Using both xylanase and laccase enzymes for pulp bleaching

Bioresource Technology 100 (2009) 2032–2039 Contents lists available at ScienceDirect Bioresource Technology journal homepage: www.elsevier.com/locate/biortech Using both xylanase and laccase enzymes for pulp bleaching Cristina Valls, M. Blanca Roncero * Textile and Paper Engineering Department, ETSEIAT, Universitat Politècnica de Catalunya, Colom 11, E-08222 Terrassa, Spain a r t i c l e i n f o Article history: Received 16 July 2008 Received in revised form 6 October 2008 Accepted 12 October 2008 Available online 26 November 2008 Keywords: Biobleaching Eucalyptus Laccase–mediator system Modeling Xylanase a b s t r a c t Two enzyme treatments involving xylanase (X) and laccase (L) were used jointly in an XLE sequence (where E denotes alkaline extraction) to bleach oxygen-delignified eucalyptus kraft pulp in the presence of 1-hydroxybenzotriazol (HBT) as mediator. The results of the XLE sequence were compared with those of an LE sequence. The application conditions for the laccase–mediator system were optimized by using a sequential statistical plan involving three variables (viz., the laccase and mediator doses, and the reaction time) with both sequences. The models used to predict the kappa number and brightness revealed that, once all accessible lignin was removed, the system altered other coloured compounds. The best conditions for the L stage involved a reduced mediator dose (0.5% odp). The xylanase pretreatment increased the accessibility of residual lignin and facilitated removal of hexenuronic acids. For a specific target brightness level of 70% ISO, the X pretreatment can save as much as 30% laccase and 80% mediator while shortening the reaction time by 45%. Ó 2008 Elsevier Ltd. All rights reserved. 1. Introduction The pulp bleaching plant, which is the most contaminating section in the paper manufacturing process, has gone through a number of changes intended to alleviate its adverse impact. Totally chlorine-free (TCF) pulping processes avoid the formation of highly toxic organochlorine compounds (AOX) during bleaching. Usually, TCF sequences include oxygen, hydrogen peroxide and ozone based stages (Pedrola et al., 2004; Roncero et al., 2003a,b; Torres et al., 2004). Biotechnology has rapidly gained ground in pulping processes; thus, enzyme stages involving xylanases (Herpoël et al., 2002; Roncero et al., 2000a, 2003c) and laccases (Fillat et al., 2007; Gamelas et al., 2007; Garcia et al., 2003; Ibarra et al., 2006; Moldes and Vidal, 2008; Oudia et al., 2007) have so far provided very promising results in pulp bleaching sequences. The use of xylanases (X) constitutes a very important technological improvement in as much as it enhances the bleaching effect of chemical reagents, thereby affording substantial savings and, more importantly, considerably diminishing the production of pollutants during bleaching (Amin, 2006; Roncero et al., 2005). A number of studies on xylanases and their use in pulp bleaching support the feasibility of their industrial application (Bajpai, 2004; Popovici et al., 2004). The use of laccases in the form of laccase–mediator systems (L) has provided an alternative to xylanases acting directly on * Corresponding author. Tel.: +34 937398210; fax: +34 937398101. E-mail address: roncero@etp.upc.edu (M.B. Roncero). 0960-8524/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2008.10.009 lignin. The mediator is typically a compound of low molecular weight capable of diffusing into cellulose fibres. Laccase–mediator systems react with the phenolic and non-phenolic fractions of lignin to delignify pulp (Freudenreich et al., 1998). Unlike xylanases, however, the industrial use of laccases has so far been impeded by their limited availability – the enzyme should be available in adequate amounts for mill production; and the high cost and potential toxicity of mediators released into effluents – which entails using them at as low as possible doses; and comparatively the long reaction times involved – which should be minimized for smooth integration into industrial bleaching processes. In this work, an experimental design was developed in order to reduce the laccase and mediator doses used in the L stage, and the reaction time. The mediator used, 1-hydroxybenzotriazol (HBT), is the most effective ever used for this purpose (Bourbonnais et al., 1997; Camarero et al., 2004; Moldes et al., 2008; Sigoillot et al., 2005); therefore, reducing its application dose would be highly beneficial. An XLE sequence and an LE sequence, E denoting an alkaline extraction stage, were studied. The target properties were the kappa number and brightness of the pulp as measured after alkaline extraction. Appropriate models for these properties allowed an optimal application point for L to be selected in both sequences. The models were compared in order to establish whether a xylanase pretreatment would be effective in reducing the laccase and mediator doses or the reaction time to be used in L. Finally, the effect of the enzymatic treatments on hexenuronic acids was evaluated. C. Valls, M.B. Roncero / Bioresource Technology 100 (2009) 2032–2039 individual reactor, using 1.5% NaOH at 90 °C at 5% consistency for 120 min, after which the liquors were recovered for pH measurement (10.5–11) and the pulp was washed for characterization. 2. Methods 2.1. Raw material The raw material used was oxygen-delignified eucalyptus kraft pulp (Eucalyptus globulus) produced by the Torraspapel S.A mill in Zaragoza, Spain. Before treatment, the pulp was washed with a buffer solution consisting of 50 mM Tris–HCl buffer at pH 7 in the laboratory at room temperature for 30 min. The initial kappa number and brightness of the pulp after washing were 8.4 and 51.2% ISO, respectively. 2.2. Xylanase pretreatment (X stage) The enzyme used in the X stage was a xylanase previously isolated and characterized by Gallardo et al. (2004). The X treatment involved using 3 U g1 odp xylanase at 10% consistency and Tris– HCl buffer at pH 7 at 50 °C for 2 h, after which the treatment liquors were recovered to measure their final pH. The resulting pulp was washed with decalcified water three times and distilled water once. The final kappa number and brightness of the pulp were 8.4 and 53.7% ISO, respectively. 2.3. Laccase–mediator treatment (L stage) Commercial laccase from Trametes villosa (NS-51002) supplied by NOVOZYMESÒ was used for this purpose. The mediator, HBT, was obtained from Sigma–Aldrich. Tests were performed in a pressurized reactor at 590 kPa, 30 °C and a stirring rate of 60 rpm, using 50 mM sodium tartrate buffer at pH 4 and 5% consistency. A few drops of 0.05% w/v of the surfactant Tween 80 were also added. The experimental design used involved the laccase dose, mediator dose and reaction time as variables. After the L stage, the pulp was washed before alkaline extraction (E stage). 2.4. Experimental design Enzyme treatments were applied in accordance with a 23 sequential statistical plan involving two levels and three variables plus three repetitions at the central point, which required a total of 11 tests. The three independent variables were changed over the following ranges: 1–20 U g1 odp (oven-dried pulp) for the laccase dose (x1), 0.5–2.5% odp for the HBT dose (x2) and 1–7 h for the reaction time (x3) (Table 1). The results of the three repetitions at the central point and their variance were used in combination with the variance of the saturated model to calculate Snedecor’s F-value in order to determine whether the variance was homogeneous or heterogeneous. Since the variance was homogeneous in all cases, a linear model was constructed its significant terms identified and potential curvature detected. Linear multiple regression was applied by using an excel spreadsheet to implement the stepwise backward regression method and discard all terms with a probability (p-value) less than 0.05. 2.5. Alkaline extraction stage (E) 2.6. Pulp properties The treated pulp samples were characterized in terms of kappa number and brightness according to ISO standards 302 (1981) and 3688 (1977), respectively. The kappa number was measured two times and brightness four in order to calculate a standard deviation, which was found to be 0.1 for both properties. The hexenuronic acid (HexA) content was determined with the method of Gellerstedt and Li (1996) as modified for UV detection by Chai et al. (2001). 3. Results and discussion Table 2 shows the kappa number and brightness of the pulp samples after the E stage in the LE and XLE sequences. In the LE sequence, the kappa number ranged from 4.7 to 6.4 and brightness from 56.1% to 68.9% ISO; in the XLE sequence, the previous properties varied over the range 3.8–5.6 and 57.9–71.4% ISO, respectively. The XLE pulp samples invariably exhibited a kappa number smaller by one unit and 1.4–6% ISO higher brightness than the LE pulp samples. 3.1. Modeling Experimental data were fitted to a second-order polynomial equation with the kappa number (YKN) and brightness (YBr) as responses. Based on previous results for flax pulp (Garcia et al., 2003), some preliminary tests (results not shown) were conducted on eucalyptus pulp in order to select tentative optimum ranges for the independent variables. In order to make the process more industrially feasible, the ranges were shrunk with respect to those previously used by Garcia et al. (2003): to 1–20 U g1 odp for x1, 0.5–2.5% odp for x2 and 1–7 h x3. The minimum effective HBT dose was 0.5% odp-no reaction occurred at lower proportions. The results for the variables were coded as 1 or +1 (Table 1), both for direct comparison of coefficients and for easier understanding of the effect of the variables on the responses. The independent variables were zeroed at the central point. A preliminary test with the four models (viz., kappa number and brightness in the LE and XLE sequences) revealed that the quadratic term was significant (p < 0.05). Two additional tests were therefore required in order to identify the variables possessing a significant term and ensure their accurate discrimination. A second analysis of the modeling equations provided the following responses, where all significant terms had p < 0.05: Kappa number model for LE Y KN LE ¼ 5:02  0:42x1  0:17x2  0:27x3 þ 0:17x1 x3 þ 0:42x21 ð2Þ 2 with R = 0.98. Kappa number model for XLE Y KN Table 1 Variables distribution with the factor of each level. ð1Þ with R2 = 0.94. Brightness model for LE Y Brð%ISOÞLE ¼ 66:9 þ 2:65x1 þ 1:07x2 þ 2:84x3  4:65x23 Pulp samples treated with the laccase–mediator system were subjected to alkaline extraction (an E stage) in a Datacolor Easydye x1 x2 x3 2033 XLE ¼ 4:04  0:30x1  0:28x2  0:31x3 þ 0:47x23 ð3Þ 2 with R = 0.94. Brightness model for XLE Variables 1 0 +1 Laccase dose (U g1 odp) HBT dose (% odp) Reaction time (h) 1 0.5 1 10.5 1.5 4 20 2.5 7 Y Brð%ISOÞXLE ¼ 70:15 þ 2:63x1 þ 1:09x2 þ 3:07x3  0:77x1 x2  0:99x1 x3  2:35x21  1:18x22 2 with R = 1.00. ð4Þ 2034 C. Valls, M.B. Roncero / Bioresource Technology 100 (2009) 2032–2039 Table 2 Applied conditions of L experiences and kappa number and brightness results after LE and XLE sequences. x1 x2 x3 Laccase dose (U g1) HBT dose (%) Time (h) 1 1 1 1 1 1 1 1 0 0 0 1 0 1 1 1 1 1 1 1 1 0 0 0 0 1 1 1 1 1 1 1 1 1 0 0 0 0 0 1 20 1 20 1 20 1 20 10.5 10.5 10.5 20 10.5 0.5 0.5 2.5 2.5 0.5 0.5 2.5 2.5 1.5 1.5 1.5 1.5 0.5 1 1 1 1 7 7 7 7 4 4 4 4 4 In the above equations, x1 = (L10.5)/9.5 (L denoting the laccase dose, in U g1), x2 = (M1.5)/1, (with M the HBT dose, as a percentage) and x3 = (t4)/3 (t being the reaction time, in hours). As can be seen from the R2 values obtained, the fit was quite good in all cases. 3.2. Models for the LE sequence The kappa number model for LE sequence (Eq. (1)) was influenced by the three variables (viz., the laccase dose (x1), mediator dose (x2) and reaction time (x3)); thus, kappa number decreased with increase in any of the three. The binary interaction between the laccase dose and reaction time (x1x3), and the quadratic term for the laccase dose (x21 ) also affected the kappa number, both with a positive coefficient. Fig. 1a shows the calculated surfaces for the kappa number at a constant reaction time. As can be seen, kappa number decreased substantially with increase in the laccase dose up to 13 U g1 odp (x1 = 0.26) and then levelled of up to the highest enzyme dose studied (20 U g1 odp). This result suggests that some lignin was inaccessible to the reagents and remained in the pulp, thereby preventing the L treatment from further reducing the kappa number. In fact, some authors believe that all lignin in pulp is available for reaction, while others think that a certain amount always remains unreacted (Eriksson and Gierer, 1985; Kaneko et al., 1983; Tessier and Savoie, 2000). Some authors have studied the influence of the treatment time with laccase–mediator systems on delignifi- a XLE Kappa number Brightness (% ISO) Kappa number Brightness (% ISO) 6.4 5.2 6.2 5.1 5.5 5.2 5.3 4.7 5.1 4.8 5.0 4.9 5.5 56.1 60.7 57.2 63.7 60.9 66.7 63.9 68.8 68.2 66.7 66 68.9 66.5 5.6 4.7 4.8 4.1 4.6 4.2 4.2 3.8 3.8 4.1 4.2 4.3 4.4 57.9 66.6 62.0 67.8 66.4 71.3 69.7 71.4 70.3 70.1 70.2 70.4 67.8 cation, or lignin model compounds, and found the process to involve two distinct stages (Balakshin et al., 1998, 1999; Garcia et al.,2003; Potthast et al., 2001). In the first, the pulp is largely delignified to a limiting kappa number similarly as in ozone treatments. In the second, delignification is slight, but the oxygen uptake continues to be high. This suggests the presence of active chemical species not reacting with residual lignin remaining in the pulp, but rather interacting with lignin fragments via side reactions (Balakshin et al., 2001). Roncero et al. (2000b) previously obtained similar results, but using ozone as bleaching agent. Therefore, our results reflect the fact that lignin in the pulp can only be partially removed in a single bleaching stage. The fact that the kappa number levelled off at a given enzyme concentration can be ascribed to: (a) competing reactions with other wood components such as carbohydrates, (b) alternative reactions with lignin intermediate products, (c) the presence of unreactive lignin especially difficult to remove under the conditions used here, (d) the residual fraction of lignin being less readily accessed by the laccase–mediator system or (e) the enzyme losing some activity at the end of the treatment or HBT being oxidized to its inactive form. The mediator dose and reaction time exhibited a linear influence in the models; thus, increasing the mediator dose or, especially, the reaction time, always diminished the kappa number. The LE brightness model (Eq. (2)) as well as the kappa number model involved the three variables. The strongest influence was that of the reaction time, followed by the laccase dose and, consis- X=+1 X=0 X=-1 b 0.5 X2 -me diat o r 1.0 X 0.0 1 -l 3 -0.5 -1.0 -0.5 0.0 0.5 60 or -1.0 -0.5 0.0 0.5 1.0 ac ca se 4 65 55 -0.5 ed iat 5 70 0.0 0.5 X3 -t im e 1.0 1.0 X 2 -m 6 O) Brightness (% IS 75 7 Kappa number LE ¢ Fig. 1. Predicted surfaces for LE sequences. Kappa number with constant reaction times-x3 (a) and brightness with constant laccase doses-x1 (b) x = +1 (h); x = 0 ( ) and x = 1 (j). 2035 C. Valls, M.B. Roncero / Bioresource Technology 100 (2009) 2032–2039 tent with the kappa number model, the mediator dose. The quadratic term for the reaction time (x23 ) was also influential on brightness. Fig. 1b shows the predicted surface for brightness in LE sequence at a constant laccase dose. As can be seen, brightness increased with increasing laccase dose from 1 to 20 U g1 odp by contrast, the kappa number levelled off. As stated above, above a laccase dose of 13 U g1 odp, the lignin remaining in the pulp was inaccessible but laccase was able to act on other pulp components containing chromophoric groups such as carbohydrates and raise brightness as a result (Garcia et al., 2004). The mediator dose also exhibited a linear influence on brightness and kappa number; thus, raising it from 0.5% to 2.5% odp reduced the kappa number and increased brightness. These results are consistent with those of Garcia et al. (2003), who found a linear dependence at HBT doses from 0.5% to 2.64% odp and laccase to be inactivated by the mediator at doses above 2.64%, all after a P stage in flax pulp. The reaction time exhibited two types of effects; thus, from 1 to 4.6 h (x3 = 1 to 0.2), brightness increased but, beyond that point (4.6–7 h), it levelled off or even decreased (x3 = 0.2 to +1). No similar effect on the kappa number was observed; in fact, kappa number invariably decreased with increasing reaction time. These results suggest that new chromophores such as carbonyl or carboxyl groups were attacked (Garcia et al., 2004; Poppius-Levlin et al., 1999) or lignin derivatives such as quinones (Chakar and Ragauskas, 2001) formed in the pulp after 4.6 h of treatment, because of the oxidizing action of the system. ables in the kappa number model had a similar influence, the reaction time (x3) was the most influential variable in the brightness model, followed by the laccase dose (x1) and mediator dose (x2) as in the LE sequence, the last was least influential variable. The quadratic terms x21 and x22 were also important in the brightness model, but the binary interactions x1x2 and x1x3 had only a slight influence. Fig. 2b shows the predicted surfaces for brightness at a constant reaction time. A quadratic effect of the laccase dose is clearly apparent: raising it to about 12.4 U g1 odp (x1 = 0.2) increased brightness; however, further increasing it (up to 20 U g1 odp) caused brightness to level off or even decrease slightly. As can be seen, the mediator had a quadratic effect by which increasing its dose up to 1.7% odp (x2 = 0.2) resulted in increased brightness, but further raising it (up to 2.5% odp) caused this property to level off or decrease slightly. By contrast, the kappa number invariably decreased with increasing laccase and HBT doses throughout their respective ranges. This may have resulted from condensation of the mediator on the lignin structure at high laccase and HBT doses, and the potential formation of quinones, as previously suggested by Moldes et al. (2008). Unlike the kappa number, where a limiting value was reached after about 4.6 h, brightness increased throughout (7 h); this suggests that, from 4.6 to 7 h, the laccase–mediator system removed other coloured compounds containing chromophoric groups (e.g., carbohydrates) that were not detected in the kappa number measurements, as Garcia et al. (2004) also suggested. 3.4. Model fitting 3.3. Models for the XLE sequence The kappa number model for the XLE sequence (Eq. (3)), which also involved the variables laccase dose (x1), mediator dose (x2) and reaction time (x3), revealed a similar influence in reducing its value. The quadratic term for the reaction time exhibited the highest coefficient. Fig. 2a shows the calculated surfaces for the kappa number at a constant laccase dose. The laccase and mediator doses exhibited a linear influence; increasing either invariably decreased the kappa number. The quadratic effect for the reaction time exhibited two regions of influence, namely: from 1 to 4.6 h (x3 = 1 to 0.2) kappa number decreased with time; subsequently, it remained constant at 3.4 h until the end of the tests (7 h, x3 = +1). Like the kappa number model, the brightness model for XLE (Eq. (4)) involved three variables. However, while all three vari- Experimental kappa number and brightness values were compared with the calculated values provided by the models. The differences were all minimal and within the acceptable ranges in the respective ISO standards. Therefore, the proposed models provide an accurate depiction of the actual process. The brightness models exhibited better fitting than the kappa number models. The bestfitting model was that for XLE brightness, with a very low error (0.4%), and the worst-fitting model that for the XLE kappa, with an error of ±6%. 3.5. Selection of the best application conditions The proposed models were used to identify the points leading to the smallest kappa number and highest brightness. In those cases where a quadratic term for some variable had any effect, the inflec- X=+1 X=0 X=-1 a b 75 -1.0 -0.5 0.0 0.5 1.0 0.0 X2 -t im 0.5 e X -0.5 2 -m 3 1.0 70 65 -1.0 -0.5 0.0 0.5 60 55 -0.5 0.0 0.5 X2 -m ediat or 1.0 1.0 X 1 - la cc as e 4 or 5 ISO) Brightness (% 6 ed iat Kappa number 7 ¢ Fig. 2. Predicted surfaces for XLE sequences. Kappa number with constant laccase doses -x1 (a) and brightness with constant reaction times -x3 (b). x = +1 (h); x = 0 ( ) and x = 1 (j). 2036 C. Valls, M.B. Roncero / Bioresource Technology 100 (2009) 2032–2039 tion point was estimated from the derivative of the model with respect to the particular variable, and the equation was equalled to zero. The point at x1 = 0.3, x2 = +1 and x3 = +1 (0.3, 1, 1) for the LE sequence was expected to provide the smallest possible kappa number (4.5) and that at (1, 1, 0.3) the highest possible brightness (71% ISO). The corresponding points for the XLE sequence were (1, 1, 0.3) for the smallest kappa number (3.4) and (0.2, 0.2, 1) for the highest brightness (73.6% ISO). The fact that the optimum points differed between properties and also between sequences was, as noted earlier, a result of the models behaving disparately. This precluded identifying a unique application point affording the best possible values for both properties and required making some compromise. Thus, the application conditions for the L stage were selected according to several criteria. The primary objective goal was to reduce the mediator dose (HBT is an expensive chemical potentially toxic at high doses), then to shorten the reaction time in order to facilitate integration in industrial bleaching processes and, finally, to reduce the laccase dose. The ultimate goal was to obtain pulp of high brightness and the smallest possible kappa number. Because reducing the laccase dose was a minor priority, the enzyme was used at a high dose (x1 = +1) to construct a contour plot for each model based on the other two variables (Figs. 3a and b). Based on the above-described criteria and on the primary objective (viz., reducing the mediator dose), the best application points were selected. The kappa number and brightness values used in both plots were limited to those in the experimental range. Thus, the smallest kappa number and highest brightness experimentally obtained in the LE sequence were 4.7 and 69% ISO, respectively. Based on Fig. 3a, for a target brightness of 68% ISO the mediator dose can be reduced to its minimum value (x2 = 1), and the reaction time (x3) to 0.2, and still obtain an acceptable kappa number (5.2). The smallest kappa number and highest brightness experimentally obtained in the XLE sequence were 3.8 and 71.4% ISO, respectively. Based on Fig. 3b, if a kappa number of 4.0 is deemed acceptable, then the mediator dose can be reduced to its minimum value (x2 = 1) and the reaction time (x3) to 0.2 in order to obtain 69% ISO brightness. a 3.6. Comparing the LE and XLE models A xylanase pretreatment is known to induce morphological changes in cellulose fibres (Roncero et al., 2000a) by hydrolysing xylans on fibre surfaces and making fibres more readily accessible to the reagents as a result. As previously inferred from Table 2, the XLE sequence provided pulp of smaller kappa number and higher brightness than did the LE sequence. The predicted models for each sequence are compared graphically in Fig. 4 for easier viewing of the effects of the enzyme pretreatment (X stage) on the different variables of L. Based on the kappa number, laccase in the LE sequence reached a specific dose (13 U g1 odp) beyond which no further delignification was observed; on the other hand, following the X stage (XLE sequence), delignification increased with increasing laccase dose from 1 to 20 U g1 odp. This suggests that the X stage facilitates penetration of laccase into the fibres and removal of products contributing to the kappa number, consistent with the bleach boosting effect of xylanase treatments observed by some authors. The effect of the mediator dose on the kappa number was not affected by the xylanase pretreatment; in fact, both models re- b 67 68 69 4.9 0.5 70 703.8 69 69 69 649 .2 1.0 -1.0 -1.0 X 2-mediator 4.6 68 -0.5 70 4.0 68 0.0 71 70 5.0 -0.5 -0.5 5.1 5.2 -1.0 -1.0 5.0 5.1 5.2 -0.5 68 67 66 71 71 71 0.0 72 3.8 72 69 X -time 3 4.9 5.0 5.1 0.0 72 4.0 70 0.5 0.5 X 3-time 1.0 4.8 4. 9 69 5.0 5.1 68 71 1.0 The selected points (Table 3) were at the maximum laccase dose (20 U g1 odp), minimum mediator dose (0.5% odp) and a reaction time of 3.4 or 4.6 h. The maximum levels studied (viz., 20 U g1 odp laccase, 2.5% odp mediator dose and 7 h reaction) are similar to the conditions employed by other works (Herpoël et al., 2002; Garcia et al., 2003; Ibarra et al., 2006; Sigoillot et al., 2005). The experimental design carried out allowed the HBT dose to be reduced by 80% and the reaction time by 34% or 51% relative to the initial values. In recent work, some authors used an HBT dose of 1.5% odp (Ibarra et al., 2006; Moldes and Vidal, 2008); as shown here, however, 0.5% odp HBT suffices to obtain acceptable results and provides a 66% saving of mediator. The optimum application conditions established can greatly facilitate integration of the laccase–mediator system in industrial processes. Finally, the best application conditions for the L stage were virtually identical for both sequences; however, XLE provided a smaller kappa number and higher brightness than LE. 4.4 68 0.0 0.5 1.0 X 2-mediator Fig. 3. Contour lines with a constant laccase dose (x1 = +1) for kappa number (–  –) and brightness (—) of LE (a) and XLE (b) sequences. Table 3 Selected points for L stage application according to LE and XLE sequences. LE XLE Laccase dose-x1 Mediator dose-x2 Reaction time-x3 Kappa number Brightness (%ISO) +1 (20 U g1 odp) +1 (20 U g1 odp) 1 (0.5%odp) 1 (0.5%odp) 0.2 (3.4 h) 0.2 (4.6 h) 5.2 4.0 68 69 2037 C. Valls, M.B. Roncero / Bioresource Technology 100 (2009) 2032–2039 a b Br LE X 2=+1 KN LE X2=+1 Br XLE X 2=-1 Br XLE X 2=0 KN XLE X2=-1 0.5 X3 -t im e 1 -l 0.0 1.0 65 60 55 -0.5 0.0 0.5 X3 -t im e -1.0 -0.5 0.0 0.5 1.0 ac ca se 3 -0.5 -1.0 -0.5 0.0 0.5 1.0 ac ca se 4 70 1 -l 5 1.0 X 6 O) Brightness (% IS 75 X Kappa number 7 ¢ Fig. 4. Surfaces estimated for kappa number at high HBT dose in LE (h) and at low HBT dose in XLE ( ) (a). Surfaces estimated for brightness at high HBT dose in LE (h) and at low or medium dose in XLE ( and j). vealed a slight, linear influence, which, as noted earlier, is consistent with the results of previous work (Garcia et al., 2003). Based on the observed effect of the reaction time in the XLE sequence, a limiting kappa number (3.4) was reached after 4.6 h; on the other hand, kappa number in the LE sequence decreased throughout the reaction time (1–7 h), but the smallest value obtained was 4.5. Therefore, the X pretreatment facilitated access to previously unreached lignin. This effect was also observed in previous studies that used an ozone treatment instead of a laccase system (Roncero et al., 2003c). Fig. 4a shows the calculated surfaces for the kappa number in the LE and XLE sequences at a constant mediator dose that was the highest (x2 = +1, 2.5% odp) for the LE sequence and lowest (x2 = 1, 0.5% odp) for the XLE sequence. As can clearly be seen, although the HBT dose was lower in XLE than in LE, the kappa number predicted by the model was also smaller for XLE. A similar effect was observed with the laccase dose and reaction time. Based on the brightness results for the LE sequence, at concentrations from 13 to 20 U g1 odp, laccase acted on other pulp components containing chromophoric groups (e.g., carbohydrates), which resulted in increased brightness. On the other hand, increasing the laccase dose from 12.4 to 20 U g1 odp, and the mediator dose from 1.7% to 2.5% odp, in the XLE sequence caused some coloured products to be formed. Since brightness levelled off while the kappa number continued to decrease, this effect can be ascribed to condensation of HBT with lignin or the formation of quinones. Brightness in the XLE sequence always increased with increasing reaction time; between 4.6 and 7 h, however, the system released some coloured compounds other than lignin as no further reduction in kappa number was observed. By contrast, the LE sequence seemingly produced chromophoric groups such as carbonyl and carboxyl or quinones, over the same period. Fig. 4b shows the calculated surfaces for brightness in the LE and XLE sequences at a constant HBT dose (x2) at its maximum value (x2 = +1, 2.5% odp) for LE, and its medium and minimum value in XLE (x2 = 0, 1.5% odp; x2 = 1, 0.5% odp). As can be seen, even the mediator dose presented high doses in LE more brightness was obtained in XLE at medium and low doses of HBT. A similar effect was observed with the laccase dose and reaction time. Therefore, the kappa number and brightness obtained suggest the possibility of using milder conditions with the laccase–media- tor system if a xylanase pretreatment is employed. Also, the pretreatment ultimately results in better pulp final properties. 3.7. Saving reagents and time with an X pretreatment Based on the results obtained with the laccase–mediator system under optimal conditions in the LE and XLE sequences, xylanase alters the behaviour of the variables in the L treatment by facilitating access to cellulose fibres. Therefore, an X pretreatment allows milder conditions to be used in the L stage. In order to assess the savings in reagent (laccase or mediator dose) or reaction time to be expected from the X pretreatment, a contour plot was constructed from the kappa number and brightness models for the XLE sequence (Fig. 5). The mediator dose was fixed at its minimum value (x2 = 1) because it was the primary target of the saving. The contour lines were limited to the smallest kappa number (4.6) and highest brightness (69%ISO) experimentally obtained in LE. As can be seen from the figure, reducing the mediator dose to its lowest value in XLE allowed similar or even smaller kappa numbers and higher brightness to be obtained. An application point was chosen by favouring a reduction in reaction time over one in laccase dose. One compromise point between both properties could be that of maximum brightness (69% ISO) crossing the kappa number line at 4.2, which corresponded to the conditions x1 = 0.4, x2 = 1 and x3 = 0. Table 4 shows the conditions of the variables at the chosen point, and also the LE conditions needed to obtain the same bright- 1.0 4.6 4.2 71 4.4 0.5 X 3-time ¢ 70 4.0 69 4.2 0.0 4.4 4.6 -0.5 -1.0 -1.0 -0.5 0.0 0.5 X1-laccase 1.0 Fig. 5. Contour lines for kappa number and brightness of XLE sequence when x2 = 1. Contour lines are limited to the lowest kappa number and to the highest brightness obtained in LE sequence. 2038 C. Valls, M.B. Roncero / Bioresource Technology 100 (2009) 2032–2039 Table 4 Saving of laccase, mediator and time with an X pretreatment. LE XLE Saving Saving (%) Laccase dose-x1 (U g1 odp) Mediator dose-x2 (% odp) Reaction time-x3 (hours) Kappa number Brightness (% ISO) (+1) 20 (0.4) 14.3 5.7 U g1 odp 28.5 (+1) 2.5 (1) 0.5 2% odp 80 (+1) 7 (0) 4 3 43 4.7 4.2 – – 68.8 68.9 – – ness level. As can be seen, for a target brightness of 69% ISO, an X pretreatment allows the laccase dose to be reduced from 20 to 14.3 U g1 odp, the mediator dose from 2.5% to 0.5% and the time from 7 to 4 h; these reduced values represent a saving of 29%, 80% and 43%, respectively. Moreover, XLE pulp exhibited a smaller kappa number than LE pulp at such a point. 3.8. Effects of the enzyme treatments on hexenuronic acids HexA are formed during the alkaline cooking of wood, where some 4-O-methyl glucuronic acids present in xylans are converted into unsaturated, hexenuronic acids through the loss of methanol (Danielsson et al., 2006). HexA are important inasmuch as they can adversely affect pulp bleaching by facilitating brightness reversion (Forsström et al., 2007). Also, they contribute to increasing the kappa number (Vuorinen et al., 1999). The initial eucalyptus pulp contained 38.0 ± 1 lmol HexA g1 odp. The HexA content was measured at high (laccase dose 20 U g1 odp, HBT dose 2.5% odp and reaction time 7 h) and low values (laccase dose 1 U g1 odp, HBT dose 0.5% odp and reaction time 7 h) of the variables. A long reaction was used with the low laccase and mediator doses in order to ensure that the acid medium in the L stage would not effect the removal of HexA (Chai et al., 2001). Using low values of the variables in the LE sequence removed HexA by only 10%, whereas using high values removed them by 22%. Moreover, the X stage substantially boosted HexA removal (by 44% of HexA were removed at high conditions in the XLE sequence). Therefore, using an X pretreatment prior to an LE sequence not only affords milder application conditions in the L stage, but also provides pulp with a reduced hexenuronic acid content which will provide more durable paper as a result. 4. Conclusions The operating conditions for a laccase–mediator treatment (an L stage) on oxygen-delignified eucalyptus kraft pulp were optimized in the presence and absence of a xylanase pretreatment by using a statistical plan for the laccase dose, HBT dose and reaction time. In this way, an LE sequence and an XLE sequence were conducted, X denoting the xylanase pretreatment and E an alkaline extraction stage. The models obtained for the kappa number and brightness showed that the system reached a point were some lignin in the pulp remained inaccessible and the system started to remove or alter other chromophoric compounds present in the pulp. The optimum points for the LE and XLE sequences were those involving the lowest HBT dose, highest laccase dose and a reaction time of 3.4 or 4.6 h (i.e., a reduction of 80% in HBT dose and 30–50% in reaction time). Also, the mediator dose was reduced by 66% with respect to the values used in previous studies. The xylanase pretreatment facilitates access to cellulose fibres, thereby boosting the effect of the laccase-mediator system in reducing the content of residual lignin and releasing more hexenuronic acids. Moreover, a similar brightness and smaller kappa number can be obtained by using 30% less laccase, 80% less HBT and a 45% shorter reaction time. Acknowledgements This work was funded within the framework of Spain’s MEC ENZPULP Project CTQ2005-08925-C02-01 and the BIORENEW Integrated European Project NMP2-CT-2006-026456. Dr. C. Valls is grateful to UPC for the research fellowship to finalize the PhD studies. Torraspapel S.A. (Zaragoza, Spain) is gratefully acknowledged for supplying the pulp used, and also the UB (Microbiology Department) and Novozymes (Bagsvaerd, Denmark) for supplying the xylanase and laccase used, respectively. References Amin, H.M., 2006. Extended usage of xylanase enzyme to enhance the bleaching of softwood kraft pulp. Tappi J. 5, 23–26. Bajpai, P., 2004. Biological bleaching of chemical pulps. Crit. Rev. Biotechnol. 24, 1– 58. Balakshin, M., Chen, C.L., Gratzl, J.S., Kirkman, A.G., Jakob., H., 1998. Kinetics of pulp delignification with laccase-mediator system. In: 5th Europ. Work. on Lignocel. and Pulp. Conf. Proc., pp. 585–588. Balakshin, M., Chen, C.L., Gratzl, J.S., Kirkman, A.G., Jakob, H., 1999. Biobleaching of pulp with dioxygen in the laccase–mediator system. Part 1. Kinetics of delignification. Holzforschung 54, 390–396. Balakshin, M., Chen, C.L., Gratzl, J.S., Kirkman, A.G., Jakob, H., 2001. Biobleaching of pulp with dioxygen in laccase–mediator system-effect of variables on the reaction kinetics. J. Molec. Catal. B: Enzyme 16, 205–215. Bourbonnais, R., Paice, M.G., Freiermuth, B., Bodie, E., Borneman, S., 1997. Reactivities of various mediators and laccases with kraft pulp and lignin model compounds. Appl. Environ. Microbiol. 63, 4627–4632. Camarero, S., Garcia, O., Vidal, T., Colom, J., del Rio, J.C., Gutierrez, A., Gras, J.M., Monje, R., Martinez, M.J., Martinez, A.T., 2004. Efficient bleaching of non-wood high-quality paper pulp using laccase–mediator system. Enzyme Microb. Technol. 35, 113–120. Chai, X.S., Zhu, J.Y., Li, J., 2001. A simple and rapid method to determine hexenuronic acid groups in chemical pulps. J. Pulp Pap. Sci. 27, 165–170. Chakar, F.S., Ragauskas, A.J., 2001. Formation of quinonoid structures in laccase– mediator reactions. ACS Symp. Series 785, 444–455. Danielsson, S., Kisara, K., Lindstrom, M.E., 2006. Kinetic study of hexenuronic and methylglucuronic acid reactions in pulp and in dissolved xylan during kraft pulping of hardwood. Ind. Eng. Chem. Res. 45, 2174–2178. Eriksson, T., Gierer, J., 1985. Studies on the ozonation of structural elements in residual kraft lignins. J. Wood Chem. Technol. 5, 53–84. Fillat, U., Vidal, T., Roncero, M.B., 2007. Application of an experimental design to modeling the laccase mediator stage in bleaching of flax pulp. 10th ICBPPI 119. Forsström, A., Wackerberg, E., Greschik, T., Jour, P., Holtinger, L., 2007. High brightness stability in standard ECF sequences. O Papel 68, 101–112. Freudenreich, J., Amann, M., Fritz-Langhals, E., Stohrer, J., 1998. Understanding the lignozym-process. IPBC 1, 71–76. Gallardo, O., Díaz, P., Pastor, F.I.J., 2004. Cloning and characterization of xylanase A from the strain Bacillus sp. BP-7: comparison with alkaline pI-low molecular weight xylanases of family 11. Curr. Microbiol. 48, 276–279. Gamelas, J.A.F., Pontes, A.S.N., Evtuguin, D.V., Xavier, A.M.R.B., Esculcas, A.P., 2007. New polyoxometalate–laccase integrated system for kraft pulp delignification. Biochem. Eng. J. 33, 141–147. Garcia, O., Camarero, S., Colom, J.F., Martinez, A.T., Martinez, M.J., Monje, R., Vidal, T., 2003. Optimization of a laccase–mediator stage for TCF bleaching of flax pulp. Holzforschung 57, 513–519. Garcia, O., Camarero, S., Colom, J.F., Martinez, A.T., Vidal, T., 2004. Formation of carbonyl groups in cellulose by a laccase–mediator system. 9th ICBPPI 185. Gellerstedt, G., Li, J., 1996. An HPLC method for the quantitative determination of hexenuronic acid groups in chemical pulps. Carbohyd. Res. 294, 41–45. Herpoël, I., Jeller, H., Fang, G., Petit-Conil, M., Bourbonnais, R., Robert, J.L., Asther, M., Sigoillot, J.C., 2002. Efficient enzymatic delignification of wheat straw pulp by a sequential xylanase-laccase mediator treatment. J. Pulp Pap. Sci. 28, 67–71. Ibarra, D., Camarero, S., Romero, J., Martinez, M.J., Martinez, A.T., 2006. Integrating laccase–mediator treatment into an industrial-type sequence for totally chlorine-free bleaching of eucalypt kraft pulp. J. Chem. Technol. Biotechnol. 81, 1159–1165. ISO standard, 1977. 3688, Pulps-measurement of diffuse blue reflectance factor (ISO brightness). C. Valls, M.B. Roncero / Bioresource Technology 100 (2009) 2032–2039 ISO standard, 1981. 302, Pulps-determination of kappa number. Kaneko, H., Hosoya, S., Liyama, K., Nakano, J., 1983. Degradation of lignin with ozone: reactivity of lignin model compounds toward ozone. J. Wood Chem. Technol. 3, 399–411. Moldes, D., Vidal, T., 2008. Laccase-HBT bleaching of eucalyptus kraft pulp. Influence of the operating conditions. Bioresour. Technol. 99, 8565–8570. Moldes, D., Díaz, M., Tzanov, T., Vidal, T., 2008. Comparative study of the efficiency of synthetic and natural mediators in laccase-assisted bleaching of eucalyptus kraft pulp. Bioresour. Technol. 99, 7959–7965. Oudia, A., Meszaros, E., Simoes, R., Queiroz, J., Jakab, E., 2007. Pyrolysis-GC/MS and TG/MS study of mediated laccase biodelignification of eucalyptus globulus kraft pulp. J. Anal. Appl. Pyrol. 78, 233–242. Pedrola, J., Roncero, M.B., Colom, J.F., Vidal, T., Torres, A.L., 2004. Application of an experimental design to modeling the hydrogen peroxide stage in TCF bleaching of eucalypt pulp. Appita J. 57, 141–145. Popovici, C., Messier, M., Thibault, L., Charron, D., 2004. Multiples avantages du xylanase dans une usine nexfor papiers fraser de pâte kraft de feuillus. Pulp. Pap. Can. 105, 78–281. Poppius-Levlin, K., Wang, M., Tamminen, T., Hortling, B., Viikari, L., i Niku-Paavola, M.L., 1999. Effects of laccase/HBT treatment on pulp and lignin structures. J. Pulp Pap. Sci. 25, 90–94. Potthast, A., Rosenau, T., Fischer, K., 2001. Oxidation of benzyl alcohols by the laccase–mediator system (LMS) – a comprehensive kinetic description. Holzforschung 55, 47–56. Roncero, M.B., Torres, A.L., Colom, J.F., Vidal, T., 2000a. Effects of a xylanase treatment on fibre morphology in totally chlorine-free bleaching (TCF) of Eucalyptus pulp. Proc. Biochem. 36, 45–50. 2039 Roncero, M.B., Colom, J.F., Vidal, T., Queral, M.A., 2000b. The role of oxalic acid in the ozone bleaching kinetics of an XO-kraft pulp. J. Wood Chem. Technol. 20, 147–167. Roncero, M.B., Torres, A.L., Colom, J.F., Vidal, T., 2003a. TCF bleaching of wheat straw pulp using ozone and xylanase. Part A: paper quality assessment. Bioresour. Technol. 87, 305–314. Roncero, M.B., Torres, A.L., Colom, J.F., Vidal, T., 2003b. TCF bleaching of wheat straw pulp using ozone and xylanase. Part B: kinetic studies. Bioresour. Technol. 87, 315–323. Roncero, M.B., Torres, A.L., Colom, J.F., Vidal, T., 2003c. Effect of xylanase on ozone bleaching kinetics and properties of eucalyptus kraft pulp. J. Chem. Technol. Biotechnol. 78, 1023–1031. Roncero, M.B., Torres, A.L., Colom, J.F., Vidal, T., 2005. The effect of xylanase on lignocellulosic components during the bleaching of wood pulps. Bioresour. Technol. 96, 21–30. Sigoillot, C., Camarero, S., Vidal, T., Record, E., Asther, M., Perez-Boada, M., Martinez, M.J., Sigoillot, J.C., Asther, M., Colom, J.F., Martinez, A.T., 2005. Comparison of different fungal enzymes for bleaching high-quality paper pulps. J. Biotechnol. 115, 333–343. Tessier, P., Savoie, M., 2000. Chlorine dioxide bleaching kinetics of hardwood kraft pulp. Tappi J. 83, 1–13. Torres, A.L., Roncero, M.B., Colom, J.F., Martínez, J.A., Vidal, T., 2004. Application of an experimental design to modeling of ozone bleaching stage in TCF processes. Ozone Sci. Eng. 26, 443–451. Vuorinen, T., Fagerström, P., Buchert, J., Tenkanen, M., Teleman, A., 1999. Selective hydrolysis of hexenuronic acid groups and its application in ECF and TCF bleaching of kraft pulps. J. Pulp Pap. Sci. 25, 155–162.