- Email: [email protected]
In-depth investigation of enzymatic hydrolysis of biomass wastes based on three major components: Cellulose, hemicellulose and lignin
Bioresource Technology 101 (2010) 8217–8223
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
In-depth investigation of enzymatic hydrolysis of biomass wastes based on three major components: Cellulose, hemicellulose and lignin Lili Lin a,b, Rong Yan b,*, Yongqiang Liu b, Wenju Jiang a a b
College of Architecture and Environment, Sichuan University, Chengdu 610065, PR China Institute of Environmental Science and Engineering, Nanyang Technological University, Innovation Center, Block 2, Unit 237, 18 Nanyang Drive, Singapore 637723, Singapore
a r t i c l e
i n f o
Article history: Received 10 March 2010 Received in revised form 20 May 2010 Accepted 26 May 2010 Available online 19 June 2010 Keywords: Enzymatic hydrolysis Biomass components Pretreatment Artificial biomass
a b s t r a c t The artificial biomass based on three biomass components (cellulose, hemicellulose and lignin) were developed on the basis of a simplex-lattice approach. Together with a natural biomass sample, they were employed in enzymatic hydrolysis researches. Different enzyme combines of two commercial enzymes (ACCELLERASE 1500 and OPTIMASH™ BG) showed a potential to hydrolyze hemicellulose completely. Negligible interactions among the three components were observed, and the used enzyme ACCELLERASE 1500 was proven to be weak lignin-binding. On this basis, a multiple linear-regression equation was established for predicting the reducing sugar yield based on the component proportions in a biomass. The hemicellulose and cellulose in a biomass sample were found to have different contributions in staged hydrolysis at different time periods. Furthermore, the hydrolysis of rice straw was conducted to validate the computation approach through considerations of alkaline solution pretreatment and combined enzymes function, so as to understand better the nature of biomass hydrolysis, from the aspect of three biomass components. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction During the past years, researches on converting lignocellulosic biomass into bioethanol are actively undertaken, aiming to produce the 2nd generation biofuel which has no competition with food and is thus sustainable. The most common processing of lignocellulosics to bioethanol consists of four major unit operations: pretreatment of raw materials, enzymatic hydrolysis of pretreated materials into fermentable sugars, fermentation of fermentable sugars into ethanol, and ethanol separation or purification. It has been well known that hydrolyzing lignocellulosics to reducing sugars is one of the crucial factors to affect the cost of bioethanol production. Lignocellulosic biomass is mainly composed of cellulose, hemicellulose, and lignin, along with smaller amounts of pectin, protein, extractives, and ash (Blasi et al., 1999). The composition of these constituents can vary from one plant species to another. Normally, cellulose, hemicellulose, and lignin cover 40–60, 20–40, and 10–25 wt.% of biomass materials on dry basis, respectively. Among these components in biomass materials, only hemicellulose and cellulose can be converted into fermentable sugars by using microbial cellulolytic and xylanolytic enzymes which both have demonstrated promising results for the subsequent fermentation. * Corresponding author. E-mail address: [email protected] (R. Yan). 0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2010.05.084
However, the enzymatic digestion of cellulose and hemicellulose in biomass substrates is always limited by the complex structure and components of natural biomass (Chang and Holtzapple, 2000; Laureano-Perez et al., 2005; Puri, 1984). Among the primary substrate features impacting enzymatic conversion of biomass, lignin is believed to impede enzyme access to glucan chains by its protective sheathing and also reduce cellulase effectiveness as a result of unproductive binding and steric hindrance (Chang and Holtzapple, 2000; Mansfield et al., 1999). For lignocellulosic substrates, unproductive binding and inactivation of enzymes by the lignin component appear to be important factors limiting catalytic efficiency (Berling et al., 2005). Therefore, to achieve an effective hydrolysis of lignocellulosics, it is necessary and important to deeply understand whether enzyme preparations (cellulase and hemicellulase) are weak lignin-binding or strong lignin-binding; and whether the different combinations of biomass components in substrates have a significant influence on enzyme activity during hydrolysis of biomass wastes. Although hydrolysis performance of single biomass component (such as cellulose or hemicellulose) has been largely reported, systematic studies on evaluation of enzyme activity in hydrolyzing different components of biomass wastes (cellulose, hemicellulose, lignin) at different ratios have rarely been seen. As the three components consist of the major portion of biomass and each component has different chemical/physical structures, it would be interesting to find out if there is any interaction between the three
8218
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
components during hydrolysis and if there is a linkage between the enzyme activity and combinations of biomass components of any specific biomass waste. This could substantively conduce to determination on what kind of pretreatment method should be adopted, whether lignin content in raw material should be effectively reduced and what would be the most rigorous factor to inhibit the efficiency of enzymatic hydrolysis. In this work, cellulose, hemicellulose, and lignin were used as pure biomass components. The enzymatic hydrolysis behaviors of both cellulose and hemicellulose were first investigated by commercial enzyme preparation. Secondly, artificial biomass samples containing two or three of the biomass components (cellulose, hemicellulose and lignin) were developed on the basis of a simplex-lattice approach and thereafter hydrolyzed by enzyme preparation. The aim of this work is to understand the different behaviors of three biomass components in hydrolysis and their potential interaction. On this basis, the enzymatic hydrolysis of natural biomass rice straw was carried out. Alkaline pretreatment of rice straw was adopted and the change of biomass components in rice straw before and after the treatment was identified. Thereafter, the fundamental difference of natural biomass and artificial samples during hydrolysis was compared. 2. Methods 2.1. Materials The basic structure of lignocellulose consists of cellulose (C5H10O5)x, hemicellulose such as xylan (C5H8O4)m, and lignin [C9H10O3(OCH3)0.91.7]n (Demirbas, 2005). In this study, microcrystalline cellulose was used to study enzymatic hydrolysis performance of cellulose. It is in a microcrystalline powder form from wood pulp. Commercial hemicellulose can hardly be purchased, xylan, which is the dominant component of hemicellulose from agricultural plants, was chosen as a representative of hemicellulose in this research. Here, xylan is in a yellow powder form processed from birch wood. Lignin used in this work is alkali lignin in a brown powder form. The characteristics of celluloses, hemicellulose and lignin are shown in Table 1. Rice straw was used as a typical biomass in this work. It was washed with tap water to remove dust and dried in a 105 °C oven till constant weight. Then it was ground in a laboratory-scale centrifugal mill (Rocklabs, New Zealand) and sieved in a Retsch test sieve with a 40-mesh screen (Retsch, Fisher Scientific Company, USA). Furthermore, the dried rice straw was pretreated by alkalineperoxide method (Mishima et al., 2006). A mixed solution of 2% (w/ v) NaOH and 2% (v/v) H2O2 (1:1) was used. Each gram of raw sample was soaked in 10 ml of the mixture solution at room temperature for 24 h. After filtration, the solid residue was rinsed repeatedly with distilled water until the pH value came to neutral. After that, it was dried at 105 °C till a constant weight reached and then sieved through a 40-mesh screen. These two rice straw samples, un-treated and treated, were taken as the feedstocks for enzymatic hydrolysis.
Table 1 The characteristics of cellulose, hemicellulose and lignin used in the experiment. Sample
Source
Particle size (lm)
Brand
Cellulose, microcrystalline Xylan (hemicellulose)
Wood pulp
<50
Alfa
Birch wood
100
Lignin
Birch wood
100
Sigma– Aldrich Alfa
2.2. Chemical analysis of natural biomass To determine the amount of extractives in biomass (rice straw), solvent extraction (60 ml acetone for 1 g of dried biomass sample) was used, and the temperature was held at 90 °C for 2 h. After that, the sample was dried at 105 °C until a constant weight was obtained. The weight difference before and after the extraction is the amount of the extractives (Blasi et al., 1999). To determine the amount of hemicellulose, 10 ml 0.5 mol/l of sodium hydroxide solution was added to 1 g of extractive-free dried biomass, and the temperature was held at 80 °C for 3.5 h. After that, the sample was washed using DI water until pH value of the solution approach 7, then it was dried to a constant weight. The difference between the sample weight before and after this treatment is hemicellulose content (Blasi et al., 1999). To determine the amount of lignin, 30 ml of 98 wt.% sulfuric acid was added for each extractive-free dried biomass. After the sample was held at ambient temperature for 24 h, it was boiled at 100 °C for 1 h. The mixture was filtered, and then the residue was washed until the sulfate ion in the filtrate was undetectable (via titration of a 10% barium chloride solution); it was then dried to a constant weight. The weight of the residue was recorded as the lignin content (Blasi et al., 1999). The content of cellulose was calculated by difference, assuming that extractives, hemicellulose, lignin, and cellulose are the only components of the entire biomass. (Blasi et al., 1999; Li et al., 2004). 2.3. Preparation of artificial biomass samples Artificial biomass samples were synthesized using the three major components (microcrystalline cellulose, hemicellulous and lignin), on the basis of the simplex-lattice design devised by Scheffe (Gorman and Hinman, 1962). The method has two key features: (1) the fractions of components making up any mixture must add to unity, and hence factor space may be represented by a regular simplex; and (2) points of the composition in this simplex are explored in accordance with a lattice arrangement. The key issue is to choose a representative and reproducible proportion of the three components. For a mixture sample, the component proportions should satisfy the constraints, that is:
8 1 2 m > < X i ¼ 0; m ; m ; . . . ; m ; i ¼ 1; . . . ; s X1 þ X2 þ . . . þ Xs ¼ 1 > : k ¼ ðsþm1Þ! m!ðs1Þ!
where s is the number of components under consideration and m is an integer related to the spacing of the points and therefore to the number of mixture. The proportions of any component Xi can take value from zero to unity, and all possible mixtures with these proportions are used. To support a polynomial model of degree m in s components over the simplex, the lattice, referred to as a {s, m} simplex lattice, consists of all possible combinations of the components. The number of design points in the {s, m} simplex lattice is equal to k. For the mixture of cellulose, hemicellulose, and lignin, s = 3 (three components). To guarantee the accuracy and keep the computing simple, m was set to 4, so the total number of point k was 15. They are shown in Fig. 1 (solid dots). Only three artificial samples involved all the three components while others involved one or two components. The biomass samples were synthesized by dry mixing the three components according to the specific ratio. In the following text, H:C indicated the weight ratio of hemicellulose to cellulose, L:C and L:H represented those of lignin to cellulose and lignin to hemicellulose, respectively, and L:H:C was the weight ratio of a mixture of lignin, hemicellulose and cellulose.
8219
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
0.00
(w t,% )
Hydrolysis rate of substrate ð%Þ
se
0.50
0.25
1.00 0.25
0.50
0.75
0.00 1.00
Hemicellulose (wt, %) Fig. 1. Proportion of cellulose, hemicellulose, and lignin in the artificial biomass samples.
2.4. Enzyme Two types of enzymes, ACCELLERASE 1500 and OPTIMASH™ BG were gifts from Genencor International Incorporation. ACCELLERASE 1500 (referred as Cellulase) enzyme contains a mixture of cellulase, hemicellulase and higher level of b-glucosidase enzyme activities. It is a brown liquid and the activity of ACCELLERASE 1500 enzyme complex is expressed in carboxymethylcellulose (CMC U) activity units. The activity of endoglucanase reported was 2200–2800 CMC U/g, and one CMC U unit of activity liberated 1 lmol of reducing sugar (expressed as glucose equivalents) in one minute under specific assay conditions of 50 °C and pH 4.8. The reported Beta-glucosidase activity was 525–775 pNPG U/g, with one pNPG unit denoted 1 lmol of Nitrophenol liberated from paranitrophenyl-b-D-glucopyranoside per minute at 50 °C and pH 4.8. OPTIMASH™ BG (referred as Xylanase) is a mixture of Xylanase and b-Glucanase which was used together with ACCELLERASE 1500 to enhance enzymatic hydrolysis efficiency of hemicellulose. It is in an appearance of light amber liquid and has an activity of 10300 CMSC U/g. 2.5. Enzymatic hydrolysis Enzymatic hydrolysis experiments of biomass were carried out in 250 ml of stoppered conical flasks. A specific amount of enzyme was added into the substrates in a total 100 ml reaction volume. The pH of reaction system was kept at 5.0 with 0.05 N citric acid–sodium citrate buffer. The flasks were incubated at 50 °C on an orbital shaker agitated at 150 rpm. In all experiments, the citric acid buffer was supplemented with antibiotics tetracycline (40 lg/ ml) to prevent microbial contamination. Liquid samples were withdrawn from the reaction media periodically, centrifuged and filtered. The filtrate was used to determine the concentration of reducing sugar based on the DNS method using 3,5-dinitrosalicylic acid reagent (Miller, 1959). The natural biomass experiment was carried out at 50 °C and pH 5.0 for 72 h. Firstly, the un-treated rice straw sample was hydrolyzed by Cellulase at the dosage of 0.3 ml for each gram of total cellulose and hemicellulose in substrate. Secondly, in order to make cellulose and hemicellulose more accessible to enzyme, the natural biomass was pretreated by alkaline-peroxide solution and subsequently hydrolyzed by Cellulase at the dosage of 0.3 ml for total cellulose and hemicellulose in each gram substrate. Thirdly, the pretreated rice straw sample was hydrolyzed by combined enzymes of Cellulase and Xylanase at the dosage ratio of Cel-
¼
Reducing sugar 0:9 100 Cellulose þ hemicelluloseðin substrateÞ
ð1Þ
3. Results and discussion 3.1. Influence of enzyme loading on the yield of reducing sugar based on cellulose or hemicellulose substrates ACCELLERASE 1500 was used at specific concentrations to investigate the enzymatic hydrolysis performances of cellulose and hemicellulose. The yields of reducing sugar from the two samples generated at different enzyme loadings after 72 h duration were determined. It was found that hemicellulose can also be effectively hydrolyzed by enzyme ACCELLERASE 1500 and has slightly lower reducing sugar yield than microcrystalline cellulose at the same conditions of enzyme loading and reaction time. On the other hand, the hydrolysis efficiency of substrates (cellulose or hemicellulose) increased sharply at beginning but gradually flatted out when the enzyme concentration exceeds a certain level (0.3 ml/g substrate) where 20% of substrates were still available. This phenomenon can be explained in two aspects: (1) the surface area of cellulose or hemicellulose is not sufficient, and thus, the excess enzyme molecules are adsorbed to form multiple layers. Only the enzyme adsorbed in the first layer participates in hydrolysis; (2) the surface area of enzyme is composed of active and inactive fractions, only those biomass molecule adsorbed on the active fraction participate in the hydrolysis (Lee and Fan, 1982). The optimal Cellulase loading for celluloses or hemicellulose was in the scale from 0.2 to 0.4 ml/g substrate. Furthermore, hemicellulose was hydrolyzed by the mixture of Cellulase and Xylanase enzymes. Here, the concentration of Cellulase was referred to 0.3 ml/g hemicellulose in substrate according to previous findings. The concentration of Xylanase was first tested at two different levels: lower level of 0.05 ml/g hemicellulose and higher level of 0.15 ml/g hemicellulose. The reducing sugar yield of hemicellulose hydrolyzed by the mixture of Cellulase and Xylanase is shown in Fig. 2. It was evidenced that the reducing sugar yield of hemicellulose was effectively enhanced by adding Xylanase. When
Yield of reducing sugar (mg/g Xylan)
0.50
) t,% (w
nin
0.75
o llul
Lig
Ce
0.25
0.75
0.00
lulase to Xylanase equal to 0.3 ml:0.05 ml according to previous experiments. Each experiment was performed in triplicates. The hydrolysis rate of substrate was calculated from the following formula (Soto et al., 1994):
1.00
1000 800 600 400 Cellulase: Xylanase = 0.3ml:0.15ml Cellulase: Xylanase = 0.3ml:0.05ml Cellulase = 0.3ml
200 0 0
24
48
72
Hydrolysis time (h) Fig. 2. The influence of additive BG on the hydrolysis of xylan. Substrate concentration 1% (w/v), temperature 50 °C, pH 5.0.
8220
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
that of hemicellulose in the first 24 h and it increased gradually up to 48 and 72 h. This is possibly because hemicellulose has a random amorphous structure with little strength and it can be easily hydrolyzed by enzyme into glucose (6-carbon sugar) and pentose (5-carbon sugars). In contrast, the cellulose molecule is a very long polymer of glucose units without branches, and it is crystalline, strong, and resistant to hydrolysis. (2) there is no obvious interaction observed between hemicellulose and cellulose during enzymatic hydrolysis. The presence of one component in artificial biomass did not exert a negative impact on the hydrolysis of the other. At the time point of 72 h, the artificial samples of cellulose and hemicellulose at different ratios had demonstrated a decreasing order of hydrolysis efficiency as: C:H = 1:0 > C:H = 3:1 > C:H = 1:1 > C:H = 1:3 > C:H = 0:1.
the volume ratio of Cellulase to Xylanase was equal to 0.3:0.05 for one gram hemicellulose substrate, the reducing sugar yield was elevated more than 15% during initial 8 h reactivity. Moreover, the reducing sugar yield of hemicellulose tended to increase as the concentration level of Xylanase increased. Therefore, the combination of two different enzymes demonstrated a better hydrolysis efficiency with a potential of completely hydrolyzing hemicellulose into reducing sugars. 3.2. Enzymatic hydrolysis of artificial biomass In this experiment, artificial biomass substrates were synthesized by microcrystalline cellulose, hemicellulose and lignin. The Cellulase loading 0.3 ml/g of total cellulose and hemicellulose in substrates was employed. The total reaction time was controlled in 72 h.
3.2.2. Mixture of cellulose/lignin and of hemicellulose/lignin The enzymatic hydrolysis of cellulose (C) and lignin (L) mixtures and that of hemicellulose (H) and lignin (L) mixtures were also carried out. The results are similar for C + L and H + L systems and only the case of C + L is given in Fig. 4, as an example. The ratios of the mixture were decided following the simplex-lattice in Fig. 1 (i.e., both C:L and H:L are equal to 1:0, 3:1, 1:1, 1:3, and 0:1). The reducing sugar yield curves of both mixtures (C + L and H + L) showed the same trends with that of pure cellulose and hemicellulose substrates, respectively. From Fig. 4(a), the higher the ratio of cellulose or hemicellulose in both mixtures, the larger the reducing sugar yield of the substrates in a proportional pattern. The contents or ratios of lignin in the two mixtures had hardly any effect on the hydrolysis rate of the substrates as seen from Fig. 4(b).
1000
Hydrolysis rate of substrate
Yield of reducing sugar (mg/g of substrate)
3.2.1. Mixture of cellulose and hemicellulose The enzymatic hydrolysis curves of the artificial biomass samples of cellulose (C) and hemicellulose (H) are plotted in Fig. 3(a) and (b). Here, the artificial biomass samples were from the simplex-lattice (3, 4) design as given in Fig. 1 (C:H = 1:0, 3:1, 1:1, 1:3, and 0:1), in total 5 samples. The enzyme dosage at 0.3 ml/g of substrate was added to each sample. Two distinct observations were found from Fig. 3(a) and (b): (1) the hydrolysis rate of hemicellulose was much faster than that of cellulose during initial 24 h and almost >80% of the total reducing sugars (800 mg/g at 72 h) were generated in the first 24 h. In terms of cellulose hydrolysis, it showed relatively lower yield of sugar and hydrolysis rate than
800 600 400
C:H=1:0 C:H=3:1 C:H=1:1 C:H=1:3 C:H=0:1
200 0
0
24
48
72
80%
60%
40% C:H=1:0 C:H=3:1 C:H=1:1 C:H=1:3 C:H=0:1
20%
0%
0
Hydrolysis time (h)
24
48
72
Hydrolysis time (h)
(a) Yield
(b) Hydrolysis rate
1000 C:L=1:0 C:L=3:1 C:L=1:1 C:L=1:3
800 600 400 200 0 0
24
48
Hydrolysis time (h)
(a) Yield
72
Hydrolysis rate of substrate
Yield of reducing sugar (mg/g of substrate)
Fig. 3. Mixture of cellulose and hemicellulose at different ratios. Substrate concentration 1% (w/v), enzyme loading 0.3 ml/g of cellulose and hemicellulose, temperature 50 °C, pH 5.0.
80% 60% 40% C:L=1:0 C:L=3:1 C:L=1:1 C:L=1:3
20% 0% 0
24
48
72
Hydrolysis time (h)
(b) Hydrolysis rate
Fig. 4. Mixture of cellulose and lignin at different ratios. Substrate concentration 1% (w/v), enzyme loading 0.3 ml/g of cellulose in substrate, temperature 50 °C, pH 5.0.
8221
1000
Hydrolysis rate of substrate
Yield of reducing sugar (mg/g of substrate)
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
C:H:L=2:1:1 C:H:L=1:2:1 C:H:L=1:1:2
800 600 400 200 0 0
24
48
80% 60% 40% 20%
72
0%
C:H:L=2:1:1 C:H:L=1:2:1 C:H:L=1:1:2
0
24
Hydrolysis time (h)
48
72
Hydrolysis time (h)
(a) Yield
(b) Hydrolysis rate
Fig. 5. Mixture of cellulose, hemicellulose and lignin at different ratios. Substrate concentration 1% (w/v), enzyme loading 0.3 ml/g of cellulose in substrate, temperature 50 °C, pH 5.0.
The hydrolysis rate of both mixtures (C + L and H + L) was clearly determined by the cellulose or hemicellulose content in the substrates. It is obvious that the existence of free lignin has no inhibition on the enzymatic hydrolysis of cellulose and hemicellulose. This is possibly that lignin can not bind with the active sites of enzyme so cellulose or hemicellulose in the substrates is fully exposed to enzyme during enzymatic hydrolysis. Therefore, it is postulated for artificial biomass that lignin in substrate has negligible effect on the enzymatic hydrolysis of cellulose or hemicellulose using the enzyme complex ACCELLERASE 1500.
y ¼ b0 þ
3 X
bi xi þ
3 XX
i¼1
bij xi xj þ b123 x1 x2 x3
ð2Þ
ij
where x1, x2, and x3 are the proportions of cellulose, hemicellulose, and lignin, respectively, in biomass and y is the reducing sugar yield of substrate (mg/g of substrate) at different reaction times. If assume there is no interaction among the three components and lignin has no influence on the activity of enzyme, the Eq. (2) can be simplified to Eq. (3)
y ¼ b0 þ b1 x1 þ b2 x2 3.2.3. Mixture of cellulose, hemicellulose, and lignin at different ratios According to the proportion of cellulose (C), hemicellulose (H), and lignin (L) given in Fig. 1, the C:H:L ratios should be 2:1:1, 1:2:1, and 1:1:2 in the artificial biomass samples containing three components. The reducing sugar yields of the artificial samples containing the three components are shown in Fig. 5. Quite similar pattern of the hydrolysis rate curves were observed in Fig. 5, in comparison to Fig. 3 where only two components (hemicellulose and cellulose) were in the mixtures. A comparison of each counterpart in Figs. 5 and 3(b) showed that the addition of lignin into the mixture of hemicellulose and cellulose didn’t cause an evident change of enzymatic hydrolysis rate. Although in Fig. 3 the superior hydrolysis of hemicellulose over cellulose was observed, the yields of two artificial biomass (2:1:1 and 1:2:1) in Fig. 5 were close to each other and their hydrolysis rates showed negligible difference. Nevertheless, the yield of reducing sugar was dramatically reduced with the increasing ratio of lignin from 25% to 50% in substrate in Fig. 5(a). At the time point of 72 h, the artificial samples of cellulose, hemicellulose and lignin at different ratios had demonstrated a decreasing order of reducing sugar yield as: C:H:L = 2:1:1 > C:H:L = 1:2:1 > C:H:L = 1:1:2. From the above analysis, it is clear that almost no significant interaction occurred between cellulose, hemicellulose and lignin during enzymatic hydrolysis of artificial biomass. Consequently, it could be concluded that the enzymatic hydrolysis performance of biomass based on ACCELLERASE 1500 enzyme could be regarded as a simple superposition of cellulose and hemicellulose. The presence of lignin in biomass sample even with high fraction has not noticeable inhibition on enzyme activities of ACCELLERASE 1500 during hydrolysis. 3.3. Computing approach for predicting reducing sugar yield of substrate A general regression function was derived to fit the data collected from the designed {3, 4} simplex lattice. The polynomial equation for the three components can be rewritten as
ð3Þ
A linear multiple regression model was introduced to obtain the coefficients (b0, b1, b2), using the data from the artificial biomass. The values of coefficients obtained at different times are shown in Table 2. The correlation coefficient R2 was calculated using Eq. (4), and the results are also given in Table 2
P
2
R ¼1
ðyexp ycalcd Þ2 P 2 yexp
ð4Þ
where yexp is the experimentally observed reducing sugar yield and ycalcd is the calculated reducing sugar yield. From Table 2 it can be observed that a linear relationship was well established between the reducing sugar yield of artificial biomass based on three biomass components and the composition of cellulose and hemicellulose in the biomass, with all the R2 larger than 0.99. The assumption of negligible interaction among three biomass components and little contribution from lignin to sugar productivity was validated from the results of artificial biomass samples. In a closer look, the coefficients (b0, b1, b2) could represent, respectively, the background, weights of contribution from cellulose and hemicellulose. In the initial 24 h duration of hydrolysis, hemicellulose (x2), with a higher coefficient than cellulose (b2 > b1) played a more important role towards the formation of reducing sugar of biomass substrate. Afterwards, in the extended reaction time up to 72 h, cellulose (x1), with a higher coefficient than hemicellulose (b1 > b2), made a more significant contribution towards the formation of reducing sugar of biomass substrate. This
Table 2 The regression coefficients and correlation coefficient R2 at different reaction times. Reaction time (h)
b0
b1
b2
R2
3 8 24 48 72
13.2 1.3 1.43 1.1 3.1
257.9 425.4 609.4 762.1 818.3
463.0 577.3 668.3 743.1 784.3
0.9958 0.9991 0.9994 0.9999 0.9998
1000 800 600 400 200 0 0
24
48
72
Hydrolysis rate of substrate
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
Yield of reducing sugar(mg/g of substrate)
8222
100% 80% 60% 40% 20% 0% 0
Hydrolysis time (h)
24
48
72
Hydrolysis time (h)
(a) Yield
(b) Hydrolysis rate
Fig. 6. Enzymatic hydrolysis of rice straw. Substrate concentration 1% (w/v), enzyme loading 0.3 ml/g of cellulose in substrate, temperature 50 °C, pH 5.0.
result indicated the different enzyme activities of ACCELLERASE 1500 during staged hydrolysis of artificial biomass in degrading subsequently hemicellulose and cellulose.
3.4. Enzymatic hydrolysis of natural biomass Nevertheless, the artificial biomass samples of the three components could still be different from the natural biomass. In terms of natural biomass structure, cellulose and hemicellulose are usually found in close association with lignin. Lignin appears to limit cellulose and hemicellulose hydrolysis by forming a physical barrier that restricts enzyme access (Mooney et al., 1998). To further understand the hydrolysis behavior of natural biomass from the aspect of three biomass components, rice straw was used as a typical raw material hydrolyzed by Cellulase and Xylanase. The reducing sugar yield and the hydrolysis rate of natural biomass samples are plotted in Fig. 6(a) and (b), and the component analysis of un-treated rice straw and the reducing sugar yield after 72 h reaction time are given in Table 3. There yexp is the experimentally observed reducing sugar yield and ycalcd is the calculated reducing sugar yield according to Eq. (3) and Table 2. For raw natural biomass, yexp was much lower than ycalcd because the complex structure is unfavorable to the contact of enzyme and cellulose and hemicellulose, although the same ratio of the three components (cellulose, hemicellulose and lignin) as in natural samples was adopted in calculation. As mentioned before, these three components in natural biomass are associated closely with each other, which make cellulose and hemicellulose less exposed to enzyme digestion than in the case of artificial biomass where the structure combination of three components could be weak and they were almost completely exposed to cellulose and hemicellulose. For the pretreated rice straw, both the reducing sugar yield and the hydrolysis rate were largely improved than un-treated sample. This can be explained by two aspects: (1) the significant change of contents of the three major components (cellulose, hemicellulose and lignin) in natural biomass sample before and after pretreat-
Table 3 The components and reducing sugar yield of rice straw before and after pretreatment. Sample
Rice straw (raw)
Rice straw (pretreated)
Extractives (wt.%, dry) Hemicellulose (wt.%, dry) Lignin (wt.%, dry) Cellulose (wt.%, dry) ycalcd (mg/g) yexp (mg/g)
7.0 43.9 29.3 19.8 503 221
3.3 29 16.2 51.5 646 714
ment. From Table 3, the content of cellulose was significantly enhanced after pretreatment and the content of lignin and hemicellulose was decreased. The increase of total cellulose and hemicellulose content (from 63.7% to 80.5%) in natural biomass substrate after pretreatment demonstrated a positive effect on the enhancement of reducing sugar yield of substrate, which was in coincidence with the results of artificial biomass; (2) the alteration of the structure of biomass sample after pretreatment. The sample in the original form was relatively resistant to microbial attack, but the removal of substantive lignin and partial hemicellulose after alkaline pretreatment caused extensive changes in the structure and accessibility of cellulose which could allow more open to swelling on contact with enzyme (Mansfield et al., 1999). Therefore, as evidenced in this study again, the reducing sugar yield of pretreated rice straw sample by alkaline-peroxide pretreatment was 3-folders higher than the un-treated one and the hydrolysis rate of the former was 2-folders higher than the latter. Moreover, dilute alkaline solution pretreatment for biomass not only caused separation of structural linkages between the three major components each other, but also led to an increase in internal surface area of biomass sample, a decrease in the degree of polymerization, and a decrease in crystallinity (Kumar et al., 2009). For pretreated biomass, the yexp was more close to ycalcd than the case of un-treated sample since the influence of lignin in biomass was diminished by alkaline solution pretreatment, but still slightly higher due to possibly the above-mentioned structure difference. In the case of combination of different-purpose enzymes, the addition of Xylanase largely increased the final yield of reducing sugar nearly up to 95% in 24 h due to further hydrolysis of hemicellulose in treated biomass sample. In conclusion, the combine of Cellulase and Xylanase was efficient in enhancing both hydrolysis rate of biomass and reducing sugar yield. Lignin showed negligible effect on the activity of enzyme ACCELLERASE 1500 during hydrolysis of cellulose and hemicellulose, which was another proof of the earlier assumption of ‘‘weak lignin-binding” enzyme ACCELLERASE 1500. 4. Conclusions The enzymatic hydrolysis behaviors of biomass wastes were investigated from the aspect of three biomass components (cellulose, hemicellulose and lignin). Based on experiments of the systematically-designed artificial biomass samples, negligible interactions among the three components were observed. The prediction of biomass hydrolysis yield was therefore made with the assist of a multiple linear-regression model, which was well con-
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
sistent with experimental results. Hemicellulose and cellulose were found to play different roles in staged hydrolysis at different time periods. These findings were validated through hydrolyzing rice straw, a natural biomass. The change of hydrolysis behavior of rice straw before and after alkaline pretreatment was well corresponding to the model prediction, in terms of biomass components contributions. Acknowledgements The author, Lili Lin, is grateful for the collaboration program between the Institute of Environmental Science and Engineering, Nanyang Technological University, Singapore and the College of Architecture and Environment, Sichuan University, China, which provides her the opportunity to pursue her PhD in Singapore under a research scholarship. References Berling, A., Gilkes, N., Kurabi, A., Bura, R., Tu, M., Kilburn, D., Saddler, J., 2005. Weak lignin-binding enzymes. Appl. Biochem. Biotechnol. 121 (1–3), 163–170. Blasi, C.D., Signorelli, G., Di Russo, C., Rea, G., 1999. Product distribution from pyrolysis of wood and agricultural residues. Ind. Eng. Chem. Res. 38 (6), 2216– 2224. Chang, V.S., Holtzapple, M.T., 2000. Fundamental factors affecting biomass enzymatic reactivity. Appl. Biochem. Biotechnol. 84–86, 5–37.
8223
Demirbas, A., 2005. Estimating of structural composition of wood and non-wood biomass samples. Energy Sour. 27 (8), 761–767. Gorman, J.W., Hinman, J.E., 1962. Simplex lattice designs for multicomponent systems. Technometrics 4 (4), 463–487. Kumar, R., Mago, G., Balan, V., Wyman, C.E., 2009. Physical and chemical characterizations of corn stover and poplar solids resulting from leading pretreatment technologies. Bioresour. Technol. 100 (17), 3948–3962. Laureano-Perez, L., Teymouri, F., Alizadeh, H., Dale, B.E., 2005. Understanding factors that limit enzymatic hydrolysis of biomass. Appl. Biochem. Biotechnol. 124 (1–3), 1081–1099. Lee, Y.H., Fan, L.T., 1982. Kinetic studies of enzymatic hydrolysis of insoluble cellulose: analysis of the initial rates. Biotechnol. Bioeng. 24 (11), 2383–2406. Li, S., Xu, S., Liu, S., Yang, C., Lu, Q., 2004. Fast pyrolysis of biomass in free-fall reactor for hydrogen-rich gas. Fuel Process Technol. 85 (8–10), 1201–1211. Mansfield, S.D., Mooney, C., Saddler, J.N., 1999. Substrate and enzyme characteristics that limit cellulose hydrolysis. Biotechnol. Progr. 15 (5), 804– 816. Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal. Chem. 31 (3), 426–428. Mishima, D., Tateda, M., Ike, M., Fujita, M., 2006. Comparative study on chemical pretreatments to accelerate enzymatic hydrolysis of aquatic macrophyte biomass used in water purification processes. Bioresour. Technol. 97 (16), 2166–2172. Mooney, C.A., Mansfield, S.D., Touhy, M.G., Saddler, J.N., 1998. The effect of initial pore volume and lignin content on the enzymatic hydrolysis of softwoods. Bioresour. Technol. 64 (2), 113–119. Puri, V.P., 1984. Effect of crystallinity and degree of polymerization of cellulose on enzymatic saccharification. Biotechnol. Bioeng. 26 (10), 1219–1222. Soto, M.L., Domínguez, H., Núñez, M.J., Lema, J.M., 1994. Enzymatic saccharification of alkali-treated sunflower hulls. Bioresour. Technol. 49 (1), 53–59.
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
In-depth investigation of enzymatic hydrolysis of biomass wastes based on three major components: Cellulose, hemicellulose and lignin Lili Lin a,b, Rong Yan b,*, Yongqiang Liu b, Wenju Jiang a a b
College of Architecture and Environment, Sichuan University, Chengdu 610065, PR China Institute of Environmental Science and Engineering, Nanyang Technological University, Innovation Center, Block 2, Unit 237, 18 Nanyang Drive, Singapore 637723, Singapore
a r t i c l e
i n f o
Article history: Received 10 March 2010 Received in revised form 20 May 2010 Accepted 26 May 2010 Available online 19 June 2010 Keywords: Enzymatic hydrolysis Biomass components Pretreatment Artificial biomass
a b s t r a c t The artificial biomass based on three biomass components (cellulose, hemicellulose and lignin) were developed on the basis of a simplex-lattice approach. Together with a natural biomass sample, they were employed in enzymatic hydrolysis researches. Different enzyme combines of two commercial enzymes (ACCELLERASE 1500 and OPTIMASH™ BG) showed a potential to hydrolyze hemicellulose completely. Negligible interactions among the three components were observed, and the used enzyme ACCELLERASE 1500 was proven to be weak lignin-binding. On this basis, a multiple linear-regression equation was established for predicting the reducing sugar yield based on the component proportions in a biomass. The hemicellulose and cellulose in a biomass sample were found to have different contributions in staged hydrolysis at different time periods. Furthermore, the hydrolysis of rice straw was conducted to validate the computation approach through considerations of alkaline solution pretreatment and combined enzymes function, so as to understand better the nature of biomass hydrolysis, from the aspect of three biomass components. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction During the past years, researches on converting lignocellulosic biomass into bioethanol are actively undertaken, aiming to produce the 2nd generation biofuel which has no competition with food and is thus sustainable. The most common processing of lignocellulosics to bioethanol consists of four major unit operations: pretreatment of raw materials, enzymatic hydrolysis of pretreated materials into fermentable sugars, fermentation of fermentable sugars into ethanol, and ethanol separation or purification. It has been well known that hydrolyzing lignocellulosics to reducing sugars is one of the crucial factors to affect the cost of bioethanol production. Lignocellulosic biomass is mainly composed of cellulose, hemicellulose, and lignin, along with smaller amounts of pectin, protein, extractives, and ash (Blasi et al., 1999). The composition of these constituents can vary from one plant species to another. Normally, cellulose, hemicellulose, and lignin cover 40–60, 20–40, and 10–25 wt.% of biomass materials on dry basis, respectively. Among these components in biomass materials, only hemicellulose and cellulose can be converted into fermentable sugars by using microbial cellulolytic and xylanolytic enzymes which both have demonstrated promising results for the subsequent fermentation. * Corresponding author. E-mail address: [email protected] (R. Yan). 0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2010.05.084
However, the enzymatic digestion of cellulose and hemicellulose in biomass substrates is always limited by the complex structure and components of natural biomass (Chang and Holtzapple, 2000; Laureano-Perez et al., 2005; Puri, 1984). Among the primary substrate features impacting enzymatic conversion of biomass, lignin is believed to impede enzyme access to glucan chains by its protective sheathing and also reduce cellulase effectiveness as a result of unproductive binding and steric hindrance (Chang and Holtzapple, 2000; Mansfield et al., 1999). For lignocellulosic substrates, unproductive binding and inactivation of enzymes by the lignin component appear to be important factors limiting catalytic efficiency (Berling et al., 2005). Therefore, to achieve an effective hydrolysis of lignocellulosics, it is necessary and important to deeply understand whether enzyme preparations (cellulase and hemicellulase) are weak lignin-binding or strong lignin-binding; and whether the different combinations of biomass components in substrates have a significant influence on enzyme activity during hydrolysis of biomass wastes. Although hydrolysis performance of single biomass component (such as cellulose or hemicellulose) has been largely reported, systematic studies on evaluation of enzyme activity in hydrolyzing different components of biomass wastes (cellulose, hemicellulose, lignin) at different ratios have rarely been seen. As the three components consist of the major portion of biomass and each component has different chemical/physical structures, it would be interesting to find out if there is any interaction between the three
8218
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
components during hydrolysis and if there is a linkage between the enzyme activity and combinations of biomass components of any specific biomass waste. This could substantively conduce to determination on what kind of pretreatment method should be adopted, whether lignin content in raw material should be effectively reduced and what would be the most rigorous factor to inhibit the efficiency of enzymatic hydrolysis. In this work, cellulose, hemicellulose, and lignin were used as pure biomass components. The enzymatic hydrolysis behaviors of both cellulose and hemicellulose were first investigated by commercial enzyme preparation. Secondly, artificial biomass samples containing two or three of the biomass components (cellulose, hemicellulose and lignin) were developed on the basis of a simplex-lattice approach and thereafter hydrolyzed by enzyme preparation. The aim of this work is to understand the different behaviors of three biomass components in hydrolysis and their potential interaction. On this basis, the enzymatic hydrolysis of natural biomass rice straw was carried out. Alkaline pretreatment of rice straw was adopted and the change of biomass components in rice straw before and after the treatment was identified. Thereafter, the fundamental difference of natural biomass and artificial samples during hydrolysis was compared. 2. Methods 2.1. Materials The basic structure of lignocellulose consists of cellulose (C5H10O5)x, hemicellulose such as xylan (C5H8O4)m, and lignin [C9H10O3(OCH3)0.91.7]n (Demirbas, 2005). In this study, microcrystalline cellulose was used to study enzymatic hydrolysis performance of cellulose. It is in a microcrystalline powder form from wood pulp. Commercial hemicellulose can hardly be purchased, xylan, which is the dominant component of hemicellulose from agricultural plants, was chosen as a representative of hemicellulose in this research. Here, xylan is in a yellow powder form processed from birch wood. Lignin used in this work is alkali lignin in a brown powder form. The characteristics of celluloses, hemicellulose and lignin are shown in Table 1. Rice straw was used as a typical biomass in this work. It was washed with tap water to remove dust and dried in a 105 °C oven till constant weight. Then it was ground in a laboratory-scale centrifugal mill (Rocklabs, New Zealand) and sieved in a Retsch test sieve with a 40-mesh screen (Retsch, Fisher Scientific Company, USA). Furthermore, the dried rice straw was pretreated by alkalineperoxide method (Mishima et al., 2006). A mixed solution of 2% (w/ v) NaOH and 2% (v/v) H2O2 (1:1) was used. Each gram of raw sample was soaked in 10 ml of the mixture solution at room temperature for 24 h. After filtration, the solid residue was rinsed repeatedly with distilled water until the pH value came to neutral. After that, it was dried at 105 °C till a constant weight reached and then sieved through a 40-mesh screen. These two rice straw samples, un-treated and treated, were taken as the feedstocks for enzymatic hydrolysis.
Table 1 The characteristics of cellulose, hemicellulose and lignin used in the experiment. Sample
Source
Particle size (lm)
Brand
Cellulose, microcrystalline Xylan (hemicellulose)
Wood pulp
<50
Alfa
Birch wood
100
Lignin
Birch wood
100
Sigma– Aldrich Alfa
2.2. Chemical analysis of natural biomass To determine the amount of extractives in biomass (rice straw), solvent extraction (60 ml acetone for 1 g of dried biomass sample) was used, and the temperature was held at 90 °C for 2 h. After that, the sample was dried at 105 °C until a constant weight was obtained. The weight difference before and after the extraction is the amount of the extractives (Blasi et al., 1999). To determine the amount of hemicellulose, 10 ml 0.5 mol/l of sodium hydroxide solution was added to 1 g of extractive-free dried biomass, and the temperature was held at 80 °C for 3.5 h. After that, the sample was washed using DI water until pH value of the solution approach 7, then it was dried to a constant weight. The difference between the sample weight before and after this treatment is hemicellulose content (Blasi et al., 1999). To determine the amount of lignin, 30 ml of 98 wt.% sulfuric acid was added for each extractive-free dried biomass. After the sample was held at ambient temperature for 24 h, it was boiled at 100 °C for 1 h. The mixture was filtered, and then the residue was washed until the sulfate ion in the filtrate was undetectable (via titration of a 10% barium chloride solution); it was then dried to a constant weight. The weight of the residue was recorded as the lignin content (Blasi et al., 1999). The content of cellulose was calculated by difference, assuming that extractives, hemicellulose, lignin, and cellulose are the only components of the entire biomass. (Blasi et al., 1999; Li et al., 2004). 2.3. Preparation of artificial biomass samples Artificial biomass samples were synthesized using the three major components (microcrystalline cellulose, hemicellulous and lignin), on the basis of the simplex-lattice design devised by Scheffe (Gorman and Hinman, 1962). The method has two key features: (1) the fractions of components making up any mixture must add to unity, and hence factor space may be represented by a regular simplex; and (2) points of the composition in this simplex are explored in accordance with a lattice arrangement. The key issue is to choose a representative and reproducible proportion of the three components. For a mixture sample, the component proportions should satisfy the constraints, that is:
8 1 2 m > < X i ¼ 0; m ; m ; . . . ; m ; i ¼ 1; . . . ; s X1 þ X2 þ . . . þ Xs ¼ 1 > : k ¼ ðsþm1Þ! m!ðs1Þ!
where s is the number of components under consideration and m is an integer related to the spacing of the points and therefore to the number of mixture. The proportions of any component Xi can take value from zero to unity, and all possible mixtures with these proportions are used. To support a polynomial model of degree m in s components over the simplex, the lattice, referred to as a {s, m} simplex lattice, consists of all possible combinations of the components. The number of design points in the {s, m} simplex lattice is equal to k. For the mixture of cellulose, hemicellulose, and lignin, s = 3 (three components). To guarantee the accuracy and keep the computing simple, m was set to 4, so the total number of point k was 15. They are shown in Fig. 1 (solid dots). Only three artificial samples involved all the three components while others involved one or two components. The biomass samples were synthesized by dry mixing the three components according to the specific ratio. In the following text, H:C indicated the weight ratio of hemicellulose to cellulose, L:C and L:H represented those of lignin to cellulose and lignin to hemicellulose, respectively, and L:H:C was the weight ratio of a mixture of lignin, hemicellulose and cellulose.
8219
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
0.00
(w t,% )
Hydrolysis rate of substrate ð%Þ
se
0.50
0.25
1.00 0.25
0.50
0.75
0.00 1.00
Hemicellulose (wt, %) Fig. 1. Proportion of cellulose, hemicellulose, and lignin in the artificial biomass samples.
2.4. Enzyme Two types of enzymes, ACCELLERASE 1500 and OPTIMASH™ BG were gifts from Genencor International Incorporation. ACCELLERASE 1500 (referred as Cellulase) enzyme contains a mixture of cellulase, hemicellulase and higher level of b-glucosidase enzyme activities. It is a brown liquid and the activity of ACCELLERASE 1500 enzyme complex is expressed in carboxymethylcellulose (CMC U) activity units. The activity of endoglucanase reported was 2200–2800 CMC U/g, and one CMC U unit of activity liberated 1 lmol of reducing sugar (expressed as glucose equivalents) in one minute under specific assay conditions of 50 °C and pH 4.8. The reported Beta-glucosidase activity was 525–775 pNPG U/g, with one pNPG unit denoted 1 lmol of Nitrophenol liberated from paranitrophenyl-b-D-glucopyranoside per minute at 50 °C and pH 4.8. OPTIMASH™ BG (referred as Xylanase) is a mixture of Xylanase and b-Glucanase which was used together with ACCELLERASE 1500 to enhance enzymatic hydrolysis efficiency of hemicellulose. It is in an appearance of light amber liquid and has an activity of 10300 CMSC U/g. 2.5. Enzymatic hydrolysis Enzymatic hydrolysis experiments of biomass were carried out in 250 ml of stoppered conical flasks. A specific amount of enzyme was added into the substrates in a total 100 ml reaction volume. The pH of reaction system was kept at 5.0 with 0.05 N citric acid–sodium citrate buffer. The flasks were incubated at 50 °C on an orbital shaker agitated at 150 rpm. In all experiments, the citric acid buffer was supplemented with antibiotics tetracycline (40 lg/ ml) to prevent microbial contamination. Liquid samples were withdrawn from the reaction media periodically, centrifuged and filtered. The filtrate was used to determine the concentration of reducing sugar based on the DNS method using 3,5-dinitrosalicylic acid reagent (Miller, 1959). The natural biomass experiment was carried out at 50 °C and pH 5.0 for 72 h. Firstly, the un-treated rice straw sample was hydrolyzed by Cellulase at the dosage of 0.3 ml for each gram of total cellulose and hemicellulose in substrate. Secondly, in order to make cellulose and hemicellulose more accessible to enzyme, the natural biomass was pretreated by alkaline-peroxide solution and subsequently hydrolyzed by Cellulase at the dosage of 0.3 ml for total cellulose and hemicellulose in each gram substrate. Thirdly, the pretreated rice straw sample was hydrolyzed by combined enzymes of Cellulase and Xylanase at the dosage ratio of Cel-
¼
Reducing sugar 0:9 100 Cellulose þ hemicelluloseðin substrateÞ
ð1Þ
3. Results and discussion 3.1. Influence of enzyme loading on the yield of reducing sugar based on cellulose or hemicellulose substrates ACCELLERASE 1500 was used at specific concentrations to investigate the enzymatic hydrolysis performances of cellulose and hemicellulose. The yields of reducing sugar from the two samples generated at different enzyme loadings after 72 h duration were determined. It was found that hemicellulose can also be effectively hydrolyzed by enzyme ACCELLERASE 1500 and has slightly lower reducing sugar yield than microcrystalline cellulose at the same conditions of enzyme loading and reaction time. On the other hand, the hydrolysis efficiency of substrates (cellulose or hemicellulose) increased sharply at beginning but gradually flatted out when the enzyme concentration exceeds a certain level (0.3 ml/g substrate) where 20% of substrates were still available. This phenomenon can be explained in two aspects: (1) the surface area of cellulose or hemicellulose is not sufficient, and thus, the excess enzyme molecules are adsorbed to form multiple layers. Only the enzyme adsorbed in the first layer participates in hydrolysis; (2) the surface area of enzyme is composed of active and inactive fractions, only those biomass molecule adsorbed on the active fraction participate in the hydrolysis (Lee and Fan, 1982). The optimal Cellulase loading for celluloses or hemicellulose was in the scale from 0.2 to 0.4 ml/g substrate. Furthermore, hemicellulose was hydrolyzed by the mixture of Cellulase and Xylanase enzymes. Here, the concentration of Cellulase was referred to 0.3 ml/g hemicellulose in substrate according to previous findings. The concentration of Xylanase was first tested at two different levels: lower level of 0.05 ml/g hemicellulose and higher level of 0.15 ml/g hemicellulose. The reducing sugar yield of hemicellulose hydrolyzed by the mixture of Cellulase and Xylanase is shown in Fig. 2. It was evidenced that the reducing sugar yield of hemicellulose was effectively enhanced by adding Xylanase. When
Yield of reducing sugar (mg/g Xylan)
0.50
) t,% (w
nin
0.75
o llul
Lig
Ce
0.25
0.75
0.00
lulase to Xylanase equal to 0.3 ml:0.05 ml according to previous experiments. Each experiment was performed in triplicates. The hydrolysis rate of substrate was calculated from the following formula (Soto et al., 1994):
1.00
1000 800 600 400 Cellulase: Xylanase = 0.3ml:0.15ml Cellulase: Xylanase = 0.3ml:0.05ml Cellulase = 0.3ml
200 0 0
24
48
72
Hydrolysis time (h) Fig. 2. The influence of additive BG on the hydrolysis of xylan. Substrate concentration 1% (w/v), temperature 50 °C, pH 5.0.
8220
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
that of hemicellulose in the first 24 h and it increased gradually up to 48 and 72 h. This is possibly because hemicellulose has a random amorphous structure with little strength and it can be easily hydrolyzed by enzyme into glucose (6-carbon sugar) and pentose (5-carbon sugars). In contrast, the cellulose molecule is a very long polymer of glucose units without branches, and it is crystalline, strong, and resistant to hydrolysis. (2) there is no obvious interaction observed between hemicellulose and cellulose during enzymatic hydrolysis. The presence of one component in artificial biomass did not exert a negative impact on the hydrolysis of the other. At the time point of 72 h, the artificial samples of cellulose and hemicellulose at different ratios had demonstrated a decreasing order of hydrolysis efficiency as: C:H = 1:0 > C:H = 3:1 > C:H = 1:1 > C:H = 1:3 > C:H = 0:1.
the volume ratio of Cellulase to Xylanase was equal to 0.3:0.05 for one gram hemicellulose substrate, the reducing sugar yield was elevated more than 15% during initial 8 h reactivity. Moreover, the reducing sugar yield of hemicellulose tended to increase as the concentration level of Xylanase increased. Therefore, the combination of two different enzymes demonstrated a better hydrolysis efficiency with a potential of completely hydrolyzing hemicellulose into reducing sugars. 3.2. Enzymatic hydrolysis of artificial biomass In this experiment, artificial biomass substrates were synthesized by microcrystalline cellulose, hemicellulose and lignin. The Cellulase loading 0.3 ml/g of total cellulose and hemicellulose in substrates was employed. The total reaction time was controlled in 72 h.
3.2.2. Mixture of cellulose/lignin and of hemicellulose/lignin The enzymatic hydrolysis of cellulose (C) and lignin (L) mixtures and that of hemicellulose (H) and lignin (L) mixtures were also carried out. The results are similar for C + L and H + L systems and only the case of C + L is given in Fig. 4, as an example. The ratios of the mixture were decided following the simplex-lattice in Fig. 1 (i.e., both C:L and H:L are equal to 1:0, 3:1, 1:1, 1:3, and 0:1). The reducing sugar yield curves of both mixtures (C + L and H + L) showed the same trends with that of pure cellulose and hemicellulose substrates, respectively. From Fig. 4(a), the higher the ratio of cellulose or hemicellulose in both mixtures, the larger the reducing sugar yield of the substrates in a proportional pattern. The contents or ratios of lignin in the two mixtures had hardly any effect on the hydrolysis rate of the substrates as seen from Fig. 4(b).
1000
Hydrolysis rate of substrate
Yield of reducing sugar (mg/g of substrate)
3.2.1. Mixture of cellulose and hemicellulose The enzymatic hydrolysis curves of the artificial biomass samples of cellulose (C) and hemicellulose (H) are plotted in Fig. 3(a) and (b). Here, the artificial biomass samples were from the simplex-lattice (3, 4) design as given in Fig. 1 (C:H = 1:0, 3:1, 1:1, 1:3, and 0:1), in total 5 samples. The enzyme dosage at 0.3 ml/g of substrate was added to each sample. Two distinct observations were found from Fig. 3(a) and (b): (1) the hydrolysis rate of hemicellulose was much faster than that of cellulose during initial 24 h and almost >80% of the total reducing sugars (800 mg/g at 72 h) were generated in the first 24 h. In terms of cellulose hydrolysis, it showed relatively lower yield of sugar and hydrolysis rate than
800 600 400
C:H=1:0 C:H=3:1 C:H=1:1 C:H=1:3 C:H=0:1
200 0
0
24
48
72
80%
60%
40% C:H=1:0 C:H=3:1 C:H=1:1 C:H=1:3 C:H=0:1
20%
0%
0
Hydrolysis time (h)
24
48
72
Hydrolysis time (h)
(a) Yield
(b) Hydrolysis rate
1000 C:L=1:0 C:L=3:1 C:L=1:1 C:L=1:3
800 600 400 200 0 0
24
48
Hydrolysis time (h)
(a) Yield
72
Hydrolysis rate of substrate
Yield of reducing sugar (mg/g of substrate)
Fig. 3. Mixture of cellulose and hemicellulose at different ratios. Substrate concentration 1% (w/v), enzyme loading 0.3 ml/g of cellulose and hemicellulose, temperature 50 °C, pH 5.0.
80% 60% 40% C:L=1:0 C:L=3:1 C:L=1:1 C:L=1:3
20% 0% 0
24
48
72
Hydrolysis time (h)
(b) Hydrolysis rate
Fig. 4. Mixture of cellulose and lignin at different ratios. Substrate concentration 1% (w/v), enzyme loading 0.3 ml/g of cellulose in substrate, temperature 50 °C, pH 5.0.
8221
1000
Hydrolysis rate of substrate
Yield of reducing sugar (mg/g of substrate)
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
C:H:L=2:1:1 C:H:L=1:2:1 C:H:L=1:1:2
800 600 400 200 0 0
24
48
80% 60% 40% 20%
72
0%
C:H:L=2:1:1 C:H:L=1:2:1 C:H:L=1:1:2
0
24
Hydrolysis time (h)
48
72
Hydrolysis time (h)
(a) Yield
(b) Hydrolysis rate
Fig. 5. Mixture of cellulose, hemicellulose and lignin at different ratios. Substrate concentration 1% (w/v), enzyme loading 0.3 ml/g of cellulose in substrate, temperature 50 °C, pH 5.0.
The hydrolysis rate of both mixtures (C + L and H + L) was clearly determined by the cellulose or hemicellulose content in the substrates. It is obvious that the existence of free lignin has no inhibition on the enzymatic hydrolysis of cellulose and hemicellulose. This is possibly that lignin can not bind with the active sites of enzyme so cellulose or hemicellulose in the substrates is fully exposed to enzyme during enzymatic hydrolysis. Therefore, it is postulated for artificial biomass that lignin in substrate has negligible effect on the enzymatic hydrolysis of cellulose or hemicellulose using the enzyme complex ACCELLERASE 1500.
y ¼ b0 þ
3 X
bi xi þ
3 XX
i¼1
bij xi xj þ b123 x1 x2 x3
ð2Þ
ij
where x1, x2, and x3 are the proportions of cellulose, hemicellulose, and lignin, respectively, in biomass and y is the reducing sugar yield of substrate (mg/g of substrate) at different reaction times. If assume there is no interaction among the three components and lignin has no influence on the activity of enzyme, the Eq. (2) can be simplified to Eq. (3)
y ¼ b0 þ b1 x1 þ b2 x2 3.2.3. Mixture of cellulose, hemicellulose, and lignin at different ratios According to the proportion of cellulose (C), hemicellulose (H), and lignin (L) given in Fig. 1, the C:H:L ratios should be 2:1:1, 1:2:1, and 1:1:2 in the artificial biomass samples containing three components. The reducing sugar yields of the artificial samples containing the three components are shown in Fig. 5. Quite similar pattern of the hydrolysis rate curves were observed in Fig. 5, in comparison to Fig. 3 where only two components (hemicellulose and cellulose) were in the mixtures. A comparison of each counterpart in Figs. 5 and 3(b) showed that the addition of lignin into the mixture of hemicellulose and cellulose didn’t cause an evident change of enzymatic hydrolysis rate. Although in Fig. 3 the superior hydrolysis of hemicellulose over cellulose was observed, the yields of two artificial biomass (2:1:1 and 1:2:1) in Fig. 5 were close to each other and their hydrolysis rates showed negligible difference. Nevertheless, the yield of reducing sugar was dramatically reduced with the increasing ratio of lignin from 25% to 50% in substrate in Fig. 5(a). At the time point of 72 h, the artificial samples of cellulose, hemicellulose and lignin at different ratios had demonstrated a decreasing order of reducing sugar yield as: C:H:L = 2:1:1 > C:H:L = 1:2:1 > C:H:L = 1:1:2. From the above analysis, it is clear that almost no significant interaction occurred between cellulose, hemicellulose and lignin during enzymatic hydrolysis of artificial biomass. Consequently, it could be concluded that the enzymatic hydrolysis performance of biomass based on ACCELLERASE 1500 enzyme could be regarded as a simple superposition of cellulose and hemicellulose. The presence of lignin in biomass sample even with high fraction has not noticeable inhibition on enzyme activities of ACCELLERASE 1500 during hydrolysis. 3.3. Computing approach for predicting reducing sugar yield of substrate A general regression function was derived to fit the data collected from the designed {3, 4} simplex lattice. The polynomial equation for the three components can be rewritten as
ð3Þ
A linear multiple regression model was introduced to obtain the coefficients (b0, b1, b2), using the data from the artificial biomass. The values of coefficients obtained at different times are shown in Table 2. The correlation coefficient R2 was calculated using Eq. (4), and the results are also given in Table 2
P
2
R ¼1
ðyexp ycalcd Þ2 P 2 yexp
ð4Þ
where yexp is the experimentally observed reducing sugar yield and ycalcd is the calculated reducing sugar yield. From Table 2 it can be observed that a linear relationship was well established between the reducing sugar yield of artificial biomass based on three biomass components and the composition of cellulose and hemicellulose in the biomass, with all the R2 larger than 0.99. The assumption of negligible interaction among three biomass components and little contribution from lignin to sugar productivity was validated from the results of artificial biomass samples. In a closer look, the coefficients (b0, b1, b2) could represent, respectively, the background, weights of contribution from cellulose and hemicellulose. In the initial 24 h duration of hydrolysis, hemicellulose (x2), with a higher coefficient than cellulose (b2 > b1) played a more important role towards the formation of reducing sugar of biomass substrate. Afterwards, in the extended reaction time up to 72 h, cellulose (x1), with a higher coefficient than hemicellulose (b1 > b2), made a more significant contribution towards the formation of reducing sugar of biomass substrate. This
Table 2 The regression coefficients and correlation coefficient R2 at different reaction times. Reaction time (h)
b0
b1
b2
R2
3 8 24 48 72
13.2 1.3 1.43 1.1 3.1
257.9 425.4 609.4 762.1 818.3
463.0 577.3 668.3 743.1 784.3
0.9958 0.9991 0.9994 0.9999 0.9998
1000 800 600 400 200 0 0
24
48
72
Hydrolysis rate of substrate
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
Yield of reducing sugar(mg/g of substrate)
8222
100% 80% 60% 40% 20% 0% 0
Hydrolysis time (h)
24
48
72
Hydrolysis time (h)
(a) Yield
(b) Hydrolysis rate
Fig. 6. Enzymatic hydrolysis of rice straw. Substrate concentration 1% (w/v), enzyme loading 0.3 ml/g of cellulose in substrate, temperature 50 °C, pH 5.0.
result indicated the different enzyme activities of ACCELLERASE 1500 during staged hydrolysis of artificial biomass in degrading subsequently hemicellulose and cellulose.
3.4. Enzymatic hydrolysis of natural biomass Nevertheless, the artificial biomass samples of the three components could still be different from the natural biomass. In terms of natural biomass structure, cellulose and hemicellulose are usually found in close association with lignin. Lignin appears to limit cellulose and hemicellulose hydrolysis by forming a physical barrier that restricts enzyme access (Mooney et al., 1998). To further understand the hydrolysis behavior of natural biomass from the aspect of three biomass components, rice straw was used as a typical raw material hydrolyzed by Cellulase and Xylanase. The reducing sugar yield and the hydrolysis rate of natural biomass samples are plotted in Fig. 6(a) and (b), and the component analysis of un-treated rice straw and the reducing sugar yield after 72 h reaction time are given in Table 3. There yexp is the experimentally observed reducing sugar yield and ycalcd is the calculated reducing sugar yield according to Eq. (3) and Table 2. For raw natural biomass, yexp was much lower than ycalcd because the complex structure is unfavorable to the contact of enzyme and cellulose and hemicellulose, although the same ratio of the three components (cellulose, hemicellulose and lignin) as in natural samples was adopted in calculation. As mentioned before, these three components in natural biomass are associated closely with each other, which make cellulose and hemicellulose less exposed to enzyme digestion than in the case of artificial biomass where the structure combination of three components could be weak and they were almost completely exposed to cellulose and hemicellulose. For the pretreated rice straw, both the reducing sugar yield and the hydrolysis rate were largely improved than un-treated sample. This can be explained by two aspects: (1) the significant change of contents of the three major components (cellulose, hemicellulose and lignin) in natural biomass sample before and after pretreat-
Table 3 The components and reducing sugar yield of rice straw before and after pretreatment. Sample
Rice straw (raw)
Rice straw (pretreated)
Extractives (wt.%, dry) Hemicellulose (wt.%, dry) Lignin (wt.%, dry) Cellulose (wt.%, dry) ycalcd (mg/g) yexp (mg/g)
7.0 43.9 29.3 19.8 503 221
3.3 29 16.2 51.5 646 714
ment. From Table 3, the content of cellulose was significantly enhanced after pretreatment and the content of lignin and hemicellulose was decreased. The increase of total cellulose and hemicellulose content (from 63.7% to 80.5%) in natural biomass substrate after pretreatment demonstrated a positive effect on the enhancement of reducing sugar yield of substrate, which was in coincidence with the results of artificial biomass; (2) the alteration of the structure of biomass sample after pretreatment. The sample in the original form was relatively resistant to microbial attack, but the removal of substantive lignin and partial hemicellulose after alkaline pretreatment caused extensive changes in the structure and accessibility of cellulose which could allow more open to swelling on contact with enzyme (Mansfield et al., 1999). Therefore, as evidenced in this study again, the reducing sugar yield of pretreated rice straw sample by alkaline-peroxide pretreatment was 3-folders higher than the un-treated one and the hydrolysis rate of the former was 2-folders higher than the latter. Moreover, dilute alkaline solution pretreatment for biomass not only caused separation of structural linkages between the three major components each other, but also led to an increase in internal surface area of biomass sample, a decrease in the degree of polymerization, and a decrease in crystallinity (Kumar et al., 2009). For pretreated biomass, the yexp was more close to ycalcd than the case of un-treated sample since the influence of lignin in biomass was diminished by alkaline solution pretreatment, but still slightly higher due to possibly the above-mentioned structure difference. In the case of combination of different-purpose enzymes, the addition of Xylanase largely increased the final yield of reducing sugar nearly up to 95% in 24 h due to further hydrolysis of hemicellulose in treated biomass sample. In conclusion, the combine of Cellulase and Xylanase was efficient in enhancing both hydrolysis rate of biomass and reducing sugar yield. Lignin showed negligible effect on the activity of enzyme ACCELLERASE 1500 during hydrolysis of cellulose and hemicellulose, which was another proof of the earlier assumption of ‘‘weak lignin-binding” enzyme ACCELLERASE 1500. 4. Conclusions The enzymatic hydrolysis behaviors of biomass wastes were investigated from the aspect of three biomass components (cellulose, hemicellulose and lignin). Based on experiments of the systematically-designed artificial biomass samples, negligible interactions among the three components were observed. The prediction of biomass hydrolysis yield was therefore made with the assist of a multiple linear-regression model, which was well con-
L. Lin et al. / Bioresource Technology 101 (2010) 8217–8223
sistent with experimental results. Hemicellulose and cellulose were found to play different roles in staged hydrolysis at different time periods. These findings were validated through hydrolyzing rice straw, a natural biomass. The change of hydrolysis behavior of rice straw before and after alkaline pretreatment was well corresponding to the model prediction, in terms of biomass components contributions. Acknowledgements The author, Lili Lin, is grateful for the collaboration program between the Institute of Environmental Science and Engineering, Nanyang Technological University, Singapore and the College of Architecture and Environment, Sichuan University, China, which provides her the opportunity to pursue her PhD in Singapore under a research scholarship. References Berling, A., Gilkes, N., Kurabi, A., Bura, R., Tu, M., Kilburn, D., Saddler, J., 2005. Weak lignin-binding enzymes. Appl. Biochem. Biotechnol. 121 (1–3), 163–170. Blasi, C.D., Signorelli, G., Di Russo, C., Rea, G., 1999. Product distribution from pyrolysis of wood and agricultural residues. Ind. Eng. Chem. Res. 38 (6), 2216– 2224. Chang, V.S., Holtzapple, M.T., 2000. Fundamental factors affecting biomass enzymatic reactivity. Appl. Biochem. Biotechnol. 84–86, 5–37.
8223
Demirbas, A., 2005. Estimating of structural composition of wood and non-wood biomass samples. Energy Sour. 27 (8), 761–767. Gorman, J.W., Hinman, J.E., 1962. Simplex lattice designs for multicomponent systems. Technometrics 4 (4), 463–487. Kumar, R., Mago, G., Balan, V., Wyman, C.E., 2009. Physical and chemical characterizations of corn stover and poplar solids resulting from leading pretreatment technologies. Bioresour. Technol. 100 (17), 3948–3962. Laureano-Perez, L., Teymouri, F., Alizadeh, H., Dale, B.E., 2005. Understanding factors that limit enzymatic hydrolysis of biomass. Appl. Biochem. Biotechnol. 124 (1–3), 1081–1099. Lee, Y.H., Fan, L.T., 1982. Kinetic studies of enzymatic hydrolysis of insoluble cellulose: analysis of the initial rates. Biotechnol. Bioeng. 24 (11), 2383–2406. Li, S., Xu, S., Liu, S., Yang, C., Lu, Q., 2004. Fast pyrolysis of biomass in free-fall reactor for hydrogen-rich gas. Fuel Process Technol. 85 (8–10), 1201–1211. Mansfield, S.D., Mooney, C., Saddler, J.N., 1999. Substrate and enzyme characteristics that limit cellulose hydrolysis. Biotechnol. Progr. 15 (5), 804– 816. Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal. Chem. 31 (3), 426–428. Mishima, D., Tateda, M., Ike, M., Fujita, M., 2006. Comparative study on chemical pretreatments to accelerate enzymatic hydrolysis of aquatic macrophyte biomass used in water purification processes. Bioresour. Technol. 97 (16), 2166–2172. Mooney, C.A., Mansfield, S.D., Touhy, M.G., Saddler, J.N., 1998. The effect of initial pore volume and lignin content on the enzymatic hydrolysis of softwoods. Bioresour. Technol. 64 (2), 113–119. Puri, V.P., 1984. Effect of crystallinity and degree of polymerization of cellulose on enzymatic saccharification. Biotechnol. Bioeng. 26 (10), 1219–1222. Soto, M.L., Domínguez, H., Núñez, M.J., Lema, J.M., 1994. Enzymatic saccharification of alkali-treated sunflower hulls. Bioresour. Technol. 49 (1), 53–59.