Bagasse is an important source of raw material for pulp and paper manufacture in the countries that produce large amounts of this type of waste and have scantly wood resources. Iran is one of these countries which have produced about 4.1 million tones of sugar cane annually, during the resent past years (FAO, 2004). An estimation based on a planted area of 41000 ha, and a production rate of dry sugar cane biomass of 89 ton/ha.year (15 ton/ha dry bagasse), shows that Iran sugar cane industry produced approximately 615000 ton of air dry lignocellulosic biomass in 2003 (Rezayati-Charani et al., 2006).
Plants as raw materials of pulp-making are usually cooked at high temperature and pressure in the conventional pulping process. There are major problems associated with this process such as energy consumed and large amount of chemicals leading to environmental pollution (Hongzhang et al., 2002). Therefore, the need to decrease pollution from cellulose pulp mills has promoted the search for alternative bleaching and especially ozone, hydrogen peroxide and biobleaching processes (Jiménez et al., 1999; Roncero et al., 2003; Khristova et al., 2006).
The biobleaching process is based on the action of the microorganisms and/or enzymes. Microbial xylanases that are thermostable and cellulose-free are generally preferred for biobleaching of paper pulp. The interest for xylan degrading enzyme and its applications in the pulp and paper industries has advanced significantly during past few years (Adachi and Chen, 2007; Han et al., 2007; Rajasekar, 2007; Shen Liu, 2007; Bajpai et al., 1994; Garg et al., 1998; Christov et al., 1999; Srinivasan and Rele, 1999). In this process, the bond between lignin and hemicelluloses is primarily between lignin and xylan which can be removed by xylanase. Once this layer of hemicellulose is removed, the lignin layer is easily available for degradative action of the ligninolytic enzymes (Eriksson, 1993). However, the cellulases produced with xylanases in the microbial based processes simultaneously, alter cellulose fibers and diminish pulp quality. Therefore, xylanases with high degree of purity are required (Jiménez et al., 1997). The studies on xylanase bleaching of hardwood and softwood pulps have received much attention, but studies on bagasse are few (Neeta and Mala, 2002; Charin et al., 2003; Sandrim et al., 2005). Prior to this research, the bagasse biodegradation with commercial xylanase enzyme Cartazyme HS (Sandoz) was investigated by Denise and his colleague (Denise and Adilson, 2002); but the effect of operation variables in enzymatic pulp bleaching was not published. The main operational variables in enzymatic pulp bleaching processes are temperature, time, pH, enzyme level and pulp consistency. Danault examined the operating conditions used by several authors and found enzyme concentration ranging from 1 to 15 IU/g dry pulp, temperature from 40 to 70°C, time from 0.5 to 24 hrs., pH from 3.5 to 8, and consistency from 2.5 to 12% (Daneault et al., 1994). However, each enzyme processes specific catalytic properties and requires special operating conditions, so no generalization can be made in this respect.
In few studies on operational variables of the pulp biobleaching proess, a factorial design has been used to develop empirical models. These models are involved in several independent variables regarding performance examining and predicting pulp properties and paper sheets in terms of the conditions used in the enzymic step (Allison and Clark, 1994; Abdul-Karimand and Rab, 1995; Ziaie-Shirkolaee et al., 2007a,b). In this work, a main composition design was used to study the influence of bleaching variables in bagasse with the xylanase and peroxide hydrogen on different obtained properties such as yield, viscosity, brightness, breaking length, bear index and tear index.
MATERIALS AND METHODS
Raw material : The bagasse used in this study, obtained from the local sugar field in central area of Iran. Before pulping, the raw material was cleaned, cut to pieces, approximately 3cm length and sun-dried. The chemical composition of bagasse was determined as follows: 50.33% cellulose, 20.29% lignin, 74.27% holocellulose, 46.05% α-cellulose, 7.07% ash and 1.57% ethanol/ dichloromethane extractable, on an oven-dry weight basis (moisture content 9.8%). Deviations of these contents from their respective means were less than 10%.
Analysis of raw material, pulp and paper sheets: Analysis of raw material and pulp made according to Tappi Standard Methods (TAPPI, 2002) with the exception of hemicellulose that determined by decreasing of cellulose content from holocellulose (holocellulose determined by Wise’s sodium chlorite method) (Wise and Murphy, 1946), cellulose according to Kurscher and Hoffner’s nitric acid method (Rowell, 1984) and viscosity of pulp was measured in cupri-ethylenediamin (CED) solution according to SCAN-CM 15:88 standard (SCAN, 1998).
Pulp yield was determined gravimetrically following drying at 105°C ± 2 for 24 h.
Pulping: Pulps were made in a 21-l batch cylindrical mini digester (stainless steel 321), that described by Ziaie-Shirkolaee et al. (Ziaie-Shirkolaee et al., 2007a,b). Wheat straw was cooked in the reactor, using the following conditions: temperature 160°C, time 80min, sulfidity 20%, Active alkali NaO2 14% and solid/liquor ratio, 1/10 (d. w.).
Bleaching process: The pulp was bleached using an XP sequence. The enzymatic treatment, based on the enzyme Cartazyme HS (Sandoz), was carried out in the transparent plastic bags containing the pulp, under different conditions.
For hydrogen peroxide treatment, enzyme-treated pulp was placed with H2O2 at 70°C for 3 hrs. In addition, in order to reduce the role of transition metals on alkaline decomposition of peroxide and also, to limit the degradation and deterioration of carbohydrates, the materials such as EDTA (diethylene-diamine-tetra-acetic acid), Epson salt (magnesium sulfate) and DTPA (diethylene-triamine-penta-acetic acid) are used in bleaching process (Table 1). Finally, after each bleaching stage, the pulp was washed by deionized water.
Experimental design: The tested model uses a series of points (experiments) around a central one (central experiment), and several additional points (additional experiments), to estimate the first- and second-order interaction terms of a polynomial. This design needs the general requirement that each parameter in the mathematical model can be estimated from a fairly small number of experiments (Montgomery, 1991).
The total number of observations (experiments) required for the five independent variables (viz. temperature -T-, cooking time -t-, pH – pH-, enzyme level -D- and pulp consistency –C-) was calculated from the following equation (Akhnazarovaand and Kafarov, 1982):
N=2k + 2k +1 (1)
and found to be 43. In this equation, k is the number of independent variables.
In this work, all independent variables (temperature, time, pH, enzyme level and pulp consistency) normalized from -1 to +1 according to following formula (Rodríguez et al., 1998):
This normalization also results in more accurate estimates of the regression coefficients as it reduces interrelationships between linear and quadratic terms (Montgomery, 1991). Normalized independent variables and experimental data of pulp properties were used for the development of empirical models, in
which the dependent variables were evaluated by the following general equation:
where Z is the response or dependent variable [viz. Pulp yield following step X (YieldX), Pulp brightness following step X (BrightnessX), Final yield (Yield), Viscosity, Kappa number (K.N), Final brightness, Breaking length (BL), Burst index (BI) and Tear index (TI)]; Xn is the normalized value of the independent variable concerned; and a0, bi, ci and dij are unknown characteristic constants estimated from the experimental data.
The values of responses obtained allow the calculation of mathematical estimation models for each response, which were subsequently used to characterize the nature of the response surface.
Table 2 and 3 show the average experimental results of the pulp and paper properties studied in the 43 bleaching experiments. By correlating the experimental pulp yield values obtained from the enzyme treatment step (Table 2) with those of independent values (Table 1) via Eq. (4), the following equation was obtained once all the terms with R-sq at a level of more than 82% were eliminated using the step wise method:
Yield (X)= 89.56 + 1.96 T2- 0.89 D + 0.52 pH - 0.46 t*D - 0.48 T - 0.44 C + 0.44 T*pH - 0.41 pH*C -0.38 D*C - 0.30 t (4)
This equation reproduces the experimental results with error less than 3% (Table 4,5, 6). The S, R-sq, R-sq (adj) and P-Value are statistical items of obtained Eq. (3) which all of them has been added and explained in the margin of Table 4 and experimental design section. Appling steepest ascent method (Press et al., 1992) to Eq. (4) provided the values of independent variables that maximized the yield. Within the range of normalized values for the independent variables from -1 to +1, the maximum yield (93.28%) was achieved at high pH (normalized value of +1 for it), in addition to low temperature, enzyme dosage and short time (normalized values of -1 for three variables).
Eq. (4) allows the estimation of changes in the yield with one of the independent variables over the range considered on constancy of all other variables. With the most suitable values for all the independent variables that are to be kept constant (viz. high pH, low temperature and enzyme dosage, short time and each pulp consistency), if a high yield is to be achieved, then the greatest variation of the yield is obtained by altering the time (1.52 units, 1.6%) and the smallest by the changing the enzyme level (0.08 units, 0.1%); an intermediate variation is obtained by altering temperature (1.18 units, 1.2%), pH (0.98 units, 1.1%) or pulp consistency (0.82 units, 0.9%). Figures 1-4 support these conclusions. From the foregoing analysis, it follows that, by high pulp consistency, the yield decreases to 73.43%, i.e., by only 0.9% with respect to its highest level (74.33%). Therefore, low temperature, enzyme level, short time , high pulp consistency and pH may be used in order to decrease operating costs and immobilized capital depreciation.
Using the same products, the following equations were obtained for the other response variables:
Brightness (X)= 43.68 + 0.62 t + 0.63 T*C + 0.49 T*pH - 0.94 T2 + 0.31 D + 0.28 T +0.20T*D (5)
Total Yield= 73.00 - 0.75 T*D - 0.58 pH*C - 0.54 t* D - 0.94 T2 - 0.41 D*pH - 0.28 T*pH - 0.17D (6)
Viscosity= 705 -55.4 T2 -21.4 T - 13.4 D*pH -13.3 T* D -10.6 t + 10.6 T*pH -9.6 pH + 8.8 D (7)
K.N= 15.15 - 0.459 T + 0.365 pH - 0.181 T*pH - 0.156 t*C - 0.141 D - 0.131 t*pH - 0.125 T*t - 0.25 T2 + 0.100 pH*C - 0.088 D*pH (8)
Total brightness= 71.98 + 0.31 pH – 0.3 D – 0.7 pH2 + 0.3 t*pH – 0.4 D*pH + 0.4 C*pH (9)
BL= 5715 + 65 pH – 56 t*pH + 51t*D + 82 T2 -35 T*pH (10)
BI= 3.592 + 0.0456 T - 0.0109 pH - 0.0071 D + 0.0065 t - 0.0037 t*pH (11)
TI= 5.998 + 0.546 T - 1.11 T2 - 0.231 T*t - 0.157 T*C + 0.142 T*pH + 0.133 C + 0.113 t*pH (12)
Eqs. (5)-(12) reproduce the experimental pulp brightness following step X, and the final yield, viscosity, kappa number, final brightness, breaking length, burst index and tear index which errors less than 3, 5, 8, 2, 6, 4, 10 and 9% respectively.
The S, R-sq, R-sq (adj) and P-Values for the fits of Eqs. (5)-(12) are given in Table 4.
The results of Tables 5 and 6 were obtained by using a procedure, similar to that applied to Eq. (4) above.
Varying one of the independent variables between the normalized values from -1 to +1 while keeping the others constant at the values shown in Table 5, resulted in the greatest changes in the dependent variables relative to the optimum values (also shown in Table 5) given in Table 6. As it can be seen, temperature changes was specially influential on the viscosity and burst index; time changes on yield in step X and tear index; D changes on the final yield; pH changes on the Kappa number and breaking length of the paper sheets and consistency changes on brightness in step X and Final brightness. On the other hand, temperature changes had little effect on the brightness in step X, Final brightness and final yield; time changes on the Kappa number and breaking length; D changes on the Yield following step X, Brightness following X, burst and tear index; pH changes on brightness in step X, Final brightness and tear index; and consistency changes on the viscosity, breaking length and burst index.
The data in Table 6 and previous conclusions show that the dependent variables vary little from their optimum values if low temperature and D values, high pH and consistency values in addition to short time are used. The estimates for the dependent variables are given in Table 7, which also shows the percentage variations of such values with respect to their optimum levels (Table 5). These new operating conditions are more suitable than those in Table 6 in which they are milder. This reduces energy and immobilized capital expenses, with only slight losses in some properties of the pulps and paper sheets obtained from them. In summary, using a temperature of 35 (°C), an enzyme concentration of 6 (IU/g) by dry pulp weight, pH 5 and a consistency of 12% for 2 (h) provides the best results.