Get ebook downloads in full at ebookname.

com

Cd Te and Related Compounds Physics Defects Hetero

and Nano structures Crystal Growth Surfaces and
Applications Physics Cd Te based Nanostructures Cd
Te based Semimagnetic Semiconductors Defects 1st
Edition Edited By
https://ebookname.com/product/cd-te-and-related-compounds-
physics-defects-hetero-and-nano-structures-crystal-growth-
surfaces-and-applications-physics-cd-te-based-
nanostructures-cd-te-based-semimagnetic-semiconductors-
defects-1st-e/

OR CLICK BUTTON

DOWNLOAD EBOOK

Explore and download more ebook at https://ebookname.com

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Te auahatanga me te ara auaha Creativity and creative

process 2nd edition. Edition Debbie Bright

https://ebookname.com/product/te-auahatanga-me-te-ara-auaha-
creativity-and-creative-process-2nd-edition-edition-debbie-
bright/

Conversation and Cognition 1st Edition Hedwig Te Molder

https://ebookname.com/product/conversation-and-cognition-1st-
edition-hedwig-te-molder/

Tao Te Ching Stephen Mitchell

https://ebookname.com/product/tao-te-ching-stephen-mitchell/

Libya since Independence Oil and State building 1998

1st ed. 1998 Edition Dirk Vandewalle

https://ebookname.com/product/libya-since-independence-oil-and-
state-building-1998-1st-ed-1998-edition-dirk-vandewalle/
Des e le ments aux traces elements and traces Ruyter

https://ebookname.com/product/des-e-le-ments-aux-traces-elements-
and-traces-ruyter/

Hollinshead s Functional Anatomy of the Limbs and Back

9e 9th Edition David B. Jenkins

https://ebookname.com/product/hollinshead-s-functional-anatomy-
of-the-limbs-and-back-9e-9th-edition-david-b-jenkins/

Programming Computer Vision with Python Tools and

algorithms for analyzing images 1st Edition Jan Erik
Solem

https://ebookname.com/product/programming-computer-vision-with-
python-tools-and-algorithms-for-analyzing-images-1st-edition-jan-
erik-solem/

Physics Pharmacology and Physiology for Anaesthetists

Key Concepts for the FRCA 1st Edition Matthew E. Cross

https://ebookname.com/product/physics-pharmacology-and-
physiology-for-anaesthetists-key-concepts-for-the-frca-1st-
edition-matthew-e-cross/

The Submerged Plot and the Mother s Pleasure from Jane

Austen to Arundhati Roy 1st Edition Kelly A. Marsh

https://ebookname.com/product/the-submerged-plot-and-the-mother-
s-pleasure-from-jane-austen-to-arundhati-roy-1st-edition-kelly-a-
marsh/
Jimmy Stewart Bomber Pilot 1st Edition Starr Smith

https://ebookname.com/product/jimmy-stewart-bomber-pilot-1st-
edition-starr-smith/
Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands
Linacre House, Jordan Hill, Oxford OX2 8DP, UK

First edition 2010

Copyright # 2010 Elsevier Ltd. All rights reserved.

No part of this publication may be reproduced, stored in a retrieval system or

transmitted in any form or by any means electronic, mechanical, photocopying,
recording or otherwise without the prior written permission of the publisher.

Permissions may be sought directly from Elsevier’s Science & Technology Rights
Department in Oxford, UK: phone (þ44) (0) 1865 843830; fax (þ44) (0) 1865 853333;
email: [email protected]. Alternatively you can submit your request
online by visiting the Elsevier web site at http://elsevier.com/locate/permis-
sions, and selecting Obtaining permission to use Elsevier material.

Notice
No responsibility is assumed by the publisher for any injury and/or damage to
persons or property as a matter of products liability, negligence or otherwise, or
from any use or operation of any methods, products, instructions or ideas
contained in the material herein. Because of rapid advances in the medical
sciences, in particular, independent verification of diagnoses and drug dosages
should be made.

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

ISBN: 978-0-08-046409-1

For information on all Elsevier publications

visit our Web site at www.books.elsevier.com

Printed and bound in Great Britain

10 11 12 10 9 8 7 6 5 4 3 2 1
 FOREWORD

Thirty years after the remarkable monography of K. Zanio and the numer-
ous conferences and papers dedicated since that time to CdTe and
CdZnTe, after all the significant progresses in that field and the increasing
interest in these materials for their extremely attractive fundamental
properties and industrial applications, the editors have thought timely
to edit a book on CdTe and CdZnTe, covering all their most prominent,
modern, and fundamental aspects. The subject has become so wide and
enriched during these 30 years that we have decided to call in well-known
specialists and experts of the field. The editors would like to thank them
deeply for their valuable contributions, with special acknowledgments to
Dr Henri Mariette for his pertinent recommendations and his continued
help and support.
This part covers the topics Physics, CdTe-Based Nanostructures,
Semimagnetic Semiconductors, and Defects. The topics Crystal Growth,
Surfaces, and Applications will be covered in Part II.

R. Triboulet
P. Siffert

xi
 LIST OF CONTRIBUTORS

C.R. Becker
Experimentelle Physik III, Universität Würzburg, Am Hubland, D-97074
Würzburg, Germany.
M.A. Berding
SRI International, 333 Ravenswood Avenue, Menlo Park, CA 94025, USA.
V. Consonni
CEA-LETI, Minatec, 17 rue des Martyrs, 38054 Grenoble Cedex 9, France.
K. Durose
Science Laboratories, University of Durham, South Road, Durham DH1
3LE, UK.
P. Fochuk
Chernivtsi National University, 2 vul. Kotsiubinskoho, Chernivtsi 58012,
Ukraine.
R.R. Galazka
Institute of Physics, Polish Academy of Sciences, Al. Lotnikow 32/46,
02-668 Warszawa, Poland.
K.A. Jones
School of Mechanical and Materials Engineering, Washington State
University, P.O. Box 642920, Pullman, WA, USA.
K. Lynn
School of Mechanical and Materials Engineering, Washington State
University, P.O. Box 642920, Pullman, WA, USA.
Y. Marfaing
Retired from Centre National de la Recherche Scientifique (CNRS),
Groupe d’Etude de la Matière Condensée (GEMaC), 1 Place A. Briand,
92195 Meudon Cedex, France.
H. Mariette
CEA-CNRS-UJF, Laboratoire de Spectrométrie Physique, UMR 5588
CNRS/Université Joseph Fourier Grenoble, Equipe CEA-CNRS-UJF
“Nanophysique et Semi-Conducteurs”, 140 Avenue de la Physique,
BP 87, 38402 Saint Martin d’Hères, France.

ix
x List of Contributors

B.N. Mavrin
Institute of Spectroscopy, Russian Academy of Sciences, 142190 Troitsk,
Moscow, Russia.
J.C. Moosbrugger
Department of Mechanical and Aeronautical Engineering, Center for
Advanced Materials Processing, Clarkson University, Potsdam,
NY 13699-5725, USA.
J.B. Mullin
EMC-HooTwo, 22 Branksome Towers, Westminster Road, Poole, Dorset
BH13 6JT, UK.
J.-O. Ndap
II-VI Inc., eV Products, 373 Saxonburg Blvd., Saxonburg, PA 16056, USA.
O. Panchuk
Chernivtsi National University, 2 vul. Kotsiubinskoho, Chernivtsi 58012,
Ukraine.
A. Sher
Sher Consulting, San Carlos Hills, CA 94070, USA.
E.A. Vinogradov
Institute of Spectroscopy, Russian Academy of Sciences, 142190 Troitsk,
Moscow, Russia.
T. Wojtowicz
Institute of Physics, Polish Academy of Sciences, Al. Lotnikow 32/46,
02-668 Warsaw, Poland.
 CHAPTER I
Introduction
J.B. Mullina,b,c

CdTe and CdZnTe are iconic examples of II–VI compounds. They epito-
mise the difficulties that need to be surmounted for successful device
exploitation in this class of semiconductors. How can this exploitation be
achieved? Put simply it can be achieved only by dominating the materials
technology. Or, as Dr Sekimoto an eminent Japanese businessman and
scientist so elegantly put it “Who dominates materials dominates technol-
ogy”. This book is about materials domination and the resulting device
exploitation. It concerns the knowledge, the abundant practical experi-
ence and the valuable device technologies that have emerged in
the development of CdTe and its close relative CdZnTe as a result of the
impressive research efforts of countless dedicated scientists.
The II–VIs have a long evolutionary history; indeed some II–VIs have
had applications from the dawn of antiquity. But our interest is their role
as semiconductors. The advent of the semiconductor era (circa 1948)
created a new standard in materials science – the need for semiconductor
quality. This standard ideally required material that was completely
single crystal and essentially defect-free and whose significant impurity
content had been reduced to the ppb level. What then is the status of CdTe
and its related alloys with respect to this goal?
The achievement of this goal for CdTe can be seen as a much more
demanding and problematic process when compared with the group IV
elements and the III–V compounds. The initial stage of semiconductor
quality was reproducibly met for germanium in about a decade and
within the next decade silicon had assumed a role as the dominant
semiconductor. In the case of the III–Vs comparable development

a
Electronic Materials Consultancy, 22 Branksome Towers, Poole BH13 6JT, UK
b
Editor in Chief, Progress in Crystal Growth and Characterization of Materials
c
Associate Editor, Journal of Crystal Growth

CDTE and Related Compounds # 2010 Elsevier Ltd.

DOI: 10.1016/B978-0-08-046409-1.00001-0 All rights reserved.

1
2 J.B. Mullin

followed a very much slower process. An exception has been InSb a

semiconductor blessed with benign preparative properties. It achieved
semiconductor quality status within 3–4 years following serious interest
in its development around 1953. Two important controlling parameters
can be identified as driving forces in the evolution of all these semicon-
ductors, the perceived significance and importance of its device applica-
tions and the intrinsic material properties especially its preparative
properties.
Why is Cadmium Telluride evolving relatively slowly as a device-
quality semiconductor even though it has a readily accessible melting
point and manageable vapour pressure over the melt? From a device
point of view its applications have industrial significance. Thus it has
gradually attracted increasing research support following its role as a
solar cell material. Its close relative HgCdTe relative has received
immense support in view of its infrared detector capabilities and its
consequential unique military role. ZnCdTe has found a critical role as a
substrate material for that difficult alloy. ZnCdTe also enjoys a crucial role
as a precursor g-ray detector material. Of course many more device
applications of CdTe and its related II–VI relatives are attracting more
support and study. But the dominating problem in their rate of evolution
has been and to some extent still is the lack of structural material knowl-
edge together with the intrinsic difficulties in the preparation of the
compounds and alloys.
The supreme importance then of this book is that it addresses this
central problem of crystal structural knowledge. This has been achieved
by the editors who have brought together a most impressive selection of
authors who clearly show how unwavering dedication to material science
and technology can bring about the understanding and control of these
most intransigent materials and the effective development of valuable
device technologies.
A fundamental feature of the II–VIs is the ionic component of the
bonding between adjacent atoms in contrast to the covalent bonding of
the group IVs. This departure from covalent character gives rise to a range
of native point defects and their associated complexes and is central to
understanding and controlling the properties of these materials. The
behaviour of the point defects has been studied since the pioneering
work of de Nobel. The main problem however in studying native point
defects is the lack of methods for their direct investigation. But whilst
there is a good working knowledge of their behaviour a definitive under-
standing is still lacking. Nevertheless very useful progress in the control
of point defects is reported.
To understand and control the II–VIs one must have a working knowl-
edge of the properties of these native point defects. This requires the
ability to control the composition with the imperative need to understand
 Introduction 3

the phase diagram. Without the application of specific precursor compo-

sition control non-stoichiometric CdTe with significant concentrations
of native point defects can result. An excess Te concentration typically
in the region of 0.0010% has been a common problem. The knowledge
assembled in the reviews on material preparation provides a splendid
account of the growth procedures for dealing with this problem.
As with all semiconductors the identification and removal of impu-
rities in and from the component elements and their compounds and
alloys is an essential requirement.
Great progress is reported in this area. This leads directly to the meth-
ods for preparing the crystals of CdTe and CdZnTe. This has been handled
admirably by Dr Triboulet who has spent a good proportion of his scientific
career in championing the development and applications of these semi-
conductors. Indeed it is the raison d’être for the book. The established
methods of melt growth, solution and vapour growth are rigorously
reviewed with a fine balance between fundamental and practical consid-
erations. This and related studies by other authors on the preparation of
doped materials and their potential relationship to structural complexes
have led to remarkable progress. Indeed the understanding of doped
material provides a necessary key to device exploitation.
A distinctive feature of CdTe and related materials, which contrasts
with the group IVs and the III–Vs, is the omnipresence of grain bound-
aries. This together with the ease of formation of dislocations, stacking
faults and inclusions of second phases creates formidable challenges to
the material scientists. The reviews give an up to date account of their
identification, behaviour and control methodology without which device
development would be truly problematic. Such advances have been nur-
tured as a result of essential research on the fundamental understanding
of the optical and physical properties of CdTe and related compounds
and alloys. These topics are also reviewed as are the physics of surfaces
and compensation.
The advances in materials knowledge are clearly demonstrated by the
development of valuable environmental, medical and opto-electronic
devices. One cannot but be impressed by the range of device activities
associated with CdTe CdZnTe and related materials. At the forefront of
these activities the importance of CdTe as a classic thin film solar cell
material is well recognised and its role is discussed in depth. The optical
detector role also covers X-ray detectors, a role of increasing importance
not only in the safety and security aspect of monitoring of nuclear emis-
sions but also in its value as a detector in medical scanning systems
involving computer tomography. The unique properties of doped mate-
rials such as V doped for their photo refractive properties and Mn-doped
materials in connection with conventionally grown nano-structures and
as semi-magnetic semiconductors. The latter application is making a
4 J.B. Mullin

significant contribution to the physics of spintronics. They offer a splen-

did insight into the future of these materials.
The reviews presented in this compilation provide an essential study
for anyone involved in II–VI development. The extensive materials
knowledge reviewed provides the key to device exploitation. Indeed the
unique device applications confirm the leading roles of CdTe, CdZnTe
and related materials in the history of semiconductors. The reader is
cordially invited to explore and assess their fascinating role for him or
herself.
 CHAPTER II
Physics

Contents IIa. Zinc Blende Alloy Materials: Band Structures and Binding
Properties
1. Introduction 7
2. Survey of First Principles Status 8
3. Crystal Structures, Binding, and Elastic Constants 9
3.1. Crystal structures 9
3.2. Insights from the bond orbital approximation 11
4. Concentration Fluctuations 15
4.1. Introduction 15
4.2. General statistical theory 17
4.3. Examples 18
5. Conclusions 19
Acknowledgments 20
References 20
IIb. Optical Phonon Spectra in CdTe Crystals and Ternary
Alloys of CdTe Compounds
1. Introduction 22
2. Phonon Spectra of CdTe 22
2.1. Bulk crystal 22
2.2. Films 25
2.3. Nanostructures 25
3. Localized Modes of Impurities in CdTe 27
4. Ternary Alloys of the CdTe Compounds 29
References 36
IIc. Band Structure
1. Basic Parameters at 300 K and Lower Temperatures 38
1.1. Band structure 38
1.2. Complex loss function 42
1.3. The CdTe(001) surface 44
1.4. Donors and acceptors 47
2. Electrical Properties 50
2.1. Carrier concentration limits 51
2.2. Mobilities 55
2.3. Carrier diffusion lengths and lifetimes 55
References 56

5
6 Chapter II

IId. Optical Properties of CdTe

1. General Features 59
1.1. Different radiative recombination processes 59
1.2. Theoretical determination of ionization
energy and orbital radius 62
1.3. Evolution of photoluminescence with
experimental parameters 64
2. Undoped and Doped Cadmium Telluride 65
2.1. Undoped CdTe 65
2.2. p-Type doped CdTe 68
2.3. n-Type doped CdTe 73
3. The Special Case of Chlorine Doping 76
3.1. Cl-doped monocrystalline CdTe 76
3.2. Cl-doped polycrystalline CdTe 78
4. Prospects 79
Acknowledgments 81
References 81
IIe. Mechanical Properties
1. Elasticity Properties 85
2. Inelastic Behavior 87
3. Fracture Properties 93
4. Optoelectronic-Mechanical Couplings 94
4.1. Photoplastic effect 94
4.2. Piezoelectric constant and
stress/strain-dependence of band characteristics 95
5. Summary 95
References 96
 CHAPTER IIA
Zinc Blende Alloy Materials:
Band Structures and Binding
Properties
A. Sher1

1. INTRODUCTION

This chapter concentrates on theories of the band structures and crystal

structural properties of zinc blende-structured II-VI compounds and
their alloys. The focus will be on CdTe-based materials, in particular
Hg(1
x)CdxTe.
We will begin with a brief survey of the most advanced first principles
approaches. These approaches are computationally intensive, but now
their predictions are remarkably accurate even for far more complex
materials than zinc blende-structured semiconductor compounds.
For example, they have been employed to predict wave functions, band
structures including proper band gaps and effective masses, ground state
crystal structures, cohesive energies, and elastic constants. These power-
ful methods still have not been applied to random alloys, but it is only a
matter of time before someone does it.
Aspects of the bond orbital approximation (BOA) [1] will also be
introduced for the insight they offer into semiconductor properties.
These methods are not as accurate as the first principles methods, but
they present a more transparent view into physical mechanisms.
Some results for band structures derived from the hybrid pseudopo-
tential tight-binding (HPT) method can be found in Ref. [2]. HPT starts

SRI International and Stanford University, Stanford, (Consulting Professor). CA, USA
1
Retired.

CDTE and Related Compounds # 2010 Elsevier Ltd.

DOI: 10.1016/B978-0-08-046409-1.00002-2 All rights reserved.

7
8 A. Sher

from a universal pseudopotential for zinc blende-structured solids that

includes long-ranged interactions. Then suitable local tight-binding terms
are added for each semiconductor that, once selected to fit, enable the
band structures for the different materials to match experimental results.
After this first fitting numerous physical properties of the materials,
including coherent potential approximation (CPA) treatments of alloys,
are calculated in agreement with experiments with no further adjustable
parameters. Following this procedure turns out to generate accurate wave
functions, as indicated, for example, by subsequent predictions of trans-
port properties [2–4], and the temperature dependence of band gaps [5].
Once the wave functions are known, matrix elements of interaction
Hamiltonians can be calculated leading to transition probabilities per
unit time. Then depending on the feature under investigation, sums
over the whole Brillouin zone are done to reach answers. Comparing
these answers with those obtained from models with parameters chosen
to fit experimental results often reveals that these fitted parameters are
unphysical. When this occurs it is usually because a mechanism has been
overlooked in the model, which is compensated by the choice of the fitted
parameters. While the HPT calculations depend on numerical methods,
they are far simpler and faster that the first principles methods. Actually if
one were to calculate the parameters from first principles that are
obtained in the HPT method by fitting them to experimental symmetry
point energies, I suspect the first principles parameters would closely
agree with the fitted ones, since both theories predict results in agreement
with experiments.
Finally a previously unpublished theory of mesoscopic concentration
fluctuations in alloys will be presented. In alloys like Hg(1
x)CdxTe,
or In(1
x)GaxAs statistics guarantees that there will be mesoscopic sized
regions (10-1000 Å) where the concentration x varies from the average. In
lattice constant-matched alloys like Hg(1
x)CdxTe, if there is a small
region with a concentration x, imbedded in its surroundings with concen-
tration
x, there is little strain energy penalty associated with such a
fluctuation. In a lattice constant-mismatched alloy, like In(1
x)GaxAs or
Cd(1
x)ZnxTe, there is a strain energy penalty that tends to suppress
concentration fluctuations [6]. In Section 4, we will introduce the under-
lying theory. While the theory is incomplete, it is never-the-less obvious
that these fluctuations can affect band edges at the fundamental gap,
defect states, and transport properties.

2. SURVEY OF FIRST PRINCIPLES STATUS

Starting from the density functional (DF) theory of Kohn and Sham, first
principles methods have advanced to a point where they reliably predict
numerous properties of solids. The local density approximation (LDA)
 Zinc Blende Alloy Materials: Band Structures and Binding Properties 9

coupled to Hedin’s GW approximation (GWA) solution methods have

been around for many years, but it is only now that serious difficulties
with its implementation are being overcome (G stands for Green’s func-
tion and W is a screened potential). The theory now treats full atom core
Coulomb potentials, and all electrons, not just the valence electrons and
ionic core pseudopotentials. Generally the choice of bases states in which
to expand the wave functions makes a comparatively small but significant
difference in the predictions, but a huge difference in the computational
time required to reach answers. For some bases choices, this presents a
practical limit on the complexity of the problems that can be approached.
At present the full-potential linerized muffin-tin orbitals (FP-LMTO) set is
one of the fastest and most accurate choices. One difficulty most recently
addressed is the number of higher state unoccupied orbitals needed to
cause some predictions to converge. A more complete status of the field
can be found in a paper by van Schilfgaarde et al. [7].
A typical set of results are presented in Table 1. The calculation of
band gaps is one of the more difficult features to get right because it
depends sensitively on the exchange interaction. Without a proper
accounting of this interaction predicted band gaps from first principles
are always too small. However, for example, the GW
Snn0 values for C,
Si, Ge, and GaAs are within 1.0, 12, 11, and 4.0%, respectively, of the zero
point energy adjusted experimental values.

3. CRYSTAL STRUCTURES, BINDING, AND ELASTIC CONSTANTS

3.1. Crystal structures

The focus of this chapter is on crystals in the zinc blende structure.
This structure consists of two interpenetrating face centered cubic sub-
lattices with one shifted from the other by 1/4 of the distance along the
(111) direction. p
Ifffiffi“a”
ffi is the length of a cube edge, then the near neighbor
distance is d ¼ 3a=4. In the group IV materials (C, Si, Ge, and gray Sn),
both sublattices are occupied by the same type of atom. In III-V and II-VI
(AB) compounds, one sublattice is occupied by the cations and the other
by the anions. In all cases, each atom site is fourfold coordinated with near
neighbors in a tetrahedral arrangement.
Pseudobinary alloys of the compounds are either cation substituted,
A(1
x)BxC, or anion substituted CA(1
x)Bx, where x is the fraction of B type
atoms on their sublattice. When the AC (or CA) compound bond length is
nearly the same as that of the BC (or CB) compound, then the system is
called “lattice matched.” If the bond lengths differ, they are “lattice mis-
matched.” In lattice-mismatched alloys, the cube edge distance of the AB
10 A. Sher

Table 1 Fundamental gap (in eV)

LDA GW GW GW Expt D/3 ZP Adj.

Z ¼1 Snn0
C 4.09 5.48 5.74 5.77 5.49 0 0.34 5.83
Si 0.46 0.95 1.10 1.09 1.17 0.01 0.06 1.24
Ge
0.13 0.66 0.83 0.83 0.78 0.10 0.05 0.93
GaAs 0.34 1.40 1.70 1.66 1.52 0.11 0.10 1.73
wAIN 4.20 5.83 6.24 6.28 0 0.20 6.48
wGaN 1.88 3.15 3.47 3.45 3.49 0 0.20 3.69
wInN
0.24 0.20 0.33 0.69 0 0.16 0.85
wZnO 0.71 2.51 3.07 2.94 3.44 0 0.16 3.60
ZnS 1.86 3.21 3.57 3.51 3.78 0.03 0.10 3.91
ZnSe 1.05 2.25 2.53 2.55 2.82 0.13 0.09 3.04
ZnTe 1.03 2.23 2.55 2.39 0.30 0.08 2.77
CuBr 0.29 1.56 1.98 1.96 3.1 0.04 0.09 3.23
CdO
0.56 0.10 0.22 0.15 0.84 0.01 0.05 0.90
CaO 3.49 6.02 6.62 6.50 7 0
wCdS 0.93 1.98 2.24 2.50 0.03 0.07 2.60
SrTiO3 1.76 3.83 4.54 3.59 3.3
ScN
0.26 0.95 1.24 0.96 0.9 0.01
NiO 0.45 1.1 1.6 4.3
Cu
2.33
2.35
2.23
2.18
2.78
Cu
2.33
2.85
2.73
2.18
2.78
Gd"
4.6
5.6
6.2
4.1
7.9
Gd# 0.3 0.2 1.8 1.5 4.3
For Gd, QPE corresponds to the position of the majority and minority f levels relative to EF; for Cu, QPE
corresponds to the d level. Low-temperature experimental data were used when available. QPEs in the GW
column are calculated with usual GWA Eqs. (6) and (7). In the Z ¼ 1 column the Z factor is taken to be unity.
In the Snn0 column the off-diagonal parts of S are included in addition to taking Z ¼ 1. k-meshes of 8  8  8k
and 6  6  6 were used for cubic and hexagonal structures, respectively (symbol w indicates the wurtzite
structure). GW calculations leave out spin-orbit coupling and zero-point motion effects. The former is
determined from D/3, where D is the spin splitting of the G15u level (in the zinc blende structure); it is shown in
the D/3 column. Contributions to zero-point motion are estimated from table 2 in Ref. [45] and are shown
in the ZP column. The “adjusted” gap adds these columns to the true gap and is the appropriate quantity
to compare to GW.

sublattice is very nearly a concentration-weighted average of those of the

AC and BC compounds, that is, following Vegard’s law [2, 8]. However,
the near neighbor AC and BC bond lengths remain within 70% of their
respective compound lengths. To accomplish this, the C atoms move off
their ideal fcc positions to accommodate to their four neighbor local
environment, for example, A4, A3B1, A2B2, A1B3, or B4. In lattice-matched
alloys, all bond lengths remain nearly constant, but to the extent that there
are small differences they follow Vegard’s law.
 Zinc Blende Alloy Materials: Band Structures and Binding Properties 11

3.2. Insights from the bond orbital approximation

Start by letting the crystal be in a zinc blende structure. Then the simplest
form of the bond orbital method employs wave functions and local
energies derived from pseudopotentials. The local cation and anion ener-
gies for their respective valence s- and p-states are denoted: eCs , eCp , eA
s ,
3
and eA p . Four sp hybrid wave functions pointing from one atom to
each near neighbor, for example, one pointing in the (111) direction has
the form:
 
jhi ¼ jsi þ jpx i þ jpy i þ jpz i =2: ð1Þ
A hybrid pointing in the (
1
1
1) direction has all the signs in front of
the p-wave functions reversed. The hybrid energies of the cations and
anions are given by:
   
  C C
C A A
A es þ 3ep
eh ¼ : ð2Þ
4
The interaction matrix element between two hybrids, one from an anion
and the other from a near neighbor cation, Harrison calls the “covalent
energy,” V2, and shows it has the form:
24:5
V2  hhC jHjhA i ¼ ðeVÞ; ð3Þ
d2
where the bond length, d, has units of Å
Next transform to a local molecular orbital basis given by:
jbi ¼ UbA jhA i þ UbC jhC i;
ð4Þ
jai ¼ UaA jhA i þ UaC jhC i:
Taking matrix elements of the Hamiltonian produces a 2  2 matrix,
and solving its secular determinate yields:
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
UbA ¼ ð1 þ ap Þ=2 ¼
UaC ;
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð5Þ
UbC ¼ ð1
ap Þ=2 ¼ UaA :
where ap is the “polarity” and is defined as:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ap  V3 = V22 þ V32 ð6Þ

and the binding and antibinding energies per electron are:

 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
þ
e a  ¼
eh V22 þ V32 : ð7Þ
b


12 A. Sher

Another major energy concept has now entered; it is the polar energy V3,
defined as:
 
V3  eCh
eA
h =2: ð8Þ
This energy tends to transfer electrons from the shallower cation to the
deeper anion hybrid states. (Note: both hybrid energies are negative so V3
is a positive energy.) It is ultimately responsible for the ionic contribution
to the solid’s binding energy.
The last contribution to the binding stems from the interaction cou-
pling each bond to its adjacent bonds. These interactions are called the
“metallization energies” and have been shown by Harrison to be:
 
V1C ¼ eCs
eCp =4
  ð9Þ
V1A ¼ eAs
ep =4:
A

They delocalize the molecular binding and antibinding states and

broaden them into the valence and conduction bands. In second-order
perturbation theory, these energies contribute a metallic term, Deb, to the
binding energy per electron of the form:
X jhbjHja0 ij2
Deb ¼ ; ð10Þ
a0
e b
e a0

where ja0 i is an antibinding state of a “molecule” adjacent to the one with

binding state b.
Finally collecting terms produces the net binding energy per unit cell:
Eb ¼ 2eb þ 2Deb þ u0
2
eatom ; ð11Þ
where eb contains the covalent and ionic components of the binding, Deb is
the metallic contribution, u0 ¼ C/d4 is the form of the screened Coulomb
repulsive energy between the ions, and
eatom is the average electron
energy per A and C atom. The constant C in u0 is chosen so the Eb has
its minimum at the observed equilibrium bond length do. The relation
between the binding energy and the equilibrium bond length do becomes:
8
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
A 2 <
A ; V  V
2 3
E b
x ¼
V2 þ V3 þ
2 2 q ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi 2d2o ; ð12Þ
2d4o V22 þ V32 :

V3 ; V2  V3
where
x 
eh þ reb
2eatom
: ð13Þ
A  24:5 ðeV
ðA Þ2 Þ
 Zinc Blende Alloy Materials: Band Structures and Binding Properties 13

The second equality in Eq. (12) states that the bond orbital model predicts
that the binding energy of pure covalent materials becomes progressively
stronger as the inverse square of the bond length, while the binding
energy of a pure ionic material is independent of the bond length.
This trend is also found experimentally. Thus one expects the variation
of the group IV compounds to vary fastest with bond length, the some-
what ionic III-V compound variations to be less steep, and the still more
ionic II-IV compounds to be even less steep. Obviously other parameters
besides bond length that impact values of x, and V3, also enter.
A number of different elastic constant models are treated in detail in the
“Semiconductor Alloys” book [2] and a review article [9]. Here once again
because of the insight it lends, the discussion will be confined to the bond
orbital model. The general relation between a displacement vector x in the
unstrained material and x0 in a strained material in component form is:
X
x0a ¼ xa þ ab xb ; ð14Þ
b

where ab are components of a 9  9 strain tensor. Then for small ab
values the energy density U is a quadratic function of the ab components:
1X
U¼  cabmn mn : ð15Þ
2 abmn ab

Because of rigid translation and rotation invariance there are in general

only 21 independent values of cabmn, the elastic stiffness tensor compo-
nents. But for zinc blende compounds their symmetry reduces the num-
ber of distinct components from 21 to 3 [10]. This permits a more compact
notation to be used in which there are only six components to ab, labeled
1-6 corresponding, respectively, to xx, yy, zz, yz, xz, and xy. In this reduced
notation the energy density becomes:
1X
U¼ Cij ei ej ; ð16Þ
2
where i ¼ ab and j ¼ mn, ei  eab  12 ðab þ ba Þ, and Cij  cabmn. By
examining three different simple distortions, relations can be found
between the bulk modulus B, C11-C12, and C44 and the bond orbital
parameters [2]:
2 3
2V2 ð9=8Þac ð5ac
4ÞðV1C þ V1A Þ5
3 2 2 2
B ¼
pffiffiffi 4a2c
;
3 d3 V22 þ V32
pffiffiffi
5:00 3a2c ð17Þ
C11
C12 ¼ ;
d5
9 6 4
¼ þ :
C44 C11
C12 B
14 A. Sher

Here the covalency ac is defined as:

qffiffiffiffiffiffiffiffiffiffiffiffiffi V2
ac  1
a2p ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi : ð18Þ
V22 þ V32

The last equality for C44 is quite instructive. It says that in the BOA there is
an analytic relation between the three elastic coefficients. Because this
relationship arises from a quantum treatment that is a reasonable approx-
imation to reality, it serves as a justification for treatments like the valence
force field model (VFF) of Keating [11] that involves only two elastic
constants, bond stretching and bond bending constants.
In a nearly pure covalent material where V2  V3 or ac ffi 1,
the expressions in Eq. (17) reduce to:
2 3
2A 4 9 2 
B ffi pffiffiffi 1
V1C þ V1A 2
=V22 5;
3d5 8
pffiffiffi
5:00 3
C11
C12 ¼ ;
d5 ð19Þ
pffiffiffi  2  2
3:00 3A 1
ð9=8Þ V1C þ V1A
2
=V2
C44 ¼ 8   9
< A 1
ð9=8Þ V 2
þ V 2
=V 2 =
1C 1A 2
2d5 1 þ :
: 5d2 ;

For this nearly covalent case all the elastic constants to first order vary
proportional to 1/d5. In the opposite extreme of a nearly ionic bonded
compound, where V2  V3, or ac ffi V2/V3  1, all of the elastic constants
in Eq. (17) are small but the leading term is proportional to 1/d9.
The agreement between the experimental and BOM predictions of the
elastic constant trends is good but the absolute values differ by factors of
2–4. When Coulomb interaction terms that are absent from the simple
form of the BOM are added, the agreement improves. When a full calcu-
lation based on HPT band structures and wave functions are done the
agreement is still better. Finally when an LMTO-based calculation is done
the agreement falls within 2-3% [2].
Because the bond length difference between HgTe and CdTe is small
(see Table 2), when they are alloyed into Hg(1
x)CdxTe, the alloy elastic
constants are very nearly a simple concentration-weighted average of the
constituents. However, when alloys of CdTe or HgTe with ZnTe are
formed, the stiff ZnTe bonds on the low ZnTe concentration side of alloys
significantly increase the alloy elastic constants above their concentration-
weighted average. This is of particular importance to dislocation densities
in these alloys. Since most of the excess energy associated with dislocations
 Zinc Blende Alloy Materials: Band Structures and Binding Properties 15

Table 2 Experimental values of the bond length and experimental and BOM results for
the bonding energy and the three independent elastic constants for three compounds

d (Å) Eb (eV) B C11-C12 C44

Material Exp. Exp. BOM Exp. BOM Exp. BOM Exp. BOM

ZnTe 2.643
1.20
1.04 5.090 1.82 3.060 1.88 3,120 1.50
CdTe 2.805
1.10
0.96 4.210 1.24 1,680 1.22 2.040 0.96
HgTe 2.797
0.82
0.48 4.759 1.23 1.817 1.33 2.259 1.11
The units for the elastic constants are 1011 dynes/cm2.

is in shears, for a given stress the dislocation density tends to be lower for a
short bond length material.
While the bond length difference between HgTe and CdTe is small, it
still causes troublesome misfit and treading dislocations in epitaxial
layers of Hg(1
x)CdxTe alloys grown on CdTe substrates. This is mitigated
somewhat by shortening the CdTe bonds by adding enough Zn to a
Cd(1
x)ZnxTe alloy substrate material to lattice match the substrate and
the epitaxial layer. In the best of circumstances threading dislocations still
propagate from the substrate into the epitaxial layers, and also long-
ranged cross hatch patterns persist [12]. These and other point and meso-
scopic sized defects degrade the performance of devices made from these
materials, especially in the long (LWIR) and very long (VLWIR) wave
length spectral regions where small fluctuations have an important effect
on the band gap.
A typical IR detector device structure is comprised of a conducting
substrate, followed by a semi-insulating p-type base layer, a more heavily
n-doped graded composition layer forming a heterojunction, and finally
an insulating cap passivation layer. Performance of these devices is sensi-
tive to the quality of the lattice matching at each interface.

4. CONCENTRATION FLUCTUATIONS
4.1. Introduction
In Hg1
xCdxTe alloys with concentration x, there is a near lattice constant
match between HgTe and CdTe bonds. Thus there is little strain energy
contribution to the mixing enthalpy [6]. Moreover, if there is a cluster
containing N cations (1 < N < many thousands) embedded in the average
medium, there is no strain penalty if the cation concentration in the cluster
is x, differing from the average alloy concentration
x. The net statistical
consequence is that there must be such concentration fluctuation clusters
16 A. Sher

and for each moderately large cluster of size N, they have a probability
distribution [6] given approximately by:
!
1 ðx


xÞ 2
PN ðxÞ ¼ exp
; ð20Þ
NN 2s2N

where the variance sN is:

xð1


xÞ x
s2N ¼ f  ; ð21Þ
N N
where f is a strain-induced reduction factor given by:

1
27BVc
f  1þ xð1
xÞ ; ð22Þ
kT
with B being the bulk modulus, and Vc the primitive cell volume.
The normalization factor is:
ð0
pffiffiffiffiffiffi
N ¼ 2psN
dx expð
ðx

xÞ2 =2s2N Þ
ð1
1 ð23Þ
pffiffiffiffiffiffi

dx expð
ðx

xÞ2 =2s2N Þ  2psN N∗N :
1

The quantity x is defined in Eq. (21), and N defined in the second equality
in Eq. (22) are both of order unity. For small clusters the Bernoulli distri-
bution must be used in place of the Gaussian. When there is a lattice
constant mismatch between the constituents, there is an extra multiplica-
tive factor f, less than unity, which enters into x [6]. For an alloy with
a large lattice constant mismatch the extra factor can be quite small. Some
representative numbers are given in Table 3. The shapes of clusters
are undoubtedly erratic, but if they are assumed to be spherical then the
relation between their radius rN, and N is just:
4p 3
N¼ r r; ð24Þ
3 N c
where rc is the density of cation sites in the solid. The effect of the
spherical cluster shape approximation needs to be examined in a refined

Table 3 Values of the reduction factor f for two alloys at two cutoff wave lengths

lc(mm) 10 20
Alloy x f x f

Hg(1
x)CdxTe 0.22 1.0 0.19 1.0

Hg(1
x)ZnxTe 0.18 0.39 0.15 0.42
 Zinc Blende Alloy Materials: Band Structures and Binding Properties 17

calculation, along with ellipsoidal, cubic, and tetragonal shapes that could
be more appropriate in some cases.

4.2. General statistical theory

In a refinement of the presentation in Ref. [6] for an alloy of average
concentration
x, we can deduce how fluctuations modify average values
and the mean deviation of a phenomenon whose concentration variation
is represented by G(x). The refinement has two major improvements.
First, the coarse graining approximation is eliminated, so cluster sizes
vary continually modulo a small fundamental cluster size. Second, some
unphysical cluster overlaps no longer occur. Examples of phenomena
sensitive to fluctuations include the band gap, the density of states
at the band edges, tunneling currents in reverse-biased p-n junctions,
and mobilities.
As before let N be the number of cations in a cluster, and ND be the
number in a large domain. Depending on the problem being addressed
the domain can be the whole volume of a pixel, its area times the deple-
tion layer thickness, or any other useful volume. Then the ratio ND/N is
the total number of clusters of size N in the domain. Next, define a new
quantity n(N, x) the number of clusters of size N and concentration x, to be
[13]:
ND
nðN; xÞ  PN ðxÞ; ð25Þ
N
where PN(x) is defined in Eq. (20) as a Gaussian. It is the probability of
finding a cluster of size N with concentration x.
Now we can ask the question, how large must a domain of size MN be
to contain on average 1 cluster of size N and concentration x? Thus we
require for ND ¼ MN(x) that n(N, x) ffi 1, or:
N
MN ðxÞ ¼ : ð26Þ
PN ðxÞ
Notice that since PN(x) < 1, we always have MN(x) > N, so the class of
clusters with size N and concentration x never overlap as they did in
Muller and Sher [6]. However, clusters with size N0 may overlap those
with size N. The density rN(x) of a specific class of clusters is:
rc
rN ðxÞ ¼ ; ð27Þ
MN ðxÞ
where rc is the density of cations in the solid. Depending on the shape
chosen for domains, the average separation between clusters of size N and
concentration x, is (rN)1/3. Small clusters are spaced more closely than
large ones. A spatial variation of the concentration profile will have large
18 A. Sher

volume bumps and dips with smaller volumes generally with larger
amplitudes, superimposed on them.
The probability in a domain of size ND of finding a cluster having size
N and concentration x is:
nðN xÞ 1
PN ðxÞ
ðN; xÞ  X ð ¼ NX : ð28Þ
ND 1
dx P N 0 ðxÞ
N0
N0 N0
N0

Because there will be small clusters superimposed on larger ones, there is

a possibility of double counting. The normalization used in Eq. (28) elim-
inates the consequences of this double counting. Therefore, an average of
a general phenomenon with a concentration variation G(x) over all cluster
sizes and concentrations is:
X1ð
D ND ð 1
dx GðxÞPN ðxÞ
 E X N
N
GðxÞ N ¼ dxðN xÞGðxÞ ¼ X1 ; ð29Þ
N¼NF 0
N
N

where NF is the size of the selected fundamental cell. One reasonable

choice is a four cation, one anion cell containing 16 bonds where NF ¼ 4.
Then N varies modulo 4 from a minimum of 4 to ND. Another choice is a
primitive cell where NF ¼ 1.
We still face the problems of setting functional x dependences of
physical quantities for small clusters (quantum dots). There are also
problems with the shapes of clusters, for example, the difference if they
are nearly spheres or ellipsoids, in the bound state energies of clusters of
size N that introduce local mesoscopic-sized potential wells.

4.3. Examples
Let us do a couple of simple examples to demonstrate the averaging
process. In this exercise we will let the integral over x vary from minus
to plus infinity rather than renormalizing PN(x).
Example (A): GðxÞ ¼ x

x
Then we have:
hx


x iN ¼ 0 ð30Þ
D  E
so hxiN ¼

x, and x N ¼

x.
Example (B): GðxÞ ¼ ðx

x Þ2
Now the integral becomes:
D E

xð1

xÞ
ðx


xÞ 2 ¼ s2N ¼ : ð31Þ
N N
 Zinc Blende Alloy Materials: Band Structures and Binding Properties 19

Replacing the sum over N by an integral, the second average over the
pixel volume is:
D E ð ND  
2

xð1


xÞ
xð1

xÞ 1 1
ðx


xÞ N ¼ dN 2 ¼
: ð32Þ
NF N lnðND =NF Þ lnðND =NF Þ NF ND
For a pixel whose volume is 40  40  10 mm3 and for Hg0.79Cd0.21Te
where rc ¼ 7.41  1021 cm
3, we find ND ¼ 1.18  1014, and ln ND ¼ 32.4.
If we take NF ¼ 2, the smallest value for which s2N =
x2 has the same answer
for both the Gaussian and Bernoulli distributions, and
x ¼ 0:21 then:
D  E
ðx

xÞ 2 N 1 1

x
D  E2 ¼ ¼ 5:93  10
2 : ð33Þ
2lnðND =2Þ
x
x N

This says the average x variation among pixels of the above volume is
0.21 0.059. This is probably too large a variation. One problem is that
the averages for small N are large and they get emphasized in the conver-
sion from a sum to an integral in the second average over the pixel. This
example needs to be redone using a proper distribution for small N and
retaining the sum until N is large. However, even when this is done the
rms value is still too high.
As it stands, this theory only accounts for strain reductions through the f
factor. Values of f are given in Table 3 for the Hg(1
x)CdxTe and Hg(1
x)ZnxTe
alloys for concentrations where the cut off wave lengths are 10 and 20 mm.
For Hg(1
x)CdxTe the reduction factor is near unity. But it is small enough in
Hg(1
x)ZnxTe alloys to significantly suppress concentration fluctuations.
A good part of the reason this theory predicts large rms fluctuation is that
a mechanism is left out of the formalism that also reduces their amplitudes.
A cluster with a concentration x differing from the average
x will have a net
cohesive energy and a mean deviation of its cohesive energy that is different
from the average. This effect still must be built into the theory.

5. CONCLUSIONS

Starting about 20 years ago and accelerating, materials science theory is
undergoing a major change. The methods have progressed from approx-
imations that often led to insights into the underlying cause of phenom-
ena, but depend on parameters fitted to experiments to produce accurate
predictions, to parameter free first principles computational methods.
These new computational intensive methods not only help uncover the
true causes of phenomena, but are proving to be reliable engineering tools
that when used properly can speed developments. The combination of
first principles predicted parameters, and processing and performance
models are powerful device development tools.
20 A. Sher

In this chapter, three levels of theories have been referenced. The BOA
relies on a simplified fundamental-based calculation with few adjusted
parameters to lend insight into the caused of a broad range of phenomena,
but the accuracy of the predictions is often in the range of factors of two or
so. The HPT method is closer to a first principles computationally inten-
sive method, but it does require some fitting parameters. However, once
that initial fitting is done the method allows predictions of a wide variety
of phenomena for compounds and alloys with no further fitted para-
meters to accuracies of 10%. First principles theories like the self-
consistent full potential GWA method predicts the behavior of complex
materials to within a few percent with no adjusted parameters.
However, even with these powerful techniques in hand it is still
necessary to ask them the right questions. That brings us to the example
of concentration fluctuations. The presentation above is a bare start into a
study of the impact these fluctuations have on the properties of alloys.
These fluctuations are expected to be especially troublesome in alloys
where the band gap is small, that is, cases where the cutoff wave length
of devices is in the 10-20 mm range.
I have long advocated that for these VLWIR devices the active materi-
als should be Hg(1
x)ZnxTe rather than Hg(1
x)CdxTe because in that alloy
system not only concentration fluctuations, but also dislocation tend to be
suppressed. Some early 20 mm cutoff, LPE grown Hg(1
x)ZnxTe arrays
were tested [14]. They displayed excellent bake stability and had detec-
tivities D* at 64 K within a few percent of BLIP (D∗ BLIP ¼ 1:3  10 Jones).
11

Because so much effort has gone into attempting to perfect Hg(1
x)CdxTe

technology (with a bit of Zn added in some cases), little effort to has gone
into continuing the development of Hg(1
x)ZnxTe-based devices despite
their promise.

ACKNOWLEDGMENTS

I would like to thank Professor Mark van Schilfgaarde for guidance on the
current status of first principle calculations and for a preprint of his paper.
I also wish to thank Professor An-Ban Chen for deducing Eq. (25) and
other suggestions. I am indebted to Dr. E. Patten who supplied data on
Hg(1
x)ZnxTe alloy based devices.

REFERENCES
[1] W.A. Harrison, Electronic Structure and the Properties of Solids, W. H. Freeman and
Company, San Francisco, 1980.
[2] A.-B. Chen, A. Sher, Semiconductor Alloys, Plenum Press, New York, 1995.
[3] S. Krishnamurthy, A. Sher, J. Appl. Phys. 75 (1994) 7904–7909.
 Zinc Blende Alloy Materials: Band Structures and Binding Properties 21

[4] S. Krishnamurthy, A. Sher, J. Electronic Mater. 24 (1995) 641–646.

[5] S. Krishnamurthy, A.-B. Chen, A. Sher, M. van Schilfgaarde, J. Electronic Mater.
24 (1995) 1121–1125.
[6] M.W. Muller, A. Sher, Appl. Phys. Lett. 74 (1999) 2343–2345.
[7] M. van Schilfgaarde, T. Kotani, S.V. Faleev, Phys. Rev. B 74 (2006) 245125–245140.
[8] J.C. Mikkelsen Jr., B. Boyce, Phys. Rev. Lett. 49 (1982) 1412–1415.
[9] A-B. Chen, A. Sher, W.T. Yost, Elastic constants and related properties of semiconduc-
tor compounds and their alloys” Chapter I in “Semiconductors and semimetals, in:
K.T. Faber, K. Malloy (Eds.), The Mechanical Properties of Semiconductors, vol. 37,
Academic Press, Inc., Boston, 1992.
[10] J.P. Hirth, J. Lothe, Theory of Dislocations, second ed., John Wiley & Sons, New York,
1982.
[11] P.N. Keating, Phys. Rev. 149 (1966) 674–678.
[12] M.A. Berding, W.D. Nix, D.R. Rhiger, S. Sen, A. Sher, J. Electronic Mater. 29 (2000)
676–679.
[13] A.-B. Chen, 2006 (private communication).
[14] E. Patten, 1997 (private communication).
 CHAPTER IIB
Optical Phonon Spectra in CdTe
Crystals and Ternary Alloys of
CdTe Compounds
B.N. Mavrin and E.A. Vinogradov

1. INTRODUCTION

The structure of the cubic face-centered unit cell of CdTe crystal is char-
acterized by the space group Td2 (F4
3m) and contains four formula units.
The primitive unit is one fourth as many. The vibrational representation
of optical phonons consists of 1 threefold-degenerated mode F2 which is
active in IR and Raman spectra. The dipole mode F2 is split into the
transverse (TO) and longitudinal (LO) modes in the vibrational spectra.

2. PHONON SPECTRA OF CdTe

2.1. Bulk crystal
First, the optical properties of CdTe crystal in IR region from 20 to 400 mm
were studied from the reflection spectra R(o) [1–7]. The Reststrahlen band
in R(o) is at 154 cm
1 with the peak reflectivity of 98% at 80 K and at
150 cm
1 with the peak reflectivity of 78% at 273 K (Fig. 1) [8, 9]. Using the
Kramers-Kronig analysis, the frequency dependences of both refraction
index n(o) and extinction coefficient k(o) can be found from equations:
pffiffiffiffi
1
R 2 R sin y
nðoÞ ¼ pffiffiffiffi ; kðoÞ ¼ pffiffiffiffi ; ð1Þ
1 þ R
R cos y 1 þ R
R cos y

Institute for Spectroscopy of Russian Academy of Sciences, Fizicheskaya Str.5, 142190, Troitsk, Moscow, Russia

22
 Optical Phonon Spectra in CdTe Crystals and Ternary Alloys of CdTe Compounds 23

1.0 R

0.8 4

Δ-1
x-2
0.6
0-3

0.4

0.2

100 150 200 250 cm–1

Figure 1 Infrared-reflection spectra of pure CdTe with the different polish of surface:
1—chemical treatment (273 K), 2—on pitch (273 K), 3—on cloth (273 K), 4—on pitch (80 K).

where
ð
o lnRðnÞ
yðoÞ ¼ dn: ð2Þ
p n2
o 2
The value n(o) and k(o) are related with dielectric function e(o) ¼ e1(o) þ
ie2(o) as follows:
e1 ðoÞ ¼ n2 ðoÞ
k2 ðoÞ; e2 ðoÞ ¼ 2nðoÞkðoÞ;
 
1 2nðoÞkðoÞ ð3Þ
Im
¼ 2
:
eðoÞ ½n ðoÞ þ k2 ðoÞ
2

On the other hand, using the general dispersion relation for e(o) with the g
damping, one can obtain:
 
ðe0
e1 Þo2TO g 1 ðe
1
1
1
e0 ÞoLO g
2
oe2 ðoÞ ¼ ; oIm
¼ : ð4Þ
ðo
oTO Þ2 þ g2 eðoÞ ðoLO
oÞ2 þ g2
The maximum of the curve oe2(o) corresponds to the TO-mode frequency
oTO and the oIm(
1/e(o)) maximum to the LO-mode frequency oLO.
The bandwidths of these curves determine the g damping of phonons.
24 B.N. Mavrin and E.A. Vinogradov

Three methods were used to find the oscillator strengths (STO ¼ e0
e1) of
TO modes:
pffiffiffiffiffiffi !2 pffiffiffiffiffiffiffi !2
R0 þ 1 R1 þ 1
e0 ¼ pffiffiffiffiffiffi and e1 ¼ pffiffiffiffiffiffiffi ;
R0
1 R1
1
ð ð5Þ
gTO 2 1 e2 ðoÞ
e0
e1 ¼ e2 ðoTO Þ; e0
e1 ¼ do;
oTO p 0 o

where R0 and R1 are the reflectivities at o  oTO and o  oLO,

respectively.
Using Eqs. (1)–(4) and R(o) (Fig. 1), one can obtain the dielectric
function (Fig. 2) as well as TO (140 cm
1) and LO (167 cm
1) frequencies
and dampings gTO (5.6 cm
1) and gLO (6.3 cm
1) for the CdTe crystal
[9, 10]. The imperfections of the crystal have an effect on the optical
properties. For example, the perfection of CdTe crystal lattice changes
by the different treatment of the reflecting surface of crystal (Fig. 1,
Table 1). In particular, from Table 1 it is seen that the frequencies of
phonons increase with a decrease of phonon dampings.
The analysis of the reflectivity spectra of n-CdTe allowed one to study
the plasmon-phonon interaction as well as to measure the concentrations
and mobilities of free carriers [11].
The TO and LO modes are active in the Raman spectra of CdTe
[12–15]. If the Raman spectra are excited in the transparency region

8 n, k
k
n
4

0
100 200 cm–1

ε1, ε2 Im (–ε–1)
100 1.0

50 0.5

0
100 200 cm–1

–50

Figure 2 Optical functions of CdTe: n(o) and k(o), e1(o), e2(o)-left axis,
and Im(
e
1 (o))-right axis, reconstructed from reflection spectra 1 of Fig. 1.
 Optical Phonon Spectra in CdTe Crystals and Ternary Alloys of CdTe Compounds 25

Table 1 Optical constants of CdTe single crystals after various polishes

T oTO gTO oLO gLO eS

Polish (K) (cm
1) (cm
1) 4pr (cm
1) (cm
1) e0 e1 e
On cloth 293 139 6.5 3 166 7.5 10.3 7.3 0.72
On pitch 293 140 5.6 3.15 167 6.3 10.4 7.3 0.73
Chemical 293 140.5 5 3.25 167.5 5.5 10.5 7.3 0.74
On pitch 80 143.5 2 3.30 171 1.7 10.5 7.2 0.77

(<1.5 eV), TO mode is dominant [12, 13]. But in the case of resonant
Raman scattering, especially at resonance with the E0 þ D0 gap
(2.54 eV), LO mode and its overtones become intense [14, 15]. Besides,
the bands near 120 and 140 cm
1 corresponding to phonons of Te on the
CdTe surface can be seen in the resonant Raman spectra.
The phonons of the CdTe crystal were studied in the whole Brillouin
zone by inelastic neutron scattering [16] and ab initio calculations [17]. As
it follows from calculations of the density of phonon states in CdTe crystal
[18], there is a gap from 123 to 128 cm
1 between acoustical and optical
density of states and the high-frequency edge of the optical density is near
173 cm
1.

2.2. Films
The direct measurements of both TO and LO phonons by infrared tech-
nique may be performed from films. Earlier it was shown [19], that one
can observe an absorption by TO and LO modes at the oblique incidence
of light on a film. As it is expected [19], only TO modes may be seen in the
s-polarized light as well as both TO and LO modes in p-polarized light.
The experimental spectra of the infrared absorption in the CdTe films on
the metallic mirror [10] correspond to these predictions (Fig. 3). Moreover,
the optical constants of CdTe measured from the reflection for bulk
crystals (Table 1) and by the absorption in the films are similar [10]. For
more detail explanation of IR optical properties of thin semiconductor
(A2B6) films see for example [20].
The Raman spectra from CdTe films were also studied (for example,
in [21]) and they had the same features as those from the bulk crystal
(Table 2).

2.3. Nanostructures
The phonon spectra of nanostructures (quantum dots, quantum wells and
heterostructures) can be different from those of bulk crystals and depend
on the nanocrystal shape and size. The interface phonons, the folded
26 B.N. Mavrin and E.A. Vinogradov

R R
1 1 1 1

1 2 1 2
1 3 1 3
1 4 1
4
0.9

0.8 0.8

0.7

0.6 0.6
cm–1 cm–1

100 150 200 100 150 200

Figure 3 Reflection-absorption spectra of CdTe films on Al substrate in s- (left) and

p-polarized (right) light for different thickness of the films (in mm): 1—0.5, 2—0.9, 3—1.6,
4—2.4.

Table 2 Optical constants of CdTe films (p-polarization)

Thickness, oTO gTO A(TO) oLO gLO A(LO)

m cm
1 cm
1 adsorption cm
1 cm
1 adsorption

0.5 140 5.5 0.01 167 6 0.03

0.8 139.5 6 0.06 167 6 0.05
0.9 139 6.5 0.1 167 6 0.06
1.6 138 7.5 0.25 167 6 0.1
2.4 137 5 0.46 167.5 5 0.16

acoustical phonons and the confined optical phonons may appear in the
nanostructure spectra [22].
For example, the Raman spectra of the CdTe quantum well show a
main peak near 173 cm
1 and smaller oscillationlike features towards
lower phonon energies, which are assigned to even-order confined pho-
nons (m ¼ 2, 4, . . ., 12) in quantum well [23]. The appearance of the
oscillations is due to the quantization of the phonon momentum along
the epitaxial growth normal to the layer:
qm ¼ mp=ðd þ a0 =2Þ; ð6Þ
where d is the layer thickness and a0 is the lattice constant of the bulk
crystal. Using the observed dispersion om (qm) of the oscillations, it is
possible not only to find the phonon dispersion at q > 0, but also to derive
and estimate the compositional profile of the quantum well [23].
 Optical Phonon Spectra in CdTe Crystals and Ternary Alloys of CdTe Compounds 27

The polariton modes of a CdTe quantum-well-embedded planar

microcavity are studied by resonant Raman scattering as a function of
cavity-exciton detuning [24]. A maximum in the Raman efficiency is
observed at the mode anticrossing, with minima in the pure photon and
exciton limits.
The optical vibrations in the quantum dots are also considered to be
confined to the dots with spherical shape and are assigned an equivalent
wave vector:
m
qm ¼ m ; ð7Þ
r
where mm is the mth node of the spherical Bessel function and r the radius
of quantum dot. The strain due to the surface-free-energy appears in a
crystal of finite size [25]. Hence, the optical-mode frequencies of CdTe
quantum dot are determined by two competing size-dependent effects,
namely, phonon confinement, causing a redshift, and compressive strain,
producing a blueshift. The second effect is essential at small r.
The transmission spectra of the CdTe quantum dots with diameters of
4.2 nm deposited on Si substrate with 1 mm thickness layer were studied
[26]. The center of the transmission band is at 152 cm
1 that was higher
than that of TO mode in bulk crystal (140 cm
1). In the transmission
spectra of the CdTe quantum dots (r ¼ 4 nm) coated by the CdSe shell
with the thickness of 1-2 nm two minima were observed at 151.2 (CdTe-
like) and 188 (CdSe-like) cm
1 [27].
The Raman spectra of the CdTe quantum dots (r  10 nm) in borosilicate
glass have shown the redshift of LO mode by 2.6 cm
1 [27]. The cavity-
induced enhancement of the Raman efficiency from a monolayer of the
CdTe quantum dots coated a dielectric microsphere was observed [28].
Far infrared reflectivity measurements of the CdTe-HgTe superlattices
allowed one to find the layer composition, thickness, and the effective
mass [14, 29]. Raman spectra from two high-quality CdTe/HgTe super-
lattices have shown CdTe, HgTe, and Hg1
xCdxTe phonon modes [30].
The locations of Hg1
xCdxTe phonon lines confirmed the degree of Hg
alloying in the nominal CdTe layers, showing the utility of Raman scat-
tering for the superlattice characterization. The IR and Raman techniques
were used to study a very complex system consisting of superlattices of
the CdTe quantum dots [31].

3. LOCALIZED MODES OF IMPURITIES IN CdTe

Substitutional impurities are known to affect the vibrational spectrum of a

crystal. In addition to modifying the frequencies of the host lattice, new
features are appeared in the spectrum [32]. When a substituent (M0 ) of an
infinitesimal amount is lighter than the host atoms (M), one expect the
28 B.N. Mavrin and E.A. Vinogradov

localized modes of the impurity with frequencies above the optic band of
the host lattice. Impurities of heavier masses may have localized mode
frequencies in the gap between the acoustical and optical bands (gap
modes). Localized modes falling inside the optical band are known as
resonance modes. The frequencies of impurity modes may be found from
equation [33]:
eo 2 X X jsj ðq; MÞj2
0

1þ ¼ 0; ð8Þ
N q j o2j ðqÞ
o0 2

where e ¼ (M
M0 )/M, o0 is the eigenfrequency of the perturbed phonon

state, N is the total number of wavevectors, sj(q, M) corresponds to the
atom of mass M in the j mode with the q wavevector, and oj(q) is the
eigenfrequency of the perfect lattice. Using Eq. (8), explicit calculations of
impurity modes have been carried out for a large class of zinc blende
crystals [33]. The eigenvalues oj(q) and eigenvectors sj(q, M) were deter-
mined by the modified rigid ion model. The calculated value of local
mode frequency for the Se impurity in CdTe has corresponded to
171.6 cm
1 [33] that was in agreement with experiment [34, 35]. Several
illustrations of an observation of local (Fig. 4) and gap (Fig. 5) modes in
system Cd1
xZnxTe are shown. Apparently, the results of a solution of
Eq. (8) depend on the lattice-dynamical model to obtain eigenvalues and
eigenvectors of perfect crystal. Alternative methods of determining impu-
rity modes using the density of phonon states are discussed in [32, 36].
The available experimental data of impurity modes in CdTe are presented
in Tables 3 and 4.

0.8 R

0.6
2

0.4

0.2
1

100 150 200 250 cm–1

Figure 4 Reflection spectra of crystals Cd1
xZnxTe at smal x: 1—CdTe, 2—Cd0.96Zn0.04Te,

3—Cd0.90Zn0.10Te.
 Optical Phonon Spectra in CdTe Crystals and Ternary Alloys of CdTe Compounds 29

0.9 R
1
2

0.7

0.5 3

0.3

0.1
cm–1

100 150 200 250 300

Im (ε) Im (−ε–1)
1.5
2
100 1 1.25

80 1.0

60 0.75

40 0.5
4
20 0.25

0 0
3
–20 –0.25
150 170 190 210 230 cm–1

Figure 5 Reflectivity spectra of gap mode in ZnTe:Cd (A) crystals with compositions:
1—ZnTe, 2—Zn0.95Cd0.05Te, 3—Zn0.80Cd0.20Te and their dielectric functions (B).

4. TERNARY ALLOYS OF THE CdTe COMPOUNDS

Single crystals of alloys of the CdTe compounds can be grown at almost

any desired composition that allows one to control their optical and
physical properties in applications. There are two extreme classes of
ternary alloys AB1
xCx, according to the compositional dependence of
their Raman and IR spectra [44, 45]. In the first class, known as one-mode
crystals, the Raman and IR spectra exhibit one TO (LO) mode whose
frequency varies continuously from oAB TO (oLO ) at x ¼ 0 to oTO (oLO )
AB AC AC
30 B.N. Mavrin and E.A. Vinogradov

Table 3 Experimental value of impurity mode frequencies (in cm
1) in CdTe

Compound Frequency Experiment References

CdTe:Zn 167 IR [34]

CdTe:Zn 173 IR [37]
CdTe:Mn 195 Raman [38]
CdTe:Fe 196 IR [37]
CdTe:Hg 130 Raman [30, 39, 40]
CdTe:Mg 250 IR, Raman [41]
CdTe:Se 173 IR [34, 35]
CdTe:S 258.7 Raman [42, 43]

Table 4 Optical constants of CdTe mixed crystals [8, 9, 34]

oTo oLo gTo gLo eS

Compound T (K) Mode (cm
1) (cm
1) (cm
1) (cm
1) e0 e1 4 pr e
CdTe0.94Se0.06 293 Main 140.5 166 10.5 6 7.2 10.2 3 0.73
CdTe0.94Se0.06 293 Local 172.4 174.6 11 12 7.1 7.2 0.1 0.6
CdTe0.94Se0.06 80 Main 144 169.5 2.4 2.5 7.1 10.3 3.2 0.79
CdTe0.94Se0.06 80 Local 175 177 6.8 6.0 7.0 7.1 0.1 0.5
CdTe0.9Se0.1 293 Main 141.3 163.5 7.5 6.0 7.2 10.1 2.9 0.74
CdTe0.9Se0.1 293 Local 172.2 179.0 10.5 11.0 6.9 7.2 0.3 0.76
Cd0.95Zn0.05Te 293 Main 140.3 168.5 6.1 9.0 7.1 10.3 3.2 0.75
Cd0.95Zn0.05Te 293 Local - - - - - - - -
Cd0.95Zn0.05Te 80 Main 142.5 171.8 2.5 5.0 6.9 10.2 3.3 0.8
Cd0.95Zn0.05Te 80 Local - - - - - - - -
Cd0.9Zn0.1Te 293 Main 141.6 175.0 6.4 9.0 7.3 10.2 2.9 0.73
Cd0.9Zn0.1Te 293 Local 167.0 163.0 8.5 9.0 7.1 7.3 0.2 0.60
Cd0.04Zn0.96Te 293 Main 177.0 207.0 5.0 4.0 6.7 9.4 2.7 0.62
Cd0.04Zn0.96Te 293 Gap 154.0 154.3 3.0 4.0 9.4 9.6 0.2 0.70
Cd0.2Zn0.8Te 293 Main 173.0 201.0 9.2 4.3 6.8 9.3 2.6 0.63
Cd0.2Zn0.8Te 293 Gap 152.5 154.0 8.0 6.0 9.3 9.8 0.5 0.70
CdSe0.9Te0.1 293 Main 170.5 211.5 9.0 9.0 6.1 9.4 3.3 0.90
CdSe0.9Te0.1 293 Gap 152.5 154.0 6.6 6.0 9.4 9.6 0.2 0.50
CdSe0.75Te0.25 293 Main 169.0 205.0 12.0 8.5 6.1 8.6 2.5 0.86
CdSe0.75Te0.25 293 Gap 151.5 155.5 10.0 9.0 8.6 9.6 1.0 0.72

at x ¼ 1. In the second class, two-mode crystals, the Raman and IR spectra

exhibit the composition dependent TO and LO phonons of two constitu-
ents (AB-like and AC-like). Several cases have been reported [45, 46]
which show some type of intermediate behavior and such crystals are
difficult to classify as one- or two-mode crystals.
 Optical Phonon Spectra in CdTe Crystals and Ternary Alloys of CdTe Compounds 31

One the most successful models which has been used to describe the
optical behavior in mixed crystals is the random element isodisplacement
(REI) model [44, 45, 47]. The REI model is based on the assumption that in
the long-wavelength limit the anion and cation of like species vibrate with
the same phase and amplitude and that interatomic forces are determined
by a statistical average of the interaction with its neighbors. A modified
REI model [47] is completely defined by the macroscopic parameters of
the pure end members. As distinct from REI models [44, 45], in incorpo-
rates second-neighbor force constant f(B-C) and a linear dependence of all
force constants on the lattice parameter without resorting to microscopic
fitting parameters.
At present there are no clear formal criteria for one- or two-mode
behavior. The conditions given earlier [44] are not valid for ternary alloys
of the CdTe compounds [8, 9]. For example, the two-mode behavior
occurs in the CdTe1
xSex crystal [34] while the conditions [44] predict
an one-mode behavior, since the Se atom mass is greater than the reduced
mass of elements of CdTe (mSe > mCdTe). The CdTe1
xSex reflection spectra
show two Reststrahlen bands [34]. The intensity of the low-frequency
band decreases continuously from a value of pure CdTe (x ¼ 0) to zero
at x ! 1 whereas the intensity of the high-frequency band increases with
the Se concentration. The dependence of optical-mode frequencies on
CdTe1
xSex composition is shown in Fig. 6 and the oscillator strengths
of TO and LO modes in Fig. 7. By analogy with the TO oscillator strength
(Eq. (5)) the oscillator strength of LO modes was found as
 
1 1 gLO 1
SLO ¼
¼ Im
: ð9Þ
e1 e0 oLO eðoLO Þ

210 cm–1 LO1 210

TO1
170 170
LO2

TO2

130 130
0.4 0.8 x
CdTe CdSe

Figure 6 Variation of LO and TO frequencies in Cd1
xSexTe as a function of composi-

tion x. The dotted lines are the fits from a MREI model.
 Another Random Document on
Scribd Without Any Related Topics
 "No, my dear. You're not going yet," said Susan, in
resolute tones. "Nannie's right to bring you in. You're not
going yet. You just take your shawl off your head, and sit
down and tell me what's the matter. And you needn't say
again that nobody cares. Come, child, sit down!"

That conquered Bess. She took the seat indicated and

laying her head on the table, broke into heavy sobs.

"Come, now—come!" repeated Susan. "Don't you be so

upset. Tell me what's wrong, and we'll see if I can't help you
put things right." Then, with a sudden thought, "Is it
anything my Nannie shouldn't hear? I'll send her away, if it
is. She don't know the bad ways of some of you girls; and I
don't choose she should, as long as I can keep her from it."

Bess choked down her sobs, and sat up with heaving

chest.

"I wouldn't tell her neither," she said earnestly. "I'm not
so bad as that. She ain't like us; and I wouldn't be the one
to make her like my sort. I'd sooner learn to be like her—if I
could."

"And of course you can," said Susan encouragingly. "I'd

begin this minute, if I was you."

"Begin to be like Nancy! This minute!" Bess said, in a

wondering tone, as if the idea were a new one.

"To be sure I would," said Susan.

Whatever Susan Dunn did or did not mean in a practical

sense, Bess evidently had some distinct notions on the
subject, for she sat more upright, gazed hard at Nancy,
then walked to a small looking-glass hung over the mantel-
shelf and surveyed herself. Susan and Nancy said nothing
for a few seconds. On the dresser stood a basin full of
water, and Bess' next move was to plunge her rough
unshawled head into this water. It came out dripping; and
she parted the flattened thick mass with her fingers,
pushing it back on either side.

"How they'll jeer!" she burst out then.

"Who will?" asked Susan.

"They! When I'm seen like this."

"The girls that you've made your friends? But you'll

have to leave them, and choose a better sort of friends,"
said Susan.

Bess walked back to her vacated seat.

"Won't you mind Nancy being seen along of me now?"

she demanded.

Susan was rather perplexed. Certain other changes

would be needed beside the possession of a neat head, if
Bess Gardiner was to be counted a fit companion for Nancy
Dunn; yet she could not bear to check the poor ignorant girl
in her first effort to take a right turn.

"What makes you want that so much?" she inquired.

"Because—" Bess' voice suddenly faltered. "Because

there ain't nobody like her—and I—I—I'd do anything in all
the world that ever I could for her—I would!"

Mother and daughter exchanged looks, tears in Nancy's

eyes, and something very like tears in Susan's.
 "If Nannie's to be the one who can help you to what's
right, I'm not the woman to hinder," she said, with a touch
of huskiness. "Seems to me it may be God's will for you
both. But look here, Bess. You've got to make your choice.
You can't do both, you know. If you want to be in and out
here, and to learn from my Nannie, you'll have to leave
your bad companions, and drop your old ways. There's to
be no bad words spoken, and no taking of God's Holy Name
in vain, and no saying of things which a pure-minded girl
shouldn't hear. And you'll have to stop going about in that
sort of dress I saw you in last Sunday. I wouldn't have my
Nannie seen in the street with a girl dressed like that—no,
not for anything you could mention. I'm not speaking
unkindly; but I do mean what I say. You've had your old
friends, and you've gone on in their ways. If Nannie's to be
your friend now, you'll have to take to new ways."

Bess's low brow was frowning anxiously.

"I'd do anything," she said; "anything you'll tell me. And

I mean it. I told mother I'd go right off and board with them
Joneses, because father did storm at me so. And I won't
now."

"No," Susan said, in decisive tones. "It's the Joneses or

Nancy. Not both."

Bess shook her head.

"Couldn't be both," she assented. "But my! Won't they

be at me!"

"It won't be an easy time for you," said Susan gravely,

"It never is, when a girl tries to leave evil and turn to good.
And mind, Bess, it isn't only just a leaving off of one thing
or another that you've got to think of. That's not enough.
You'll have to be whole and thorough—give yourself up to
serve God, and do His will. For if you haven't His Power to
keep you straight, nothing else 'll be of much use."

"Parson said so too," Bess remarked briefly. "He

stopped me one day, and I jeered, and he had his say out
and never minded. I've thought of it a many times since."

"What did he tell you, Bess?" asked Nancy.

Bess was in difficulties. She evidently retained no clear

recollection of the words spoken. Yet, as evidently, a distinct
impression had been made. By dint of questioning, Susan
came a vague remembrance of "something about God
caring."

"And he said I just hadn't ought to go on a-troubling of

Him with my bad ways," added Bess. "Him as was nailed up
on the Cross. I didn't know nothing about it troubling Him
before. And I thought—maybe—Nancy 'ud tell me what to
do."

CHAPTER XX.
TROUBLE.

"THERE's something wrong with Mr. Wilmot!" Other

people, besides Mrs. Mason, were saying this as the
summer went on.
 Annie was slower to perceive the alteration in him than
were many. For one thing, he did his best to keep up in her
presence, fearing to awake his darling's anxiety. For
another, she was young still, and had seen little of illness.
Moreover, she was extremely busy in the Parish, and was by
no means given to conjuring up troubles.

Conjuring was, however, in this matter no longer

needed, for a very real trouble lay already to hand.

The change in Mr. Wilmot had become patent to all who

knew him. A laboured and languid gait replaced the old
brisk walk; a fixed perpetual pallor replaced the old healthy
sunburn. If he had to ascend a little slope, he stood still
often to pant for breath. The exertion of preaching would
bring visible drops of moisture to his brow; and not seldom
the once clear and ringing tones were inaudible to half his
congregation.

Yet with this appearance of weakness, there existed an

unusual brightness, and this it was chiefly which helped to
blind Annie's eyes.

For a while, in the earlier part of the summer, she and

many others had thought him unwontedly grave and
depressed. The gravity and depression were gone now,
utterly. Never had his eyes shone with so calm a light,
never had his smile been so full of sunshine. There were
some who noted in his look and bearing a strange
unearthliness—noted it with mingled awe and fear. Yet they
could not have told wherein it consisted; for even while they
noted it, and thought him worn and altered, his laugh would
break out in all its old gaiety, as he paused to speak to
some little child. And how the children loved him!
 Annie's eyes remained long strangely shut. She thought
him tired unusually often, but the hot summer seemed to
account for this. By-and-by, he would take his autumn
holiday, and that would set all right.

But there came a day of awakening—sudden and

unexpected.

She had had her Bible class as usual one Sunday

afternoon, with the half-dozen girls who regularly came to
the Rectory for that purpose. It was a very interesting hour
commonly to them all; not least so to Nancy Dunn, who by
this time loved Miss Wilmot dearly. This day's lesson proved
certainly not less interesting than usual.

Annie had chosen the subject of trouble, and of how to

bear trouble. She had talked it over with her father
beforehand, and she had much to say about the bright side
of trouble, the often good effects of it, and the spirit in
which it should be borne.

"I dare say some of you remember that sermon of my

father's about being always ready," Annie said, in the
course of the class.

Nancy smiled a response.

"Perhaps we haven't any of us just now any great

troubles to bear—only just little every-day ones. But the
great troubles may come at any time; and when they do,
we ought to be ready. I suppose there is only one way of
being really ready, and that is,—" Annie went on reverently
—"that is, living always very close to Jesus. For if we are
fighting close to His side, and under His banner, then
whatever He orders we shall be ready to do—or whatever
He gives us we shall be ready to take.
 "I don't mean," she added, after a pause, "that one
wouldn't feel trouble. My father says that is a mistake.
When God sends trouble, He means it to be trouble. And He
means it to bring us near to Him, that He may comfort us. I
don't suppose He can comfort us till we are 'ready' to have
whatever He sends. Some don't learn to be ready till after
the trouble comes. But I should like to learn beforehand—
shouldn't you? I should like to be able, when it comes, to
look straight up, and say, 'Thy will, not mine, be done.'
That's what the Lord Jesus could do, even in the midst of
His great terrible struggle in Gethsemane. He could say,
'Not My will, but Thine.' For all the while He was ready—
truly ready. He could always say from His heart, 'Lo, I come
to do Thy will, O God!' And that is what we have to learn to
say."

It seemed strange afterwards to Annie herself, as well

as to those who listened, that she should have spoken just
on that particular day in this particular manner.

The class ended, Annie put away her Bible, set the
chairs straight, and went out into the garden, singing softly.
She believed her father to be at the school, where he
usually went every Sunday afternoon. A favourite Homer
rose bush near the garden-gate drew Annie's attention. She
thought she would gather a few buds, and put them on
father's study table, to refresh his eyes when he returned.
He had not seemed at all well that morning; and Annie had
almost made up her mind to ask Mr. Rawdon privately some
day soon whether Mr. Wilmot ought not to take a tonic.

Six pretty pinkish half-open blossoms were in hand,

when a movement near made her look up. "Good
afternoon," she said, smiling at the sight of Archie Stuart.
Mrs. Stuart's cottage was in Annie's district, and Annie knew
the mother and son well.
 "How is Mrs. Stuart?" she asked, as Archie seemed to
hesitate outside the gate.

Archie's mother was "pretty well," he said.

"Her foot quite recovered?" Annie asked, plucking

another rose bud. "Do you think she would like two or three
of these? Come in, and I'll give them to you."

Archie was much pleased. He stepped inside, letting the

gate swing to.

"There!" Annie said, handing him a small bunch. "Tell

Mrs. Stuart I am coming very soon to see her again." Then,
with another smile, Annie inquired, "Has she begun to like
Nancy Dunn yet?"

Archie's face fell. "No, Miss," he said. "Not as anybody

could help liking Nancy—"

"No, indeed!" put in Annie.

"But she won't hear one word of me and Nancy having

things settled between us," pursued Archie.

"Then you have to be patient a little longer," said Annie.

Between confidences from Archie, from Mrs. Stuart, and
from Nancy, she knew pretty well all about the matter.
"Nobody is the worse for a little waiting, and I think you are
pretty sure about Nancy—are you not?"

"Well, I did speak to her, Miss Wilmot, and to Mrs. Dunn

too," admitted Archie. "For I didn't seem as if I was able to
keep in any longer. And my mother she was in a great
taking. But Mrs. Dunn said, and Nancy said too, that it
wasn't to be anything settled until mother was willing. And
it does seem as if she never was going to be willing."
 "You have waited a very short time yet, and you and
Nancy are both young," said Annie. "I think you must have
patience still; and your mother has been a good mother,
hasn't she? I always notice how very fond she is of you. I
am sure you must want to be a very good son to her."

"Well, yes—that I do, Miss," assented Archie, though

perhaps not so heartily as Annie wished, for his head was
full of Nancy. Then he inquired: "Is Mr. Wilmot better,
please? Mother said I was to be sure and ask. She does set
store by Mr. Wilmot, and no mistake, and it's worried her to
see him so ill lately."

"My father! Why, he has not been ill," said Annie in

surprise. "He seemed tired this morning."

Archie looked at Annie somewhat strangely. He had

heard many remarks lately on the Rectors altered look.

"I am expecting him home from Sunday school directly,"

pursued Annie. "So I must go indoors now and be ready for
him. Good-bye. I do hope your mother will soon give way
about Nancy. But you have to be patient, haven't you, till
then?"

Annie went back into the house, wondering uneasily

what could have made Archie speak so of her father. She
would certainly got hold of Mr. Rawdon as soon as possible,
and beg him to see after Mr. Wilmot.

The study door was shut; but Annie, believing her

father to be out, went straight in, meaning to put the rose
buds on his study table. Her light and quick approach was
unheralded. The door gave no warning creak; and Annie
had a noiseless manner of turning handles.
 The room proved to be tenanted. Mr. Wilmot lay on the
sofa, and beside him sat Mr. Rawdon, bending somewhat
forward, and speaking in distinct tones—

"As for your fear of hydrophobia—"

These words struck upon Annie's ears; and no trumpet-

clang could have rung out with more startling clearness.

In a moment, the two were aware of her presence, and

Mr. Rawdon stopped short.

CHAPTER XXI.
THE GREAT DREAD!

"WHY, Annie, my child!" the Rector said, raising himself

on one elbow, and greeting her with a smile, in which some
veiled anxiety might have been detected. "We did not hear
you coming. Are those roses for me? How has the class
gone to-day?"

Annie could not answer him. She hardly knew what it

was that restrained her to some appearance of calm. That
terrible word "Hydrophobia" seemed to be ringing still
through the room, and with it an awful dread had rushed in
upon her. Yet she only stood motionless, holding the back of
a chair with one hand and clutching the rose buds fast in
the other.
 "I did not go to the school this afternoon," pursued Mr.
Wilmot. "The truth is, I was not feeling quite—well, not
quite as I should; so I took an hour's rest instead."

Annie seemed to hear herself speaking in a hollow

distant tone—"And you sent for Mr. Rawdon?"

"No, I came without being sent for," said Mr. Rawdon.

He uttered the words in a deliberate repressive manner, as
if to impress upon Annie a need for caution. "I was not
satisfied with your father's look in Church this morning."

Then a pause. Annie remained perfectly still, her eyes

fixed on vacancy. All colour had fled from her face. The two
gentlemen exchanged meaning glances.

"Come, my child—sit down here, and tell me about your

class," said Mr. Wilmot.

He made a little space on the sofa, and Annie went to it

mechanically, but not to talk. In another moment, she had
buried her face on his shoulder, and was clinging to him in a
passionate wordless agony.

For two or three minutes there was silence, which

nobody liked to break. The brisk ticking of the clock
sounded clearly; but to Annie that sound was lost in the
rapid fluctuating throb of something nearer at hand,
something fast yet broken. She could not see her father's
face, but Mr. Rawdon could, and his hand took firm hold of
Annie's arm.

"Miss Wilmot! Stop this, please. You must get up."

There was a sound of warning in the tone. Annie

obeyed, still as one in a maze. It seemed to her that the
whole world was suddenly changed with this new fearful
dread. Then she saw her father's ghastly look, and she
knew that something must be very wrong. He could not
speak—could not breathe. Mr. Rawdon had moved away to
the table, to pour out some liquid into a small glass, and
now he was administering it, holding up Mr. Wilmot. The
breathless struggle lessened slowly; and then a fresh fear
came over Annie. Had she done him harm? With that
thought she was herself again.

"It is going off now—not much this time, I hope," Mr.

Rawdon said cheerfully. "Keep still, Wilmot. Don't try to talk
yet. Miss Wilmot and I will arrange about the evening."

Annie could only look at the doctor imploringly, and Mr.

Wilmot's hand drew her back to her former position—except
that she sat upright, not leaning against him.

"Now," Mr. Rawdon said, with a glance at Annie and a

movement as if to leave the room. But Annie could not stir.
The restraining arm around her might be weak, yet it held
her fast. Mr. Rawdon had taken one step towards the door,
and he paused hesitatingly.

"The fact is, Wilmot, any kind of agitation is bad for

you, in the present state of your heart," he said. "Miss
Wilmot and I are used to each other. Better let us have our
little talk in another room."

"No," said Mr. Wilmot gently. "Here, please. It will

distress me less. Annie must be good and calm."

Mr. Rawdon took a chair, by no means with the air of a

man convinced.

Then another pause. Mr. Wilmot's eyes were on his child

lovingly; and Annie could be seen to draw one or two deep
breaths, as if mastering herself with difficulty.
 "Yes," she said at length. "Please tell me."

"I will see some one, and arrange for the service this
evening." Mr. Rawdon spoke deliberately. "I have already
warned your father that he must consent to do less work.

"I am ready," Mr. Wilmot said, in a quiet voice.

"The fact is, Miss Wilmot—the fact is, your father has
had lately a severe return of certain troublesome heart
symptoms, from which he suffered a good deal about two
years ago. You will probably remember."

"Yes—I remember," Annie found herself saying. "He had

to take a long holiday—and—But—but that is not—not—"

"It is essential that he should greatly lessen work now. I

tell you both honestly, I don't like these attacks; and this is
the third, I believe, within a fortnight."

Annie looked bewildered, even while scarcely taking in

the full sense of his words. She could only feel that
something more terrible lay behind, something not yet
touched upon. But for her father's sake, she dared not ask
questions, till she should see Mr. Rawdon alone.

Doubtless, Mr. Rawdon too purposed putting off until

then the needful telling. Mr. Wilmot was of a different mind.

"Annie, my child, did you happen to hear what Mr.

Rawdon and I were talking about when you came in?"

"Yes, father," she whispered.

"That comes of stealing in upon folks without warning,"

said Mr. Rawdon.
 "What did you hear?" Mr. Wilmot asked.

Annie hid her face in her hands.

"Wilmot, I don't like this for you," the doctor said.

"I must risk it. What did you hear, my darling?"

She lifted her face, and said in a voice quiet as his own:

"Father, Mr. Rawdon will tell me, please—not you." She

grew paler: turning to the doctor—"Was father—? Did the
dog—?"

"No, not bitten; but he had a touch from the creature's

tooth. A mere scratch," said Mr. Rawdon. "Of course the
slightest scratch should be avoided. He came to me,
however, at once, and I burnt the place out—burnt deeply. I
believe he managed to conceal from you that anything was
the matter with his wrist."

"Wrist!" Annie repeated the word, and Mr. Wilmot drew

up his sleeve.

"That is the scar of the burn," said Mr. Rawdon. "The

scratch itself was a mere nothing. The wound healed slowly
but thoroughly, as you may see. My own belief is that the
remedy was prompt and complete enough to ensure
safety."

But Annie knew that these words did not imply

certainty. She sat silent once more, hardly thinking, but
rather weighed down by a dull pressure of misery.

"And I was never told!" she murmured at length.

 "There was no need," said Mr. Rawdon. "The less said
and thought about the matter, the better. Now I must be
off, Wilmot, to arrange for your evening's work being done
by somebody else. You may leave that in my hands. Keep
very quiet, and don't exert yourself. I shall see you early to-
morrow."

He said good-bye to Mr. Wilmot, but not to Annie, and

turned away. Annie knew that she was to accompany him
out of the room, and she stood up, her father's arm relaxing
to set her free.

He said only, "Come back to me, Annie."

 CHAPTER XXII.
LIFE LESSONS.

"NOW, remember," said Mr. Rawdon authoritatively,

having stepped with Annie into the drawing-room, where he
stood pulling on his gloves—"remember, Miss Wilmot, the
less you dwell on that thought, and the less you allow your
father to dwell on it, the better for him."

"But how can I help—?" sobbed the poor girl, for the
moment entirely overcome.

"You must help it. Self-control in this matter is essential

for your father's sake. It is not merely a question of talking.
He reads every turn of your face, and if he sees you unlike
yourself, sad and unhappy, you will act as a perpetual
reminder of that which he ought to forget as much as
possible."

"I will try hard—indeed I will," said Annie brokenly. "But

if—if—"

"No; you are not to indulge in that 'if.' Understand me,

Miss Wilmot. Your father is not suffering in the remotest
degree from any premonitory symptoms of hydrophobia."

"You are quite—quite sure?"

"Perfectly sure. There is not a sign of anything of the

kind about him. Some weeks ago I confess I did feel
anxious for a time. He was under great depression, and
living in a constant expectation of ill results. You must have
remarked his depression. That has all passed off now. I
cannot say he has entirely lost the expectation—perhaps I
should rather say the distinct sense of what might come.
But it is not depression, and it is not fear. I was wrong to
use that word. He faces the matter in a wonderfully manly
and Christian spirit. I wish he could banish the subject from
his mind; but no doubt the present state of his health acts
upon him, and lessens the power of self-restraint."

"His heart?" Annie strove to say.

"Yes—the mischief is there." Mr. Rawdon spoke in a

grave tone. "I was not satisfied two years ago—but he
seemed so far to rally from the weakness, that one had
almost ceased to recall it. No doubt there has been mischief
long brewing, which must sooner or later have declared
itself. The strain and agitation of this summer have only
hastened matters."

"But he will be better—he will get stronger by-and-by,"

said Annie imploringly. "When this dreadful year is over, and
we are sure—"

"Yes, I hope so." Mr. Rawdon's voice was still more

grave. "We must check his doing too much."

"If he were to get away for change? Could he not take

his holiday sooner?"

"That has been discussed already. It is a difficult

question," Mr. Rawdon said thoughtfully. "The fact is, I don't
like his going far with only you—and he seems scarcely in a
state for much travelling. If change could mean full
occupation of mind—but too much leisure for thought is not
at all desirable. Perhaps a moderate amount of work is
better at present. But we shall see. You must try, for his
sake, to take a cheerful view of things, and do your best to
keep up his spirits. Good-bye now. I will look in to-morrow.
But mind, he is not to count himself a regular invalid."

"No," said Annie.

She found it hard to respond, hard to lift her eyes—the

trouble which had come upon her seemed so very terrible.
She dreaded going to the study to meet her father's look.
When Mr. Rawdon was gone, she turned mechanically into
the dining-room, and stood there in an attitude of hopeless
despondency.

Only half-an-hour or so earlier she had sat just here, a

light-hearted girl still, speaking to other light-hearted girls
of troubles that might one day come, and how they should
be borne. What had she known then of trouble?

Yet her words had been true, and she knew it. But she
could not feel or see their truth now. She could only bow
her head beneath the blow.

"I was dumb, I opened not my mouth: because Thou

didst it."

But she could not reach beyond "dumbness." She could

not look up and say, "It is well."

After all, was there any need—as yet? The blow had
only just fallen: and He who sent it knew its weight, knew
her weakness. Annie had only just entered the School of
Sorrow, and He who called her into it could pity her faltering
steps with all a mother's tenderness.

She had to go back to her father. That recollection came

soon, and Annie yielded to its call. Leaving the dining-room,
where she had stood alone with clasped hands and drooping
head, she crept thither.
 And she had to look bright, to seem cheerful, to wear a
face of calm unconsciousness! How could she, with this
weight upon her heart?

"I have been looking for you," her father said. "Come
here, my child."

Annie did as he told her. She knelt down beside the

couch, and laid her face against him.

"That is the right attitude for both of us, isn't it?" said
Mr. Wilmot softly.

"Father—" Annie tried to say, hardly knowing what she

meant to utter. But the broken word was taken up in quiet
accents—

"Father, Thy will, not ours, be done."

Annie shivered; and he spoke again—

"'The King's servants said unto the King, "Behold, Thy

servants are ready to do whatsoever my Lord the King shall
appoint."'"

Another little break.

"It is not for us to choose, nor for us to resist. He has

His own mighty and loving purposes. We have but to be
ready—ready to do—ready to bear—'whatsoever' Christ our
King may appoint."

"But we may pray—pray—" she sobbed.

"Yes, pray and plead as earnestly as you will; only in

the spirit always of Christ's prayer—'Thy will—not mine.'"
 "O father, I can't say that, and I thought it would be so
easy if trouble came—only not this trouble."

"Poor little Annie! My poor child! Yes, it is always so with

us, 'only not this!' But He understands and pities. No pity
was ever like His pity. He will teach you in His own good
time. He knows how, for He has gone through all the worst
of it Himself—worse agony than any of us can ever have to
bear. And it is enough meanwhile to sit at His feet, to hear
His voice. No more blessed position than that! He is so
merciful. He doesn't hurry us, like man, in the lesson
learning."

Mr. Wilmot spoke slowly, in brief sentences. "No, I am

not hurting myself. But I can speak from experience, Annie.
I have learnt much this summer—much of His exceeding
gentleness. Where He lays His hand most heavily, He brings
the sweetest balm."

"If only you had told me, father! To bear it alone!"

"Alone! I had my Master's presence."

Annie looked up, but she could not face his smile. Her
head sank anew.

"There was the battling for awhile—not easy, but close

to His side. I seem to have reached beyond the battling
lately—to a quiet spot. One of His green pastures, I
suppose. He gives rest when it is needful. But my child need
not go through all that I went through. It is not necessary.
That dread is over now. Mr. Rawdon was mistaken. I have
no fear."

"He said so," she whispered.