Bleaching

An optimization approach for the oxygen

bleaching of kraft wheat straw cellulose
Laith A. Abdul-Karim and Attila Rab

reaction and an undesirable carbo-

ABSTRACT: The objective of this work was to determine hydrate degradation. Any complete
optimum conditions for oxygen bleaching of kraft wheat straw kinetic study of oxygen bleaching
must therefore include both reac-
cellulose using designed experiments and statistical techniques. tions.
The three process variables for the delignification of kraft pulps Earlier investigators (3) studied
examined were time of delignification, added alkali oxygen bleaching at an extremely
low pulp consistency of 0.2%, so
concentration, and remaining alkali concentration. There was change in the hydroxide concentra-
good fit between calculated and experimental values for tion would be negligible. There were
bleached pulp yields using polynomial models. There are no general kinetic equations pro-
vided in the study. Indeed there are
optimum processing conditions of oxygen bleaching to produce no studies for the kinetics of carbo-
suitable wheat straw cellulose. hydrate degradation during oxygen
bleaching in published literature.
KEYWORDS: Alkali pulping, cellulose, experimental design, The aim of this investigation was
optimization, oxygen bleaching, variables, wheat straw. to determine the optimum bleaching
conditions for both the delignification
and carbohydrate degradation dur-
ing oxygen bleaching. Experiments

T here is considerable interest in

Hungary to improve technology
for production of kraft quality
ing begins at pH above 13 and fin-
ishes at pH approximately 10. To
avoid severe carbohydrate degrada-
for optimization of oxygen bleach-
ing used constant bleaching condi-
tions for each run with investigation
bleached chemical pulps derived tion, the bleaching ends after re- of different levels of the variables in
from wheat straw. There is special moval of approximately 50% of the separate experiments. The variables
concern with processes that give lignin (1). under study were reaction time,
minimum amounts of organochlorine To optimize and control the oxy- added alkali concentration, and re-
compounds in the effluents. gen bleaching stage fully, informa- maining alkali. The remaining alkali
Oxygen bleaching is a commer- tion on oxygen bleaching kinetics is is the quantity within the pulp after
cial process for the removal of lignin necessary. A mathematical kinetic cooking and washing.
from kraft pulp. The main benefit of model combined with the stoichiom-
oxygen bleaching is the high reten- etry of the process, i.e., amount of
tion of lignin within the mill. Other bleaching chemicals required to Experimental
bleaching techniques can form vari- reach a given kappa number, can
ous pollutants. This is especially true describe and control industrial What straw pulp characteristics
of chlorination. bleaching stages (2). Before undertaking the bleaching
Conventional oxygen bleaching Two competing reactions occur si- experiments, the study determined
uses medium consistency in the pres- multaneously during oxygen bleach- the approximate analysis of the pulp.
ence of sodium hydroxide. Bleach- ing. They are a delignification Table I shows the results.

Abdul-Karim is research scientist and Rab is head of the fiber technical department
at Paper Research Institute Ltd., Duna U 57, PF 86, Budapest, H–1215, Hungary.

Vol. 78, No. 1 Tappi Journal 83

 Bleaching
I. Characteristics of wheat straw pulp (4-9) II. Coded and experimental values of process variables

Property Value Coded Remaining alkali, Added alkali Time,

level mg Na2O/g pulp charge, % min
Kappa number 17.4
Brightness, % 29.0 -1.682 0.00 0.00 45
Freeness, °SR 31.0 -1.0 6.25 0.65 56
Tear index, mNm2/g 5.51 0.0 12.5 1.30 67
Tensile index, Nm/g 73.5 +1.0 18.7 1.95 78
Burst index, kPa·m2/g 5.55 +1.682 25.0 2.60 90

Bleaching process III. Central composite rotatable design

The following was the sequence of
steps to bleach the pulp: Experiment Coded variable Actual value
number X1 X2 X3 RA AA t
• The pulp experienced bleaching
in an electrically heated, stainless 1 -1 -1 -1 6.25 0.65 56
steel, rotatable digester. The 2 +1 -1 -1 18.75 0.65 56
charge consisted of 50 g moisture 3 -1 +1 -1 6.25 1.95 56
4 +1 +1 -1 18.75 1.95 56
free pulp. The chemicals under-
5 -1 -1 +1 6.25 0.65 78
went mixing to reach the calcu- 6 +1 -1 +1 18.75 0.65 78
lated charge of added alkali. 7 +1 +1 +1 18.75 1.95 78
Chemical addition to pulp was as 8 -1.682 0 0 0.00 1.30 67
an aqueous solution. Pulp to li- 9 +1.682 0 0 25.00 1.30 67
10 0 -1.682 0 12.50 0.00 67
quor ratio was 12% consistency.
11 0 +1.682 0 12.50 2.60 67
Magnesium sulfate charge was 12 0 0 -1.682 12.50 1.30 45
directly to the pulp at a level of 13 0 0 +1.682 12.50 1.30 90
0.5%. Then the pulp and chemi- 14 0 0 0 12.50 1.30 67
cals were placed in the digester, 15 0 0 0 12.50 1.30 67
the lid was closed, the tempera-
ture set to 95°C, and the electrical
heater and motor started. The di-
gester pressure was at the indi- the variables for the experimental The experimental design required
cated oxygen pressure. design were as follows: the 15 experiments conducted in the
sequence shown in Table III. Ex-
X1 = (RA – 12.5)/6.25 (1)
• The next steps were to reduce the perimental results took the form of
digester pressure to atmospheric X2 = (AA – 1.3)/0.65 (2) the mathematical relationship given
level, cool the autoclave, collect by the following equation:
the pulp on a muslin cloth, and X3 = (t – 67.5)/11 (3) Y = bo + biXi + biiXi2 + bijXiXj + . . . (4)
wash the pulp for 2 h with tap
water. where In this equation, Y can be bleached
RA = remaining alkali, mg Na2O/g yield, kappa number, freeness,
Experimental design and pulp brightness, tear index, tensile index,
nonlinear regression or burst index. Xi and Xj are the pro-
AA = added alkali charge, % NaOH in
the moisture free pulp cess variables under investigation,
The study investigated remaining
and the b values are constants com-
alkali concentration as Na2O, added t = duration of delignification, min. puted by the method of least squares.
alkali (as NaOH), and the duration
of bleaching. The ratio of the pulp to Table II shows the values of the vari-
the liquor, temperature of bleach- ables and their codes. The number of Results and discussions
ing, and pressure of oxygen remained different levels for each variable was
constant. always five despite the number of Regression analysis
To optimize the above parameters, variables. A preliminary step was Tables IV and V summarize the re-
the work used a central composite setting each coded level to the corre- sults of oxygen bleaching of wheat
rotatable design (10). The range of sponding Xj real process variable straw pulp at oxygen pressures of
where j = 1, 2, and 3. The calcula- 0.6 mPa and 0.4 mPa, respectively.
tions used Eqs. 1–3.
84 January 1995 Tappi Journal
 IV. Coded variables, actual values, and corresponding fiber characteristics for oxygen bleaching of wheat straw kraft pulp at oxygen
pressure of 0.6 mPa

Experi- Free- Tear Burst Breaking Bright-

ment Yield, Kappa ness, index, index, index, ness, Double
number X1 X2 X3 RA AA t % number °SR mNm2/g kPa·m2/g Nm/g % fold, no.

1 -1.228 -1 -1 4.15 0.65 56 94.75 8.73 24 5.38 4.59 69.77 35 423

2 -0.560 -1 -1 9 0.65 56 90.82 8.23 26 4.92 5.21 80.95 36 227

3 -1.228 +1 -1 4.15 1.95 56 90.74 7.02 23 5.74 4.66 70.49 39 389

4 -0.560 +1 -1 9 1.95 56 92.73 8.52 25 6.29 5.31 80.84 38 163

5 -1.228 -1 +1 4.15 0.65 78 93.89 10.9 23 5.55 4.82 72.66 35 295

6 -0.560 -1 +1 9 0.65 78 89.63 7.55 25 4.70 5.25 77.32 37 230

7 -0.560 +1 +1 9 1.95 78 87.38 7.00 25 5.51 5.68 73.53 40 182

8 0.312 -1.682 0 14.45 0 67 87.25 7.54 24 5.14 4.82 65.92 40 327

9 0.312 +1.682 0 14.45 2.60 67 85.42 6.37 24 5.42 4.96 66.01 39 322

10 0.312 0 -1.682 14.45 1.30 45 87.02 11.9 24 6.56 4.82 67.92 39 305

11 0.312 0 +1.682 14.45 1.30 60 82.12 7.30 26 5.49 4.59 60.94 33 211

12 0.312 0 0 14.45 1.30 67 88.12 6.39 24 6.16 4.70 74.80 44 613

13 0.312 0 0 14.45 1.30 67 62.51 6.39 24 6.93 4.67 72.51 32 493

V. Coded and real variables and corresponding fiber characteristics for the oxygen bleaching of wheat straw kraft pulp at oxygen pressure of 0.4 mPa

Experi- Free- Tear Burst Breaking Bright-

ment Yield, Kappa ness, index, index, index, ness, Double
number X1 X2 X3 RA AA t % number °SR mNm2/g kPa·m2/g Nm/g % fold, no.

1 -1.228 -1 -1 4.15 0.65 56 94.07 8.73 25 5.38 4.52 56.85 35 300

2 -0.560 -1 -1 9 0.65 56 92.45 7.82 27 4.94 5.28 73.97 35 119

3 -1.228 +1 -1 4.15 1.95 56 89.74 7.42 24 6.58 4.65 64.34 37 569

4 -0.560 +1 -1 9 1.95 56 93.44 7.72 26 4.50 6.31 80.64 38 153

5 -1.228 -1 +1 4.16 0.65 78 95.04 8.92 24 6.35 4.74 65.97 35 547

6 -0.560 -1 +1 9 0.65 78 92.56 8.23 27 7.48 5.22 87.55 35 114

7 -0.560 +1 +1 9 1.95 78 90.78 7.90 28 6.71 4.67 74.61 42 196

8 0.542 -1.682 0 15.89 0 67 84.25 10.7 23 6.09 4.72 69.57 33 370

9 0.542 +1.682 0 15.89 2.60 67 83.42 7.44 24 6.67 5.09 71.73 39 318

10 0.542 0 -1.652 15.89 1.30 45 89.32 8.06 24 5.78 4.61 67.35 36 297

11 0.542 0 +1.682 15.89 1.30 90 61.22 8.63 25 6.87 4.92 67.65 36 436

12 0.542 0 0 15.89 1.30 67 82.32 8.39 28 6.26 4.83 67.22 36 420

13 0.542 0 0 15.89 1.30 67 82.62 7.89 24 6.43 4.77 69.81 37 472

Vol. 78, No. 1 Tappi Journal 85

 Bleaching
VI. Results of regression analysis at oxygen pressure of 0.6 mPa

Bleached yield = Y = 85.180 - 6.866X1 - 0.579X 2 - 1.456X3 + 0.951X1X2 - 0.388X1X3 -0.682X2X3 -1.291X1 2 + 1.212X22 + 0.620X32

Kappa number = K = 6.479 + 0.210X1 - 0.316X2 - 1.018X3 + 0.328X1X2 - 1.460X1X3 - 0.161X2X3 + 0.974X12 + 0.111X22 + 1.046X32

Brightness = Br = 42.658 + 2.271X1 + 0.283X2 - 0.991X3 - 2.025X1X2 - 1.840X1X3 + 0.557X2X3 + 1.367X1 2 - 1.413X22 - 2.650X32

Freeness = Fr = 24.810 - 1.723X1 - 0.089X2 + 0.320X3 - 0.289X1 X2 + 0.705X1X3 + 0.151X2X3 -2.802X12 -0.00000006X22 + 0.453X32

Tensile index = Tn = 80.41 - 15.969X1 - 0.330X2 - 2.147X3 + 1.135X1X2 - 0.764X1X3 - 1.499X2X3 -16.186X12 - 2.788X2 2 - 3.331X32

Burst index = Bu = 5.063 - 0.919X1 - 0.070X2 - 0.007X3 - 0.071X1X2 - 0.191X1X3 + 0.096X2 X3 - 0.942X12 + 0.072X22 + 0.007X32

Tear index = Tr = 5.789 + 0.625X1 + 0.205X2 - 0.274X3 - 0.126X1X2 - 0.259X1X3 - 0.056X2X3 + 0.616X1 2 - 0.270X22 - 0.007X32

VII. Results of regression analysis at oxygen pressure of 0.4 mPa

Bleached yield = Y = 87.824 - 7.664X1 - 0.496X2 - 1.506X3 + 0.839X1X2 - 1.706X1X3 - 0.267X2 X3 - 4.085X12 + 0.482X22 + 0.989X32

Kappa number = K = 7.606 - 0.579X1 - 0.666X 2 + 0.190X3 - 0.389X1X2 - 0.083X1X3 + 0.098X2X3 + 0.748X12 + 0.328X22 + 0.072X32

Brightness = Br = 37.403 - 1.242X1 + 2.055X2 + 0.460X3 - 0.389X1X2 - 0.804X1X3 + 1.196X2X3 -0.784X1 2 - 0.176X22 - 0.176X32

Freeness = Fr = 25.439 - 1.780X1 - 0.364X2 - 0.238X3 + 1.084X1X2 + 1.017X1X3 - 0.886X2X3 -3.318X12 -0.000002X22 + 0.354X32

Tensile index = Tn = 79.988 - 9.760X1 - 0.118X2 + 0.522X3 + 0.982X1X2 - 0.205X1X3 - 4.741X2X3 -21.046X12 + 0.754X22 - 0.358X32

Burst index = Bu = 5.157 - 0.269X1 + 0.006X2 - 0.009X3 + 0.181X1X2 + 0.154X1X3 - 0.212X2X3 -0.720X12 + 0.037X2 2 - 0.012X32

Tear index = Tr = 5.729 + 0.688X1 - 0.009X2 + 0.465X3 + 0.059X1X2 - 0.006X1X3 - 0.308X2X3 + 0.823X12 + 0.012X2 2 - 0.007X32

VIII. Calculated and experimental optimum oxygen bleaching conditions and product characteristics at oxygen pressures of 0.6 and 0.4 mPa

Free- Tear Burst Breaking Bright-

Condi- Yield, Kappa ness, index, index, index, ness,
tion X1 X2 X3 RA AA t % number °SR mNm2/g kPa·m2/g Nm/g %

A -1.498 -1.682 1.682 1.682 0 90 99.33 15.40 21 5.15 4.46 56.69 26

B -0.932 -1.682 1.682 6.675 0 90 99.63 9.534 28 11.05 5.46 88.25 31

C -1.149 -1.682 1.682 4.88 0 90 99.24 10.96 27 7.60 6.67 68.2 33

D -1.149 -1.682 1.682 4.88 0 90 96.01 11.27 27 6.68 5.27 79.5 33

A = Optimum conditions using 0.6 mPa oxygen pressure

B = Optimum conditions using 0.4 mPa oxygen pressure
C = Calculated results using optimum values
D = Experimental results using optimum values

86 January 1995 Tappi Journal

 1. The effect of remaining alkali upon bleached pulp yield The tables show bleached yield,
kappa number, freeness, tear index,
100
burst index, breaking index, bright-
BLEACHED PULP YIELD, %

ness, and double fold as a function of

90
the process controlling parameters.
80
The computer-assisted second-or-
der polynomial analysis of the pri-
70 AA = 0 mary function under investigation
t = 90 min (bleached pulp yield) and the other
= 0.6 mPa oxygen
60
= 0.4 mPa oxygen
functions (kappa number, bright-
ness, etc.) led to the equations pro-
50 vided in Tables VI and VII
0 4.00 6.25 8.00 10.00 12.50 14.00 16.00 18.75 20.00 22.00 24.00 25.00 27.00
respectively. The F-values of these
RA, mg Na2O/g pulp equations exceed the tabulated F-
value of 2.2 determined at 10 de-
grees of freedom and are within 95%
confidence limits (11). The estimated
standard deviation values were less
2. The effect of added alkali charge upon bleached pulp yield than one. This means the equations
in Tables VI and VII adequately de-
100 scribe the process behavior through
BLEACHED PULP YIELD, %

the region studied.

95 0.4 MPa O2 pressure
Process optimization
90 0.6 MPa O2 pressure Based on the regression analysis de-
scribed above, the process optimiza-
RA = 1.68%, t = 90 min
tion involves maximizing the
85
RA = 6.67%, t = 90 min bleached yield as a function of the
parameters such as kappa number,
80
0 0.40 0.65 0.80 1.00 1.30 1.40 1.60 1.95 2.00 2.40 2.50 2.60 2.70 brightness, etc.
Investigation of the optimization
AA, % problem was undertaken using
Rosenbrock’s method (12) to deter-
mine the optimum process control-
ling parameters. A computer
3. The effect of bleaching time upon bleached pulp yield program generated the solutions to
the optimization problems using re-
100 gression equations. Table VIII pre-
BLEACHED PULP YIELD, %

0.6 MPa O2 pressure sents the results that define the

95
optimum pulping conditions. Accord-
0.4 MPa O2 pressure ing to these results, bleaching using
conditions designated as B in the
90 table provides the best results for
oxygen bleaching of wheat straw
85 RA = 1.68%, AA = 0% pulp. An experimental trial later con-
RA = 6.67%, AA = 0%
firmed the calculated data. As the
data in the table indicates, there was
80
45 50 56 60 67 70 75 78 80 85 90 100 satisfactory agreement between the
t, min calculated and experimental opti-
mum values.

Process variable effect on

bleached pulp characteristics
Figures 1–6 show the effects of re-
maining alkali concentration, concen-

Vol. 78, No. 1 Tappi Journal 87

 Bleaching
tration of added alkali, and duration 4. The effect of remaining alkali upon kappa number
of bleaching on oxygen bleached pulp
yield and kappa number using dif- 20
ferent oxygen partial pressures for AA = 0%
kraft wheat straw oxygen

KAPPA NUMBER
15 t = 90 min
delignification. The bleached pulp = 0.6 mPa oxygen
= 0.4 mPa oxygen
yield of kraft wheat straw pulp con-
10
tinuously decreased under 0.6 and
0.4 mPa oxygen partial pressure as
the remaining alkali concentration, 5
concentration of added alkali, and
duration of delignification increased. 0
The carbohydrate degradation and 0 4.00 6.25 8.00 10.00 12.50 14.00 16.00 18.75 20.00 22.00 24.00 25.00 27.00

rate of delignification during oxy- RA, mg Na2O/g pulp

gen bleaching increased with in-
creasing process variables.
The bleached pulp yield decreased
by 30% and 40% under 0.6 and 0.4
mPa oxygen partial pressure, re-
5. The effect of added alkali charge upon kappa number
spectively, over the range of remain-
ing alkali concentration from zero
through 25 mg Na2O/g pulp as shown 18
in Fig. 1. This occurred at the condi- 16 RA = 1.68%, t = 90 min
tions of 0% added alkali charge and
KAPPA NUMBER

RA = 6.67%, t = 90 min
14
90 min duration of delignification.
The bleached pulp yield also de- 12
0.6 MPa O2 pressure
creased by 11% and 4.5% under 0.6 10
and 0.4 mPa oxygen pressure and
8
1.68 and 6.67 mg Na2O/g pulp, re-
spectively, for the same duration of 0.4 MPa O2 pressure
6
delignification (90 min) over the 4
range of added alkali charge from 0 0.40 0.65 0.80 1.00 1.30 1.40 1.60 1.95 2.00 2.40 2.50 2.60 2.70
zero through 1.95%. Minimum AA, %
bleached pulp yield occurred with
leveling off after pulping with 1.95%
added alkali charge, as Fig. 2 indi-
cates.
Small losses of 1% and 2.5% in
bleached pulp yield resulted in kraft
wheat straw pulp with an 11% in-
crease in duration of delignification
under 0.6 and 0.4 mPa oxygen pres-
sure and 1.68 and 6.68 mg Na2O/g
pulp, respectively. Minimum ble-
ached pulp yield occurred with lev-
eling off after pulping for 56 min as
shown in Fig. 3. 0% added alkali charge and 90 min under 0.6 and 0.4 mPa and 1.68 and
The same trend existed for the duration of delignification which Fig. 6.68 mg Na2O/g pulp, respectively,
kappa number. The kappa number 4 indicates. for the same duration of delign-
decreased by 60% and 30% under Figure 5 shows that the delig- ification (90 min) over the range of
0.6 and 0.4 mPa oxygen partial pres- nification rate during the rapid ini- added alkali charge from zero to
sure, respectively, over the range of tial phase and in the slow final phase 2.6%.
remaining alkali concentration from of oxygen bleaching increases with Figure 6 shows the decrease in
zero through 25 mg Na2O/g pulp. increased alkali concentration. kappa number with reaction time
This occurred at the conditions of Kappa number decreased by 26% during oxygen bleaching for two

88 January 1995 Tappi Journal

 6. The effect of bleaching time upon kappa number alkali charge, and 90 min bleach-
ing at 0.4 mPa oxygen pressure,
20 as Figs. 1–6 indicate.
RA = 1.68%, AA = 0% • Oxygen bleached pulp from wheat
KAPPA NUMBER

15 RA = 6.67%, AA = 0% straw meets the requirements

necessary as a raw material for
0.6 MPa O2 pressure
10 the production of bleached cellu-
0.4 MPa O2 pressure lose.
5 • A pulping experiment using the
predicted optimum conditions
verified the validity of the calcu-
0
45 56 67 78 90 100 lated process variables. The dif-
t, min ference between the experimental
and calculated results was insig-
nificant. TJ

Literature cited
7. Effect of kappa number upon bleached reaction had been first order for the
pulp yield lignin content in the pulp, as some- 1. Markham, L., Courchene, C., Elton, E.,
TAPPI 1981 Pulping Conference Pro-
times assumed, there would have ceedings, TAPPI PRESS, Atlanta, p. 165.
100
been a straight line relationship. Re- 2. Axegard, P., Moldenius, S., and Olm, L.,
ductions of 40 and 30% in the degree Svensk Papperstid. 82(5): 131(1979).
3. Jarrehult, B. and Samuelson, O., Svensk
95 of delignification occurred over the Papperstid. 81(17): 533(1978).
range of oxygen pressures of 0.6 and 4. TAPPI T 236 cm-85, “Kappa number of
0.4 mPa. pulp,” TAPPI PRESS, Atlanta.
90 5. TAPPI T 217 wd-77, “Brightness of pulp,”
Figure 7 shows that with oxygen TAPPI PRESS, Atlanta.
BLEACHED PULP YIELD, %

bleaching prolonged to kappa num- 6. TAPPI T 227 om-85, “Freeness of pulp,”

85 bers below 10 the yield curve starts TAPPI PRESS, Atlanta.
7. TAPPI T 414 om-88, “Internal tearing
to drop faster, i.e., there is higher resistance of paper (Elmendorf–type
80
yield loss per kappa number unit. method),” TAPPI PRESS, Atlanta.
This means that the process becomes 8. TAPPI T 494 om-88, “Tensile breaking
properties of paper and paperboard (us-
less selective. For best yield and se- ing constant rate of elongation appara-
75
lectivity, the oxygen bleaching tus),” TAPPI PRESS, Atlanta.
should have an approximate degree 9. TAPPI T 404 om-87, “Tensile breaking
strength and elongation of paper and pa-
70 of delignification of 40–45% as the perboard (using pendulum-type tester,”
example of changing from kappa no. TAPPI PRESS, Atlanta.
17 to 10 shows. The specific yield 10. Cochran, W. G. and Cox, G. M., Experi-
65 = 0.6 mPa O2 pressure mental Design, Wiley, New York, 1957,
= 0.4 mPa O2 pressure
values are 94–96% at kappa number pp. 343–360.
10. Prolonging oxygen bleaching to 11. Mickley, H. S., Sherwood, T., and Reed,
60 kappa numbers below 7 gives spe- C. E., Applied Mathematics in Chemical
0 5 7 9 11 15 19 Engineering, McGraw Hill, New York,
cific yield values that are too low. An 1957, p. 145.
KAPPA NUMBER
optimal kappa number level is ap- 12. Rosenbrock, H. H., Computer J. 3:
proximately 10 as the results indi- 175(1960).
cate. Received for review Jan. 14, 1994.

kraft pulps of kappa no. 15.6 and Accepted June 6, 1994.

12.5. The kappa number decreased Conclusions
with reaction time whatever the ini-
Using the results reported here from
tial kappa number. The oxygen par-
our experimental work, it is possible
tial pressure and remaining alkali
to draw the following conclusions:
concentration varied during these
conditions. The added alkali concen- • Optimum bleached pulp yield oc-
tration and the temperature re- curs at remaining alkali level of
mained constant, however. If the 6.675 mg Na2O/g pulp, 0% added

Vol. 78, No. 1 Tappi Journal 89