFFECT OF MOISTURE ADSORPTION/DESORPTION

ON HEAD RICE YIELD
A. Moisture Adsorption/Desorption Studies
Kondo and Okamura (22) were perhaps the first to recognize the influence
that moisture adsorption had on fissure development in rice kernels. Their
studies showed that cracks developed after about 1 hour of adsorbing
moisture and that the percentage of cracks tended to increase with time.
Stahel (22) showed that addition of moisture when the rice MC was more
than 15% had no effect on the HRY, but at MCs below 15%, significant
drops in HRY were observed. He concluded that it was not rapid drying
that caused development of fissures in the kernel and consequent breakage
but the reabsorption (through adsorption) of moisture by the kernel, and
that these effects manifested themselves only when the grain MC was be-
low about 14%. Similar results were obtained by Grant in Burma and
Coyaud in Vietnam (22), Desikachar and Subramanyan in India (23), and
Kunze (24) in the United States.
Kunze and Choudhary (25) studied the effect of moisture adsorption on
tensile strength of rice. They hypothesized that when a kernel adsorbed
moisture at the surface, the swelling of cells in the surface layers produced
compressive stresses on the kernal surface. These stresses were balanced by
development of tensile stresses in the inner portions of the kernel. If,
however, the compressive stresses developed at the surface resulted in
tensile stresses in the center far in excess of the tensile strength of the
kernel, failure would occur.
Recent research (26) corroborates this hypothesis. Lague used a finite
element model to simulate preharvest stress cracking of rice kernels. Simu-
lation results showed that cracking was initiated when the kernel was under
tensile stresses during moisture adsorption. Results also showed that
endosperm cracking first occurred at the center of the kernel and that it
could progress both in the radial and longitudinal directions. This also
explained the observations made by Henderson (27) using X-ray films,
which showed that rice fissures began at the center of the kernel and
progressed toward the minor circumference and that there could be frac-
tures in many directions.
One of the earliest analyses of the effect of artificial drying on HRY was
by Schmidt and Jebe (28). Their results indicated that it was drying rate
356 SIEBENMORGEN

rather than drying air temperature that resulted in HRY reduction, a con-
clusion consistent with later observations.
Sharma and Kunze (2) studied postdrying fissure development in rough
rice. Rough rice was dried from MCs of 20% and 14% using air at 60°C for
2, 10, and 24 hours. Dried rice was stored in air-tight vials, and appearance
of fissures in rice kernels was monitored for a period of 144 hours. In
general, an increase in drying time resulted in fissuring more kernels, and
more kernels fissured when the MC of the rough rice was low.
A study by Stermer (29) showed that cracks caused by desorption were
irregular and those due to adsorption were straight and more severe. Thus,
fissures resulting from drying would proceed from the surface toward the
center of the kernel. This kind of damage is, however, seldom seen. There-
fore, postdrying adsorption of moisture appears to be a more important
factor than drying per se (2).

B. Other Factors Influencing Fissure Development

Although moisture adsorption is an important factor in the development of
fissures in kernels, other factors such as variety and chemical composition
also influence the extent of fissuring. Rice varieties and types have differ-
ent fissuring characteristics. For instance, short-grain varieties are more
susceptible to fissure development than long-grain varieties (30). No experi-
mental data are available to quantify the effect of chemical composition of
kernels on fissure tendency, though the observation that varieties similar in
physical characteristics had differing number of fissures does indicate some
kind of dependence (31).

C. Current Research
Since there have been few studies on the effects of rate of moisture
adsorption on HRY, recent research has focused on quantifying these
relationships.
1. Moisture Adsorption Rates of Rough Rice
Banaszek and Siebenmorgen (32) exposed Newbonnet long-graint rice at
MCs of 9.0, 10.5, 12.0, 13.5, and 15% to an airstream with temepratures of
12.5, 20, and 30°C and RHs of 70 and 90%. Samples were removed from
the airstream after durations of 1, 2, 4, 8, 16, 24, 48, 72, 96, 120, 144, and
168 hours. The samples were immediately stored in sealed plastic bags and
MC determined after at least one week in storage.
An important finding of the study was that rice at different initial MCs
reached different equilibrium MCs. Banaszek and Siebenmorgen (33) hy-
pothesized that irreversible changes occurred whent he rice was dried to
MOISTURE CONTENT AND HEAD RICE YIELD 357

the various initial MCs, which resulted in the rice at each level acting as a
different material, with each having its own equilibrium MC.
The rate of moisture adsorption was found to be dependent on the initial
MC and the air RH and temperature. Air RH had a significant effect on the
rate of moisture adsorption as well as on the final MC of the sample. From
Figures 9 and 10 it can be seen that the rate of moisture adsoprtion was
greater at a RH of 90% than at 70%. The effect of temperature on the rate of
moisture adsorption is depicted in Figures 9 and 11. 'The figures show that
RH had more effect on moisture adsorption than did temperature.
The following equation was found to accurately describe the moisture
adsorption data:
MR = exp( -k*f') (1)
where:
MR = (M1-M,)/(M0 -M,), moisture ratio, dimensionless
t =time (hr)
k,n = regression coefficients
The parameter k of the equation was found to be a function of initial MC
and temperature, while the parameter n was determined to be a constant.

20

18
'"'
..ci
~
16
...,
N

+I 14
c
OJ
+I
c 12 .____. 9. 15% 1MC
0 +---+ 9. 92% 1MC
u ~·,..-<~ 10. 94% 1MC
OJ - - - 12. 43% IMC
L
:J
10 H- -f# 13. 53% IMC
+I 1----1 15.35% IMC
(J)
.....
0 8
::::E

6 0 20 40 60 80 100 120 140 160 180

Time (hours)

Figure 9 Moisture adsorption curves for rough rice exposed to an airstream at

20°C and 70% RH. (From Ref. 32.)
358 SIEBENMORGEN

20
o----o 9. 48% JMC
>1- - i < 11.12% JMC
18

16

_,) 14
c(II
_,)
c 12
0
u
OJ
L
:J
10
_,)

....
(/)

0 8
:=<:

6 0 20 40 60 80 100 120 140 160 180

Time (hours)

Figure IO Moisture adsorption curves for rough rice exposed to an airstream at

20oc and 90% RH. (From Ref. 32.)

20

18

...ci
~
16
N
.....,
_,) 14
c
(II

---
_,)
c _____,. 9. 04%
0 12 IMC
u +---+ 10.32% IMC
o---o 10. 90% IMC
(II
12. 43% JMC
L
:J
10 N--Il 13.53% IMC
_,) 1----1 15.49% IMC
....
(/)

0 8
:=<:

6
0 20 40 60 80 100 120 140 160 180
Time (hours)

Figure II Moisture adsorption curves for rough rice exposed to an airstream at

30°C and 90% RH. (From Ref. 32.)
MOISTURE CONTENT AND HEAD RICE YIELD 359

Banaszek and Siebenmorgen (32) list the equation describing k and the
results of using other equations to describe the data. The moisture adsorp-
tion rate was found to be highest when the initial MC of the rice was low.
The rate of moisture adsorption was also found to be directly proportional
to temperature.

2. Head Rice Yield Reductions Resulting from Moisture

Adsorption
Rice samples removed at the times listed above were milled to determine
HRY. Figure 12 shows that HRY reduction (HRYR), defined as the differ-
ence between the HRY at time 0 and the HRY at a given time, was not
significant unless adsorption was initiated below a MC of approximately
12%, which is consistent with the field results for Newbonnet reported by
Siebenmorgen et al. (21). The HRYR at 20°C and 90% RH was almost
identical to that at 30°C and 90% RH (Fig. 12), showing that air tempera-
ture had little effect on HRYR. Comparison of Figures 12 and 13 shows
that the reduction in HRY when the RH was 70% was much less than when
at 90%. Banaszek and Siebenmorgen (19) present complete HRYR data
for this study.

24
22

c 20 *
.....0 *
...,,.... 18
u (J)
16 .----. 9.04% 1MC
:::J+'
lJC
* o-- .<J 10. 32% 1MC
01 ..... 14 +-·-+ 10.90% 1MC
,_ _ __.., 12. 43X 1MC
L 0
Q.
12 * fl--11 13.53% IMC
lJ i-··-1 15.49% IMC
_, 01
01 m 10
..... 0
>-+' 8
c 0 0
01 01
uu 6 +
·.-< L +
L 01 4
Q.
lJ'-"
0 2
01
I
0
@

2468 10 12 14 16 18 20 22 24 26

Time (hours)

Figure 12 Experimental data and head rice yield reduction curves for rough rice
exposed to an airstream at 30°C and 90% RH. (From Ref. 19.)
360 SIEBENMORGEN

24
22

c 20
......,,..,
0
18 .----. 9.48%
o---o 11.12% IMC
IMC

u (J)
16
::>...,
"U c
Ql ... 14
L 0
c.. 12
"U
-w
wm 10
... 0
>....., B
c
Ql Ql
uu 6
... L
L Ql
*
4
"U'-J
c..
* *
0
Ql
:c
2
*-*- * 0
--
0
0
0
0
-2
0 2 4 6 B 10 12 14 16 1B 20 22 24 26

Time <hours)

Figure 13 Experimental data and head rice yield reduction curves for rough rice
exposed to an airstream at 20°C and 70% RH. (From Ref. 19.}

An empirical equation was used to describe the data:

HRYR=--- (2)
a + b*t

where:

HRYR = (Ho- H,) (%)

H, = HRY at timet(%)
H0 = HRY of control (time= 0) (%)
t =time (hr)
a, b = regression coefficients

The values of coefficients a and b were:

IMC = initial moisture content(%)

RH = relative humidity(%)
a= a 2 * IMC
b = IMC2 * (a 3 + b3 * RH + c3 *IMC) + d3
MOISTURE CONTENT AND HEAD RICE YIELD 361

where:
a2 = 0.011674512
a3 = -0.027676183
b3 = -0.011442723
c3 = 0.003100051
d3 = 1.073776072
Equation 2 was compared with a "crack generation" equation developed by
Nishiyana et al. (34). The crack generation equation requires the lowest
asymptotic HRY obtained during a test as an input to the equation and is
thus analogous to equilibrium MC used in the moisture ratio term of Eq. 1.
In terms of mean square error, the crack generation equation was superior
to Eq. 2 in predicting HRYR due to moisture adsorption. However, since
the lowest asymptotic HRY is usually not known, Eq. 2 was deemed more
appropriate at this time for general prediction of HRYR due to moisture
adsorption.
3. Head Rice Yield Reduction Rates
To obtain the rate of HRYR, the partial derivative of Eq. 2 with respect to
time was used:
d(HRYR) a
(3)
dt [a+ b*t] 2
Rate of HRYR curves were generated using Eq. 3 and are shown in Figures
14 and 15. Figure 14 displays the HRYR rates for various levels of initial
MC at an air condition of 30°C and 90% RH. The lowest initial MCs
produced the highest HRYR rates. Increasing the initial MC level corre-
spondingly decreased HRYR rates. For example, the 9% initial MC level
in Figure 14 showed a HRYR rate of 1 percentage pointlhr at an exposure
time of 6 hours, whereas with the higher initial MCs, the reduction rates
had reached near minimum within 6 hours.
Figure 15 shows the effect that RH had on HRYR rates at the conditions
of 9% initial MC and 30°C. At the 90% RH, the rate of HRYR is initially
high and then reduces exponentially. The 70% RH level resulted in a
similar trend but with much less drastic results. The 90% RH showed
HRYR rates at least five times greater than the 70% level for the first 10
hours. The rates converged after 24 hours of exposure.
4. Effects of Adsorption Rate on Head Rice Yield Reduction
The curves presented in Figure 16 were generated using Eqs. 1 and 2 and
represent the amount of HRYR incurred in a time interval as a function of
 362 SIEBENMORGEN

c 5
0 - - 9.07. 1MC
- -10.57. 1MC
+' - · -12. Oi. 1MC
u
:J ----13.57. 1MC
""Or..
CII..C 4
L'-..
(/)
lJ+J
-c
(II .....
..... 0 3
>-.0...
(II (II
u m
..... D
L+J
c 2
lJ (II
D U
Cll L
..c (II
0...
4-'-'
0
(II
+'
D
a:::

Time (hours)

Figure 14 Rate of head rice yield reduction as affected by initial moisture con-
tent at the condition of 30°C and 90% RH. (From Ref. 19.)

c 5
0
+' - -707. RH
u --907. RH
:J
""Or..
ai..C 4
L'-..
(/)
lJ+J
-c
(II .....
..... 0
>-.0... 3
Cll (II
um
.,... D
L+'
c 2
lJ(II
D U
al L
..c Cll
0...
4-'-'
0 1
Cll
+'
D
a:::
0
0 2
-- 4 68 10 12 14 16 18 20 22 24
Time Chours)

Figure 15 Rate of head rice yield reduction as affected by relative humidity at

the condition of 9% initial moisture content and 30°C. (From Ref. 19.)
MOISTURE CONTENT AND HEAD RICE YIELD 363

7 9.0% lMC
c
....
0
6
"':J+'
u"' "(}I
"'
"U c
w.... 5
L 0
a.
"U
...... w 4
WIJJ
.... 0 10.5% lMC
>-.+'
c 3
ww
uu
.,.. L
L W 2
a.
"U>J 12. 0% lMC
0
w
:r:

0 0 •4 .8 1.2 1.6 2

Average adsorption rote C7.MC/h)

Figure 16 Head rice yield reduction as affected by adsorption rates for 1-hour
intervals at the condition of 30°C and 90% RH. (From Ref. 19.)

the corresponding average adsorption rates for 1-hour intervals at the condi-
tion of 30°C and 90% RH. Average adsorption rates were calculated as
follows:
. %MC(t+ll - %MC<,>
Average adsorption rate = __..:...._..:....___~ (4)
..1t
The first point at the right of each curve represents the first hour interval.
The lowest initial MC of 9% resulted in the highest average adsorption rate
(1.6% MC/hr) and associated HRYR (7 percentage points) during the first
hour. During the second hour, the average rate of adsorption decreased
dramatically to approximately 0. 7% MC/hr but resulted in an additional
HRYR of 4 percentage points. As the initial MC level was increased, the
corresponding average adsorption rate and the amount of HRYR for each
time interval decreased. For a given average adsorption rate, there was not
a single level of HRYR. Thus, it appears that not only is HRYR a function
of adsorption rate, but that it is also a function of the initial MC at which
adsorption is occurring.
In summary, the work on adsorption effects revealed that most of the re-
ductions in HRY occurred fairly rapidly, generally within the first 8 hours of
exposure to adsorptive environments. It appears that significant HRY reduc-
364 SIEBENMORGEN

tions occurred in the long-grain variety Newbonnet only when initial MCs
were below 12%. Initial MC played an important role in determining the
extent of HRY damage and also directly determined the equilibrium MC.

rv. MOISTURE CONTENT EFFECTS DURING RICE

MILLING
A. Factors Affecting Rice-Milling Quality
It is well documented that MC at the time of milling has a significant effect
on HRY and DOM (35-40). Webb and Calderwood (36) determined that
as MC decreased, HRY increased and DOM decreased. They stated that in
order to obtain a DOM at lower MCs equivalent to that observed at higher
MCs, increased milling with associated HRY reduction was necessary.
Banaszek et al. (35) determined HRY and DOM for rice milled at MCs
ranging from 10 to 16%. A McGill #2 miller was used with milling time
held constant at 30 seconds. They stated that within the range in which rice
was classified as being well milled, MC accounted for more than 10 percent-
age points of change in HRY. The above-cited studies indicate that, in
general, milling at lower MCs produces higher HRYs than at high MCs, if
milling time and miller settings are held constant. Recommendations con-
cerning the proper MC of rice at milling are not apparent in the literature,
although rice standards as given by USDA (41) state that milling yield shall
not be determined when the MC of the rice exceeds 18.0%.
Velupillai and Pandey (42) associated DOM with the milling time used
on a McGill #2 miller. They determined that 65-73% of the bran was
removed in the first 20 seconds of milling. In the case of Newbonnet vari-
ety, as much as 84% of the breakage occurred in the first 20 seconds of
milling. No recommendations for a standard milling time to be used in
routine HRY determinations were made.
Head rice yield and DOM can be influenced by the pressure applied to
the rice during milling. The pressure applied to rice in the milling chamber
of laboratory millers is controlled by the amount and location of the weight
placed on the weight lever. Another factor thought to influence HRY and
DOM is the amount of brown rice placed in the miller.

B. USDA Standards for Rice Milling

The USDA rice-milling standards (1,41) specify the use of the McGill #3
miller or an "approved miller that produces the same results for determin-
ing head rice yield." However, the standards are written for use of the
McGill #3 miller. An initial rough rice sample of 1 kg is required (41).
Since the weights for the McGill #3 miller hang from the end of the weight
MOISTURE CONTENT AND HEAD RICE YIELD 365

lever, there is no adjustment for the position of the weights. Total weight
on the weight holder is adjusted according to the type (short-, medium-, or
long-grain) of rice. The milling duration is usually set at 30 seconds for all
types of rice (41).

C. Current Research
Although the McGill #3 miller is specified by the USDA rice-milling stan-
dards, the McGill #2 is becoming more popular in the rice industry due to
lower initial costs and lower power and sample size requirements.
A study was conducted by Andrews et al. (43) to determine the effects
of rough rice MC, milling time (residence time of rice in the milling cham-
ber), sample size (amount of rice to be used in the miller), and pressure
applied to the rice (location of weight on the weight lever) on HRY and
DOM. The HRYs attained under the various experimental conditions in
the McGill #2 miller were compared to those attained by milling with the
McGill #3 miller.
1. Experimental Design and Procedure
A McGill #2 miller, equipped with an automatic timer and an adjustable
weight to attain different degrees of milling, was used. The pressure on the
rice during milling was controlled by the placement of a 1.5-kg weight on
the weight lever (lever arm). The positions originally chosen for testing
were 6, 12, 18, and 24 em from the center of the saddle to the center of the
weight. However, it was discovered that when the 1.5-kg weight was placed
at 24 em from the center of the saddle, the pressure applied to the rice was
too great for the miller to start.
Rough rice MCs of 9.5, 11, 12.5, and 14% were chosen. Rough rice
samples of 100, 125, and 150 g were milled for durations of 0, 15, 30, 45,
and 60 seconds. Two long-grain varieties, Lemont and Newbonnet, were
used, and the experiment was replicated. Samples milled in the McGill #2
miller were graded for HRY and DOM by the Federal Grain Inspection
Service (FGIS).
As a control for the experiment and a basis of comparison of the McGill
#2 and #3 millers, 40 1-kg samples (five samples from each of the four
MCs for both varieties) of rough rice were milled by the FGIS. The FGIS
milled the rough rice samples in a McGill #3 miller according to the USDA
standard procedure and determined HRY and DOM.
2. McGill #3 Milling Results
The 40 1-kg samples milled by the FGIS in the McGill #3 miller were used
as the basis of comparison for the HRYs obtained under the experimental
366 SIEBENMORGEN

Table 4 Data from 40 1-kg Samples Milled by the FGIS Using

a McGill #3 Miller
Moisture Milling Head rice Standard
content yield3 yield3 deviationb
Variety (% w.b.) (%) (%) (%)
Lemont 9.5 73.3 64.6 0.553
Lemont 11.0 72.4 62.9 0.793
Lemont 12.5 71.6 60.4 0.785
Lemont 14.0 70.7 57.4 0.532
Newbonnet 9.5 72.6 67.9 0.262
Newbonnet 11.0 71.4 65.7 0.380
Newbonnet 12.5 70.4 64.0 0.385
Newbonnet 14.0 69.1 61.7 0.174
•Values are the average of five replications.
bValues are the standard deviation of the five head rice yields compris-
ing the average.
Source: Ref. 43.

settings of the McGill #2 miller. These control HRYs are listed in Table 4,
which shows that HRY was inversely related to MC in both varieties. The
change in MC from 9.5 to 14% accounted for a difference of 7.2 HRY
percentage points in the Lemont and 6.2 HRY percentage points in the
Newbonnet. Table 4 indicates that milling yield was also inversely propor-
tional to MC. This indicates that the performance of the #3 miller in
removing bran, and possibly some endosperm, is greater at higher MCs.
3. McGill #2 Milling Results
The HRY and DOM classification for each combination of variables for the
Lemont and Newbonnet varieties are shown in Figures 17 and 18, respec-
tively. The tables show that MC and HRY were inversely related, just as
they were with the McGill #3 miller. The tables also show that as MC
decreased, the rice became more difficult to mill to a "well-milled" degree,
as indicated by the increasing number of samples that were not "well
milled."

D. Statistical Analysis
Because there were significant higher-order interactions between variety
and the other independent variables, models were fitted independently for
the two varieties. The details of the quadratic response surface for the four
independent variables (MC, milling time, weight placement, rough rice
weight) used to predict HRY are given in Table 5. All of the effects in Table
5 were statistically significant at the 5% significance level. The linear ef-
MOISTURE CONTENT AND HEAD RICE YIELD 367

Figure 17 Milling data for Lemont rice milled in the McGill #2 miller. Head rice
yields are shown as the first number in each block and are the average of two
replications. Degree of milling is classified as: UM, undermilled; LIM, lightly
milled; RWM, reasonably well milled; WM, well milled. Number in parentheses is
the difference between the HRYs obtained with the McGill #2 miller and those
obtained with the McGill #3 miller (see Table 4). Blocks in grey are below the WM
classification; blocks in black are WM, and the HRYs deviate less than two percent-
age points fromt he McGill #3 HRYs; blocks in white are well milled but deviate
more than two percentage points from the #3 miller HRYs. (From Ref. 43.)
368 SIEBENMORGEN

Figure 18 Milling data for Newbonnet Rice Milled in the McGill #2 Miller.
Nomenclature is the same as in Figure 17. (From Ref. 43.)
MOISTURE CONTENT AND HEAD RICE YIELD 369

Table 5 Coefficients of Equation Predicting Head Rice Yields for Lemont and
Newbonnet Varieties Milled in a McGill #2 Miller

Rice variety
Independant Experimental Lemont Std. Newbonnet Std.
variables settings coefficient error coefficient error
Intercept 21.1352 6.2708 106.7517 5.947
Linear comp.
MC(%) (9.5,11,12.5,14) 13.226 0.7329 -1.4099 0.5126
MT(sec) (15,30,45,60) 0.2784 0.0434 0.2088 0.043
WP(cm) (6,12,18) 0.3371 0.1576 -0.1104 0.1565
RRW(g) (100,125,150) -0.4363 0.0606 -0.4299 0.0644
Quadratic comp.
MC2 -0.5983 0.0289 0.041 0.019
MT2 0.0013 0.0003 0.0005 0.0003
wp2 O.ol 0.0038 0.0436 0.0039
RRW2 0.0025 0.0002 0.0019 0.0002
Interactions
MC*MT -0.0175 0.0023 -0.0153 0.0019
MC*WP -0.0194 0.0079 -0.1037 0.0066
MC*RRW -0.0055 0.0019 0.0028 0.0017
MT*WP -0.0024 0.0008 -0.0027 0.0008
MT*RRW -0.0022 0.0002 -0.0011 0.0002
WP*RRW -0.0051 0.0007 -0.0019 0.0006
Overall mean y = 58.5332 y = 60.8527
Residual mean square 1.22 with 273 d.f. 1.25 with 273 d.f.
r square ,:z = 96.10% ,:z = 96.38%
MC = moisture content; MT = milling time; WP = weight placement; RRW =rough rice
weight.
Source: Ref. 43.

fects of the four variables explained most of the variability in the HRYs
when using the McGill #2 miller.
The influence that each of the independent variables had on HRY is
illustrated in Figure 19. Moisture content was the most important factor
affecting HRY while rough rice weight was the least important. Figure 19
illustrates that the relative influence of the experimental variables on HRY
was different for the two varieties. In an effort to explain this difference,
100 kernels were measured from each variety. The average length, width,
and thickness for the Lemont variety were 7.08, 2.35, and 1.73 mm, respec-
tively. The average length, width, and thickness for the Newbonnet variety
were 6.82, 2.06, and 1.59 mm, respectively. At-test showed that there were
370 SIEBENMORGEN

Lemont

Rough Rice Weight

5%

Weight Placement
13%

Milling Time
20%

Newbonnet

Rough Rice Weight

2%

Milling Time
11%

Weight Placement
33%

Figure 19 Contributions of experimental variables in accounting for changes in

HRY as indicated by sum of squares. (From Ref. 43.)
MOISTURE CONTENT AND HEAD RICE YIELD 371

differences at the 5% significance level between corresponding dimensions

of the two varieties. From this, it is postulated that the difference in contri-
butions shown in Figure 19 are due to differences in kernel dimensions
between the varieties.
To plot the response surface for both varieties, two variables (weight
placement and rough rice weight) were fixed. The surfaces are shown in
Figures 20 and 21 for the Lemont and Newbonnet varieties, respectively. In
general, the contours show that for both varieties, HRY decreases as either
MC or milling time is increased.

E. Rough Rice Weight Effects

In general, increasing rough rice weight resulted in fewer broken kernels

and thus an increase in HRY. The rotor action in the McGill #2 miller
is such that the more kernels present in the milling chamber, the less
each kernel is milled per unit time. Thus as rough rice weight is in-
creased, the interactions from one kernel to another are decreased. This
is supported by Figures 17 and 18 showing that as rough rice weight is
increased, the number of well-milled samples decreases. However, there
were interactive effects between rough rice weight and the other experi-
mental variables.
Figure 22 reveals the results of a physical limitation of the McGill #2
miller. The graphs indicate that for the 100-g rough rice weight, varying the
milling time minimally affected HRY. When using 100 g of rough rice,
which resulted in approximately 80 g of brown rice to be milled after
hulling, it was observed that the lever arm of the miller rested on the
milling chamber frame instead of entirely on the rice. It was concluded that
the McGill #2 miller requires a brown rice sample larger than 80 g in order
to prevent the miller from "bottoming out." It is to be noted, however, that
the data in Figure 22 were obtained with a weight placement of 18 em,
which was the greatest weight placement setting, and thus produced the
greatest pressure on the rice. As a result, Figure 22 represents the worst-
case scenario for illustrating the bottoming-out action.
Figure 20 also illustrates the interactive effects of milling time and
rough rice weight. The relative changes in HRY obtained when milling
time was varied were dependent on rough rice weight. As rough rice
weight was increased from 125 to 150 g, HRYs increased for low milling
times, indicating that there was less milling action per kernel. However,
as milling time was increased, HRY for the 150-g rough rice level de-
creased by relatively greater amounts than the 125-g level. This decrease
in HRY reflects the increase in the exposure of the rice to milling or
frictional forces.
 372 SIEBENMORGEN

WP•6 em, RRW•125 g WP•6 em, RRW•150 g

z
80

170
>-
!leo
a:
iso
:r:
40
8.5 n.o 12.1 14.0 n.o 12.1 14.0
Moloturo Content (.,) Moloturo Content ('4)

WP•12 em, RRW•125 g WP•12 em, RRW•150 g

8.5 n.o 12.1 14.0 n.o 12.1 14.0

Molaturo Content (.,) Moisture Content ('4)

WP•18 em, RRW•125 g WP•18 em, RRW•150 g

zj1oeo
>-
~eo
...leo
a:

:r:
40
n.o 12.1 14.0 8.5 n.o 12.1 14.0
Moloture Content ('4) Molature Content ('4)

Figure 20 Head rice yield response surfaces for Lemont rice milled in a McGill
#2 miller. WP = weight placement; RRW = rough rice weight. (From Ref. 43.)
 MOISTURE CONTENT AND HEAD RICE YIELD 373
WP•6 em, RRW•125 g WP•6 em, RRW•150 g

80
~
i70
>-
~eo

:.o
a:

:r::
40
8.1 tt.O 12.1 14.0 11.0 12.1 14.0
Molature Content ("') Molature Content ("')

WP•12 em, RRW•125 g WP•12 em, RRW•150 g

80
~
}ro
>-
1!eo
:.o
a:

:r::
40
... n.o 12.1
Molature Content ("')
14.0 ... 11.0 12.1
Moiature Content ("')
14.0

WP•18 em, RRW•125 g WP•18 em, RRW•150 g

40
11.0 12.1 14.0 11.0 12.1 14.0
Moisture Content ("') Molature Content Cor.)

Figure 21 Head rice yield response surfaces for Newbonnet rice milled in a McGill
#2 miller. WP = weight placement; RRW = rough rice weight. (From Ref. 43.)
 MC=9.5%, WP=18 em MC=11.0%, WP=18 em ~
65 65 ~

l
:!! 55
60
-~
l 601
:!! 55
~
~
~ ~
!l
0:
.,so ~
:g j 50
:1:
milling time milling time
45
1- 15 sec -+- 30 sec --- 45 sec -e- 60 sec J
45
1--- 15 sec -+- 30 sec -liE- 45 sec -i3- 60 sec I
40 40
100 125 150 100 125 150
Rough Rice Weight (grams) Rough Rice Weight (grams)

MC=12.5%, WP=18 em MC=14.0%, WP=18 em

---------
55 55
miUingtime

60 60 ,..,._15sec +:lOMe: +45MC-+ 6osec J

l l
I ss
>
I 55
>
i

~
;;
150 ~
)50
milling time J t:=
45
1--- 15 sec -+- 30 sec -+- 45 sec -i3- 60 sec I 45 ~
z
40 40
125
~
100 125
Rough Rice Weight (grems)
150 100
Rough Rice Weight (grams)
150
0
==
~
Figure 22 Head rice yields produced with the McGill #2 miller for the Lemont variety and the ~
given variables. MC =moisture content; WP =weight placement. (From Ref. 43.) z
MOISTURE CONTENT AND HEAD RICE YIELD 375

F. Moisture Content Effects

Figures 19, 20, and 21 show that MC is a major factor in determining
HRYs. A change in MC from 9.5 to 14% accounted for a change in HRY of
as much as 14.6 percentage points in the Lemont and a change of 17
percentage points in the Newbonnet variety. These figures reinforce the
observations made from the McGill #3 milling results. Figures 17 and 18
show that as MC decreases, so do the number of "well-milled" samples.
This is an indication that bran removal becomes more difficult at lower
MCs.

G. Weight Placement and Milling Time Effects

As weight placement was increased, HRY decreased (Figs. 21 and 22). A
change in weight placement from 6 to 18 em accounted for a reduction in
HRY of 8. 7 percentage points in the Lemont variety and a reduction of
12.2 percentage points in the Newbonnet variety. This inverse relationship
was consistent throughout the data. An analogous trend was found for
milling time in that as milling time was increased, HRYs decreased. How-
ever, milling for given time periods was necessary to attain a well-milled
sample. Tables 5 and 6 show that many of the samples milled for the 15-
second duration were not well-milled samples.
A sample milled in the #2 miller is defined herein to be "equivalent" to
the #3 miller if it was not only well milled, but also deviated less than 2
percentage points from the HRY determined using the #3 miller. Two
percentage points was assumed to account for typical variations associated
with grading. As previously stated, rough rice weights of 100 g were below
the lower limit of the McGill #2 miller and therefore were not included in
making a comparison between the two millers. In the Newbonnet variety,
only two experimental combinations produced equivalent results at the
12.5% MC level. These two combinations were at a weight placement of 6
em, a milling time of 45 seconds, and a rough rice weight of 125 or 150 g.
These same two combinations will produce equivalent results to the #3
miller at the 11.0 and 14% MC levels. In the Lemont variety, more combi-
nations produced equivalent results, but no single combination was suffi-
cient at all four MC levels.
A weight placement of 6 em tended to produce equivalent results in
more experimental combinations than the 12- or 18-cm weight placements.
Aside from weight placement, MC plays such a large role in determining
the ease or difficulty in bran removal that a consistent milling time did not
produce equivalent results across all MCs. At 12.5% MC, a weight place-
ment of 6 em and a milling time of 45 seconds produced equivalent results
to the #3 miller for both varieties. In the Lemont variety, a milling time of
376 SIEBENMORGEN

30 seconds also produced equivalent results. When the milling time was set
at 30 seconds in the Newbonnet variety and at a weight placement of 6 em,
there was less than 2 percentage points difference in HRY from the McGill
#3 at the 12.5% MC level, but the samples were only reasonably well
milled.

H. Summary
In summary, MC was found to be the most significant variable affecting
HRY in both the #2 and #3 millers. As MC decreased, bran removal
became more difficult and HRYs increased. The interactive effects of MC
with milling time, weight placement, and rough rice weights were found to
be significant and greatly influenced HRY and DOM in both varieties.
The effects of milling time, sample size (rough rice weight), and the
pressure applied to the rice (weight placement) were found to be critical
when determining HRYs. Milling time was found to have an inverse effect
on HRY and played an important role in determining DOM. A rough rice
weight of 100 g was found to produce an insufficient amount of brown rice
for the McGill #2 miller. rough rice weights of 125 and 150 g were adequate
for the #2 miller, and 150 g usually produced higher HRYs for equivalent
weight placement settings and milling time. Weight placement was in-
versely related to HRY in both varieties and had a greater effect on HRYs
than did milling time in the Newbonnet variety.
When comparing HRYs from the McGill #2 miller to those of the Mc-
Gill #3 miller, it was determined that equivalent results could be obtained
with the proper settings. However, no one combination was found to pro-
duce equivalent results across all four MCs for both varieties.

V. FUTURE RESEARCH OPPORTUNITIES

Many research opportunities exist throughout the overall spectrum of areas
affecting rice processing, from rice maturation to final milling. It is often
the case that a technological advance in equipment or techniques will allow
researchers to investigate topics that once were only abstractly pictured.
Such was the case in the studies reported herein using the individual kernel
moisture meter. While it was speculated that kernel-to-kernel MC differ-
ences existed in rice samples, this meter enabled the quantification of MC
variation, which has in turn provided insight to practical problems such as
HRY reduction prior to harvest.
Another advancement that is beginning to be used more frequently in
the food-processing industry is the science associated with nuclear magnetic
resonance (NMR) equipment and techniques. Magnetic resonance imaging
MOISTURE CONTENT AND HEAD RICE YIELD 377

(MRI), an extension of NMR spectroscopy, has been used in the medical

industry for some time and is being increasingly applied in food-processing
research (44). A significant advantage of the use of NMR is that the tech-
nique is nondestructive and noninvasive. The NMR signal is sensitive to
the number of nuclei in the sample, the mobility of the nuclei, and the local
chemical and electronic environment of the nuclei (44). Thus, NMR is
readily adaptable for applications of measuring MC and oil concentrations.
Further, it can be used to quantify the state or binding energies of water in
a product.
It is this later capability that prompts the author to speculate about the
potential of NMR for rice-drying and -processing research. If a full or
better understanding of moisture migration within kernels is obtained,
better equipment, controls, and processes can be developed to maximize
drying efficiency and product quality. It would appear that NMR could
also be used to evaluate physical and chemical constituent differences
among varieties in terms of potential for food product use and milling
performance.
Research is currently being conducted at the University of Arkansas in
applying NMR techniques in the rice-processing area. Projects being inves-
tigated at the time of this writing include measuring MC of bulk samples
using NMR (45) and determining the state of water in rice over the full
range of MC levels encountered in rice drying and processing. In addition,
NMR signals from samples at the same MC, but reached through adsorp-
tion or desorption, are being evaluated to investigate hysteresis effects in
rice. Projected work in this area will use MRI to visually represent the
moisture migration process in rice kernels.
It is also believed that another area worthy of investigation is that of
kernel size effects as related to milling performance. The milling study of
Andrews et al. (43) described above indicated that kernel size may have
accounted for some of the varietal differences in milling behavior. It ap-
pears that further understanding of the kernel-to-kernel interactions during
the milling process, and the associated effects of kernel size distribution in
samples, could be beneficial in maximizing HRY. Also, differences in
chemical and physical characteristics of the various rice size fractions could
be potentially useful in rice product development.
Yet another area for future investigation is that of controlling milling
processes to produce desired degrees of bran removal. Although several
techniques, including color comparisons, are being used to measure DOM,
improvements are needed, especially in light of the recent interest in rice
bran and oil as a food product. Measurements from near-infrared reflec-
tance and image processing techniques need to be correlated to accurate
DOM measurement techniques, such as lipid extraction tests.
378 SIEBENMORGEN

ACKNOWLEDGMENTS
The author wishes to thank Mr. Virendra Bhumbla and Dr. Renfu Lu for
the excellent assistance in preparing and editing this chapter.

REFERENCES
1. USDA. (1979). Inspection Handbook for the Sampling, Inspection, Grading,
and Certification of Rice. Agricultural Marketing Service, HB 918-11. Wash-
ington, D.C.
2. Sharma, A. D., and Kunze, 0. R. (1982). Post-drying fissure developments in
rough rice. Trans. ASAE, 25(2):465.
3. Wright, R. E., and Warnock, W. K. (1983). Effect of temperature and vapor
pressure on rice milling quality. ASAE Paper No. 83-3513. Am. Soc. Agric.
Eng., St. Joseph, MI.
4. Fryar, E. 0., Parsch, L. D., Holder, S. H., and Tugwell, N. P. (1986). The
economics of controlling peck in Arkansas rice. Arkansas Farm Res., 35(3):7.
5. Kocher, M. F., Siebenmorgen, T. J., Norman, R. J., and Wells, B. R. (1990).
Rice moisture content variation at harvest. Trans. ASAE, 33(2):541.
6. Wadsworth, J. 1., and Matthews, J. (1985). Milling yields and moisture distri-
bution in rough rice. Trans. ASAE, 28(4):1346.
7. Chau, N. N., and Kunze, 0. R. (1982). Moisture content variation among
harvested rice grains. Trans. ASAE, 25(4):1037.
8. Wadsworth, J. 1., Matthews, J., and Spadaro, J. J. (1982). Moisture content
variation in freshly harvested rice associated with kernel thickness. Trans.
ASAE, 25(4):1127.
9. Nelson, S. 0., and Lawrence, K. C. (1989). Evaluation of a crushing-roller
conductance instrument for single-kernel corn moisture measurement. Trans.
ASAE, 32(2):737.
10. Siebenmorgen, T. J., Banaszek, M. M., and Kocher, M. F. (1990). kernel
moisture content variation in equilibrated rice samples. Trans. ASAE,
33(6):1979.
11. Smith, W. D., and Jones, J. W. (1937). When to cut rice. Leaflet No. 148:1-5.
U.S. Department of Agriculture, Washington, D.C.
12. McNeal, X. (1950). When to harvest rice for best milling quality and germi-
nation. Bulletin 504. Agri. Expt. Station, University of Arkansas, Fayette-
ville, AR.
13. Kester, E. B., Lukens, H. C., Ferrel, R. E., Mohammad, A., and Finrock, D.
C. (1963). Influence of maturity on properties of western rices. Cereal Chern.,
40:323.
14. Steffe, J. F., Singh, R. P., and Miller, G. E. (1980). Harvest, drying and
storage of rough rice, Rice: Production and Utilization (B.S. Luh, ed.), AVI,
Westport, CT.
15. Kunze, 0. R., Turner, E. K., and Turner, F. T. (1988). Fissured rice related to
grain moisture, weather, and fertilization rates. ASAE Paper No. 88-6510.
Am. Soc. Agric. Eng., St. Joseph, MI.
MOISTURE CONTENT AND HEAD RICE YIELD 379

16. Counce, P. A., Siebenmorgen, T. J., Vories, E. D., and Pitts, D. J. (1990).
Time of draining and harvest effects on rice grain yield and quality. J. Prod.
Agric., 3(4):436.
17. Siebenmorgen, T. J., and Jindal, V. K. (1986). Effects of moisture adsorption
on the head rice yields of long-grain rice. Trans. ASAE, 29( 6): 1767.
18. Kunze, 0. R., and Prasad, S. (1978). Grain-fissuring potential in harvesting
and drying of rice. Trans. ASAE, 21(2):361.
19. Banaszek, M. M., and Siebenmorgen, T. J. (1990). Head rice yield reduction
rates caused by moisture adsorption. Trans. ASAE, 33(4):1263.
20. Shizuoka Seiki Co., Ltd. (1987). Instruction manual for Shizuoka's single
kernel moisture tester, Model CTR-800 and CTR-800A. 4-1 Yamana, Fu-
kuroi, Shizuoka, Japan.
21. Siebenmorgen, T. J., Counce, P. A., Lu, R., and Kocher, M. F. (1991).
Correlation of head rice yield to individual kernel moisture content. ASAE
Paper No. 91-6060. Am. Soc. Agric. Eng., St. Joseph, MI.
22. Rhind, D. (1962). Breakage of rice in milling: A review. Trap. Agric., 39:19.
23. Desikachar, H. S. R., and Subramanyan, V. (1961). The formation of cracks
in rice during wetting and its effect on the cooking characteristics of the cereal.
Cereal Chern., 38:356.
24. Kunze, 0. R. (1964). Environmental conditions and physical properties which
produce fissures in rice. Ph.D. dissertation, Michigan State University, East
Lansing, Michigan.
25. Kunze, 0. R., and Choudhary, M.S. U. (1972). Moisture adsorption related
to the tensile strength of rice. Cereal Chern., 49(4):684.
26. Lague, C. (1989). Modeling pre-harvest stress cracking of rice kernels. Ph.D.
dissertation, University of California, Davis, CA.
27. Henderson, S.M. (1954). The causes and characteristics ofrice checking. Rice
J., 57:16,18.
28. Schmidt, J. L., and Jebe, E. H. (1959). The effect of artificial drying on the
yield of head rice and germination of rice. Trans. ASAE, 2(1):26,31.
29. Stermer, R. A. (1968). Environmental conditions and stress cracks in milled
rice. Cereal Chern., 45:365.
30. Kunze, 0. R., and Hall, C. W. (1965). Relative humidity changes that cause
brown rice to crack. Trans. ASAE, 8(3):396, 405.
31. Kunze, 0. R. (1985). Effect of variety and environmental factors on milling
quality of rice. Paper presented at the International Rice Research Confer-
ence, 1-5 June, 1985.
32. Banaszek, M. M., and Siebenmorgen, T. J. (1990). Moisture adsorption rates
ofrough rice. Trans. ASAE, 33(4):1257.
33. Banaszek, M. M., and Siebenmorgen, T. J. (1990). Adsorption equilibrium
moisture contents of long-grain rough rice. Trans. ASAE,33(1):247.
34. Nishiyama, Y., Satoh, M., and Shimizu, H. (1979). Crack generation of rough
rice after drying. J. Faculty Agric., 14(3):277.
35. Banaszek, M. M., Siebenmorgen, T. J., and Sharp, R. N. (1989). Effect of
moisture content at milling on head rice yield and degree of milling. Arkansas
Farm Res., 38(2):15.
380 SIEBENMORGEN

36. Webb, B. D., and Calderwood, D. L. (1977). Relationship of moisture con-

tent to degree of milling in rice. Cereal Foods World, 22(9):484.
37. Pominski, J., Schultz, E. F., and Spadaro, J. J. (1961). Effect of storage time
after drying on laboratory milling yields of southern rice. Rice J., 68(5):20, 43.
38. Wratten, F. T. (1960). Effects of milling rice at low moisture levels. Proceed-
ings of the Rice Technical Working Group Meeting, June 29-July 1, Lafa-
yette, LA.
39. Wasserman, T. (1960). Heated air drying of western rice. Proceedings of the
8th Rice Technical Working Group Meeting, June 29-July 1, Lafayette, LA.
40. Wasserman, T. (1961). Low moisture milling. Proceedings of the 2nd Confer-
ence on Rice Utilization, May 18-19, Albany, CA.
41. USDA. (1982). Rice Inspection Handbook. Agricultural Marketing Service,
Washington, D.C.
42. Velupillai, L., and Pandey, J. P. (1987). Color and bran removal in rice process-
ing. Louisiana Agricultural Experiment Station Manuscript No. 87-07-1342,
Baton Rouge, LA.
43. Andrews, S. B., Siebenmorgen, T. J. and Mauromoustakos, A. (1992). Eval-
uation of the McGill #2 rice miller. Cereal Chern., (in press).
44. McCarthy, M. J., and Kauten, R. J. (1990). Magnetic resonance imaging
applications in food research. Trends Food Sci. Techno/., (December):134.
45. Bhumbla, V. K., Siebenmorgen, T. J., and Komoroski, R. A. (1992). Mois-
ture content determination of rice constituents using nuclear magnetic reso-
nance. Proceedings of the Rice Technical Working Group Meeting, February
23-26, Little Rock, AR.
 16
Processing and Utilization of Rice Bran in the
United States
Keith L. Hargrove, Jr.
Farmers' Rice Cooperative, Sacramento, California

I. INTRODUCTION
Dr. Robin M. Saunders would generally include Figure 1 in his many
presentations and discussions of rice bran. Although the translation to En-
glish may not be technically precise, the Chinese symbols predate most of the
current research on rice bran by about 2000 years, give or take a few centu-
ries.
Actually, the international scientific research community has recog-
nized the underutilization and potential of rice bran for many years. Much
has been published to this effect (1). However, only recently has the
United States rice-processing industry begun to explore in earnest and
develop the full commercial potential of this relatively abundant byprod-
uct of rice milling.
This interest, at least in the United States, was initially and primarily
spurred by the work of Saunders and his associates at the United States
Department of Agriculture, Agricultural Research Service, Western Re-
gional Research Center (USDA, ARS, WRRC). Their research described
the first commercially feasible method of overcoming the inherent prob-
lems of stability and storability of rice bran (2-5). Extrusion type stabiliza-
tion systems were subsequently installed in a few major rice mills in the
mid-1980s, and rice bran was stabilized on a limited basis for commercial
export to Japan, Korea, and Taiwan for oil extraction.

381
382 HARGROVE

(a) Rice: Health or Sound

Body

(b) Rice: Waste or What

Remains After the
Good Portion is
Taken

Figure 1 Chinese characters depicting (a) rice bran and (b) white rice.

Rice bran feeding studies were conducted on hamsters by the WRRC,

and the results indicated that rice bran was equal to oat bran in its ability
to lower serum cholesterol in hamsters (6). Further studies with hamsters
were carried out to confirm this finding (7). With the commercial success
that oat bran was having at the time, these new findings attracted the
immediate attention of the rice industry along with the food industry in
general. Several major companies began studying the various potential
applications of stabilized rice bran in food products. Subsequent clinical
trials with human subjects in Australia (8), California (9), and Louisiana
(10), appeared to confirm the beneficial effects of rice bran in controlling
cholesterol in humans. Additional studies in Europe (11) and the United
States (J. W. Lampe, unpublished) appear to confirm the laxation benefits
associated with rice bran and its dietary fiber components.
Today, a number of high quality food products containing stabilized rice
bran are being produced and marketed successfully. The U.S. rice industry
has been relatively conservative in its approach to the research and market-
ing of rice bran for food products. Additional clinical studies are needed
and are being planned, and much remains to be learned. Market growth,
consumer awareness, and demand is steadily increasing.
RICE BRAN 383

II. COMPOSITION OF RICE BRAN IN THE UNITED STATES

Rice bran in the United States can generally be divided into two catego-
ries: food grade and feed grade. Most U.S. rice mills are of the multiple
break design, and the hulls are removed in a process that is separate from
the actual milling or bran removal process. Feed grade bran composition
can vary widely depending upon the mill design, dehulling and paddy
separation efficiencies, and whether calcium carbonate is used as a mill-
ing aid.
Weed seeds, straw, hulls, dust, and calcium carbonate may be present in
varying amounts, as feed grade bran is generally sold against very basic
minimum guaranteed specifications, including crude fiber, crude protein,
and ash content. Many rice mills are not equipped with the processing and
handling systems necessary to keep these byproducts of the milling process
separate from the actual bran and germ. In order to produce a food quality
rice bran, it is necessary to minimize the commingling of the the various
byproducts to assure a consistently pure and wholesome bran. In many
instances, this requirement may necessitate significant mill redesign and
may involve significant capital investment.
A thorough review of the physical properties and other characteristics of
·rice bran has been done by Barber and Benedito de Barber (1) and Luh et
al. (12). The descriptions of particle distribution, sources of contamination
or dilution, and composition are quite comprehensive.
Due to the naturally occurring enzymatic activity and subsequent
hydrolytric rancidity that occurs rapidly in the bran after milling, it is
further necessary to stabilize the bran or to inactivate the indigenous lipase
either before milling or very shortly afterwards. Although there are several
potential methods of accomplishing stabilization, most are associated with
some type of heating process. Parboiling the rough rice or paddy prior to
milling and dry or moisture-added extrusion of the bran after milling are
the most common methods in use today in the United States.
Food grade and stabilized rice bran can be further categorized as full fat
stabilized or parboiled, defatted, or partially defatted. Although these
variations can be customized or modified somewhat to meet certain end use
requirements, typical proximate compositions found in the United States
are included in Table 1.

A. Protein
Rice bran protein is of relatively high nutritional value. Reported PER
values generally range from 1.6 to 1.9 compared to a casein value of 2.5,
while for protein concentrates extracted from bran with dilute alkali, values
from 2.0 to 2.5 were found. Digestibility of protein in rice bran is reported
384 HARGROVE

Table 1 Proximate Composition and Caloric Content of Stabilized, Parboiled,

and Defatted Rice Brans
Crude
Moisture Protein Fat fiber Ash Calories/
(%) (%) (%) (%) (%) oz.

Rice bran 8-12 12-16 17-22 8-12 7-10 90-100

Parboiled rice bran 7-9 17-20 25-32 12-15 8-10 100
(without calcium
carbonate)
Parboiled rice bran 6-9 14-18 23-27 10-13 10-13 100
(with calcium
carbonate: 4-6%)
Defatted bran 6-9 15-20 0.5-1.5 10-15 9-12 60
Defatted parboiled bran 6-9 23-27 0.5-1.5 16-20 11-14 40
(without calcium
carbonate)
Source: Ref. 13.

to be 73%, whereas in the extracted concentrates, protein digestibility was

greater than 90% (13).

B. Lipids
Rice bran typically contains 16-32% oil. Three major fatty acids, palmitic,
oleic, and linoleic make up more than 90% of the total fatty acids (13).

C. Minerals
The ranges of expected mineral contents in rice bran are depicted in Table 2.
The mineral content is impacted by variety, soil conditions and growing
environment, and by the milling process used. In the United States, silicon
would be at the low end of the range due to the absence of hulls.

D. Vitamins
Vitamins found in rice bran are listed in Table 3. The range in content reflects
rice variety, degree of milling and processing, and possible contamination
with hulls. In the United States, bran produced for domestic food consump-
tion is virtually devoid of hulls, and therefore the vitamin content would be
at the high end ofthe range listed. Rice bran contains at least 78% ofthe rice
kernel thiamine, 47% of the riboflavin, and 67% of the niacin. Parboiled
RICE BRAN 385

Table 2 Minerals in Rice Bran

Mineral Content (ppm)
Aluminum 53-369
Calcium 140-1310
Chlorine 510-970
Iron 190-530
Magnesium 8650-12300
Manganese 110-877
Phosphorus 14800-28700
Potassium 13650-23900
Silicon 1700-16300
Sodium 0-290
Zinc 80
Source: Ref. 13.

Table 3 Vitamins in Rice Bran

Vitamin Content (ppm)
Vitamin A 4
Thiamine 10-28
Riboflavin 2-3
Niacin 236-590
Pyridoxine 10-32
Pantothenic acid 28-71
Biotin 0.2-0.6
Myoinositol 4600-9300
Choline 1300-1700
p-Aminobenzoic acid 0.7
Folic acid 0.5-1.5
Vitamin B12 0.005
Vitamin E 150
Source: Ref. 13.

bran would be expected to contain lower levels of the soluble vitamins due to
leaching of these vitamins during the parboiling process (13).

E. Carbohydrates
The major carbohydrates present in commercial rice bran are cellulose,
hemicelluloses (or pentosans), and starch. Starch is not botanically present
386 HARGROVE

in the outer pericarp layers, but because of endosperm breakage during

milling, it appears in the bran. The quantity varies according to the amount
of breakage and degree of milling, but values of 10-20% could be ex-
pected. Amylose and amylopectin in the starch depend upon the rice vari-
ety. In general, amylose content is 2% or less in waxy (sweet) varieties,
about 10-20% in short- and medium-grain varieties, and 20-30% in long-
grain varieties (13).
Hemicelluloses have been reported to comprise 8.7-11.4% of the bran.
Cellulose in rice bran is reported to range from 9.6 to 12.8%. Beta-glucans
in rice bran are present at less than 1%. These polysaccharides comprise
part of the dietary fiber complex (13).
Sugars in rice bran range from 3 to 8%. Glucose, fructose, sucrose, and
raffinose have been reported, although sucrose is by far the predominant
sugar. Sugars would be lower in parboiled bran due to the heat and mois-
ture used in the parboiling process (13).

F. Dietary Fiber
The dietary fiber content of stabilized and parboiled rice bran varies with the
degree of milling and with the amount of starch and sugars present in the
brans. The ranges for soluble and total dietary fiber are given in Table 4.

G. Antinutritional Compounds
The occurrence of some antinutritional compounds has been reported.
Among these are trypsin inhibitor (14), pepsin inhibitor (15), hemaglutinin
(14), phytates, and an antithiamine factor (12). Fortunately, activity of
these compounds is relatively low and can be inactivated by heat treatment
(12). There appears to be little data available regarding the impact, if any,
that current commercial stabilization processes may have on the an-
tinutritional factors present in raw bran.

Table 4 Dietary Fiber Contents of Stabilized and

Parboiled Rice Bran (dry basis)

Total Soluble
dietary fiber dietary fiber
Bran type (%) (%)
Stabilized bran 20-25 1.8-2.6
Parboiled bran 31-33 2.0-2.5
Defatted stabilized bran 24-28 2.0-2.4
Defatted parboiled bran 44-51 2.4-2.9
Source: Ref. 13.
RICE BRAN 387

H. Full-Fat Stabilized Rice Bran

The majority of the commercially developed and available food grade
stabilized rice bran is considered to be full fat stabilized rice bran, which is
stabilized bran in which the natural lipid content has not been reduced by
further processing or extraction. The amount of lipids or fats present are
partially dependent upon the nature of the milling process. The impact that
the milling process imparts to the quality of the bran is explored further in a
later section of this chapter.
In an effort to protect and assure the integrity of food grade rice bran in
the United States, the Rice Millers Association (16) in February 1990
adopted a definition and minimum standard for full-fat stabilized and par-
boiled rice bran for food use (Table 5). Future standards and definitions
will likely be established for defatted and partially defatted rice bran, as
these products are developed commercially.

Table 5 The Rice Millers Association Full-Fat Stabilized Rice Bran For
Human Food

DEFINITION: Rice bran is the brown outer layer of the brown rice kernel that is
removed when milling brown rice to milled or white rice. The bran is comprised
primarily of the pericarp, aleurone, and subaleurone layers of the kernel, and
typically includes the embryo or germ and a small amount of the starchy
endosperm.
Full-Fat Stabilized Rice Bran For Human Food shall be suitably treated after
milling to deactivate the naturally occurring lipase enzymes, which, if untreated,
would rapidly deteriorate the oil in the bran.
To assure its purity and suitability for human consumption, Full-Fat Stabilized
Rice Bran For Human Food shall meet or exceed the following minimum specifica-
tions as determined by the corresponding analytical methods:
PRODUCT ANALYSIS APPROVED METHODS
AOAC AACC
Fat Min 16% 7.056 30-25
Protein (Nx5.95) Min 13% 2.057
Total dietary fiber Min 20% JAOAC 71:1017
Crude fiber Max 9% 7.061-7.065 32-10
Ash Max 10% 14.006 08-01
in parboiled rice bran Max 15%
Moisture Max 12% 14.004 44-15
FFA in crude fat extract Max 4% 28.029 02-01A
Silica (Si02) Max 0.1% 3.005 40-21
Calcium carbonate Max 2%
in parboiled rice bran Max 6%
Source: The Rice Millers' Association, Washington, D.C.
388 HARGROVE

I. Defatted or Low-Fat Stabilized Rice Bran

Currently there are few, if any, U.S. producers of sizable commercial quan-
tities of rice bran oil and defatted or partially defatted rice bran. Most rice
oil currently being marketed in the United States is imported. However, a
significant potential commercial market for these products appears to exist,
and several major companies are actively developing plans for commercial
production.

ill. RICE BRAN AND THE HEALTHY DIET

Due to rice bran's overall composition, nutritional profile, functional char-
acteristics, and apparent hypoallergenicity, rice bran has many applications
in a healthy diet that is high in dietary fiber and low in saturated fat. It may
be particularly beneficial to those individuals who are allergic to other
cereal grains. However, there are strong indications from recent research
that the consumption of rice bran may be specifically beneficial in reducing
the risk of cardiovascular disease and colon cancer.
Kahlon et al. (6, 7) found that rice bran lowered serum cholesterol in hy-
percholesterolemic hamsters to the same extent as did oat bran (Table 6).
This observation is indirectly supported by earlier work from Japan (17-
19) and India (20,21), where several studies indicated cholesterol-lowering
activity by a number of components present in rice bran (Table 7). The
extent of the contribution by each individual component to the lowering of
serum cholesterol has not been firmly established, but results available
point to rice bran oil as containing the bulk of hypercholesterolemic activ-
ity (13).

Table 6 Plasma Cholesterol of Hamsters Consuming Rice

Bran or Oat Bran Diets for 3 Weeks
Plasma cholesterol
Treatment (mg/dl)
Control (10% cellulose) 394.9 ± 15.63
Stabilized bran (47.8% of diet) 269.5 ± 13.5b
Parboiled bran (31.8% of diet) 297.1 :±; 18.3b
Defatted stabilized bran (31.1% of diet) 347.2 ± 16.oac
Defatted parboiled bran (19.6% of diet) 376.8 ± 21.030
Oat bran (53.7% of diet) 288.9 ± 16.4b

n = 10, values with different superscripts (p < 0.05). All diets con-
tained 0.5% cholesterol, 10.7% fat, and 10% dietary fiber.
Source: Ref. 6.
RICE BRAN 389

Table 7 Factors in Rice Bran That May Contribute to

Lowering Cholesterol
1. Tocotrienols
2. Oryzanols (ferulic acid ester)
3. {3-Sitosterol
4. Hemicelluloses
5. {3-Glucan
6. Oil unsaturation
7. Protein

Source: R. M. Saunders, personal communication.

Nestel and Kestin (8) conducted a double-blind clinical study with 24

male subjects to compare the effects of rice bran, oat bran, and wheat bran
on plasma lipids and laxation. In the study, only oat bran led to a significant
reduction in total plasma cholesterol, but rice bran equaled oat bran in
increasing the high density lipoprotein cholesterol (HDL-C) to total choles-
terol (TC) ratios, and the rice bran demonstrated the highest laxative effect
of the three brans in terms of stool bulking and decreased intestinal transit
time.
Gerhardt and Gallo (9) concluded that:
1. Rice bran and germ produced significant cholesterol reduction in mod-
erately hyperlipidemic subjects consuming their usual prudent diet,
and the decline in cholesterol was entirely due to decreased low density
lipoprotein cholesterol (LDL-C) levels.
2. HDL-CffC and LDUHDL ratios improved with rice bran and germ.
3. There was no statistical difference between the effectiveness of rice
bran and germ and oat bran in producing cholesterol reduction.
4. Lipid levels for males and females responded equally to the addition of
fiber products in the diet.
5. No conclusion could be drawn about which specific nutrient in the two
bran products caused the cholesterol reduction.
Another study by Hegsted et al. (10) on 11 human subjects concluded that
rice bran and oat bran were equally effective in lowering TC and LDL-C
levels.
Feeding studies on nonhuman primates conducted by Nicolosi et al. (22)
indicated that the content of rice bran oil in the diet was the predominant
factor influencing serum TC, LDL-C, and apolipoprotein B causing up to a
40% reduction in LDL-C without affecting HDL-C when rice bran oil was
the sole dietary oil fed, and the cholesterol-lowering capabilities or rice
bran oil were not explained by its fatty acid composition. (For further
390 HARGROVE

information on the cholesterol-lowering properties of rice bran oil, see

Chapter 18.)
Lampe et al. (unpublished) compared the laxative properties of rice
bran to those of wheat bran in male human subjects. This research sup-
ported previous work in Europe (11) and Australia (8), which showed that
rice bran supplementation had effects similar to wheat bran in improving
laxation and decreasing transit time. In this study, rice bran was slightly
more effective in increasing stool weight and decreasing transit time than
was wheat bran on a per gram of total dietary fiber consumed basis.
Additional work is underway in identifying and quantifying the active
components present in rice bran and oil. Additionally, the impact that
processing has on the variability and concentration of these components
needs further study.

IV. STABILIZATION AND PROCESSING

A. Milling and Purification
All U.S. rice mills of commercial significance utilize distinctly separate
machines and processes for removing the hull prior to removing the bran
layers and germ. In many less developed areas of the world, these pro-
cesses are performed by the same machine, rendering it nearly impossible
to produce a rice bran that is suitable for human consumption. However,
merely separating these functions is only one of many steps necessary in
producing rice bran that is truly food grade quality. The systems and meth-
ods used for harvesting, drying, storing, and shipping the rough rice to the
mill should be considered first.
In addition to the effects on the milling yield and quality of the rice
itself, improper drying, inadequate aeration during storage, and abusive
handling has a profound impact on the quality of the bran. Excessive and
prolonged heat during the drying process can negatively modify the flavor
of the bran, and there are indications that some of the natural antioxidants
such as the tocopherols can be modified or destroyed even at this early
stage of the process (23). It is generally recommended that multipass drying
processes be utilized and that grain temperatures not exceed 90°F, particu-
larly as moisture content falls below 20% (w.b.)(24). Adequate cooling and
equilibration time should be provided between passes to minimize thermal
and moisture gradient stresses.
Proper aeration during storage helps to prevent insect and microbiologi-
cal activity, which can damage the kernel, affecting the quality and flavor.
There is ample research literature available on proper and commercially
acceptable drying and aeration concepts. However, most of these relate to
RICE BRAN 391

preserving the quality and milling yield of the rice as opposed to addressing
the impact on the quality of the bran. It would seem that the two are
somewhat interrelated and inseparable, and that a process that preserves
the quality and freshness of the rice will also do the same for the bran.
Objectionable odors or flavors that appear in the milled rice are usually the
result of initial degradation or decomposition of the bran and its lipids due
to improper storage conditions (25).
Abusive material-handling systems for transporting and conveying can
result in partial or total dehulling of some of the rough rice kernels. Al-
though some dehulling is unavoidable in conventional systems, precautions
should be taken to minimize the occurrence. Mechanical damage to the
bran layers of some of the kernels promotes the hydrolytic and oxidative
rancidity processes while in storage and prior to actual milling and could
result in unacceptable bran quality. The author has generally found that
rice that is harvested, dried, handled, and stored in a proper commercial
manner for up to 18 months will produce rice bran that contains less than
2% free fatty acids in the lipid fraction at the time of milling.
As rice is a field crop generally grown under flooded conditions, the
rough rice or paddy must undergo thorough precleaning prior to milling.
Defects and foreign materials such as stones, mud balls, straw, and weed
seeds must be separated and removed with the proper use of screens,
aspirators, and gravity separators.
After the rough rice is thoroughly cleaned, special attention should be
paid to the combined shelling (or de hulling) and paddy separation pro-
cesses. Most modern rice mills utilize what are generically termed rubber
roll shellers for removing the hull from the rice kernel to produce brown
rice, although the rolls may be covered with natural rubber or synthetic
material such as urethane. These machines typically come equipped with,
or are placed above, aspirators that remove the loose hulls from the stream
after shelling.
Shelling efficiency is a term generally used in the industry to describe the
percentage of the rough rice that is actually dehulled or shelled in the
process. It is difficult if not undesirable economically to operate shellers at
a shelling efficiency of greater than 90-95%. Exceeding this level usually
results in reduced roll life, increased rice breakage, and reduced capacity,
each of which can have a significant economic impact on the operation and
profitability of the mill. Therefore, it is generally more efficient overall to
utilize paddy separators in the process after the shellers to separate, by
density, the unshelled kernels from the brown rice stream prior to the bran
removal or whitening process. The unshelled paddy kernels are recycled
back to the shelling operation, and the relatively paddy free brown rice is
then milled. It is very important to utilize a very efficient combined
392 HARGROVE

shelling-paddy separation process to assure that a minimum amount of

hulls or paddy enter the whitening process. It is recommended that the
system be operated in a manner that will maintain fewer than five paddy
kernels per 500 g of brown rice. This level is significantly lower than the
standards for a U.S. No. 1 brown rice for reprocessing but will result in a
much purer and consistent bran product. Additionally, a secondary aspira-
tion system should be considered between the shellers and the paddy sepa-
rators to supplement the aspiration and loose hull removal capability that is
typically provided with the shelling machine.
Two basic types of whitening machines are used in the United States: the
abrasive type and the friction type. There are many variations within these
categories, and even some that combine the two principles within the same
machine. There are a number of manufacturers worldwide, but most gener-
ally follow what is considered modem Japanese rice-milling technology
(see Chapter 10). The selection of the proper arrangement and combina-
tion of whitening machines in a rice mill is typically dictated by the type and
class of the rice that is processed and whether white rice or parboiled rice is
being milled. The number of consecutive passes or breaks installed in U.S.
mills vary widely and range from one to six, although there is an evolution
toward using more breaks as opposed to fewer, with three to four breaks
most prevalent.
The overwhelming emphasis in selecting the most efficient arrangement
is in preserving the whole grain yield and minimizing the production of
broken kernels, while achieving the milling degree or quality specification
at the lowest energy requirement. Achieving this goal also enhances the
consistency and purity of the rice bran by minimizing the amount of small
broken pieces of rice that are removed with the bran.
Some rice mills, particularly those milling parboiled rice, utilize calcium
carbonate as a milling aid to assist in the polishing of the kernel, and to
aid the conveyability of the oily bran after removal. To assure that the
usage of calcium carbonate when milling to produce food grade rice bran
does not exceed the minimum required, the Rice Millers Association in-
cluded maximum limits in the Full Fat Stabilized and Parboiled Rice Bran
Standards (16).
As was mentioned in an earlier section, rice bran in the United States
generally consists of the pericarp, aleurone, and subaleurone layers of the
brown rice kernel, along with the germ or embryo, and a small amount of
the starchy endosperm (Fig. 2). The endosperm may be present in the bran
either as small broken pieces or as a flour resulting from the abrasive or
friction whitening process. It is technically possible, although difficult due
to the high oil content, to sift the raw bran through fine screen mesh sizes
to remove virtually all of the small broken rice. However, the germ is
RICE BRAN 393

Figure 2 Rice kernel cross section. (From Ref. 13.)

generally removed from the kernel as a relatively intact particle and is

typically large enough to be retained on a U.S. 18 mesh screen. Therefore,
sifting the raw bran through screens finer than 18 mesh will result in re-
moval of most of the germ where a large percentage of the desirable miner-
als, lipids, and vitamins reside. Although processes are available and in use
in Europe to separate the germ from the bran and broken rice, these
processes are not widely employed in the U.S. as no significant efforts have
been made to market the germ separately.
A more desirable and economical approach to minimizing the grittiness
resulting from the presence of the small broken pieces of rice may be to
394 HARGROVE

grind or reduce the particle size after stabilization. Again, due to the high
oil content, this is not easily accomplished, but it is possible. Stabilized
rice bran ground to pass through 20, 40, and even 100 mesh sieves are
commercially available today and may be more suitable for many food
products.
Most rice milled today in the United States is milled to a degree that is
defined as well milled, which means that most of the bran and germ and
some of the endosperm is removed, and the resulting lipid content of the
milled kernel will be less than 1% and even as low as 0.5% by weight.
Milling white rice to this degree will generally result in rice bran that
contains between 16 and 24% total lipids. The amount of lipids present in
the bran is inversely proportional to the degree of milling and can be
lowered additionally by the presence of hulls, calcium carbonate, and ex-
cess broken rice pieces. Therefore lipid content is a good indicator of the
quality and purity of the bran.
As the bran is milled from the kernel, it is typically removed from the
milling chamber by air suction and conveyed pneumatically. The bran is
then separated from the air by primary cyclone separators and then secon-
dary cloth filters. It is generally not desirable to utilize the fine portion of
the bran that is collected by the cloth filter, as the residence time and
sanitary aspects of such filters may lead to quality degradation. Addition-
ally, the very fine portion that escapes the primary separator typically
consists of polish, which is primarily the flour portion of the endosperm and
contains fewer of the fiber-related traits generally associated with bran.
The freshly milled bran is then typically screened through a size 16 or 18
mesh to remove the majority of the small broken pieces of rice. Once the
bran is removed from the kernel, the lipase enzymes present are liberated
and immediately begin to hydrolyze the lipids into free fatty acids. This
process can be very rapid, and accordingly the bran should be stabilized
immediately to inactivate the lipase. As the free fatty acids can progress to
levels exceeding 5% in just a few hours, it is impractical and undesirable to
accumulate or store raw bran for subsequent stabilization in a batch pro-
cess. (For a more detailed description of the equipment involved in rice
milling, see Chapter 10.)

B. Stabilization
Due to the rapid hydrolysis that occurs after milling, a proper stabilization
process should be a continuous process, close coupled to the mill, and
ideally the bran should be treated within a few minutes after removal from
the kernel. Conveying systems should be self-cleaning or otherwise de-
signed to avoid accumulations and buildup of old product.
RICE BRAN 395

Although there are many potentially suitable means to stabilize or inacti-

vate the lipase in rice bran, most commercial systems currently utilized in
the United States use the moisture-added or dry extrusion methods. These
systems were generally selected due to relatively low energy requirements,
low capital cost, and ease of installation and operation.
Currently, a substantial amount of research is being performed evaluat-
ing alternative methods of stabilization which might better preserve or
enhance the desirable characteristics of rice bran (see Chapter 2). Stabiliza-
tion by dry extrusion utilizes shear, friction, and pressure to generate the
heat required to inactivate the lipase. Depending upon the particular char-
acteristics of and retention time in the extruder, the temperature of the
bran must reach a minimum of 130-140°C for up to 3 seconds to assure
inactivation of the lipase (2-4).
Although this system will adequately destroy or inactivate the lipase,
there is growing evidence (23) that these conditions may also modify or
destroy some of the naturally occurring antioxidants, such as the tocoph-
erols and tocotrienols, which are so abundantly present and desirable in
rice bran. Since rice bran contains fairly high levels of sucrose and other
sugars, excessive shear and heat may also initiate certain Maillard reactions
and even scorch the bran resulting in undesirable or bitter flavor attributes.
Acceptable extrusion stabilization can be achieved under less harsh con-
ditions by the direct addition of water or steam and the addition of a means
to maintain the bran temperature at 90-100°C after extrusion for 2-3 min-
utes prior to cooling. The lipase appears more heat sensitive at higher
moisture and can therefore be inactivated at somewhat lower extrusion
temperatures (R. M. Saunders, personal communication). The addition of
steam or moisture can also be beneficial in attempting to form shapes or
texture in the extruded bran. Due to the relatively low level of starch and
the high lipid content of purer rice brans, very little expansion occurs upon
exiting the extruder, and the extrudate tends to be relatively fragile in
terms of maintaining flakes or pellets.
Research in both the private and public sectors continues regarding the
exploration and evaluation of alternative means to stabilize rice bran. As
more is learned about the composition as it relates to nutritional and health
benefits and product functionality, other systems are likely to evolve and
become commercially practical.
Residual peroxidase activity is generally used as the standard measure to
assure that lipase has been deactivated in stabilized rice bran. Peroxidase is
generally considered to be more heat stable than lipase, and peroxidase
activity tests are typically easier and more reliable than those procedures
currently available for lipase. The process conditions required to inactivate
peroxidase as well as lipase may also cause modifications to or loss of some
396 HARGROVE

or all of the antioxidants naturally present in the bran, leading to a product

that is shelf stable to the formation of free fatty acids, but which may be
more subject to oxidative rancidity than raw bran (R. M. Saunders, per-
sonal communication). More research and evaluation ofthe optimum time,
temperature, and moisture relationships in stabilization is needed. Addi-
tionally, more accurate methods of lipase analysis would assist in refining
and optimizing the stabilization processes.
Whatever the method of stabilization, as with most nutritious food prod-
ucts, rice bran is subject to invasion by mold, yeast, and bacteria. The
stabilization process must effectively sterilize the product, and the further
processing and packaging systems used should prevent the recontamination
of the product.

C. Storage and Shelf Life

Currently, the predominant goals in stabilizing rice bran are microbial
sterilization and destruction of the lipase enzyme to prevent the formation
of free fatty acids. Lipase and peroxidase activity, along with free fatty acid
content, are the current standard measures of stability. Following the extru-
sion stabilization conditions outlined, bran can be produced that will re-
main stable for up to 18 months or longer under normal warehouse storage
conditions in terms of free fatty acid formation.
However, the bran lipids are subject to oxidation and decomposition
during storage, and the effective shelf life of stabilized rice bran can be
affected by oxidation as well as the formation of free fatty acids. Generally,
the maximum recommended shelf life is six months under normal ware-
house conditions. Refrigerated storage will extend the shelf life signifi-
cantly (R. M. Saunders, personal communication).

D. Packaging, Shipping, and Transportation

Due to the high lipid content, rice bran poses some challenges in bulk
material handling and storage systems. Depending upon the granulation, it
may have a relatively high minimum chute angle and is subject to ratholing
and forming stable arches in bulk storage systems. Minimum chute angle
defines the minimum angle at which a material will slide on a chute.
Ratholing and stable arches define no-flow conditions in a bin that result
from the formation of stable vertical holes or arches over the bin outlet.
However, these difficulties can generally be overcome by utilizing storage
and handling systems that promote mass flow. Rice bran can be conveyed
readily in conventional mechanical and pneumatic conveying systems but,
again due to the oil and waxes present, tends to form deposits in those
systems. Provisions should be employed to allow routine cleaning.
RICE BRAN 397

Although it may be possible to transport stabilized rice bran in bulk by

truck or rail, most stabilized rice bran is currently marketed in the United
States in 50-lb multiwall paper bags or in semi-bulk bags or totes containing
up to 2000 lb.

V. USE OF RICE BRAN IN FOOD PRODUCTS

There are virtually unlimited ways to incorporate rice bran into a healthy
and balanced diet. These can range from simply sprinkling rice bran on a
dinner salad to using it as a major ingredient in a complex shredded or
extruded, multigrain ready-to-eat cereal. The USA Rice Council for Mar-
ket Development lists the development and market introduction of 24 new
products containing rice bran in 1989 in the United States, along with 45
new rice bran products introduced in 1990 (26).
The functional characteristics and performance of rice bran in processed
food products are somewhat different and unique among the cereal brans.
This is particularly true of full-fat rice bran. As the utilization of rice bran
in commercial food products is a relatively new and developing phenome-
non, there is a limited amount of information about specific formulations
and processes available in the public domain. As with the rice kernel itself,
a great deal more needs to be learned about its composition as it relates to
functionality and processing characteristics.
Even within the definition and new industry standards for rice bran
(Table 5), there appears to be considerable variation in functional character-
istics among the various supplies available, due to variations in processing
and varieties. Along with the need for further research in this area, more
refinements to the definition and standards appear desirable.
However, several excellent reviews are available that depict some of the
more common measures of functionality. Luh et al. (12) describe flavor,
color, protein, extractability, solubility, water absorption, fat absorption,
and foaming capacity as important factors. Water absorption in full-fat
bran is reported to be approximately 200 g water/100 g bran. Fat absorption
using corn germ oil was reported to be around 150 g oil/100 g bran, and the
bran exhibited very low foaming capacity. Stabilized rice bran has a charac-
teristic but relatively bland flavor described as slightly toasted and nutty
and is generally light tan to light brown in color.

A. Ready-to-Eat Cereals
Currently, most of the stabilized rice bran in the United States is used as an
ingredient in ready-to-eat cereals. Due primarily to its relatively high lipid
content, rice bran normally cannot be substituted for another ingredient in
398 HARGROVE

an already developed and formulated product. Some adjustments in formula-

tion and processing are necessary to assure the texture, flavor, and durability
desired in most products.
Although the specific formulations for products on the market are gener-
ally proprietary and unavailable, rice bran is and can be utilized success-
fully in many rolled or flaked, shredded, extruded, and mixed grain or
nugget-type cereals. Formulations utilizing rice bran at levels of up to 40%
of dry ingredients have been developed and marketed.

B. Baked Products
The nutritional and functional properties of rice bran appear well suited for a
variety of baked products such as multigrain or high fiber breads and muf-
fins, as well as cookies, crackers, pastries, and pancakes or waffles. Again,
experimentation and formula adjustments may be necessary to achieve the
desired product characteristics. Proteins from rice bran do not have the
properties of wheat gluten, and therefore the amount rice bran incorporated
into breads may be limited by the final volume of the loaf (27).
In a recent study (H. Bright and W. T. Doerry, unpublished), stabilized
rice bran was compared to a high quality oat bran in several baked prod-
ucts. The stabilized rice bran was found to be a suitable high fiber ingredi-
ent for pan bread, cookies, and crackers. A ground sample of stabilized
rice bran was test-baked in several bakery products and compared with a
commercially available oat bran. Although the rice bran contributed more
color, aroma, and taste than oat bran did, the pan bread and cake muffins
prepared with rice bran had good appearance and eating qualities. Similar
results were obtained in cookies and in a cheese and bacon-flavored
cracker, where the performance of both ground and unground rice bran
was compared to oat bran. Both samples of rice bran produced good crack-
ers but were slightly darker in color than the oat bran cracker. In chocolate
chip and chewy cookie trials, the oat bran chocolate chip cookies spread
slightly more than the rice bran cookies, and there was no significant differ-
ence between oat bran and rice bran chewy cookie quality.
Sharp and Kitchens (28) replaced 15 and 30% of the wheat flour with
rice bran in yeast bread and concluded that rice bran can be substituted
successfully up to the 15% replacement level without affecting loaf weight,
height, or volume. Sensory ratings for breads containing 15% rice bran
were not significantly different from the control.

C. Pasta Products
Due to texture and availability of very finely ground stabilized rice bran, it
can be incorporated into pasta products with relative ease to produce a
RICE BRAN 399

higher fiber product. The finest grinds of rice bran are generally more
suitable for extruded or thin and fragile pasta products. Additionally, rice
bran has been added to the seasoning component as a flavor and fiber
enhancer for seasoned pasta and rice mixes.

D. Healthy Snacks and Desserts

With its abundance of vitamins and minerals and overall nutritional proper-
ties, along with the wide variety of granulations available, stabilized rice
bran is becoming a popular ingredient in a number of high fiber and nutri-
tious snacks and desserts. Products currently being developed or marketed
include extruded snacks, granola bars and cereals, pastry fillings and pud-
dings, and health drinks.

E. Stabilized Brown Rice Flour

With the growing emphasis on more healthy diets, brown rice and brown rice
flour have many applications where a nonallergenic starch or flour is desired.
Although brown rice is reasonably stable if carefully dehulled and handled,
once it is ground into flour, hydrolytic rancidity occurs rapidly as it does in
raw bran. Therefore, brown rice flour is difficult to obtain and utilize.
A stable, reconstituted brown rice flour may be produced by combining
and blending approximately one part of full-fat stabilized rice bran to six
parts of white rice flour. The resulting product will have essentially the
same nutritional profile as brown rice, assuming that the flour and the bran
are derived from the same variety of rice.

VI. RICE BRAN AS A RAW MATERIAL FOR DERIVATIVES,

OTHER FOOD INGREDIENTS, AND
PHARMACEUTICALS
In addition to being a source of high quality vegetable oil, rice bran is an
excellent, and in some cases unique, source of many valuable derivatives
and concentrates. Although a significant amount of activity is underway in
the United States towards developing a few of these products, Japan ap-
pears to be the current world leader in these processes and products.

A. Rice Bran Oil

Perhaps the most significant product, in terms of potential volume and
demand, is rice bran oil. Rice bran oil is generally considered to be one of
the highest quality vegetable oils available in terms of its cooking qualities,
shelf life, and fatty acid composition (29). As more is known about the role
400 HARGROVE

Table 8 Typical Fatty Acid Composition of

Rice Bran Oil
Fatty acid %

Myristic (C14:0) tr
Palmitic (C16:0) 16
Stearic (C18:0) 2
Oleic (C18:1) 42
Linoleic (C18:2) 38
Linolenic (C18:3) 1.4
Arachidic (C20:0) 0.6

Source: Ref. 13.

Table 9 Characteristics of Refined Rice Bran Oil

Iodine value: 99-108

Saponification value: 180-190
Smoke point: 213°C
Fire point: 352°C
Cloud index: 17°C
Refractive index 25°C: 1.470-1.473
Specific gravity 25/25°C: 0.916-0.921
Unsaponifiable matter: 3-5%
Total tocopherols: 200 mg/kg (0.02%)
Source: Ref. 13.

of oils and fats in the diet, rice bran oil utilization and demand in the
United States will likely increase significantly. The only limit to the demand
may lie in its cost of extraction and refining compared to other readily
available vegetable oils. Tables 8 and 9 depict the typical fatty acid composi-
tion and characteristics offully refined rice bran oil.
The predominant commercial process for producing rice bran oil is
solvent extraction with hexane followed by refining steps including bleach-
ing, dewaxing, winterization, and deodorization (see Chapter 18). How-
ever, there appears to be an interest developing in evaluating other means
of extraction such as cold pressing or expelling, propane extraction, and
supercritical C02 • As more is known about the desirable components of
rice bran oil and how they are affected by extraction and refining, other
processes may be better able to preserve and enhance the nutritional
characteristics.
RICE BRAN 401

B. Defatted or Low-Fat Rice Bran

The predominant byproduct of rice bran oil extraction in terms of volume
is, of course, the defatted bran. The marketability and value of the defatted
bran plays an important, if not overwhelming, role in the economics and
resulting unit cost of the oil itself. Since there are currently few, if any,
major U.S. producers of rice oil, the market potential for defatted or low-
fat bran has not been fully developed in the United States.
However, in many instances, the defatted bran may be more function-
ally desirable in many food products and is an good source of insoluble
dietary fiber and protein. Further, its shelf life is greatly extended as a
result of removing all or most of the lipids. The flavor of defatted rice bran
is also generally more bland and accordingly somewhat easier to incorpo-
rate into many products.

C. Secondary Products from Rice Bran and Oil

The distillates, soaps, and waxes produced in the oil-refining processes are
sources for many potentially valuable and useful products used in pharma-
ceutical and other applications. Among these are fatty acids, oryzanol,
tocopherols, tocotrienols, gums and waxes, and inositol. Additionally, the
defatted bran is a source of such compounds as dibasic calcium phosphate,
phytic acid, calcium-magnesium phytate, inositol, vitamin B, and protein
concentrates.

VII. FEED AND OTHER USES OF RICE BRAN

Until the recent development of commercial stabilization processes for rice
bran, the predominant use and market for rice bran in the United States
has been as an ingredient in livestock feed, particularly for dairy cattle,
swine, and poultry. It is still very much in demand for those uses and
provides a good source for protein, fiber, minerals, vitamins, and vegetable
fat. The nutritional and feed value of rice bran has been reviewed by
Juliano (30) and Luh and Benedito de Barber (12).
In addition to the more traditional feed uses, rice bran is also being used
in high quality dog and horse rations. Although not yet fully explained or
scientifically validated, there are indications that the rice bran contributes
to significant improvement in the luster and quality of the coat in these
animals (J. Baldiviez, personal communication). The oryzanol present in
rice bran is reported to have functions similar to vitamin E in promoting
growth, facilitating capillary growth in the skin, and improving blood circu-
lation, along with stimulating hormonal secretion (12). Additionally, there
402 HARGROVE

is growing interest in using rice bran as an ingredient in commercial catfish

food (J. Baldiviez, personal communication).
When blended with sawdust or other components, rice bran is used as a
culture medium for the growing of mushrooms (30,32). Koji malt obtained
from deoiled bran is suitable for the manufacture of miso (bean paste) and
shoy (soy sauce) (30,31). Additionally, a paste made from rice bran and salt
water is used in Japan for pickling vegetables and fish (12,32).

VIII. CONCLUSIONS
The foundation appears well established for the growth of rice bran as a
food and as a source for many potentially valuable byproducts and com-
pounds. The evolution from an underutilized byproduct of one of the
world's largest cereal grain crops to a commodity of great commercial
importance is long overdue and continues slowly but surely. However, the
processing and utilization of rice bran is far from a mature technology.
Many opportunities appear to exist to expand upon present knowledge in
areas such as processing, functionality, nutrition and health benefits, prod-
uct development, and analytical methodology.

ACKNOWLEDGMENTS
The U.S. rice-milling industry owes a debt of gratitude to the many preemi-
nent researchers in the field who have dedicated a significant portion of
their careers to expand the utilization of rice bran. The list is long and
includes such notable authorities as S. Barber and C. Benedito de Barber
of the Instituto de Agroquimica y Tecnologia de Alimentos, Valencia,
Spain; M. M. Bean, A. A. Betschart, D. F. Houston (retired), T. S.
Kahlon, J. M. Randall, R. M. Saunders (deceased), and R. N. Sayre of the
USDA, ARS, WRRC, Albany, California; B. 0. Juliano of the Interna-
tional Rice Research Institute, Los Banos, Laguna, Philippines; B. S. Luh
of the University of California, Davis, California; J. M. Harper and R. E.
Tribblehorn of Colorado State University-Fort Collins, Colorado; among
many others in many nations.
In addition to acknowledging the contributions to the field by those
listed above, the author wishes to extend his gratitude to Robert N. Sayre
and Frank Orthoefer for kindly reading this chapter.

DEDICATION
This chapter is dedicated to the memory of Robin M. Saunders, Ph.D.
(1940-1990).
RICE BRAN 403

REFERENCES
1. Barber, S., and Benedito de Barber, C. (1985). Rice Bran: An Underutilized
Raw Material. United Nations Industrial Development Organization (UNIDO)
Publication E, 84. II. B.3.
2. Enochian, R. V., Saunders, R. M., Schultz, W. G., Beagle, E. C., and
Crowley, P. R. (1981). Stabilization of Rice Bran With Extruder Cookers and
Recovery of Edible Oil. U.S. Dept. of Agric., Agric. Res. Service, Marketing
Research Report No. 1120: 18 pp.
3. Sayre, R.N., Saunders, R. M., Enochian, R. V., Schultz, W. G., and Beagle,
E. C. (1982). Review of rice bran stabilization systems with emphasis on
extrusion cooking. Cereal Foods World, 27:318.
4. Randall, J. M., Sayre, R.N., Schultz, W. G., Fong, R. Y., Mossman, A. P.,
Tribblehom, R. E., and Saunders, R. M. (1985). Rice bran stabilization by
extrusion cooking for extraction of edible oil. J. Food Sci., 50:361.
5. Saunders, R. M. (1985). Rice bran: Composition and potential food uses.
Food Rev. Int., 1:465.
6. Kahlon, T. S., Saunders, R. M., Chon, F. 1., Chiu, M. C., and Betschart, A.
A. (1989). Effect of rice bran and oat bran on plasma cholesterol in hamsters.
Cereal Foods World, 34:768.
7. Kahlon, T. S., Saunders, R. M., Chow, F. 1., Chiu, M. M., and Betschart, A.
A. (1990). Influence of rice bran, oat bran, and wheat bran on cholesterol and
triglycerides in hamsters. Cereal Chern., 67:439.
8. Nestel, P., and Kestin, M. (1990). Comparative effects of three cer~al brans
on plasma lipids, blood pressure, and glucose metabolism in mildly hypercho-
lesterolemic men. Am. J. Clin. Nutr., 52:661.
9. Gerhardt, A. L., and Gallo, N. B. (1989). Effect of a processed medium grain
rice bran and germ on hypocholesterolemia. American Association of Cereal
Chemists Annual Meeting, Washington D.C. (poster presentation).
10. Hegsted, M., Windhauser, M. M., Lester, F. B., and Morris, S. K. (1990).
Stabilized rice bran and oat bran lower cholesterol in humans. FASEB J.,
4:A386.
11. Tomlin, J., and Read, N. W. (1988). Comparison of the effects on colonic
function caused by feeding rice bran and wheat bran. Eur. J. Clin. Nutr.,
42:857.
12. Luh, B.S., Barber, S., and Benedito de Barber, C. (1991). Rice bran: chemis-
try and technology, Rice Production and Utilization, Vol. II (B.S. Luh, ed.),
Van Nostrand Reinhold, New York, p. 313.
13. Saunders, R. M. (1990). The properties of rice bran as a foodstuff. Cereal
Foods World, 35:632.
14. Benedito de Barber, C., and Barber, S. (1978). Toxic constituents of rice bran.
Rev. Agroquim. Techno/. Aliment., 18:89.
15. Mitsuda, H., Kawai, F., Suzuki, A., and Hondo, J. (1977). Studies on the
production of a protein rich fraction from rice bran by means of fractional
sedimentation in n-hexane, Rice Report 1976 (S. Barber, H. Mitsuda, H.S.R.
Desikachar, and E. Tortosa, eds.), International Union of Food Science and
404 HARGROVE

Technology. Working Party on Rice Utilization. Institute for Agricultural

Chemistry and Food Technology, Valencia, Spain.
16. (1990). Committee of Rice Millers Association defines and standardizes rice
bran. Cereal Foods World, 34:614.
17. Suzuki, M. (1982). Repressive effect of dietary fiber fractions in unpolished
rice on the increase of cholesterol and triglyceride.! Nutr. Food, 35:155.
18. Suzuki, S., Tezuka, T., and Oshima, S. (1984). Cholesterol lowering effect of
rice bran oil. Oil Fat Japan, 37:59.
19. Aoe, S., Ohta, F., and Ayano, Y. (1989). Effect of rice bran hemicellulose on
cholesterol metabolism in rats. Nippon Eiyoshokuryo Gakkaishi, 42:55.
20. Seetharamaiah, G. S., and Chandrasekara, D. (1988). Hypocholesterolemic
activity of oryzanol in rats. Nutr. Rep. Int., 38:927.
21. Sharma, R. D., and Rukmini, C. (1987). Hypocholesterolemic activity of
unsaponifiable matter of rice bran oil. Indian l Med. Res., 85:278.
22. Nicolosi, R. J., Ausman, L. M., and Hegsted, D. M. (1991). Rice bran oil
lowers serum total and low density lipoprotein cholesterol and apo B levels in
nonhuman primates. Atherosclerosis, 88:133.
23. Herting, D. C., and Drury, E. E. (1969). Alpha-tocopherol content of cereal
grains and processed cereals.l Agric. Food Chern., /7:785.
24. Wang, C. Y., and Luh, B. S. (1991). Harvest, drying, and storage of rough
rice, Rice Production and Utilization, Vol. I (B. S. Luh, ed.), Van Nostrand
Reinhold, New York, p. 311.
25. Juliano, B. 0. (1985). Criteria and tests for rice grain qualities, Rice Chemis-
try and Technology (B. 0. Juliano, ed.) AACC, St. Paul, MN, p. 495.
26. Rice as an Ingredient, 1989-1990. Gorman's New Prod. News. Gorman Pub-
lishing Co., Chicago, Ill.
27. Barber, S., Benedito de Barber, C., and Martinez, J. (1981). Rice bran pro-
teins. II. Potential value of rice bran fractions as protein food ingredients.
Rev. Agroquim. Techno. Aliment., 21:247.
28. Sharp, C. Q., and Kitchens, K. J. (1990). Using rice bran in yeast bread in a
home baker. Cereal Foods World, 35:1021.
29. Sayre, R.N., and Saunders, R. M. (1990). Rice bran and rice bran oil. Lipid
Techno/., 2:72.
30. Juliano, B. 0. (1985). Rice bran, Rice Chemistry and Technology (B. 0.
Juliano, ed.), AACC, St. Paul, MN, p. 647.
31. Yokochi, K. (1977). Rice bran processing for the production of rice bran oil
and characteristics and uses of the oil and deoiled bran. Proceedings of Rice
By-Products Utilization, International Conference, 1974, Valencia, Spain.
Vol. III. Rice Bran Utilization (S. Barber and E. Tortosa, eds.), Institute for
Agricultural Chemistry and Food Technology, Valencia, Spain.
32. Sakurai, J. (1977). Utilization of rice by-products in Japan. Proceedings ofRice
By-Products Utilization, International Conference, 1974, Valencia, Spain. Vol.
IV. Rice Bran Utilization: Food and Feed (S. Barber and E. Tortosa, eds. ), Inst.
Agroquim. Techno. Aliment., Valencia.
 17
Food Applications for Modified Rice Starches
Roy N. Sharpt and Carolyn Q. Sharp
University of Arkansas, Fayetteville, Arkansas

I. INTRODUCTION
Starch accounts for the greatest quantity of all the food consumed by
mankind. Its functional properties exert great influence on the consumer's
acceptance of the food. The way food starch functions in a food system
depends on its source. There are four general groupings of starch primarily
based on their source: tubers, cereal grains, waxy cereal grains (starch
fraction contains less than 2% amylose) grains, and legumes. Functional
properties of cereal starches differ distinctly from those of legumes and
tubers. Moreover, differences have been reported not only among cereals
but also among varieties of the same nonwaxy cereal.
The ability of starch to form a viscous paste when heated in water is its
most important property. Therefore, starch performs several different func-
tions in foods:

1. Binding agent for batters, breadings, and ice cream cones

2. Coating and glazing agent for nut meats and candies
3. Colloidal stabilizer for oil-in-water emulsions such as salad dressing
4. Dry-dusting of bakery products and certain candies
5. Fat substitute in salad dressings
6. Gel-forming agent in such confections as gum drops

'Deceased.

405
406 SHARP AND SHARP

7. Moisture-retention agent, as in toppings and icings for cakes

8. Thickening agent for pie-fillings, sauces, and soups.
The two major sources of starch in the United States are corn, which is
native, and cassava (starch known as tapioca), which is imported. Early
food processors sought starch supplies according to the commodity source.
As the consumer's textural expectations and quality consciousness in-
creased, functional properties of the starch became more important. The
need to improve the functionality of starch and expand its usefulness
brought about the search for ways to modify the properties of native starch.
Modified starches of today are designed to overcome the limitations of
native starch in providing controlled hydration, water-holding capacity, gel
formation, clarity and stability during aging.

II. MODIFIED STARCHES

Modification is the altering of the physical and/or chemical properties of
starch. Modified starches are often referred to, in general, as converted
starch, but may be defined by source, prior treatment, amylose and amylo-
pectin content or ratio, degree of polymerization or molecular weight, de-
gree of substitution, physical form, type of derivative, and associated con-
stituents. Because rice is the focus of this publication, other sources of starch
will not be addressed except where appropriate for comparative purposes. It
is outside the scope of this chapter to describe the detailed mechanisms of
starch modification. We will concentrate here on procedures to accomplish
certain changes and the applications of those alterations in food products.
As indicated above, rice is not one of the primary sources of modified
food grade starch. On a global basis, the majority of rice is consumed for
food as intact kernels in the country of origin. While the United States
produces only 1.3% (6.5 million metric tons [MMT]) of the world's rice, it
is a major exporter and supplies almost 20% of the rice that is traded
internationally (1). The amount of corn produced in the United States
(approximately 181.5 MMT) and the utilization pattern (2) present a very
different economic picture. Approximately 80% of the corn is used for
animal feed, providing a situation especially suitable to value-added pro-
cessing techniques, such as starch modification for food purposes.
It is understandable that the literature is lacking in information report-
ing specific processes for the modification of rich starch. This in itself does
not mean that techniques have not been developed for rice starch modifica-
tion. No doubt there have been proprietary investigations that were not
published, and many of the processes developed to modify other starch
sources are appropriate for rice.
FOOD APPLICATIONS FOR MODIFIED RICE STARCHES 407

In accordance with paragraph 172.892, Part 21, Code of Federal Regula-

tions (3), food starch modified under certain specified conditions "may be
safely used in food. The quantity of any substance employed to effect such
modification shall not exceed the amount reasonably required to accom-
plish the intended physical or technical effect nor exceed any limitation
prescribed." Food starch may be modified by treatments prescribed in
Table 1.

Table 1 CFR Requirements for Effecting the Modification of Food Starch

Maximum
allowable
Reactant reactant Limitations

Acid modified
Hydrochloric acid and/or Sulfu-
ric acid
Bleached (may use one or more
reactants)
Active oxygen from hydrogen 0.45% active
peroxide and/or peracetic acid oxygen
Ammonium persulfate and sul- O.o75%
fur dioxide 0.05%
Chlorine as calcium hypochlo- 0.036% Finished starch may be
rite used only as a compo-
nent of batter for com-
mercially processed
foods.
Chlorine as sodium hypochlorite 0.0082 lb. per.
lb dry starch
Potassium permanganate 0.2% 50 ppm Mn in finished
starch
Sodium chlorite 0.5%
Oxidized
Chlorine, as sodium hypochlo- 0.055 lb. per lb.
rite dry starch
Esterified
Acetic anhydride 2.5% acetyl groups in fin-
ished starch
Adipic anhydride and acetic 0.12% 2.5% acetyl groups in fin-
anhydride ished starch
Monosodium orthophosphate 0.4% Pin finished starch
1-0ctenyl succinic anhydride 3%
1-0ctenyl succinic anhydride 2%
and aluminum sulfate 2%
Table 1 Continued
Maximum
allowable
Reactant reactant Limitations
Phosphorus oxychloride 0.1%
Phosphorus oxychloride fol- 0.1% 2.5% acetyl groups in fin-
lowed by acetic anhydride 8% ished starch
Phosphorus oxychloride fol- 0.1% 2.5% acetyl groups in fin-
lowed by vinyl acetate 7.5% ished starch
Sodium trimetaphosphate 0.4% P in finished starch
Sodium tripolyphosphate and so- 0.4% Pin finished starch
dium trimetaphosphate
Succinic anhydride 4%
Vinyl acetate 2.5% acetyl groups in fin-
ished starch
Etherified
Acrolein 0.6%
Epichlorohydrin 0.3%
Epichlorohydrin and propylene 0.1% 5 ppm propylene chlorohy-
oxide 10% in combi- drin in finished starch
nation or in
any sequence
Epichlorohydrin followed by 0.1% 5 ppm propylene chlorohy-
propylene oxide 25% drin in finished starch
Propylene oxide 25% 5 ppm propylene chlorohy-
drin in finished starch
Esterified and etherified
Acrolein and vinyl acetate 0.6% 2.5% acetyl groups in fin-
7.5% ished starch
Epichlorohydrin and acetic anhy- 0.3% 2.5% acetyl groups in fin-
dride ished starch
Epichlorohydrin and succinic an- 0.3%
hydride 4%
Phosphorus oxychloride and 0.1% 5 ppm propylene chlorohy-
propylene oxide 10% drin in finished starch
Modified
Chlorine as sodium hypochlo- 0.055 lb per lb 5 ppm propylene chlorohy-
rite, active oxygen from hydro- dry starch drin in finished starch
gen peroxide and propylene 0.45%
oxide 25%
Sodium hydroxide 1%
Other
Modification may also be accomplished by combinations of the treatments
shown under the above sections of acid-modified and/or bleached and any
one of oxidized, esterified, etherified, esterified and etherified, or modified,
provided all limitations are respected.
Source: Ref. 3.
FOOD APPLICATIONS FOR MODIFIED RICE STARCHES 409

III. STARCH MODIFICATION METHODS

A. Cross- Linking
Starch is a polymer of anhydroglucose units; therefore, hydroxyls (-OH)
are abundantly available for potential chemical reactions. There are sev-
eral compounds capable of forming chemical bonds with starch at these
sites. Specific compounds, such as epichlorohydrin, phosphorus oxy-
chloride, and sodium trimetaphosphate (Fig. 1), contain two or more ac-
tive sites capable of reacting with hydroxyls and therefore may react with
hydroxyls from different starch molecules to yield a cross-linking effect
(Fig. 2).

NaO 0
\~
p
I\
0 0
I \
O=P-0-P=O
I I
ONa ONa

(A)

0
II
C l - P - Cl
I
Cl

(B)

(C)
Figure 1 Molecular structure of (A) sodium trimetaphosphate, (B) phosphorus
oxychloride, and (C) epichlorohydrin.
410 SHARP AND SHARP

starch Epichlorohydrin

-o0o<r~o0o-OH OH 0 OH
I I
CH 2 -CH-CH 2
I + HCl

-o,)o;-f)ro~o 1 ~o-
~o
CH 0H
1---o
CH 2 0H
~o/
CH 0H
1
2 2

Cross-linked starch

Figure 2 Chemical reaction between starch and the cross-linking agent epichloro-
hydrin. (From Ref. 4.)

Cross-linked rice starch products are not easily found. The process of
cross-linking rice starch should follow the same procedure as any other
starch. However, there is little mention in the literature of studies using
rice starch. This leads one to conclude either that rice starch does not
respond to typical cross-linking conditions and/or that the majority of
the rice starch cross-linking experiments were conducted as proprietary
work.
The most highly publicized rice cross-linking experiments were con-
ducted on milled, intact rice kernels to effect greater canning stability (4-
7). Milled long-grain rice (Starbonnet, cv.) is rinsed with water to remove
any talc and glucose applied for commercial purposes, then treated as
shown in Figure 3 to achieve cross-linking of starch in the rice kernel.
Treatment of the rice with NaOH increases the pH sufficiently to activate
the hydroxyl groups of the starch. The entire procedure is conducted at
room temperature and is quite time consuming in order to achieve swelling
without gelatinizing the starch granules. A closed system is required to
prevent loss of the volatile epichlorohydrin. The high pH required for
activation of the starch causes a yellow discoloration of the rice kernel. This
412 SHARP AND SHARP

result in excessive leaching of solids. Once the cross-linking process is

complete, the treated rice exhibits approximately 68% less starch leaching
than the untreated rice at pH 5 (6).
Rutledge et al. (4) explained that many starches are cross-linking by
reacting epichlorohydrin with hydroxyl groups of starch to form mono- and
diethers, which may be either inter- or intramolecular. Therefore, they
assumed that the chemical reactions would be the same for rice. Rice so
treated is reported to have marked changes in the swelling behavior and
overcooking resistance, with 78% lower solids loss than normally found in
parboiled rice. These changes reportedly eliminate the need for parboiling
rice prior to thermal processing of canned soups.
Phosphorus oxycholoride and sodium trimetaphosphate are ineffective
cross-linking reactants for white rice (7). These same reactants are effec-
tive when used on parboiled rice, with sodium trimetaphosphate- and
epichlorohydrin-treated parboiled rice being superior to that treated with
phosphorus oxychloride (5).
A greater degree of crystallinity in the ungelatinized kernel of non-
parboiled rice, in comparison to parboiled rice, reportedly is why sodium
trimetaphosphate and phosphorus oxychloride fail as reactants for cross-
linking nonparboiled rice (8). Two factors have been proposed to explain
why epichlorohydrin effectively cross-linked nonparboiled rice kernels
while the other two compounds failed: (a) the relative molecular structure
of the three reactants (see Fig. 1), and (b) the condition of the rice when it
is subjected to the reactant.
Storage studies of canned cross-linked rice indicated that a considerable
amount of solids continues to leach from the kernel during storage. Further-
more, the loss is greater at pH 5 than at pH 7 (9). Whether rice is hydrated
before or after the can is sealed, the cross-linked rice loses less solids and is
less sticky than the untreated rice (10).
In addition to the compounds mentioned above, polyhydric alcohols
have been used to cross-link rice starch. Patents were issued for a proce-
dure for preparing canned rice that included treatment of milled rice with
polyhydric alcohols partially esterified with higher fatty acids (11,12). This
treatment probably modifies the starch on the surface of the rice kernel by
chemical esterification to reduce the degree of stickiness.
Rice flours modified by either acetic or succinic anhydride treatment
yielded products with varied functional characteristics (13). Increased loaf
volume, softer crumb texture, and better-tasting bread were reported as
advantages attained by using modifed rice flour. Amylographic viscosity
profiles produced from modified rice starch slurries (Fig. 4) indicate that
functional properties of rice starch can be manipulated to increase the
viscosity, but greater degrees of cross-linking decrease the viscosity (14).
 FOOD APPLICATIONS FOR MODIFIED RICE STARCHES 413

t-EATING HOLDING CODLING HOLD!r-13

1400 ................ REACTION A

1300 ..··· /
......-·
'....,"(/)" 1200 -----·-------····REACT I ON 8
1100
c
. ..····.·.•'<::>';'>-···- .-----·---·RER:Tl(J< c
/
:::J
1000
L

"0
QJ 900 ( ... ··. / ......-----·-·
c
D
QJ 800 _,./
0
700 : ! //
"""' I !/ ---------· REACT I ON 0
L
m
500
>-
1- 500 .,~~~/<~~---STARCH
, .. / ___ -:>-=-
400 . ft /---------
(f)
0
u
(f)
300 1 I :•
> i iii
200 i i if
100
025 ·c
_//if..l
92.5 c 92.5 c 30 c 30 c
Temperature (C)

Figure 4 Amylographic viscosity profiles of rice starches modified by adipic

anhydride and acetic anhydride esterification. Adipic acid contents of the modified
starches from reactions A, B, C, and D were 0.008, 0.023, 0.065, and 0.167%,
respectively.

The starch was modified by a joint acetylation and cross-linking reaction

using adipic acid and acetic anhydride.

B. Retrogradation
Retrogradation is the recrystallization of gelatinized starch. This term is
normally used to indicate the thickening phenomenon that occurs during
the cooling of cooked starch. Retrogradation occurs in most starches and in
general is not considered as a starch modifying process. In the context of
altering the starch component of food (such as a rice kernel), retro-
gradation can be used as a method to change the functionality of the starch.
During cooling, carbohydrate molecules become physically aligned in a
somewhat parallel manner due to hydrogen bonding among the hydroxyl
groups in one molecule of amylose and those in another (15). Therefore,
recrystallization of starch is essentially a cross-linking process driven by
414 SHARP AND SHARP

electrostatic interactions rather than covalent bonding. In this sense, delib-

erate, controlled retrogradation of the starch in rice kernels may be consid-
ered as starch modification, especially when it is used for a specific effect.
Domoto (16) induced a firming action on milled rice kernels by chilling
cooked rice to approximately 37°F (3°C). Roseman (17) reported that struc-
tural alteration of rice resulted from the freezing of rice, which had been
hydrated to a minimum of 60% water. Such treatment changes the water
uptake of the rice. Retrogradation of the rice starch promotes the develop-
ment of a firmer kernel surface maintaining a greater degree of kernel
integrity and individuality. This surface firming allows the development of
a method for canning rice that employs a freezing step immediately follow-
ing thermal processing to prevent the rice from forming a solid body in the
can and provides a product that pours more easily from the can (18).
Freezing and thawing the rice before canning (19) results in canned rice
with retarded water uptake during and increased solids loss following can-
ning. The net effect of freezing and thawing treatments is canned rice with
a more acceptable appearance and less stickiness.
The ability to alter rice functionality by retrogradation depends on the
amylose content of the rice. The higher the amylose content of the rice
starch, the greater the degree of retrogradation as evidence by amylo-
graphic cold paste viscosity (20) and amylographic consistency (21) of rice
flour (Table 2).
X-ray diffraction has shown that freezing induces the same effect on rice
starch as found in normal retrogradation of cooked rice (22).

C. Parboiling
The method most used to alter the functional properties of rice starch may
be the one that is most overlooked. Parboiling (partially cooking) is a

Table 2 Influence of Amylose Content on Amylographic•

Consistency and Cold Paste Viscosity (50°C)

Cold pastec
Amylose type Consistencyb viscosity

High 560 1135

Intermediate 275 765
Low 185 555
Waxy 73 450
•10% slurry.
bCold paste (50°C) viscosity minus hot paste (95°C) viscosity.
cSlurry cooled to 50°C.
Source: Ref. 21.
FOOD APPLICATIONS FOR MODIFIED RICE STARCHES 415

process by which the texture of rice is physically altered by applying heat to

high-moisture rough rice. Procedures by which parboiling may be accom-
plished are extremely varied. There are several excellent reviews of parboil-
ing (23-25) addressing the development of the technology and its impact
on human nutrition. Parboiling is thought to have originated in ancient
India (24) and spread throughout the general area of India, Bangladesh,
and Sri Lanka as a home preparation technique and no doubt was later
adapted to the village level to meet needs of trading.
The high incidence of beriberi (thiamine deficiency) in areas of the
Orient where the diet was composed largely of milled rice, and the absence
of the disease in specific areas where parboiled rice was consumed, called
attention in the early nineteenth century to the nutritional benefits of the
process. An improved parboiling process developed in England became
known as the "conversion process." The use of this process was acquired
and put into production in Houston, Texas, about 1941. War World II
greatly influenced the development of the parboiling industry in the United
States. Military operations needed rice with greater resistance to adverse
storage conditions and increased nutritional value. Parboiling imparted
these qualities to the rice.
Modification of rice starch due to parboiling specifically results in the
entrapment of protein, B vitamins, and minerals that were translocated
from the outer layers of the rice kernel and dispersed throughout the
endosperm. Up to 80% of the original thiamine has been found in the
endosperm of milled parboiled rice (26). Parboiling changes the visual
appearance of the rice kernel from opaque and dull to translucent and
shiny. The grain structure is more compact, imparting a greater resistance
to insect attacks, greater resistance to water uptake during cooking, and a
firmer texture after cooking (24,25). Along with the these advantages come
disadvantages. Water-soluble pigments are also carried from the outer lay-
ers during soaking and are subsequently trapped in the endosperm upon
starch gelatinization. It took many years to understand that thiamine, being
water soluble, migrated from the outer layer into the endosperm of the
kernel and became trapped there when the starch became disordered (dena-
tured, modified, or also referred to as converted).
Regardless of the original purpose for parboiling rice, we can look back
on the process as being one of the earlier methods of modifying starch. In
addition to gelatinization and reassociation, thermal degradation of the
starch must be considered as a contributor to the peculiar properties of
parboiled rice (27).
Commercially prepared rice and rice flour can be obtained with varied
degrees of parboiling. The trademark of the process may vary depending of
the processor. Some of the commercial terms indicating that rice has under-
gone a parboiling process are "conditioned," "converted," and "parboiled."
416 SHARP AND SHARP

The firming effect of parboiling on the kernel was recognized early in the
history of the U.S. rice-processing industry. This provided greater incentive
for refining parboiling methodologies (28,29). Later documentation substan-
tiated that parboiling prevented the deterioration of rice kernels during
canning as evidence by reduced solids loss and increased firmness (30).
X-ray diffraction spectra and differential scanning calorimetry of par-
boiled and nonparboiled rices indicate that parboiled rice is largely amor-
phous with only minor crystallinity (31).

D. Dry Heating
Rapid heating of milled rice using preheated air, nitrogen, carbon dioxide,
or superheated steam (32) can be used to alleviate stickiness of canned rice.
The application of dry heat to parboiled rice to develop a distinct toasted
flavor results in an increased water uptake ratio (water uptake at 60°C
divided by water uptake at 96°C), a reduction in the amylographic viscosity
(32), and a greater degree of firmness in kernels of canned rice (33,34)
(Table 3).
Pyrolysis of starch to pyrodextrins, sometimes called British gums, has
applications for the food, paper, and textile industries. Pyrolysis is the ther-
mal degradation of starch without the use of added catalyst. Pyrodextriniza-
tion (35) increased the solubility of starches from wheat, jowar (maize), ragi,
and rice. The much lower solubility for rice starch is attributed to the nature
of associative bonding forces within the granules, the composition of starch
(particularly amylose:amylopectin ratio), and their crystalline structures.
Prolonged pyrodextrinization (greater than 10 hours) results in reduced {3-

Table 3 Effects of Dry Heat on Water Inhibition, Firmness, and

Amylographic Viscosity of Parboiled Rice

Toasting Water Shear Amylographic viscosity (BU)c

time at 170oc uptake valueb
(min) ratioa (lb) Peak Hot paste Cold paste

0 .41 44 520 465 1165

15 .46 59 185 250 570
30 .47 67 155 240 535
45 .49 70 125 215 500
60 .51 76 110 180 450
•Water uptake 60°C/water uptake 96°C.
blOO g canned rice product using a 13 blade Kramer shear cell.
cBrabender units.
Source: Ref. 33.
FOOD APPLICATIONS FOR MODIFIED RICE STARCHES 417

amylolysis in all four starches indicating the likelihood of increased branch-

ing due to a transglycosidation reaction during pyrolysis.

E. Acidification
The knowledge that dispersed starch can be acid hydrolyzed to produce
sugars and syrups dates back to the early 1800s. The use of controlled acid
concentrations and processing temperatures to produce modified starches
on a commercial basis began about the turn of this century (36).
Comparative acid modification of various starches (37) has shown that
the susceptibility of starch granules to acid attack differs widely depending
on the original source. These findings suggest that cereal starches behave
almost as one group and yield thinner (lower viscosity) liquids than starch
from chickpea or potato under similar treatment conditions. The degree of
vulnerability is also suggestive of the differences in the molecular and
granular makeup of different starches.
The most significant application of the acidification process for rice ap-
pears to have been to the rice kernel rather that to dispersed starch. Nor-
mally the purpose of acidification is to reduce the hot paste viscosity of native
starch. However, when used on intact rice kernels it can provide a very
different effect. Willison (38) prepared milled rice for canning by rinsing it
with cold water, then boiling it for 20 minutes in acidified (phosphoric acid)
water. Rice treated in this manner retains kernel individuality. Gallenkamp
(39) reversed the steps of Willison and still attained the desired effect of a
firmer textured kernel. Tollefson and Bice (40) steeped parboiled rice in
acidified (glacial acetic acid) water to attain a 55-60% moisture content with
no actual inactivation step to prepare rice for canning and attained the same
effect as Willison. Gallenkamp (39) and Wilbur (41), working to produce a
commercially sterile canned rice product, attained the same benefit of firmer
texture and greater kernel individuality though decreased pH.
Parboiled rice canned in acidified (food grade citric acid) water, to yield
increasingly lower final product pH, exhibits the greatest degree of firmness
in the 6.1-5.2 pH range; thereafter, the shear value decreases steadily as
final pH decreases (42). When the acidulant is malic acid, a similar trend is
found but the texture values are not statistically different. When parboiled
rice is toasted to develop a prominent toasted flavor, the initial firming effect
due to pH decrease is not demonstrated using either citric or malic acid.

IV. FUTURE FOR NEW RICE STARCH MODIFICATIONS

Both academic and industrial research have focused on the modification of
corn starch. Therefore, a great variety of modified starches from corn are
418 SHARP AND SHARP

already commercially available. Price comparison of corn and rice indicates

that rice is at an economic disadvantage for direct competition. However,
as food manufacturing becomes more complex, starch products with differ-
ent characteristics are needed. Differences in the functional quality of
nonmodified rice starches signify potential for great variability in modified
starches, and marketing opportunities can be enormous if new modified
rice starches are developed to meet specific needs.

V. CONCLUSIONS
Rice starch provides significant nutritional benefits in many cultures. The
modification of rice starch can extend its nutritional benefit and provide
products with very different functional properties. New uses for modified
rice starch continue to be developed, with functions ranging from colloidal
stabilizers to fat replacers to dusting agents. Although not a major source
of starch in the United States, many rice products are the result of rice
starch modification. The most important modification from an economic
standpoint and often the most overlooked is the process of parboiling. Rice
is one of the few commodities in which starch modification is practical in
both the intact kernel and in the extracted starch.

REFERENCES
1. Sharp, R. N. (1991). Rice: production, processing, and utilization, Handbook
of Cereal Science and Technology (K. J. Lorenz and K. Kulp, eds.), Marcel
Dekker, Inc., New York, p. 301.
2. OTA. (1989). U.S. Congress, Office of Technology Assessment, Enhancing
the Quality of U.S. Grain for International Trade, OTA-F-399, Washington,
DC.
3. FDA. (1988). Food starch-modified. Code of Federal Regulation, Title 21,
Sec. 172.892.
4. Rutledge, J. E., Islam, M. N., and James, W. H. (1972). Improved canning
stability of rice by chemical modification. Cereal Chern., 49:430.
5. Rutledge, J. E., Islam, M. N., and James, W. H. (1974). Improved canning
stability of parboiled rice through cross-linking. Cereal Chern., 51:46.
6. Rutledge, J. E., and Islam, M. N. (1973). Canning and pH stability of
epichlorohydrin-treated parboiled rice. J. Agr. Food Chern., 21:458.
7. Islam, M. N. (1973). Improved canning stability of rice by chemical modifica-
tion. Diss. Abstr. Int. B, 33 (12, Part I):5900.
8. Islam, M. N., Rutledge, J. E., and James, W. H. (1974). Influence of rice
crystallinity on cross-linking. Cereal Chern., 51:51.
9. Rutledge, J. E., and Islam, M. N. (1976). Improved shelf-life at room tempera-
ture of canned rice modified by cross-linking. Cereal Chern., 53:862.
FOOD APPLICATIONS FOR MODIFIED RICE STARCHES 419

10. Rutledge, J. E., and Islam, M. N. (1976). Limited moisture canning of rice-
Effects of cross-linking. Cereal Chern., 53:982.
11. Ferrel, R. E. (1959). Procedure for preparing canned rice. U.S. Patent
2,898,214. August 4, 1959.
12. Ferrel, R. E. (1959). Method of canning rice. U.S. Patent 2,898,215. August
4, 1959.
13. Alhusaini, S. S. (1985). Modification of rice flour and its potential in the food
industry. Diss. Abstr. Int. B, 46 (6):1758.
14. VanDevender, D., and Sharp, R.N. (1990). Development of a freeze-thaw
stable cross-linked rice flour. Proc. Rice Tech. Working Group, Biloxi, MS,
February 26-28, 1990, p. 80.
15. Dengate, H. N. (1984). Swelling, pasting, and gelling of wheat starch, Ad-
vances in Cereal Science and Technology (Y. Pomeranz, ed.), American Asso-
ciation of Cereal Chemists, Inc., St. Paul, MN, p. 49.
16. Domoto, T. (1936). Process of canning rice. U.S. Patent 2.055,120. September
22, 1936.
17. Roseman, A. S. (1958). The effect of freezing on the hydration characteristics
of rice. Food Techno/., /2:464.
18. Verity, N. S., and Allen, R. C. (1964). Method of canning rice. U.S. Patent
3,132,030. May 5, 1964.
19. Loustaunau, J. H. (1976). Retrogradation of rice starch for improved canning
stability. Dis. Abstr. Int. B, 36 (12, Part I):6050.
20. Webb, B. D. (1985). Criteria of rice quality in the United States, Rice Chemis-
try and Technology (B. 0. Juliano, ed.), American Association of Cereal
Chemists, Inc., St. Paul, MN, p. 403.
21. Juliano, B. 0. (1985). Criteria of rice quality, Rice Chemistry and Technology
(B. 0. Juliano, ed. ), American Association of Cereal Chemists, Inc., St. Paul,
MN, p. 443.
22. Roseman, A. S., and Deobald, H. J. (1959). Effect of freeze-processing on
amyloclastic susceptibility, crystallinity, and hydration characteristics of rice.
J. Agr. Food Chern., 7:774.
23. Kik, M. C., and Williams, R. R. (1945). The nutritional improvement of
white rice. Bull. No. 112, National Research Council, Washington, D.C.
24. Bhattacharya, K. R. (1985). Parboiling of rice, Rice Chemistry and Technol-
ogy (B. 0. Juliano, ed.), American Association of Cereal Chemistry, Inc., St.
Paul, MN, p. 289.
25. Luh, B.S., and Mickus, R. R. (1991). Parboiling rice, Rice: Vol. II, Utiliza-
tion, 2d ed. (B.S. Luh, ed.), Van Nostrand Reinhold, New York, p. 51.
26. Kik, M. C. (1943). The story of rice conversion. Rice J., 46 (3):7.
27. Mahanta, C. L., and Bhattacharya, K. R. (1989). Thermal degradation of
starch in parboiled rice. Starch, 41:91.
28. Yonan-Malek, M. (1943). Control system of boiling and canning rice. U.S.
Patent 2,334,665. November 16, 1943.
29. Yonan-Malek, M. (1943). Control system of boiling and canning rice. U.S.
Patent 2,334,666. November 16, 1943.
420 SHARP AND SHARP

30. Demont, J. 1., and Bums, E. E. (1968). Effects of certain variables on canned
rice quality. Food Techno/., 22:1186.
31. Mahanta, C. L., Ali, S. Z., Bhattacharya, K. R., and Mukherjee, P. S.
(1989). Nature of starch crystallinity in parboiled rice. Starch, 41:171.
32. Ott, M. (1970). Treatment of rice. U.S. Patent 3,532,508. October 6, 1970.
33. Sharp, R.N., Kattan, M. W., and Sharp, C. Q. (1985). The effects of toasting
on the quality of canned rice. J. Food Sci., 50:340, 381.
34. Sharp, R. N., Kattan, A. A., and Sharp, C. Q. (1984). Effects of toasting on
rice cooking quality. Ark. Farm Res., 33 (4):11.
35. Wankhede, D. B., and Umadevi, S. (1982). Preparation and some physico-
chemical properties of pyrodextrins of ragi, wheat, jowar and rice starches.
Starch, 34:162.
36. Wurzburg, 0. B. (1986). Converted starches, Modified Starches: Properties
and Uses (0. B. Wurzburg, ed.), CRC Press, Inc., Boca Raton, FL, p. 18.
37. Singh, V., and Zakiuddin Ali, S. (1987). Comparative acid modification of
various starches. Starch, 39:402.
38. Willison, W. W. (1926). Method of preparing rice for canning. U.S. Patent
1,589,672. June 22, 1926.
39. Gallenkamp, N. (1952). Process for canning rice. U.S. Patent 2,616,810. No-
vember 4, 1952.
40. Tollefson, C. I., and Bice, C. W. (1972). Method of preparing canned cooked
rice. U.S. Patent 3,647,486. March 7, 1972.
41. Wilbur, P. C. (1940). Rice product and process. U.S. Patent 2,187,718. Janu-
ary 23, 1940.
42. Sharp, R. N. (1981). Process control requirements, quality attributes and
nutritive characteristics of thermally processed rice and its fortification with
soy protein. Ph.D. dissertation, University of Arkansas, Fayetteville, AR, p.
28.
 18
Rice Bran Oil and Its Health Benefits
Robert J. Nicolosi and Eugene J. Rogers
University of Massachusetts-Lowell, Lowell, Massachusetts

Lynne M. Ausman
Tufts University, Boston, Massachusetts
Frank T. Orthoefer
Rice/and Foods, Inc., Stuttgart, Arkansas

I. INTRODUCTION
Many studies have been performed on the effect of various oils on serum
cholesterol. An oil is often identified by its source such as rice, corn,
soybean, or coconut oil. Often the oil is insufficiently characterized as to its
composition, degree of oxidation, or nontriglyceride components. Only
recently have investigators become more aware of the non triglyceride com-
ponents beyond the fatty acid composition. This chapter attempts to bring
together the composition of rice bran oil, the effect of processing on the
non triglyceride components of the oil, and the health benefits derived from
rice bran oil, including the oil as part of the diet.

ll. RICE BRAN OIL AND ITS STABILIZATION

Rough rice contains 0.6-3.5% oil. The oil is concentrated primarily in the
germ but is present in the bran as well. About 80% of the lipids of brown
rice are in the bran and polish. About one third is in the embryo. Rice bran
includes, in the United States, the soft germ and several histologically identi-
fiable soft layers surrounding the hard starchy endosperm. Rice bran and
germ contain the oil referred to as rice bran oil. Bran contains 15-20% oil.
Rice bran also contains a highly reactive lipase enzyme. When the bran
layers are removed from the endosperm during milling, the individual cells

421
422 NICOLOSI ET AL.

are disrupted and the rice lipids come into contact with the lipase enzyme.
Hydrolysis of the ester occurs with a rapid increase in the free fatty acid
content of the bran. Bran from parboiled rice is stable due to the deactiva-
tion of the enzyme during the parboiling process.
Extrusion cooking, a low-cost stabilization technique, is increasingly
used because of its positive effects on nutritional value. Extrusion is gener-
ally practiced immediately after milling to minimize the extent of hydrolytic
rancidity. Extrusion-stabilized bran possesses several months of storage
stability. Both nonstabilized bran and extrusion-stabilized bran have been
used for oil extraction. Once extracted, the oil is stable and may be stored
for subsequent processing through the conventional steps to an acceptable
oil.

III. COMPOSITION OF RICE BRAN OIL

A. Triglycerides
The main component of crude rice bran oil, the triglycerides, make up
approximately 80% of the oil. For a typical fatty acid composition of rice
bran oil, an oleic acid oil, see Table 1. The triglyceride content varies but is
primarily dependent on the extent of hydrolysis that occurred prior to stabili-
zation of the bran (Table 2). Partial esters present as mono- and diglycerides
and the free fatty acids are a reflection of the hydrolysis that has occurred.

B. Minor Constituents
The minor constituents of an oil consist of phospholipids, glycolipids,
sterols, waxes, and tocopherols. The phospholipids present are phos-

Table 1 Composition of Rice Bran Oil

Chain length:
No. of
Fatty acid double bonds Percent

Myristic 14:0 0.1- 1.0

Palmitic 16:0 12.0-18.0
Palmitoleic 16:1 0.2- 0.6
Stearic 18:0 1.0- 3.0
Oleic 18:1 40.0-50.0
Linoleic 18:2 20.0-42.0
Linolenic 18:3 0.0- 1.0
Arachidic 20:0 0.0- 1.0
HEALTH BENEFITS OF RICE BRAN OIL 423

Table 2 Composition of Crude Rice Bran Oil

Triglycerides 80%
Phospholipids 2%
Glycolipids 1%
Sterols 5%
Waxes 2-5%

phatidylcholine, phosphatidylethanolamine, and phosphatidylinositol. The

glycolipids are mainly galactose and glucose derivatives.
The broad class of unsaponifiable matter of rice bran oil consists of
approximately 42% sterols, 24% higher alcohols, 20% ferulic acid esters,
10% hydrocarbons, and 2% unknown. The unsaponifiables are generally
represented by the sterol fraction of the oils. These include free sterols,
sterol esters, sterylglycosides, and acylsteryl glycosides. J3-Sitosterol is the
most abundant sterol present. Oxyzanol, present at 0.96-2.9% of bran oil,
is a ferulic acid ester of triterpenoid alcohols and is present in the sterol
fraction removed primarily during caustic refining of the oil. Oxyzanol is
reported to have antioxidant properties similar to that of vitamin E.
Wax concentration in the crude oil is dependent on the extraction
method and origin of the bran. Wax may comprise about 3-9% of rice bran
oil. Generally, the higher the extraction temperature, the greater the quan-
tity of wax removed from the bran when hexane is used as the extraction
solvent. Waxes have low iodine values (10-12) and high melting points
(82-84°C). Waxes have been classified into hard and soft fractions. The
fatty alcohols of the hard fraction were found to be C-24, C-26, and C-30.
The saturated fatty acids were C-22, C-24, and C-26, and normal alkanes of
C-29 and C-31. Soft wax contained C-24 and C-30 alcohols, C-16 and C-24
saturated fatty acids, and C-21 and C-29 normal alkanes.
Tocopherols are present in relatively high quantities at 0.1% (5% of the
unsaponifiable fraction). a-Tocopherol, a-tocotrienol, and y-tocotrienol
constitute the major fractions present.

IV. EXTRACTION OF RICE BRAN OIL

The process flow of rice bran for oil extraction and oil refining is shown in
Figures 1 and 2, respectively. Bran from freshly milled rice has an initial
free fatty acid level of 1.5-2% for milled rice and 2-5% for parboiled. Free
fatty acid level of the bran for oil extraction should be as low as possible for
ease of extraction and efficiency of refining of the oil as well as for yield
considerations.
424 NICOLOSI ET AL.

!
Rice Bran from mill

Stabilization

!
Drying

L
Storage

!
Hexane extraction

I
Meal desolventizing
I Hexane ~istillation
J, ~
Defatted meal Crude rice bran oil

Figure 1 Process flow chart for rice bran oil extraction starting with rice bran
from the mill.

Oil may be removed from the bran using hydraulic pressing and/or
solvent extraction. Hydraulic pressing is limited to approximately a 50% oil
recovery. Pressed oil generally contains significant quantities of fines requir-
ing filtration prior to refining of the oil.
Extraction of the oil may be carried out with a variety of solvents,
although hexane is generally used. Rice bran for solvent extraction may be
steamed for stabilization and to facilitate pellet or collet formation. The
steam treatment also assists freeing the oil from other bran components.
Bran in a pellet or collet form has a significantly higher solvent percolation
rate leading to shorter extraction times. Extrusion-stabilized bran can be
stored without degradation prior to oil removal. The extraction may be
carried out in a batch-type or continuous extractor. The temperature of
extraction influences the quantity of wax removed. The solvent plus oil,
referred to as micella, is filtered prior to distillation of the solvent. The wet
defatted bran is desolventized, dried, and cooled. Solvents are recovered
throughout the process. The micella may undergo dewaxing and refining
prior to desolventizing. Solvent dewaxing facilitates wax removal with mini-
HEALTH BENEFITS OF RICE BRAN OIL 425

Crude rice bran oil

Phosphoric acid----~~~ f
Water---..:1!~~ ~
Degumming ) Phospholipids

NaOH---)-. ~
Refining----~);~ Fatty acid soaps

Acid activated
bleaching earth
--~'• 7 ~
I
Bleached oil

~
Winterization --~)• Stearine

~
Deodorization--~)~ Vegetable oil
~ distillate

Rice bran oil

Figure 2 Process flow chart for conventional oil refining starting with crude rice
bran oil.

mal loss of neutral oil. Crystallization times are minimized also. Solvent
refining offers similar yield advantages. Desolventizing is accomplished
through distillation. The products of extraction may consist of defatted
bran, crude rice bran oil, wax, and soaps of fatty acids.

V. PROCESSING OF RICE BRAN OIL

The degree of processing of the oil depends upon its intended uses. Gener-
ally, food applications require only the triglyceride portion of the crude oil.
Various components are more difficult to remove from the crude oil and do
not detract from food utilization. These usually consist of unsaponifiables,
pigments, and limited quantities of partial esters. The processes used for
preparation of food oils consist of dewaxing, degumming, neutralization,
bleaching, winterization, and deodorization, as described below.
426 NICOLOSI ET AL.

A. Dewaxing
Dewaxing of the rice bran oil while still in micella form is the most efficient
dewaxing method. Other methods consist simply of cooling the oil and
allowing the waxes to crystallize and settle. Suspended waxes may be re-
moved by filtration or centrifugation. Micella dewaxing consists of cooling
the oil-solvent mixture to about 15°C, allowing crystallization to occur, and
centrifugation or filtration for wax removal.

B. Degumming
Gums consist of polar lipids having surface-active properties. Gums from
soybean oil when dried are the lecithin emulsifiers of industry. Rice oil
degumming generally requires the use of degumming agents such as phospho-
ric or citric acids to hydrolyze the gums, ensuring complete hydration. The
oil is heated to 60-80°C, a small amount of the acid is added along with water
for hydration. The wet gums are separated from the oil by centrifugation.
Acidic degumming is usually combined with neutralization or refining.

C. Neutralization

Neutralization or refining of the rice oil consists simply of neutralizing the

free fatty acids present. Caustic soda (NaOH) is generally used for the
neutralization. Free fatty acids present are converted to sodium soaps,
which are hydratable and removable by centrifugation. A small excess of
caustic (0.5% excess of a 18% caustic solution) ensures neutralization of
both the free fatty acids and acidic degumming agents. Temperature is kept
low, less than 65°C, for rice bran oil as severe darkening occurs at higher
temperatures.
High refining losses are often encountered with rice bran oil because of
high free fatty acid levels in the oil. Micella refining results in decreased
losses due to entrainment of oil. Rice oil, having high levels of free fatty
acids, may be double refined using conventional methods to reduce oil
losses yet ensure free fatty acid removal.

D. Bleaching

Bleaching of the oil is carried out to remove pigments, oxidized lipids, and
polar components from the oil. Acid-activated bleaching clays are added to
the oil at 1-3%, the mixture is agitated and heated under vacuum, and the
clay is removed by filtration.
HEALTH BENEFITS OF RICE BRAN OIL 427

E. Deodorization
Final removal of odors, flavors, and free fatty acids occurs during deodori-
zation. Deodorization, a steam distillation process, consists of a combina-
tion of high vacuum, steam purging, and high-heat treatment of the oil.
Continuous and batch-type systems are employed. Actual conditions used
were 220-250°C, 4-8 mmHg, and 3-5% steam. Volatile compounds re-
moved include aldehydes, ketones, and peroxides along with a portion of
the tocopherols and sterols present in the oil. After deodorization, the oil is
cooled to 70-90°F prior to storage or shipping.

F. Winterization
Winterization is performed to remove the high melting triglycerides from
that fraction that remains liquid at refrigeration temperature. The oil is
winterized by slowly cooling the oil to 5°C and holding for up to several
days. The saturated glycerides that crystallize may be removed by filtration
producing a stearine (high melting fraction) and oil (low melting fraction).
The stearine fraction may be used in formulation of high stability blended
fats. Winterization is normally performed prior to deodorization.

VI. EFFECT OF PROCESSING ON OIL COMPOSITION

Processing of rice oil can cause significant variation in the levels of unsaponi-
fiables seen in commercially available rice bran oil. Pilot plant processed oils
followed the conditions shown in Table 3. Analysis oftocotrienols (Table 4)
and oryzanol (Table 5) from commercially available rice bran oil is highly
variable. When the processing steps that lead to losses in total unsaponi-
fiables are investigated, one notes up to 50% losses of these components
from the crude to refined, bleached, deodorized, and winterized procedures
(Table 6). However, there are more striking selective losses of individual
unsaponifiable components. For example, during processing, up to 90% of
the oryzanol (Table 5) and tocotrienols (Table 7) can be lost during process-
ing. Obviously, this implies the need to develop new methods of processing
that will optimize recovery of these active ingredients.
While we have focused on the losses that can occur during the processing
of rice bran oil, it is important to note that significant differences exist in
the active unsaponifiable ingredients in the rice bran itself. For example,
during extraction of the rice bran, we observed very striking differences in
the levels of tocotrienols (Table 8), suggesting that not all full-fat rice brans
are created equally. This might suggest that differences in the stabilization
procedure of rice bran can result in variable recoveries of the unsaponi-
428 NICOLOSI ET AL.

Table 3 Conditions Used for Preparation of Pilot

Plant Quantities of Rice Oil
Rice bran
Fresh from rice mill
FFA 2.0-2.5%
Stabilization
To peroxidase negative by bubble test
Anderson VOX 88 Expander
Water to 25-28%
Sparge steam (130°C)
Drying
Forced-air, gas-fired dryer
Final moisture 8-10%
Extraction and desolventizing
Crown extractor similar to conditions for soybean oil.
Refining
Pretreat: 0.25% phosphoric
Neutralization: 0.5% excess, 18% NaOH
Temperature: 120°F
~leaching
1.5-2% acid-activated bleaching earth
Winterization
5°C, 48 hr
Deodorization
Temperature: 490-SOOoF
Vacuum: 7 mmHg
Steam: 2-3%

Table 4 Tocotrienol Content of Commercially

Available Rice Oil
Brand 1: 500ppm
Brand 2: 595 ppm
Brand 3: 353 ppm
Brand4: 78ppm
Brand 5: 407 ppm
Brand 6: 487 ppm
Brand 7: 288 ppm
HEALTH BENEFITS OF RICE BRAN OIL 429

Table 5 Oryzanol Content of Rice Oil

Crude: 12221 ppm
Commercially available: 1000 ppm (Brand 1)
2000 ppm (Brand 2)
720 ppm (Brand 3)
200 ppm (Brand 4)

Table 6 Unsaponifiable Content of Rice Oil

Oil type Percent
Crude 3.98
Refined and bleached 2.85
Refined, bleached, and deodorized 2.51
Refined, bleached, deodorized, and winterized 1.99

Table 7 Tocotrienol Content of Rice Oil

Oil type ppm
Crude 713
Refined and bleached 683
Refined, bleached, and deodorized 87
Refined, bleached, deodorized, and winterized 70

Table 8 Tocotrienol and Tocopherol Content of Two Different Rice

Bran Products
Tocotrienols (ppm) Tocopherols (ppm)
8- {3-/y- a- Total 8- {3-/y- a- Total
RB-X3 53 520 173 710 ND 98 164 262
RB-Y3 17 NDb ND 17 ND ND ND ND
•RB-X and RB-Y are two different brands of commercially available rice bran.
bNone detected.

fiable components. While processing plays a major role in determining the

final concentration of these components, one cannot rule out the contribu-
tion that varietal differences of rice can play in determining the levels of
unsaponifiables and the active ingredients that make up rice bran oil.
430 NICOLOSI ET AL.

VII. HEALTH BENEFITS OF RICE OIL

A. Effects of Fatty Acids
Numerous studies have demonstrated that diets enriched in saturated fatty
acids raise serum total cholesterol and, in particular, low density lipo-
protein cholesterol (LDL-C) levels (1-3), while those enriched in unsatu-
rated fatty acids (4-15) lower LDL-C when replacing saturated fat. The
mechanism(s) involved in the hypocholesterolemic action of the unsatu-
rated fatty acids are not well understood, although studies from our labora-
tory (16,17) and those of others (18,19) would suggest that unsaturated
fatty acids prevent the downregulation of the LDL receptor normally ob-
served during intakes of saturated fat and cholesterol. In general, the pre-
dictive equations of Keys et al. (8,9) and Hegsted et al. (10) have demon-
strated that the fatty acid component in the diet and, in particular, satu-
rated fatty acids are considered the primary determinants of diet-induced
hypercholesterolemia.

B. Effects of Unsaponifiable Components

A role for certain unsaponifiable components of dietary oils is reemerging.
For example, in a review article by Grundy (20), observations were cited
from the recent investigations by Grundy and Mok (21), Lees et al. (22),
and Heinemann et al. (23), which indicated that unsaponifiable sterols such
as phytosterols can have significant effects on LDL-C levels even at rela-
tively low intakes. Along these lines, Best et al. (24) fed the plant sterol {3-
sitosterol on top of an unrestricted diet and demonstrated sustained reduc-
tions of serum total cholesterol.
The beneficial effects of other unsaponifiables have also been demon-
strated. For example, recent studies by Qureshi et al. in humans (25) and
various animal models (26-28) suggest that tocotrienol (Fig. 3), an analog
of tocopherol (vitamin E) and an unsaponifiable component of palm oil,
inhibits cholesterol synthesis and lowers serum cholesterol. Tocotrienols
have also been shown to influence certain parameters of hemostasis (29)

Figure 3 Chemical structure for tocotrienol.

HEALTH BENEFITS OF RICE BRAN OIL 431

and alter the occurrence of chemically induced tumors in rodent models

(30,31).
Another group of unsaponifiables, '}'-oryzanol, which is a group of
ferulic acid esters of triterpene alcohols and plant sterols (Fig. 4), have also
been demonstrated to have certain beneficial effects. For example, the
feeding of the unsaponifiable fraction of soybean had a hypocholesterol-
emic effect in primary type II hyperlipoproteinemia (32). Similarly, when
the unsaponifiable fraction of soy sterol was further purified and fraction-
ated, both cyloartenol and 24-methylene cyloartenol, which are main con-
stituents of the triterpene alcohols in soybean oil unsaponifiable matter,
reduced plasma cholesterol and enhanced cholesterol excretion when fed
to rats (33). Similar lipid-lowering effects of oryzanol have been noted by
others (34-37) in both humans and animals as have inhibitory effects on
platelet aggregation (38).
Particularly germane is the finding that rice bran oil contains an unusu-
ally high content of the same unsaponifiable matter (up to 4.4%) described
above, which is severalfold greater than most other vegetable oils. In addi-
tion, rice bran oil, depending on source and degree of processing, can
contain up to 20% saturated fatty acids and approximately equal amounts
of polyunsaturated (40%) and monounsaturated fatty acids (40%), a fatty
acid profile quite different from other often-utilized hypocholesterolemic
vegetable oils (Table 9). Equally important is the recently reported studies
in which the hypocholesterolemic action of rice bran oil has been attributed
to its unsaponifiable fraction. For example, Sharma and Rukmini (39)
showed that rice bran oil fed to rats at a 10% level significantly lowered
total cholesterol ( -37% ), very low density lipoprotein (VLDL) + LDL-C
(-52%) and raised high density lipoprotein cholesterol (HDL-C) ( +82%)
compared to groundnut oil. These alterations in lipoprotein cholesterol
were associated with a 136% increase in neutral steroids and a 47% in-

Figure 4 Chemical structure for oryzanol (24-methylene-cycloartanol ester of

ferulic acid).
432 NICOLOSI ET AL.

Table 9 Distribution of Major Fatty Acids of Dietary Oils(%)

c<12 C12 C14 c16 cls:o c1s:1 c1s:2 cls:3
Butter 10 3 11 28 12 31 41
Coconut oil 15 47 17 9 372
Corn oil 10 2 27 60 1
Olive oil 8 3 84 4 1
Canola oil 4 2 62 22 10
Safflower oil 62 13 78 1
Rice bran oil 17 2 44 36 1
Sunflower oil 65 20 68 1
Soybean oil 11 4 24 53 8

crease in total bile acids in rats fed rice bran oil compared to groundnut oil.
Since the fatty acid compositions of rice bran oil and groundnut oil are
similar, the authors concluded that some fraction or fractions of the
unsaponifiable components are responsible for the lipoprotein changes and
rather striking alterations in fecal excretion of neutral sterols and bile
acids. In a similar study of the hypocholesterolemic activity of rats fed rice
bran oil compared to groundnut oil, Seetharamaih and Chandrasekhara
(40) also found rather striking decreases in serum total cholesterol (-38%)
and VLDL + LDL-C ( -39%) and a 27% increase in HDL-C. Serum
cholesterol levels were further reduced by 20% upon addition of oryzanol
to the diet. These changes in lipoprotein cholesterol levels were associated
with striking decreases in liver cholesterol content (30%) in rats fed rice
bran oil compared to groundnut oil. Thus, these two rat studies suggest that
changes in lipoprotein cholesterol levels in response to the feeding of some
unsaponifiable component of rice bran oil are associated with alterations in
liver cholesterol content and excretion of fecal sterols and bile acids.

C. Dietary Studies with Rice Bran Oil

One study in humans fed 35-40 g of either rice bran oil or some combina-
tion of coconut oil, palm oil, or groundnut oil, which showed 25-30%
reductions in total serum cholesterol levels in the rice bran oil group, is also
very interesting but needs to be qualified because of the highly saturated
fatty acid nature of coconut oil and the sparse information on peanut oil
(4). However, a preliminary report in humans that compared rice bran oil
to other vegetable oils as part of a blend of oils that are more normally
consumed, confirmed the cholesterol-lowering properties of rice bran oil
(42). They also suggested that factors other than fatty acids in rice oil were
responsible for its cholesterol-lowering effects.
HEALTH BENEFITS OF RICE BRAN OIL 433

More recently, our laboratory has investigated the effects of rice bran oil
in nonhuman primates fed 0-35% of calories as fat as part of a blend of
several vegetable oils (43). One of the diet treatments consisted of a combi-
nation of oils that had the same fatty acid composition as rice bran oil but
did not contain rice bran oil. From the studies, we concluded that (a) the
rice bran oil content of the diet was the primary determinant of the degree
of diet-induced hypocholesterolemia, (b) the reductions in serum total and
VLDL + LDL-C with rice bran oil feeding were not associated with signifi-
cant changes in HDL-C, and (c) a component of rice bran oil other than its
fatty acid composition was responsible for hypocholesterolemic action.
In a preliminary report (44) from our laboratory in nonhuman primates,
rice bran oil, corn oil, and canola oil fed at 20% of calories (total calories
from fat was 30%) when compared to a blend of oils that approximated the
average American diet, at 36% calories as fat, all reduced total and VLDL
+ LDL-C approximately 30%. HDL-C levels were least effected in the rice
bran oil group. Thus, despite the high saturated fatty acid and lower polyun-
saturated fatty acid profile of rice bran oil compared to the other unsatu-
rated vegetable oils, reductions in serum lipoprotein cholesterol levels were
comparable, suggesting an important contribution of the non-fatty acid
component of rice bran oil to its cholesterol-lowering properties. This is
supported by the observation that, while a predictive equation for choles-
terol lowering based on fatty acid composition derived from 60 diet studies
in monkeys correctly approximated the expected response to both the corn
and canola diet, it was not able to accurately predict the response to rice
bran oil {44).
Another preliminary report (45) from our laboratory demonstrated that
rice oil fed to hamsters not only showed cholesterol-lowering properties
but, in addition, antioxidant characteristics and a striking reduction in
aortic fatty streaks, the initiating event of atherosclerosis.

D. Conclusions

The explanation for the hypocholesterolemic action of rice bran oil is un-
known, but many of its unsaponifiable components have cholesterol-
lowering activity due to various mechanisms. For example, ferulate esters
of plant sterols and triterpene alcohols such as oryzanol can inhibit dietary
cholesterol absorption and enhance fecal sterol and bile acid secretion. The
high content of tocotrienols in rice bran oil suggests that cholesterol synthe-
sis may be inhibited in individuals fed rice bran oil. Either one of these
mechanistic responses can result in the unregulation of the LDL receptor,
thereby increasing hepatic uptake of circulating LDL and decreasing serum
LDL levels. This suggests the possibility of inducing mechanistic synergism
434 NICOLOSI ET AL.

in response to the feeding of blends of oils low in saturated and high in

unsaturated fatty acids, which, when combined with rice bran oil, could
alter membrane fluidity and cholesterol absorption and metabolism, and
therefore LDL receptor activity. Thus, one might speculate that a blend of
the appropriate vegetable oil with rice bran oil would yield a hypocholester-
olemic response greater than either oil alone.

REFERENCES
1. Ahrens, E. H., Jr., Hirsch, J., Insull, W., Jr., et al. (1957). The influence of
dietary fats on serum lipid levels in man. Lancet, 1:943.
2. Schaefer, E. J., Levy, R. 1., Ernst, N. D., et al. (1981). The effects of low
cholesterol, high polyunsaturated fat, and low fat diets on plasma lipid and
lipoprotein cholesterol levels in normal and hypercholesterolemic subjects.
Am. J. Clin. Nutr., 34:1758.
3. Shepherd, J., Packard, C. J., Grundy, S.M., et al. (1980). Effects of saturated
and polyunsaturated fat diets on the chemical composition and metabolism of
low density lipoproteins in man. J. Lipid Res., 21:91.
4. Connor, W. E., and Connor, S. L. (1986). Diet and Prevention of Coronary
Heart Disease and Cancer (B. Hallgren, ed.), Raven Press, New York, p. 113.
5. Kinsell, L. W., Partridge, J., Boling, L., et al. (1952). Dietary modification of
serum cholesterol and phospholipid levels. J. Clin. Endocrinol., 12:909.
6. Dreon, D. M., Vranizan, K. M., Krauss, R. M., et al. (1990). The effects of
polyunsaturated fat vs monounsaturated fat on plasma lipoproteins. J. Am.
Med. Assoc., 263:2462.
7. McDonald, B. E., Gerrard, J. M., and Bruce, V. M. (1989). Comparison of
the effect of canola and sunflower oil on plasma lipids and lipoproteins and on
in vivo thromboxane A2 and prostacyclin production in healthy young men.
Am. J. Clin. Nutr., 50:1382.
8. Keys, A., Anderson, J. T., and Grande, F. (1965). Serum cholesterol response
to change in the diet. II. The effect of cholesterol in the diet. Metabolism,
14:776.
9. Keys, A., Anderson, J. T., and Grande, F. (1957). Prediction of serum choles-
terol responses of man to changes in fats in the diet. Lancet, 2:959.
10. Hegsted, D. M., McGandy, R. D. B., and Myers, M. C. (1985). Quantita-
tive effects of dietary fat on serum cholesterol in man. Am. J. Clin. Nutr.,
17:281.
11. Mattson, F. H., and Grundy, S. M. (1985). Comparison of effects of dietary
saturated, mono-unsaturated, and poly-unsaturated fatty acids on plasma lip-
ids and lipoproteins in man. J. Lipid Res., 26:194.
12. Mensink, R. P., and Katan, M. B. (1989). Effect of a diet enriched with
monounsaturated or polyunsaturated fatty acids on levels of low-density and
high-density lipoprotein cholesterol in healthy women and men. N. Engl. J.
Med., 321:436.
436 NICOLOSI ET AL.

28. Qureshi, A. A., Premticen, Zafeer, Z., et al. (1984). Influence of culture
filtrate of trichoderma viride and barley on lipid metabolism of laying hens.
Lipids, 19:250.
29. Qureshi, A. A., Qureshi, N., Hasler-Rapacz, J., et al. (1991). Dietary
tocotrienols reduce concentrations of plasma cholesterol, apolipoprotein B,
thromboxane B2, and platelet factor 4 in pigs with inherited hyperlipidemias.
Am. J. Clin. Nutr., 53:1042S.
30. Tan, B., and Chu, F. L. (1991). Effects of palm carotenoids in rat hepatic
cytochrome P450-mediated benzo(a)pyrene metabolism. Am. 1 Clin. Nutr.,
53:1071S.
31. Gould, M. N., Haag, J. D., Kennan, W. S., et al. (1991). A comparison of
tocopherol and tocotrienol for the chemoprevention of chemically induced rat
mammary tumors. Am. J. Clin. Nutr., 53:1068S.
32. Nakashima, Y., Nakamura, T., Aramaki, Y., and Kuroiwa, A. (1983). Effect
of the purified unsaponifiable fraction of soybean on primary type II hyperlipo-
proteinemia. Artery, 12:199.
33. Kiribuchi, M., Miura, K., Tokuda, S., et al. (1983). Hypocholesterolemic
effect of triterpene alcohols with soysterol on plasma cholesterol in rats. J.
Nutr. Sci. Vitaminol., 29:35.
34. Sakamoto, K., Tabata, T., Shirasaki, K., et al. (1987). Effects of 'Y-oryzanol
and cycloartenol ferulic acid ester on cholesterol diet induced hyperlipidemia
in rats. Japan 1 Pharmacal., 45:559.
35. Yoshino, G., Kazumi, T., Amano, M., et al. (1989). Effects of 'Y-oryzanol and
probucol on hyperlipidemia. Cur. Ther. Res., 45:975.
36. Sasaki, J., Takada, Y., Handa, K., et al. (1990). Effects of 'Y-oryzanol on
serum lipids and apolipoproteins in dyslipidemic schizophrenics receiving ma-
jor tranquilizers. Cline. Ther., 12:263.
37. Shinomiya, M., Morisaki, N., Matsuoka, N., et al. (1983). Effects of 'Y-
oryzanol on lipid metabolism in rats fed high-cholesterol diet. Tohoku 1 Exp.
Med., 14/:191.
38. Seetharamaiah, G. S., Krishnakantha, T. P., and Chandrasekhara, N. (1990).
Influence of oryzanol on platelet aggregation in rats. J. Nutr. Sci. Vitaminol.,
36:291.
39. Sharma, R. D., and Rukmini, C. (1986). Rice bran oil and hypocholesterol-
emia in rats. Lipids, 21:715.
40. Seetharamaiah, G. S., and Chandrasekhara, N. (1989). Studies on hypocholes-
terolemic activity of rice bran oil. Atherosclerosis, 78:219.
41. Raghuram, T. C., Rao, U. B., and Rukmini, C. (1989). Studies on the
hypolipidemic effects of dietary rice bran oil in human subjects. Nutr. Rep.
Int., 39:889.
42. Lichtenstein, A. H., Ausman, L. M., Lenner, J. L., et al. (1991). Effect of the
consumption of rice bran, corn, canola, and olive oils on plasma lipids and
apolipoprotein concentration in humans. Arteriosclerosis, JJ:1603a.
43. Nicolosi, R. J., Ausman, L. M., and Hegsted, D. M. (1991). Rice bran oil
lowers serum total and low density lipoprotein cholesterol and apo B levels in
nonhuman primates. Atherosclerosis, 88:133.
HEALTH BENEFITS OF RICE BRAN OIL 437

44. Nicolosi, R. J., Ausman, L. M., and Hegsted, D. M. (1990). Comparative

cholesterol lowering effects of rice bran oil, canola oil and corn oil. Circula-
tion, 82:1109.
45. Nicolosi, R. J., and Liang, S. (1991). Comparative effects of rice bran oil,
soybean oil and coconut oil on lipoprotein levels, low density lipoprotein
oxidizability and fatty streak formation in hypercholesterolemic hamsters.
Arteriosclerosis, 11 :1603a.
 19
Changing Market Demands for Rice and Rice
Products
Richard A. Meyers
U.S.A. Rice Council, Houston, Texas

I. INTRODUCTION
Both total and per capita consumption of rice have increased substantially
over the last 10 years in the United States. Total domestic consumption rose
from 26.9 million hundredweight (cwt) in 1978/79 to almost 47.5 million
cwt in 1988/89, a 77% increase, based on USDA's biannual milled rice
distribution surveys, Department of Treasury data for brewers' use, and
Foreign Agricultural Service data on imports. Per capita use rose 60% from
about 12 pounds to 19.2 pounds.
Rice consumption is divided into three categories: direct food use, pro-
cessed foods, and brewers' use. However, not all categories of rice con-
sumption have grown equally. Analysis of USDA's biannual milled rice
distribution survey results from 1978/79 to 1988/89 yields several significant
changes in domestic rice-consumption patterns.
First, the percentage of domestic rice consumed in processed food has
grown from about 14% or 3.7 million cwt in 1978/79 to over 18% or 8.6
million cwt in 1988/89. Much of this growth is accounted for by new prod-
ucts such as rice cakes, pet foods, and candy. Second, specialty rices (par-
boiled, precooked, and brown rice) have grown as a share of direct food
use, more than doubling from 2.9 million cwt to over 5.7 million in 10
years. Among the specialty rices, parboiled and brown rice shares have
risen, while precooked share has declined. Third, while increasing over the

439
440 MEYERS

10-year period, brewers' use of rice has declined as a share of domestic rice
consumption, from almost 30% in 1978/79 to under 24% in 1988/89 accord-
ing to Department of Treasury statistics. Use of rice in beer rose from 7.9
million cwt to 11.1 million cwt, a much slower increase than found for
either direct food use or processed food. Finally, rice consumption is grow-
ing most rapidly in regions where per capita use is well above the national
average, such as the Pacific coast, Middle Atlantic states, and certain states
along the South Atlantic. These three regions accounted for over 72% of
direct food use of rice in the United States in 1988/89, up from about 64%
10 years earlier. Growth in per capita consumption in most other regions
has been much slower or even absent.
Use of rice in processed foods more than doubled between 1978/79 and
1988/89, from 3.7 million cwt to 8.6 million cwt, up over 130% in 10 years.
In contrast, direct food use rose about 65% over the decade. Direct food
use's share of domestic rice consumption has declined from about 60% in
1980/81 to 58% in 1988/89.
Growth in processed-food use of rice can be divided into two compo-
nents: expansion in consumption of old products and development of new
products. In the early part of the decade, most growth in processed rice use
was by established products, principally cereal, baby food, and soup. But
since the 1986/87 survey, new products such as pet food, rice cakes, and
new package mixes accounted for most of the growth in processed-food use
of rice.
Use of rice in cereals more than doubled between 1978/79 and 1986/87,
from 2 to 4.8 million cwt, and consistently accounted for over half of process
food use of rice. No other product reported this large an absolute increase in
consumption. However, reported shipments declined in 1988/89, and cereals
share of processed rice dropped to about 45%. Likewise, reported use of
soup and baby foods both increased early in the decade, peaking in the 1984/
85 survey. But use of rice in these product categories has declined since 1986/
87. Soup and baby food each account for about 1-2% of processed-food use
of rice, and their share declined over the 10 years examined.
Since 1986/87, most of the growth in processed-food use of rice has been
by new products. Pet food, which uses less expensive brewers' rice, soared
from 426,000 cwt in 1986/87 to almost 1.4 million cwt in 1988/89. Pet food is
the third largest use of processed rice behind cereal and package mixes.
Rice cakes, typically viewed as a low-calorie snack, climbed from 288,000
cwt in 1986/87 to over 700,000 cwt in 1988/89 and is the fourth largest use of
processed rice. Pet food and rice cakes have shown the largest percentage
growth of any processed rice products tracked by this survey since 1986/87.
Package mixes, available in many flavors, is the second largest processed-
food use of rice. Reported shipments were a mere 222,000 cwt in 1982/83 but
MARKET DEMANDS 441

steadily grew to over 1. 7 million cwt in 1988/89 and accounted for over 20%
of processed food use of rice. In addition, use of rice in candy and frozen
dinners has increased by over 30% since 1986/87, but their share of
processed-food use is less than 1% each.
Specialty rice's share of direct food use has grown from about 19% in
1978/79 to 23% in 1988/89. Reported shipments of specialty rice doubled in
10 years. Consumption of regular milled white rice, the only other compo-
nent of direct food use, grew 56% in 10 years, and its share of direct food
use dropped 4 percentage points.
The largest absolute increase in specialty rice shipments has been in
parboiled rice, which grew from 1.8 million cwt in 1978/79 to almost 4.4
million cwt in 1988/89. And parboiled rice's share of specialty rice ship-
ments rose from 60 to 75%. Brown rice shipments exhibited the largest
percentage increase, expanding from just 237,000 cwt to almost 700,000
cwt. Much of the increase in brown rice use occurred in the second half of
the 10-year reporting period.
In contrast to parboiled and brown rice, precooked shipments have
actually declined since 1978/79 and are only slightly above reported brown
rice shipments. Precooked shipments totaled almost a million cwt in 1978/
79 but were only 670,000 cwt in 1988/89 and declined steadily over the 10-
year period.

II. TRENDS BEHIND THE NUMBERS

The steady growth of rice consumption over the last decade is due to a
variety of factors, including:
Increased awareness of rice as a healthful food, i.e., cholesterol-free,
sodium-free, and very low in fat and calories.
Influx of recent immigrants from traditional rice-consuming countries.
Increasing visibility given rice by food service operators during a time
when more and more Americans are eating/purchasing food away from
home.
Rising popularity of ethnic foods, especially Cajun, Mexican, Indian, and
Thai.
New product introductions, e.g., microwaveable rice products, boil-in-bag,
single servings.
A growing consensus among regulators, health professionals, and consum-
ers that rice should be a staple of the dinner plate.
While rice dishes are not new to the U.S. market, recent events and trends
suggest that consumption of rice, as both a side and a main dish, will
increase significantly during the 1990s.
442 MEYERS

A. Population and Lifestyle Changes

Growing Hispanic-American and Asian-American populations, the issu-
ance of recent government nutritional guidelines, a public becoming com-
mitted to healthy lifestyles, and an increasing incidence of menuing and
serving rice as both an entree and a side dish among food service operations
all point to the growth of the rice industry in the future.
However, these demographic and attitudinal changes are just part of
what is going on in the American marketplace. Underlying product prefer-
ences is a movement toward a simpler lifestyle. While it is unlikely that
American women will leave the workforce in great numbers or that Ameri-
cans in general will find more leisure time, it is true that there are-and will
continue to be-changes in the types of products and services purchased
and the way leisure time is spent.
Americans are placing less emphasis on status and more on quality. Ac-
cording to Time magazine, "the pursuit of a simpler life with deeper meaning
is a major shift in America's private agenda" (1). The writers of the April 8,
1991 cover story support this finding with a variety of testimony from experts
and survey research that concludes that Americans want to slow down and
live a more relaxed life (69% ), spend more time with family (69% ), and find
more time for hobbies and personal interests (56%). Just 13% ofthe survey
sample thought it important to keep up with fashion and trends (1).
Those demographic changes that are altering the geographic distribution
of the U.S. population and changing the makeup of the American family
household are detailed in this chapter. Also, the psychographic and life-
style trends that are affecting the way Americans purchase, prepare and
consume food are described. Specific emphasis is placed on the direct or
indirect effects these trends will have on the growth of the rice industry in
the U.S. marketplace.

B. Demographic Trends
The most important demographic changes that will affect, either directly or
indirectly, the way rice and other foods are sold, purchased, prepared, and
eaten during the 1990s are:
1. Increasingly nontraditional families, e.g., two-income couples; later
marriages, later childbirth; increased single-parent households; greater
numbers of men as single parents; increased numbers of people living
alone.
2. The continued importance of the baby boom generation, who will be
moving into middle age and into their peak earning years during the
1990s.
MARKET DEMANDS 443

3. An aging population that brings with it a fairly predictable set of psy-

chological and physiological stages.
4. Working women.
5. Rapid growth of immigration from countries with rice-based diets;
changes in the geographic distribution of older people and Asian-
Americans and Hispanic-Americans.

ill. NONTRADITIONAL FAMILIES

While the number of adults is forecast to increase by just 10.4% between

1988 and 2000, the number of households is expected to grow by 15.9%
(2). This disparity is influenced in part by the continuing decline in house-
hold size that has decreased uninterrupted for the last 25 years. The num-
ber of nonfamily and single-person households is increasing. So, too, has
the number of male homemakers. During the next decade, male homemak-
ers will increase at triple the rate (37.2%) of female homemakers (12.1% ).
Industry analysts forecast increased sales of food products based on
research findings that smaller families tend to spend more on groceries than
large families. For example, a single-person household spends an average
of $40 a week on groceries compared to just $19 per person for a family of
five (3).

Iv. BABY BOOMERS

Baby boomers are moving into middle age. By 1995, the first group of the
76 million individuals defined as baby boomers (i.e., born between 1946
and 1964) will reach the age of 50 (4). As shown in Figure 1, the number of
middle-aged people will grow rapidly in the 1990s, with 40- to 49-year-olds
and 50- to 59-year-olds representing the growth areas of the population.

A. Health
People who are now age 45 can expect to live another 33 years (5). Conse-
quently, their health and nutrition will be of critical concern. For example,
according to the National Center for Health Statistics, chronic conditions,
including heart disease and high blood pressure, become more prevalent
after the age of 45.
Older baby boomers are expected to enjoy better health than their
predecessors because this generation has information not available before.
Also, research has shown that baby boomers have shifted attitudes about
responsibility for good health from the physician to themselves.
444 MEYERS

MIDDLE-AGED POPULATION
• INCREASING •
(Number of people aged 20 to 69. by ten-year age group. in millions. 1990-2000)

50

40

20

1990 2000 0

Figure 1 Estimated U.S. population trends for the 1990s.

B. Work/Lifestyle

This segment of the population works longer hours, juggles family obliga-
tions (most will have children living with them through the next decade),
and complains about a lack offree time (4). Increasingly, baby boomers are
seeking goods and services that save time and offer convenience, such as
businesses that offer extended shopping hours and home delivery.
A University of Maryland study (6) shows that while all age groups
perceive themselves as more rushed today than they were in 1965 and 1975,
men and women between the ages of 35 and 54 tend to feel the most
rushed. American Demographics reports (7) that half of dual-earner par-
ents and one third of single-earner parents say that shopping for products
and services is stressful.

C. Discretionary Income

The baby boomer's biggest financial advantage is education. This is espe-

cially true for women who are far more likely than their predecessors to
have gone to college. Also, baby boomers are more likely than their par-
MARKET DEMANDS 445

ents to live in dual-income homes. These factors translate into increased

discretionary income for this population.
According to American Demographics (5):

As baby boomers age, they will accumulate wealth through property

ownership and investment. In 1986, the net worth of householders under
the age of 35 averaged only $40,000 ... while the average net worth of
householders aged 35-44 was over $100,000. Those aged 45 to 54 aver-
aged nearly $175,000. And households ages 55 to 64 averaged $200,000
in net worth.

Labor shortages are expected to keep baby boomers in the workforce

longer, thus adding to their accumulated wealth. Currently, 25.6 million
people aged 45-54 are in their peak income years (5). When all the baby
boomers pass their 45th birthday in 2010, that number will increase by 47%
to 37.7 million prompting analysts to forecast "staggering" affluence
through the 1990s.

D. Food Expenditures
The population aged 35-64 will account for two thirds of U.S. food dollars
by the year 2000 (8). Overall, food spending by this group is forecast to
increase 36% during the decade, acccounting for a 24% projected rise in
the nation's overall food spending.

V. THE OLDER POPULATION

Social scientists predict that the aging of the U.S. population will have far-
reaching effects across all sectors of the U.S. economy throughout this and
the next decade. The older population has financial influence. Today, those
over 50 years old have 60% of commercial banking deposits and 80% of
thrift deposits. They own nearly half of all corporate stocks and two thirds
of all portfolios over $25,000.
Key to changing food habits is the increasing migration of Americans
aged 65 and over suburbs and cities. Through the 1970s and early 1980s,
the proportion of the U.S. population living in metropolitan areas re-
mained steady at around 64%. In 1986, however, following the Census
Bureau's redrawing of metro cases, that number increased to 73%.
Where older people relocate should have an impact on increased con-
sumption of rice in mature households. Increasing numbers of older Ameri-
cans are moving to warmer regions, areas of the United States where rice
consumption is traditionally the most prevalent (9).
446 MEYERS

As the population ages, the dwindling under-24 age group (representing

the majority of potential food service employees) is expected to diminish
by the year 2000 to less than 34% of the total population, a drop of 5 or 6
million individuals (10). Food service operators can no longer depend on a
vast labor pool. Therefore, food and food service equipment manufactur-
ers are being called upon to provide products that will minimize this labor
shortage.
According to food service operators, rice is one of the simplest foods to
prepare, filling this growing need to offer convenient and easy-to-prepare
food offerings that are not labor intensive.

VI. WORKING WOMEN

Working women will continue to represent the most significant change in
the American workforce and culture during the 1990s. Currently, 65% of
women with children under 18 have jobs-almost three times as many as in
1960. Some predict that by the year 2000, 75% of U.S. children will have
working mothers.
Women are still the primary grocery shoppers for the family, thereby
providing the greatest influence on family eating habits. However, other
family members are beginning to play an increasingly important role in
shopping and meal preparation. According to the New Traditional Home-
maker survey (11) sponsored by the American Frozen Food Institute based
in New York City, nearly two thirds of American children under age 13
prepare at least one meal on their own each week.
In an era of dual working parents, children exercise a far greater influ-
ence over meal selection. For example, according to a study (12) conducted
by Selina Guber, president of Children's Market Research:

Forty-nine percent of children either buy their own food or help select
specific brands and products.
The family shopper, i.e., mother, will often purchase brands requested by
children. However, the busy mother's choice can be driven by conve-
nience of foods, while children's choices are most often driven by taste.

Teenagers exercise an ever-greater influence over a number of household

decisions, e.g., food shopping as well as electronics and auto purchases.
This is due in large part to increasing discretionary income and greater
household responsibilities for teens.
Preparing dinner is still largely the function of women. According to a
1990 readership study (13), on weekdays and weekends, 86% and 81% of
women, respectively, prepare the evening meal. On weekends, women,
MARKET DEMANDS 447

too, are most likely to prepare breakfast (58%) and lunch (44%). Not
surprisingly, breakfast and lunch are meals individuals are most likely to
make for themselves. This nation of working mothers is expected to speed
the increase of new food products and preparation technology that offers
both convenience and health benefits for their families.

VII. IMMIGRATION/GEOGRAPHIC TRENDS

During the 1980s, immigration accounted for more than one third of U.S.
growth and should remain key to further population expansion throughout
the next decade. Probably the greatest boost to rice consumption in the
United States will come from continued increases in minority groups for
whom rice is a diet staple.

A. Hispanic-Americans
To illustrate just how significant that growth should be, consider these
recent population projections:
Hispanics now make up about 8% of the U.S. population. That share is
forecast to increase steadily to 11% by the year 2000 (14).
More Hispanics now reside in "California, Texas, New York, and Florida
than lived in the entire country in 1980" (9).
Los Angeles, New York, and Miami each are already home to over one
million Hispanics, soon to be joined by San Antonio, San Francisco, and
Houston.

B. Asian-Americans
Asian-Americans are the fastest growing ethnic group in the United States
and will double their population in the next 20 years. Since 1980, Asian-
Americans have grown from 1.7 to 2.8% of the U.S. population. Region-
ally, the numbers are even more dramatic. Fifty-six percent of Asians live
in the West, compared with 21% of all Americans. Twelve states are home
to over 100,000 Asian-Americans: California, Hawaii, New York, Illinois,
New Jersey, Texas, Massachusetts, Pennsylvania, Virginia, Florida, Michi-
gan, and Washington (15).
The impact that Asian immigrants are having on U.S. rice consumption
is dramatically apparent in reviewing the trend of rice imports, which virtu-
ally parallels that of Asian immigration. As can be seen in Figure 2, rice
imports have increased more than 10-fold during the 1980s. Approximately
90% of the rice imported into the United States comes from Thailand. This
448 MEYERS

MILLION CWTS.
4

1981182 1983/84 1985/86 1987/88 1989/90

1982/83 1984/85 1986/87 1988/89 1990/91*

~ Milled Rice *Projected

Imports (Est.J
Jasmine 90% Basmati 8-9% Glutinous 1-2%

Figure 2 Changes in rice imports for the 10-year period 1981-1990.

Jasmine aromatic variety is very much favored by recent Southeast Asian

immigrants. While the overall imported quantity of rice is still relatively
small compared to the total consumed in the domestic market, it is appar-
ent that Asian immigration will play an important role in future rice con-
sumption in this country.
In those regions where rice consumption is likely to be the heaviest,
some conversion among nonusers can be expected. For example, in regions
where large number of either Hispanics or Asians reside, you can expect to
see more restaurants, grocery stores, and new product introductions that
incorporate native tastes utilizing rice.
As shown in Figure 3, we are eating more ethnic foods. More and more
Americans are experiencing Mexican, Italian, Chinese, Cajun, and Japa-
nese cuisines, all of which incorporate a liberal amount of rice in their
dishes.
MARKET DEMANDS 449

33%
31%
"'"" ...... "' .. c.... "'" 29.5%
., .... ., . -~~ ....
,. ":«-: c" ..";:c.,
".I

..". 17%
1 .. " ..
,. "' .. c- 1 .) 1
.... ,."'",. .. c
~~ ..............
........
.... ,..,
C' .. ., .... .., ¥.& " " ' .. ,.

,..,,. .. ';.~. c<,.•:~: ., ,. "1 c 1

. . . 7": .., . ,."
" .. c "'" .. "'"'

.... ...... ~: . ~~:::

&.)>
",. c .. .," .... c,. v " .. c
" ,.
:~ .::: ........ ,.
.,,. .. At' .. ,. A<(
" " .. ., .. c ..

,.,. .... .1,.""

.. ., .,c" .. ",
)>" t',) ol '7lloy

. . .,.,... ;,.~:,. ,.
.,. . . . . . .

v"~~~o: ~\·:>: . . ,. ... ,."'."' ,.. .,.:,." ,c""""

&. A 7 C
"
",., &o A 1 4 7

F..•.···~~
..... ,fit'...... "'v ..
.'I , .v,.vlo
,. ~lop. ,. .
"'1 ....
,.• "....
.. "',... ....
c"' .Jo".a
)o
5.5%
)o A

. "'. . ,. . &.: :,. A1 >C. t" ........ c , ..

..

:. . . . ........, . . "'" ,... . ,. ,.."¥......

..,_.. ., ........ C'..
.. ,. c. ...... ..
""'""
,." •"',.
C
...
loA .. ,. .. ...,"',.: .... !~ ~: ..
",.. ., .... < ........

Mexican Italian Chinese Cajun Japanese

Figure 3 Changes in ethnic food consumption patterns.

C. Lifestyle Changes/Social Trends

Four basic trends will continue to affect the way Americans purchase and
consume food throughout the 1990s (14):
1. Faster food: Busy lifestyles will continue to foster the U.S. industry for
faster, more convenient foods.
2. Healthy foods: While Americans have not yet fully become acclimated
to healthier diets, there will continue to be strong demand for food
products with pure ingredients and less cholesterol and fat.
3. Pleasure eating: Preparing quality meals or eating away from home is
expected to become one of the luxuries of the 1990s.
4. Environmental concerns: U.S. consumers will express their "green"
behavior by continuing in larger numbers to demand food packaged in
environmentally-friendly containers.
Marketers of foods and services are starting to realize that each of these
trends can exist alone or cross lines for successful marketing strategies.

VIII. FASTER FOOD

A. Purchase Behavior
According to Roper's Public Pulse (16), three major factors are contribut-
ing to changes in the purchase behavior of U.S. consumers:
A lessening of "conspicuous consumption"
Time demands and a decrease in discretionary income
An increased "practical-mindedness" in a middle-aged nation
450 MEYERS

Specifically, conspicuous consumption of the past is being replaced by a

more calculated pattern of spending. The purchasing patterns of new
"multiprofile" consumers will reflect their values and lifestyles. For these
consumers, "thrift and luxury neatly coexist." For example, a consumer
who spends extravagantly on gourmet foods may scrimp on paper products
for the home (16).

B. Convenience Trends
Today, 70% of American mothers work full or part time versus 30% in
1980. This statistic alone accounts for much of the public demand for
products and services that offer convenience as a feature.
Probably the greatest symbol of Americans' demand for products that fit
their busy lifestyles has been the assimilation of the microwave into the
U.S. household. Currently, 8 in 10 U.S. households have a microwave, and
about half of the rest plan to purchase one within the next year (17).
Supporting this partnership of busy lifestyles and convenience appliances, a
recent survey (18) found the microwave and the convection oven to be the
appliances used most often to make dinner. According to another recent
survey (20), while the number one use of a microwave is to reheat frozen
leftovers, the second most common use of a microwave is to prepare side
dishes and snacks.
Additionally, the growing number of children of working parents has
prompted manufacturers to latch onto the convenience trend by introduc-
ing microwaveable and nutritionally correct single-serving meals. Accord-
ing to a study (12) conducted by Selina Guber, president of Children's
Market Research, 81% of children prepare food for themselves at least two
or three times a week, usually using a microwave. Thus, with the increasing
penetration of microwave ovens into U.S. households and the ready accep-
tance of consumers to use them to prepare side dishes, a category where
rice plays an important role, a promising future seems inevitable for rice
mixes and rice-based dry-frozen packaged dinners.

C. Cooking from Scratch

With the trend toward faster food, there is an ongoing decline in preparing
meals from scratch and a corresponding decline in basic cooking skills. For
example, baking a cake today for most American cooks starts with opening
a cake mix package.
A nationwide survey of newspaper food editors (20) in the major metro-
politan markets conducted by the USA Rice Council found a high level of
interest among editors in receiving quick-to-fix recipes as well as do-ahead
MARKET DEMANDS 451

recipes and salad/no-cook recipes, with only moderate to low interest in

receiving prepare-from-scratch recipes. At the same time, more and more
prepared foods are being incorporated into family meal preparation, a fact
that is driving increased consumption of rice mixes.
In the Suppertime USA Survey (18), of the 73% of women who said
they cooked the previous evening's meal, 72% used some type of prepared
food. Products used included:

Canned foods-36%
Frozen foods-36%
Packaged mixes-16%
Prepackaged foods-9%
Supermarket salad bar-3%

D. Carryout Foods and Home Delivery

The "need for speed" fosters the continuing growth of buying food away
from home. Nearly one half of the meals people buy at restaurants are
eaten somewhere else. This is up from one third in 1983 (14). During the
1990s this proportion is forecast to increase as the take-out industry contin-
ues to expand from simple to gourmet (21). New services, e.g., home
delivery, are expected to increase to meet a growing demand for ease and
convenience in food preparation.
In the early 1980s, supermarkets suffered a decline in retail sales and
blamed it on food-away-from-home establishments. Recognizing an oppor-
tunity to recapture food sales share and meeting the needs of the changing
American consumer, who demands fast, easy, and convenient ways to fit
eating into a busy lifestyle, supermarkets began investing in prepared food
sections for both hot and cold items, in-store bakeries, and expanding deli
sections.
Today, according to Progressive Grocer (22), 91% of supermarkets
have service deli departments, 19% have sit-down eating areas, 38% have
salad bars, and 32% have soup bars. Refrigerated entrees are growing in
popularity, along with take-out, home delivery, and drive-thru. Sixty-
three percent of the 65+ consumer and single people surveyed indicated
they brought partially prepared entrees in 1986. By 1990, it had jumped
up to 83%.
Rice is participating in this carryout and home delivery trend. As a food
that travels well, maintaining textual integrity, rice is well suited for use in
all applications requiring packaged meals to go. Rice is even used for travel
to outer space. NASA incorporates rice into several of the prepared meals
the astronauts take with them on the space shuttle missions.
452 MEYERS

~. HEALTHFULFOOD
In the latter part of the 1980s and continuing into the 1990s, nutrition has
become a core part of the marketing strategy for most U.S. food compa-
nies. As increasing numbers of the population move into middle age and
become more concerned and aware of health issues, nutrition has taken on
increasing importance. For example, 63% of women and 49% of men have
changed their diet in the past year to eat more healthful foods (23). Forty-
four percent of Americans say they have changed their eating habits in the
last 2 years in order to lower their cholesterol level (24). And the number
two selection as "Yesterday's Main Entree" in the Suppertime USA Survey
(18) was a vegetable dish. A study (25) of shoppers' concerns about the
nutritional content of foods, conducted by the Food Marketing Institute,
comparing attitudinal changes since 1988 shows fat and cholesterol to be
the biggest areas of concern.
A 1989 study (26) by the National Restaurant Association found that the
restaurant-going public can be categorized into three distinct groups: uncon-
cerned, committed, and vacillators. Unconcerned eaters describe them-
selves as meat-and-potato eaters, having whatever they want regardless of
its health effects. This group declined from 38% of the adult population in
1986 to 32% in 1989. Committed eaters believes that a good diet can
prevent illness. The members of the group eat healthful foods most of the
time. This is the largest group, increasing from 35% of all adults in 1986 to
39% in 1989. The last group, vacillators, is concerned about nutrition and
health but still goes for taste when eating out. This group grew slightly,
from 27% of all adults in 1986 to 29% in 1989.
Consumer interest in nutrition has not been lost on American food
processors. Table 1 shows the number of new products introduced since
1988 claiming various health benefits. With the exception of calcium,
health-food marketing trends continue to look upward. Nor is there an end
in sight to this trend. In a survey (27) of 352 U.S. food processors to
determine their intentions to develop products with health claims, 76% of
the companies plan to increase development of reduced-calorie, reduced-
fat, or diet food products, and 57% plan to increase development of
reduced-cholesterol products. Among the respondents, 46% said they
planned to increase development of "all natural" products, and 43% re-
sponded that they intended to beef up efforts to reformulate ingredients
with "additive" connotations out of existing products-a 36% increase
over those surveyed in 1989.
There can be little doubt that rice is benefiting from this growing trend
in nutritional marketing. As one of the premier complex carbohydrates in
the food chain, rice is cholesterol-free and contains only a trace of fat. Low
MARKET DEMANDS 453

Table 1 New Products Bearing Health Benefit Claims

%change
1988 1989 1990 1989-1990
Reduced/Low-calorie/Lite 475 962 1,165 +21%
Reduced/Low fat 275 626 1,024 +64%
All natural 215 274 754 +175%
Reduced/Low salt 202 378 517 +37%
No additives/No preservatives 153 186 371 +99%
Low/No cholesterol 126 390 694 +78%
Added/High fiber 56 73 84 +15%
Reduced/Low sugar 52 188 331 +76%
Added/High calcium 4 27 20 -26%
Organic 98 140 324 +131%
Category totals are not additive. New products may carry more than one claim.

in sodium, gluten-free and nonallergenic, easy to digest, with only 82 calo-

ries per half-cup serving of white rice (89 calories per half-cup serving of
brown), 90% of the calories in rice come from carbohydrates.
In the last couple of years, sales of brown rice in the United States have
increased dramatically. In addition to the excellent taste of brown rice,
much of the reason for this growth is due to the additional fiber found in
the bran. With ongoing medical research being conducted on rice bran and
rice bran oil and the publicity of the positive results of these studies, con-
sumers are becoming increasingly aware of the nutritional benefits of eat-
ing rice and rice bran.
While the introduction of rice bran for human consumption suffered a
temporary setback due to recent research denouncing oat bran, American
consumers continue to seek legitimate food fiber sources. Thus, while no
one source will likely offer a "magic bullet" cure-all and oat bran or any
other bran may never reach the sales levels of late 1989, sales of many
cereal brans, including rice, are beginning to build steadily. The ongoing
research with rice bran and rice bran oil, which is intended to continue the
steady documentation of the efficacy of rice as a healthy food rather than to
provide banner headline newspaper copy announcements, will continue to
create interest in rice as a healthy staple of the American diet, as it has long
been in much of the world.

A. Association with White Meats

The 1980s has seen the decline in per capita consumption of red meats and
the rise in popularity of the perceived more healthful white meats. The
454 MEYERS

acceptance of poultry during the past decade by the American consumer

has been remarkable. While not as dramatic, per capita consumption of
fish has also increased. At the same time, the popularity of beef, the
traditional mainstay of the American diet, has plateaued and even declined
somewhat. Losing even more favor with consumers, as evidenced by its
slide in per capita consumption, has been pork.
These changes in the center-of-the-plate protein that American consum-
ers are demanding have had important implications for rice. USA Rice
Council research with both consumers and food service operators during
this period indicates a high correlation between the association of rice with
poultry and fish and, conversely, a high correlation between meat (beef)
and potatoes.
In a survey (20) of newspaper food editors in the major metropolitan
markets, USA Rice Council Research indicated that 75% of the editors
would recommend to their readers the serving of rice and chicken together,
whereas fewer (29%) of the food editors said they would recommend serving
rice with beef. Thus, rice has benefited from its perceived association with
poultry and fish during this decade when consumers switched dramatically to
these alternative proteins. Table 2, which shows the per capita consumption
changes for various meats as well as for rice, potatoes, and pasta, demon-
strates clearly the beneficial association of rice, poultry, and seafood.

Table 2 Changes in Per Capita Consumption of

Various Foods (1987 vs. 1980)
Food %change
Meat
Beef -4
Pork -14
Poultry +28
Seafood +20
Complex Carbohydrates
Rice +46
Pasta +27
Potatoes +5

B. Consumers Are Torn Between Nutrition and Convenience

While Americans remain unsure about the makeup of a proper diet, they
increasingly embrace the concept of wellness and will continue to take
more responsibility for their own health during the 1990s. In larger num-
bers, Americans are expected to:
MARKET DEMANDS 455

Diagnose their own ailments. This trend toward self-care is due to a num-
ber of factors, including increased costs of traditional medical care and
access to drugs formerly only available through prescription but now
sold over the counter.
Exercise, particularly in a moderate but frequent fashion.
Adhere to healthier diets.
Focus on relieving stress as an element of good health.
However, as is often the case, a person's behavior is often inconsistent with
attitudes on a subject. Two recent surveys (28) conducted by the NPD
Group found that "while food attitudes have changed dramatically, the
impact on behavior has been relatively minor." It is clear that the busy
lifestyles of Americans have much to do with their inability to adhere to
healthy menus.
According to a discussion of the two NPD studies (28), "dietary con-
cerns are clearly on the rise, but competing needs-such as the need for
convenience-may lessen the impact." While Americans say they are eat-
ing right, according to industry surveys, snack foods outperformed the
average volume growth for all grocery products in 1989 (29)! For example,
sales of potato chips increased by 8.9% to approximately $3.9 billion.
There is some indication, however, that Americans, while continuing to
snack at record levels, are looking for healthier substitutes. For example,
sales of popcorn, popularly perceived as a low-calorie snack food, in-
creased 22% during 1989, totaling sales of $368 million. Growth among
healthier snack categories-coupled with Americans' desire to return to
home-based socializing-increases the likelihood of consumer interest in
snack foods such as flavored mini-rice cakes-a grocery category that has
realized strong growth in the past year.

C. Professionals Are Turning to Rice

Among nutritionists and those involved in menu planning in food prepara-
tion away from home, rice is becoming more and more the grain of choice.
In nationwide surveys (30) among 204 school food service directors, 152
college and university food service directors, and 300 dietitians, the USA
Rice Council found that rice is increasing in popularity among these audi-
ences and that rice was perceived to be healthier than either potatoes or
pasta by 38% of the school foodservice directors, 53% of the college and
university foodservice directors, and 25% of the dietitians.
When food service operators were asked to name their top concerns as
they enter the 1990s, operators ranked "offering healthful menu items" at
the top of their list (31). Seventy-nine percent indicated that rice is very
applicable to healthful menus. Similarly, in another study (32), 57% of
456 MEYERS

foodservice operators surveyed said that rice has an image of a light,

healthy food.

X. PLEASURE EATING
A. Family Dinners
The family meal, "once the centerpiece of American family life," became
slightly less popular in the mid-1980s(33). But that appears to be changing.
For example, in 1989 people prepared an average of 225 dinners per year in
the home, still down from 246 in 1984, but up slightly from 223 in 1988.
Gourmet cooking will become an affluent person's status symbol during
the 1990s. Cooking "gourmet style" for guests says that the "host has the
time and taste to spend the time preparing meals themselves" (33).
Dining must meet the needs of a population that demands time and
labor-saving convenience in a more relaxed and comfortable lifestyle.
Longer work weeks, working mothers, microwave ovens, and less leisure
time have made cooking large and time-consuming meals at home a thing
of the past. As people look to alternative ways to dine, food service has
responded beyond the traditional establishments by expanding its focus on
take-out, home delivery, catering, convenience stores, and supermarket
food service.

B. Dining Out
According to the Tastes of America survey (36), the percentage of house-
holds eating out (79%) is at the highest level it has been in recent years.
The survey also reported that 28% of households surveyed said they are
eating out more often in 1991 than in 1989. The types of restaurants that
will be most popular with Americans during the 1990s include fast food,
homestyle, and specialty or ethnic.
Working mothers wanting to spend more time with their children are
having a positive impact on family food establishments. Fifteen years ago,
54% of mothers said they often went out to eat with their children, com-
pared to 72% today (35). Less and less leisure time has prompted the
continuous growth of fast-food restaurants. Over half of Americans (56%)
now say they eat at a fast-food restaurant at least once a week.
Aging baby boomers will have a major impact on the current age distri-
bution of the population. As this segment of society grows older, the aver-
age age of the entire population will increase, affecting not only the labor
force of the food service industry, but the age of its patrons as well. As they
continue to increase their net worth, producing more discretionary income
for the economy, much of the baby boomers' food dollars will go into food
MARKET DEMANDS 457

service. By the year 2000, this 35-54 age group will account for the largest
percentage (46%) of the total population (10). Long-term estimates indi-
cate that this group will grow by 11.3 million individuals between now and
the year 2000 (an increase of 32%). This group will spend more on food
away from home than any other demographic group.
The next trend to watch is the rapid increase in the number of older,
wealthier households. The food service industry can expect to be the benefi-
ciary of discretionary dollars of 55- to 64-year-olds as the industry continues
to expand to meet their needs for convenience and good value (10).

C. Food Service Industry Overview

In the 1960s, food service was viewed as a luxury and reserved for only
special occasions or as a necessity of travel. At the time, the industry
totaled about $19 billion in sales, with 27% of all food expenditures going
to away-from-home consumption, while 73% still went to food consump-
tion at home.
By the 1990s, food service had matured into an integral part of the
American way of life. Previously, the definition of food service was "food
eaten away from home." Today, the more appropriate, accepted definition
has become "any food prepared away from home." This definition now
includes take-out foods consumed at home, which has become a way of life
in the United States.
More food dollars move through food service than ever before. In 1991,
supermarkets will compete for these dollars as food service gets about 48%
(36). Changing consumer lifestyles in America, with redefined eating hab-
its, have caused the industry to triple in total number of establishments
with sales exceeding $256 billion to fill the demand (10).
These establishments comprise no less than 18 different service seg-
ments that have emerged with their own market audience, menu needs,
and personality. The industry defines two major segment categories: com-
mercial establishments, which include restaurants, hotels, cafeterias, and
all other retail outlets. This group represents about 70% of all sales volume
generated in the industry. The second grouping is the noncommercial,
institutional feeding segments, comprised of hospitals, schools, colleges,
universities, military, and employee feeders, to name a few. This group
represents the other 30% of the industry volume (10).
The largest segment is defined as the commercial full service and fast
food restaurant segment, representing over 480,000 total unit establish-
ments. These become even more fragmented as the industry defines the
difference between the upscale, fine dining or white tablecloth segment,
the midscale, casual or theme operation, and the fast food feeders.
458 MEYERS

Hotels, colleges, schools, hospitals, and nursing homes, business em-

ployee cafeterias, theme parks, transportation food outlets, specialty retail
outlets, such as cookie or muffin stores, and even the military feeding installa-
tions are also specific segments serving the food service market. Food service
now even encompasses supermarket delis, convenience stores, and drive-
throughs to contribute to food service sales volume.

D. Segment Trends Influencing Rice Usage

Rice is among the grains capturing more billing on restaurant menus, indi-
cating that rice is growing in popularity among food service operators,
especially in the commercial segments. In the Tastes of America survey
(34), 48% of households said they ordered rice away from home in the past
year, compared to 44% in 1988.
The USA Rice Council initiated a rice benchmark tracking study (32) to
measure changes in food service usage from year to year. Data collected in
these annual studies indicate that the proportion of food service operators
offering rice on a daily basis has increased from 68% in 1988 to 74% in
1990. Additionally, the survey found that the proportion of food service
operators using rice more frequently has increased, with 54% responding
that their usage has increased in 1990 versus 43% in 1988.
Restaurant Business, in an independent study (37), found that the great-
est increase in consumption of rice has occurred in high-dollar volume, full-
service independent operations in the Northeast and South. USA Rice
Council reports (32) that 61% of rice users in the commercial restaurant,
hotel, and cafeteria segments served more rice in 1989 than the previous
year, indicating growing patron popularity of rice and rice dishes.
A 1990 study (30) among college and university food service directors
reported ranking rice over pasta or potatoes as the healthiest and most
cost-efficient complex carbohydrate choice available. Seventy-eight per-
cent of those queried confirmed that rice usage in their operations has
grown in the past year. More ethnic dishes, light and healthy entrees, and
vegetarian selections are being offered. Rice was chosen as an important
ingredient consideration by 83% of those who are preparing more ethnic
dishes, 88% of those adding light and healthy entrees, and 80% of those
serving vegetarian meals. In addition, 79% of the survey respondents re-
ported that rice plays an important role in their salad selections.
Among elementary and secondary school food service directors sur-
veyed by the USA Rice Council (30), 99% indicate they serve rice since it
is a commodity product offered by the government. However, many agree
that without positive student acceptance of rice, it would not be on the
school lunch menu. Other reasons cited for the high menu incidence, ac-
MARKET DEMANDS 459

cording to school directors, include ease of preparation, menu versatility,

nutritional benefits, and cost efficiency. Results show that 60% of school
food service directors serve rice as part of an entree rather than only a side
dish, with nearly 50% substituting rice for potatoes and 30% for pasta.
Another important survey in 1990 (30) on attitudes of dietitians towards
rice found that 44% stated they were receiving more patron and patient
requests for menu options and recipes using rice over the previous year,
compared to 39% for pasta and 19% for potatoes.

E. Menu Trends Influencing Rice Usage

According to a recent survey (38), Pacific rim cuisines, i.e., Japanese,
Chinese, and Thai, will continue to rise in popularity "due to their healthful
nature and exotic appeal." The survey also predicts that U.S. diners will
discover Philippine and Korean cuisines in the next decade.
Much of the 19.2 pounds of rice that Americans now consume per
person per year comes from food service, where ethnic and regional trends
have increased demand for rice-based main dishes, salads, and soups. The
various food service segments are forced to watch and react to menu trends
or suffer extinction in a highly saturated marketplace of food service estab-
lishments. Each segment continues to plan menu changes that offer their
patrons variety, good value, and a reason to repeat the dining experience.
The "blurring" of the segments is evident as schools compete to keep
children on school premises during lunch period by adding items like pizza,
tacos, burritos, and chicken nuggets; hospitals alter their image of offering
"institutional" food by upscaling their menus to satisfy more sophisticated
patients and visitors on-site; casual restaurants and fast-food feeders ex-
pand their breakfast offerings to include a variety of "fresh" baked goods
served all day; and full-service restaurants offer appetizers as minimeals so
that patrons can sample and taste a variety of items.
Eight out of 10 people in the world depend on rice for 40% of their
energy needs. With the influx of immigrants from Asia, Latin America,
and the Caribbean into the United States, items such as Middle Eastern
rice pilaf, Spanish paella, arroz con polio, Indian curries, and Oriental fried
rice have flourished on American food service menus.
In a survey (39) of food service operators regarding the performance of
rice in various areas, it was found that 67.6% of the operators surveyed felt
that rice was "very applicable" in its versatility for its use in ethnic
menuing, with an additional 26.7% rating rice as "somewhat applicable."
Rice dishes are integral to regional American cuisines, too. With its
many varieties and its ability to complement other ingredients and flavors,
rice makes a perfect menu choice as an accompaniment, a component of
460 MEYERS

the main dish entree, as a soup ingredient, in salads, and as a dessert. Rice
can also form the base for many vegetarian dishes. A hearty brown rice
blended with beans, seeds, and a variety of vegetables is a healthful meat-
less entree. It is especially popular on college campuses and with health
care workers.
USA Rice Council-sponsored studies (32) show that 70% of food ser-
vice operators rate rice as excellent when served with chicken and 43%
when served with seafood. Thirty-five percent report that rice is excellent
when served with pork and 26% with beef. According to these studies, over
54% of all operators surveyed used rice more frequently in 1990 than in
1989. The average number of pounds purchased weekly is about 41, com-
pared to 23 pounds of pasta. The studies also indicate that one third of all
meals served continue to use rice, with about 36% serving rice at lunch and
43% at dinner. Figures from this same research (32) also indicated that
74% of operators serve rice daily, with the highest percentage serving rice
as a side dish at 64%. Fifty-one percent serve rice as an accompaniment to
the main dish item, and 35% use rice as an ingredient in an entree or soup.
In addition to increases of on-menu incidence of parboiled white rice in
long and medium grain, Restaurants & Institutions 1989 menu census re-
search (40) reports a variety of other types of rices being served by both
commercial and institutional operations. Included in these varieties are
brown rice (19% ), fried rice (22.5% ), rice pilaf (46% ), Spanish rice (32% ),
and wild rice (32% ).

F. Other Factors Influencing Rice Usage

Operators agree that rice is easy to plate. In fact, research (32) indicates
that operators perceive rice as outperforming pasta on ability to plate since
it can be easily scooped or portioned. Rice is also perceived as being highly
suitable for batch cooking ahead of time, offering portion controllability
and the ability to be held throughout peak meal periods on a steam table.
Food service operators recognized that unused rice can be refrigerated
and used again by adding water and reheating (32). Leftover rice is added to
omelets, frittatas, quiches, soups, pancake and waffle batters, stuffings,
stews, or turned into cold salad presentations. It can be cooked with milk and
cinnamon for a new hot breakfast cereal. It may be sprinkled into soups and
salad. Rice is also used instead of bread as a base for stuffings for meats or
vegetables. Above all, rice is versatile enough to combine well with chicken,
beef, pork, seafood, vegetables, and countless seasonings. Compared to
pasta, rice is perceived by operators to have greater menu versatility.
Rice is also perceived to be an economical food carrier. As restaurateurs
experience mounting pressures to hold the line on check averages despite
MARKET DEMANDS 461

rising wholesale prices, rice is viewed as inexpensive enough for food ser-
vice operators to use as a plate extender and act as an ideal stretcher for
costly meat and fish proteins. When food service operators were asked to
rank their most important concerns entering the 1990s, "finding ways to
increase profits while controlling food costs" was also top on their list (31).
Sixty-seven percent indicate that rice is very applicable for helping to con-
trol food costs and improve plate profitability. Compared to either pasta or
potatoes, one third of all operators independently surveyed believe rice
offers a better perceived price/value relationship (32).

XI. NEW PRODUCTS

The 1980s saw a dramatic change in the supermarket rice aisle with the
emergence of entirely new rice categories. Rice mixes began to be a signifi-
cant factor in 1984 as consumers sought convenience to accommodate the
rapid increase in working women. Rice also became a key ingredient in
many of the new dry packaged dinners, a category that also evolved to
meet the need for added convenience. Rice moved to another aisle with the
introduction of rice cakes, an alternative healthy snack. The common ele-
ment among these new rice categories is that they are all value-added
processed foods. While the more traditional forms of rice presently account
for the dominant portion of the tonnage sold and will continue to be for the
foreseeable future, it is apparent that value-added rice products will con-
tinue to represent the fastest-growing segment.
In an analysis (41) prepared for the USA Rice Council, Gorman's New
Product News identified 508 new products introduced between 1989 and
1990 that contained rice as a principal ingredient. Of these, 30% were
prepared rice mixes, 24% snacks/crackers, 13% desserts/candies/cookies,
10% bakery products (either fresh-baked or mixes), 10% entrees (to which
some type of protein is added), 7% cereals/breakfast items, 2% baby
foods, 1% drinks, 1% dog foods, and 2% miscellaneous products. Re-
cently, brewers' rice has been appearing in many new premium varieties of
pet food (42). Rice is recognized for its positive nutritional image among
pet food manufacturers as well as pet owners. With 54 million dogs and 60
million cats in America, there is a substantial market here for rice.
Rice processors and spice companies have boosted rice's potential in
food service with a broad array of convenient, top-quality flavored and
premixed rice/spice blends that let any operator add an ethnic or regional
touch to the menu (31). This category has grown tremendously in the past 5
years in response to operators' need for variety and convenience. Over
50% of these operators are using prepackaged seasoned rice mixes, value-
added items that command a higher price by the manufacturer and distribu-
462 MEYERS

tor. Rice pilaf is among the most popular tlavored blends. Mexican-style
rice, Spanish rice, rice pilaf, Cajun-style rice, and wild rice/rice blends offer
fast, easy preparation, consistency, and low portion costs (12-23 cents per
half-cup serving). Operators can also create rice side dishes custom-
blended to accompany specific menu items. Adding nuts, fruits, vegeta-
bles, meats, and seafoods or unusual spices greatly increases the perceived
value of the rice dish by consumers.
According to Restaurant Hospitality's 1989 MenuTrak Research of food
service operations (43), both bulk bag and box forms are available. Forty-
eight percent of all rice products are purchased in box bulk form, with 52%
purchasing bag bulk. This compares with over 80% that purchase rice mixes
in box bulk form, with only 20% buying in bag bulk. Only about 1.5% is
offered in single-serve form. In this same study, figures indicate that 71% of
all food service operations offered rice in the past 12 months, with 27% using
rice mixes. Fourteen percent indicated that their operation had added rice
during the past 12 months, with another 14% adding a rice mix.
Research conducted by Restaurant Business (31) found that 46% of all
operators surveyed are currently blending or mixing white rice with other
rices, like brown or wild, or with pastas. Of those who serve a blended rice
mix, 50% buy it preblended, while 75% prepare these blends from scratch.
With promising ongoing reseach on rice bran (see Chapter 16) and rice
bran oil (see Chapter 18), it is likely that a variety of products incorporating
these rice byproducts as key ingredients will be introduced in the future.

XII. CONCLUSIONS
There is considerable reason for optimism about the future of rice consump-
tion within the United States. According to Supermarket Business (2):
Currently 12.1% of female homemakers say they are heavy users of rice,
and 10.0% of that same group say they are heavy users of flavored or
seasoned rice.
Rice usage is expected to increase by almost 12% by the year 2000. Heavier
usage will likely be fueled by those currently between the ages of 18 and
24 ( + 33%) and those between the ages of 35 and 44 ( + 12%).
For flavored or seasoned rice, heavy usage is forecast to increase 9.8% by
the year 2000.

ACKNOWLEDGMENT
The author wishes to thank Nathan W. Childs, USDA, Economic Research
Service, Washington, D.C., for the data used in the introductory section of
this chapter.
MARKET DEMANDS 463

REFERENCES
1. Castro, J. (1991). Goodbye to having it all: Americans are embracing the
simple life. Time, 137:58.
2. Dietrich, R. (1989). Looking toward the 1990s: What's heating up; what's
cooling off. Supermarket Bus., 44:33.
3. (1990). Lifestyles america: 1990's. Res. Alert, 8:41.
4. Naisbitt, J. (1991). As baby boomers approach mid-life, their impact contin-
ues. John Naisbitt's Trend Letter, 10:1.
5. (1990). Money's in the middle: Projected consumer spending in the 90's.
Prevention
6. Waldrop, J. (1991). The baby boom turns 45. Am. Demographics, January.
7. (1990). American's use of time project. Am. Demographics, February:33.
8. (1990). Shopping for stress. Am. Demographics, October:45.
9. Naisbitt, J. (1991). Immigration flood, subu,ban movement transforms U.S.
John Naisbitt's Trend Letter, 10:1.
10. (1991). 1991 foodservice forecast. Cahners Bureau of Foodservice Research.
11. (1990). New traditional homemaker survey. Am. Demographics, Decem-
ber:17.
12. Guber, S. (1991). Children of the 1990s. Marketing Rev., 8:16.
13. (1990). Who's cooking now! Res. Alert, 8:2.
14. (1990). Lifestyles america: 1990s. Res. Alert, 8:6, 39.
15. O'Hare, W. (1990). A new look at asian americans. Am. Demographics,
October.
16. Miller, T. A. W. (1991). Thirty-one trends shaping the future of american
business. Public Pulse, 2:1.
17. Rubbright, H., et al. (1989). The 1990s ... The microwave decade. Packag.
Strategies (Special Report).
18. (1991}. Suppertime USA survey. Food & Beverage Market, May:50, 54.
19. Mullins, M. E. (1990). Why microwaves are hot. USA Today, May.
20. (1990}. Food Editor Recipe Survey. USA Rice Council.
21. (1990}. Consumer Data: Why we use the microwave. Adweek Market. Week,
31:10.
22. (1990). Foodservice supermarkets. Prog. Grocer, April.
23. (1991}. Maritz marketing research, Fenton, Missouri. Am. Demographics,
February:4.
24. (1990}. Gallup survey for voluntary hospitals of America, Irving, Texas. Am.
Demographics, January.
25. (1991}. Nature of shopper's concern about the nutritional content of food:
1988-1990. Food Marketing Institute.
26. Marsh, E. (1990). 1989 study by the National Restaurant Association. Am.
Demographics, July.
27. Hoehberg, K. (1991). Health claims shift into high gear. Prep. Foods New
Prod. Ann. :47.
28. (1991). Consumers don't always do what they say, Am. Demographics, April:
29. (1990). State of the industry report, Snack World Mag., June:
464 MEYERS

30. USA Rice Council Survey of School Foodservice Directors, College and Uni-
versity Foodservive Directors, and Dietitians.
31. 1990. Advertising Awareness Surveys done for the USA Rice Council by
Restaurant Business and Restaurant & Institutions.
32. (1991). Rice image and usage studies 1988-1991. Com-Sci Systems, Inc.,
Chicago, IL.
33. (1989). 4th annual report on national eating trends. The NPD Group, July.
34. (1990). Tastes of America: 1990 national consumer survey. Restaurants &
Institutions.
35. Maritz Marketing Research, July 1989.
36. 1991 Annual Foodservice Forecast and Outlook, IFMA.
37. (1989). 1989 side-of-the-plate monitor. Restaurant Bus.
38. (1991). The changing mood of America. The Good Housekeeping/Roper Sur-
vey, first quarter.
39. (1990). How rice performs in various areas. Restaurants & Institutions.
40. (1989). 1989 Menu Census. Restaurants & Institutions.
41. (1990). Rice as an ingredient 1989-1990. Gorman's New Product News.
42. (1991). Pedigreed pet foods. Food & Beverage Market., May.
43. (1989). 1989 Menu Trak. Restaurant Hospitality.
 Index

Bran: (Bran)
composition, 383-388 in non-food applications, 401-402
antinutritional compounds, 386 animal feed, 401-402
carbohydrates, 385-386 derivatives and pharmaceuticals,
dietary fiber, 386 401
lipids, 384 nutritional benefits, 388-390
minerals, 384 laxation effects, 389-390
protein, 383-384 plasma cholesterol reduction,
vitamins, 384-385 388-390
dietary considerations, 37 potential as a food source, 381-383
in food products, 40, 397-401 health benefits, 382
baked products, 398 stabilized bran, 381-382
defatted or low-fat bran, 401 quality, 37
pasta products, 398-399 stabilization, 39-41, 394-396 (see
ready-to-eat cereals, 397-398 also Brown rice)
rice bran oil, 399-401 (see also enzyme inhibition, 39-40
Bran oil) lipolysis, 40-41
snacks and desserts, 399 methods of, 39
stabilized brown rice flour, 399 oxidative deterioration, 40-41
food types, 387-388 processing effects, 39
defatted or low-fat stabilized, stabilization and, processing, 390-
388 397
full-fat stabilized, 387 milling and purification, 390-394

465
466 INDEX

[Bran] [Breeding]
packaging, shipping, and trans- objectives, 87-89
portation, 396-397 disease and insect resistance, 87-
storage and shelf life, 396 88
storage study, 41-47 grain yield, 87
discussion, 46-47 quality characteristics, 88-89
materials and methods used, 41- stress resistance, 88
43 programs, 84-86
results, 43-46 Breeding, ecogeographic races, 83
Bran oil: Breeding for high milling yield, 99-
composition, 422-423 101
minor constituents, 423 Breeding for new market types, 104-
triglycerides, 422 106
effect of processing, 427-429 Brown rice:
loss of unsaponifiables, 427-429 consumption, 17
extraction, 423-425 deterrents to, 17
health benefits, 430-434 lipolytic hydrolysis, 18
dietary studies, 432-433 lipolytic oxidation, 18-20
effects of fatty acids, 430 enzymatic, 19
effects ofunsaponifiables, 430-432 nonenzymatic, 19-20
metabolic considerations, 433- nutritional value, 17
434 stabilization, 20-31
refining, 425-427 ethanol processes, 25-29
bleaching, 426 future prospects, 31
degumming, 426 heat processes, 20-23
deodorization, 427 to oxidation, 30-31
dewaxing, 426 products from, 29-30
neutralization, 426 solvent extraction process, 23-25
winterization, 427
stabilization, 421-422
extraction cooking, 422 Degree of milling:
parboiling, 422 definition, 140
Breeding: economic aspects, 140-141
market types, 89-98 grading, 140
long grain, 90, 92-94 head yield, 140-141
medium and short grains, 94-97 moisture content, 141
special purpose, 97-98 estimation of, 147-152
methods, 86-87 chemical methods, 150
induced mutation, 86-87 colored bran balance, 149
pedigree breeding, 86 differential dye staining, 148
new techniques, 102-104 New May Grunwald, 149
anther culture, 102 physical methods, 151
genome mapping, 103 qualitative descriptors, 147
protoplast fusion, 103 visual, 148
tissue culture, 102 by weight, 147
INDEX 467

[Degree of milling] [Drying]

lipid analysis, 170-172 solar, 280-288
near infrared spectroscopy, 152- drying conditions, 282-288
173 equipment, 280-282
calibration, 162-170 rough rice drying, 280
computer software, 153 use of agricultural residues, 288-
curve normalization, 154, 161- 292
162 energy potential, 288
equipment, 153 equipment, 288-289
math transformations, 155 rough rice drying, 289-292
sample preparation, 156-161
quality and nutrition, 141-147
brown rice, 142
composition, 141 Eating trends:
deep milling, 144-147 convenience foods, 449-451
insect infestation, 144 carryout and home delivery, 451
milling fractions, 141, 143, 145 convenience trends, 450
undermilled rice, 142 cooking from scratch, 450-451
uniformity of, 150 purchase behavior, 449-450
Drying: healthful foods, 452-456
breakage,299,313,329 consumer interest in rice, 452-
combined solar and biomass, 292- 453
293 nutrition versus convenience,
comparison of drying systems, 293- 454-455
294 rice as a healthy food, 455-456
continuous-flow dryers, 301 white meats and rice, 453-454
conventional, mechanism of, 302 new rice food products, 461-462
deep-bed dryers, 300 pleasure eating, 456-461
microwave-vacuum, 303-338 dining out, 456-457
drying curves, 317 family dinners, 456
drying rate, 309-311, 321-323 food service industry, 457-458
efficiency, 312-313, 323-328 food service rice usage, 458-461
equipment, 305-308
green rice, 305-315
kernel density, 328-329, 334 Food enrichment:
mechanism of, 303 dietary implications, 196
milling yields, 313-314, 329-331 enrichment versus fortification, 195
parboiled rice, 315-332 health implications, 196-197
rice quality, 333-338 nutritional benefits, 203-204
on-the-farm, 276-280
drying conditions, 277-278
energy use, 278-280
equipment, 277 Grain drying:
problems associated with, 276- energy considerations, 275-276
277 nonconventional methods, 293
468 INDEX

Head yield: [Milling]

definition, 341 bran removal, 230-234, 236
economic impact, 341 cone-shaped mill, 234
environmental effects, 350-352 effects of rotor cross section, 240
harvest moisture, 343 effects of water-mist injection,
milling moisture, 364-371 242-246
moisture adsorption/desorption, friction-type mill, 230-232
355-364 historical development, 234-236
head yield model, 360 huller-mill, 235
rates, 356-359 hull removal, 235
moisture variation effect, 347-350, humidifier, 251
352-355 mill speed, 232
multiple break, 242
pressurized moisture conditioner,
Marketing trends: 251
baby boomers, 443-445 rice washer, 258-259
discretionary income, 444-445 rotors, 242
food expenditures, 445 vertical mill, 236-242
health issues, 443 water mill, 242-249
work/lifestyle, 444 yields, 246-248
demographictrends,442-443 test milling, 364-376
immigration/geographic trends, degree of milling, 364-366
447-449 McGill mill, 364-366
Asian-Americans, 447, 448 mill settings, 375-376
Hispanic-Americans, 447 moisture content, 366, 375
non-traditional families, 443 sample size, 371-375
older population, 445-446 statistical analysis of, 366-371
population and lifestyle changes, uniformity, 143-144
442 Modified starch:
working women, 446-447 CFR requirements, 407-408
Microwave parboiling: future applications, 417-418
development of, 265-266 modification methods, 409-417
process features, 266-270 acidification, 417
cooking, 268-269 cross-linking, 409-413
drying, 269-270 dry heating, 416-417
soaking, 266-268 parboiling, 414-416
versus conventional parboiling, retrogradation, 413-414
271-273 Moisture content:
energy use, 272 adsorption model, 357
equipment costs, 271 distribution, 349
product quality, 272-273 effect on head yield, 341-380
space considerations, 271 environmental effects, 350-352
Milling: grain development, 342-243, 345
equipment, 230-249 harvest, 342-343
abrasive-type mill, 231-234 hysteresis, 377
INDEX 469

[Moisture content] [Quality]

individual kernel measurement, 345 computer data systems, 132
individual kernel variation, 345- cooking tests, 118-119
347,352-355 molecular level methods, 132-
maturity, 344 133
postharvest equilibration, 347 NIR spectrometry, 128-129
texture analysis, 120-125
viscosity analysis, 125-127
Nuclear magnetic resonance spectros- vision systems, 129-132
copy, 376-377 old methods to evaluate, 114-117
breeder tests, 114-115
FGIS standards, 115-116
Parboiling (see also Microwave par- ingredient rice tests, 117
boiling): table rice tests, 116-117
process, 263-265 washing of rice, 256-262
research needs, 264-265
Processing:
combined friction/abrasive milling, Rice:
235 composition of, 3
milling action, 230-236 consumption, 5-13, 439-441
milling pressure, 237 brewing, 439-440
miling ratio, 236 direct food use, 439, 441
moisture conditioning, 250-256 export market, 6-8
washing technology, 256-261 processed foods, 439-441
"Jiff" process, 259-261 United States, 10-13
traditional, 256 world, 5-6
waste water treatment, 260 current research and development,
water mist polishing, 242-249 14-15
benefits, 246 deterioration, 38-39
mold growth, 249 mechanism of, 38-39
storage stability, 246 development, 342-347
historical perspective, 1
moisture adsorption/desorption,
Quality: 355-364
influence of aging on, 189-190 head yield, 361-364
influence of protein on, 177-178 production, 5-13
methods and terminology, 179-180 United States, 8-10
microwave-dried rice, 333-338 world, 5-6
alkali spreading, 334 structure, 2-5, 139
amylographic analysis, 334-335 caryopsis, 3-5, 139
organoleptic evaluation, 335-338 embryo, 139
water uptake, 335 endosperm, 139
moisture conditioning, 250-256 hull,3
new methods to evaluate, 117-133 pericarp, 139
calorimetry, 127-128 Rice bran (see Bran)
470 INDEX

Rice bran oil (see Bran oil) experimental conditions, 206-

Rice enrichment: 207
application techniques, 202 practical applications, 225
coated kernel process, 201 thermal parameters, 207-208
other enrichment processes, 202 in milled rice, 208-213
powder enrichment process, 200- gelatinization models, 210-213
201 thermal curves, 208-120
rinse, cook, and wash resistance, 203 in milled rice flour, 208-213
standards, 197-200 thermalcurves,208-210
Rice grain (see Rice) monitored during rice cooking,
Rice kernel (see Rice) 205-206
Rice protein: Starch-protein interaction, 183-189
classification, 178-179 effect on starch gelatinization, 183-
effect on texture, 180-183 189
grain development, 182-183 relation to texture, 183
nitrogen fertilizer, 181-182 starch-oryzenin model, 183
grain content, 178 Storage:
protein bodies, 179 changes due to, 49, 51
starch granule associated, 179 correlations among physico-
Rice quality (see Quality) chemical and quality factors,
Rice storage (see Storage) 72-75
physicochemical changes, 51-59
carbohydrates, 51
Starch: enzymes, 53-55
function in foods, 405-406 lipids, 55-59
improvement by modification, 406 protein, 53 (see also Rice pro-
Starch gelatinization: tein)
related to grain structure, 213-225 starch, 51-53
effects of cracks and fissures, vitamins, 59
213-216 quality factor changes, 59, 61-72
effect of milling, 216-218 cooking properties, 66-67, 70
effect of particle size, 218-223 kernel hardness and swelling, 63-
structural barriers, 223-225 64
determined by DSC, 206-208, 225- odor and flavor, 70-72
226 starch properties, 62
applied to starch-based foods, storage gases and color, 59, 61
226 water uptake and solids loss, 61