- Email: [email protected]
Ultrasound-assisted alkaline pretreatment of sugarcane bagasse for fermentable sugar production: Optimization through response surface methodology
Bioresource Technology 112 (2012) 293–299
Contents lists available at SciVerse ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Ultrasound-assisted alkaline pretreatment of sugarcane bagasse for fermentable sugar production: Optimization through response surface methodology Rajendran Velmurugan, Karuppan Muthukumar ⇑ Department of Chemical Engineering, Alagappa College of Technology Campus, Anna University Chennai, Chennai 600 025, India
a r t i c l e
i n f o
Article history: Received 29 April 2011 Received in revised form 26 January 2012 Accepted 30 January 2012 Available online 8 February 2012 Keywords: Sugarcane bagasse Ultrasound Alkaline pretreatment Lignin removal RSM
a b s t r a c t Ultrasound-assisted alkaline pretreatment of sugarcane bagasse (SCB) for fermentable sugar production was carried out and the influence of particle size, liquid to solid ratio (LSR), NaOH concentration, temperature and sonication time on delignification and reducing sugar production was ascertained with Placket–Burman design. The best combination of each significant factor was determined by a central composite design (CCD) and optimum pretreatment conditions for maximum reducing sugar yield (96.27%) were particle size of 0.27 mm, LSR of 25 ml/g, NaOH concentration of 2.89% (w/v), temperature of 70.15 °C and pretreatment time of 47.42 min. Under these conditions, 92.11% of theoretical reducing sugar yield was observed experimentally. The substantial reduction in pretreatment time and temperature with improved efficiency is the most attractive features of the ultrasound-assisted alkaline pretreatment. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Enzymatic hydrolysis is regarded as the most promising technique for converting lignocellulosic compounds into fermentable sugars (Wooley et al., 1999; Wright, 1998); however, this approach requires biomass in small particles and removal of lignin (Dasari and Berson, 2007; Pan, 2008). Pretreatments for lignocellulosic materials include steam explosion (Hernandez-Salas et al., 2009), dilute acid hydrolysis (Hendriks and Zeeman, 2009), alkaline pretreatment (Zhang and Cai, 2008; Hendriks and Zeeman, 2009) and wet oxidation (Rosgaard et al., 2007). Alkaline pretreatments show less sugar degradation and furan derivatives formation (Gonzalez et al., 1986) than thermal and acid pretreatments (Fengel and Wegener, 1984). The alkaline pretreatment process can be improved further by the application of ultrasound (Filson and Dawson-Andoh, 2009). The ultrasonic treatment of aqueous media produces cavitation, which generates high temperature, pressure and extreme shear forces. The decomposition of water molecules into free radicals by cavitation aids in cleaving the linkages in lignin and xylan networks (Chuanyun et al., 2004; Yaldagard et al., 2008). Sun et al. (2004) treated sugarcane bagasse (SCB) with alkali/alkaline peroxide and ultrasound for the extraction of hemicellulose from SCB and reported 90% hemicellulose and lignin removal, but more detailed analyses of factors influencing delignification or cellulose recovery are needed. Therefore, the aim of the present study was to optimize particle size, liquid to solid ratio (LSR),
NaOH concentration and sonication time on ultrasound-assisted alkaline pretreatment by response surface methodology (RSM) with the intention to improve the delignification and reducing sugar yield. 2. Methods 2.1. Materials Sugarcane bagasse was obtained from a sugarcane juice shop in Chennai, India. The SCB was air dried, milled using a laboratory blender (Remi Anupam Mixie Ltd., Mumbai) and screened to obtain particles of different size ranges. The particle size ranges are average values of upper and lower sieves openings of ASTM and BS sieves. Determination of the main fractions (cellulose, hemicelluloses, and lignin) was carried out by detergent extraction methods (Goering and Vansoest, 1970). 2.2. Sonochemical reactor The ultrasonic treatment was carried out using a titanium probe-type sonolyzer (Hielscher UP 400S, Germany). The operating frequency and power of the sonolyzer were 24 kHz and 400 W, respectively. The amplitude was maintained at 100% and the temperature was controlled using a water bath. 2.3. Design of experiments
⇑ Corresponding author. Tel.: +91 44 22359153; fax: +91 44 22352642. E-mail address: [email protected] (K. Muthukumar). 0960-8524/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2012.01.168
The entire design was carried out in two stages. Initially, five variables were screened to identify significant factors using a
294
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299
two-level Plackett–Burman design (PBD) and in the second stage the level of these factors were optimized using a central composite design (CCD). 2.3.1. Screening of parameters The effect of particle size, LSR, NaOH concentration, temperature and sonication time on reducing sugar production was evaluated using PBD in twelve experimental runs. The software used for PBD was MINITAB 16. The variables with a confidence level higher than 90% were considered significant for both delignification and reducing sugar production. All experiments were carried out in duplicate and the averages were taken as responses. 2.3.2. Optimization of significant parameters Effects of the variables such as particle size, LSR, NaOH concentration, temperature and sonication time on response were fit with the second-order polynomial model given in Eq. (1).
Y ¼ b0 þ b1 A þ b2 B þ b3 C þ b4 D þ b5 E þ b11 A21 þ b22 A22 þ b33 A23 þ b44 A24 þ b55 A25 þ b12 AB þ b13 AC þ b14 AD þ b15 AE þ b23 BC þ b24 BD þ b25 BE þ b34 CD þ b35 CE þ b45 DE
ð1Þ
where Y is the predicted response; b0 is the constant; b1, b2, b3, b4, and b5 are linear coefficients; b11, b22, b33, b44, and b55 are quadratic coefficients; b12, b13, b14, b15, b23, b24, b25, b34, b35, and b45 are interaction coefficients; A, B, C, D, and E are factors representing particle size, liquid to solid ratio, NaOH concentration, temperature and sonication time, respectively. Five significant variables were considered for the optimization using a five-level CCD. ANOVA and regression analyses were carried out to evaluate the effects of variables and their interactive effects. Coefficients of the full model were analyzed for their significance and the insignificant ones were eliminated from the model. The predicted response and optimal levels of the variables were found by solving the second-order polynomial model for maximization of response using MINITAB 16 software (Lu et al., 2009).
SCB biomass was dispersed in 100 ml of desired concentration of NaOH solution (0.25%, 1.00%, 1.75%, 2.50% and 3.25%) in an Erlenmeyer flask and treated with ultrasound at different operating conditions based on the experimental design. The performance of ultrasound-assisted alkaline pretreatment was compared with those of alkaline pretreatment carried out at 30 ± 1 °C, ultrasound treatment in the absence of NaOH and commercial pretreatment (2% NaOH in an autoclave at121 °C for 1 h) (Zhang and Cai, 2008; Hendriks and Zeeman, 2009). After the treatment, the contents were filtered through Whatman No. 1 filter paper and the supernatant was subjected to sugar, acetic acid and furfural analyses. The solid residue was repeatedly washed with water until the filtrate was neutral. The residue was dried at 50 °C until constant weight was observed and used for saccharification experiments. Experiments were performed in triplicate and mean values were employed in the analyses. The sugar concentration in terms of gram per gram was calculated by dividing the sugar concentration in pretreated solids by the total solids recovered after pretreatment. The polysaccharide (glucan, xylan and arabinan) recovery in the solid content was calculated using the following equation:
PPT-SCB 100 PSCB
% delignification ¼
LSCB LPT-SCB 100 LSCB
ð3Þ
where LSCB is the concentration of lignin in SCB, LPT-SCB is the concentration of lignin in pretreated SCB, which was calculated by multiplying estimated lignin concentration (g/g) by recovered solids after the pretreatment. 2.5. Enzymatic convertibility Pretreated SCBs were weighed and suspended in 100 ml of 50 mM sodium citrate buffer (pH 4.8) supplemented with 0.02% sodium azide under aseptic conditions. The contents were mixed with 25 FPU/g dry matter and 0.46 CBU/g cellulose of commercially available cellulase and b-glucosidase (Sisco Research Laboratory (SRL), India), respectively. The reaction mixture was incubated in a shaking incubator at 45 °C and 150 rpm for 26 h (Aswathy et al., 2010). The release of soluble reducing sugars was periodically measured with the dinitrosalicylic acid (DNS) method (Miller, 1959). 2.6. Analytical methods 2.6.1. Physical characterization The morphological differences of SCB pretreated by various methods were examined using scanning electron microscopy (SEM, JEOL Ltd., Tokyo, Japan). The crystalline nature of the raw and treated SCB was analyzed using a Rigaku RINT-TTR3 X-ray diffractometer (Rigaku Co., Tokyo, Japan). The nickel-filtered Cu Ka radiation (k = 0.1542 nm) was applied at 50 kV and 30 mA. Samples were scanned over the range of 2h = 5–50° and the crystallinity index (CrI) was determined using Eq. (4) (Segal et al., 1959):
CrI ¼
ICrystalline IAmorphous 100 % ICrystalline
ð4Þ
where, Icrystaline = intensity at 21° and Iamorphous = intensity at 18.8°.
2.4. Experimental
% recovered ¼
where PSCB is the concentration of polysaccharide in SCB, PPT-SCB is the concentration of polysaccharide in pretreated SCB, which is calculated by multiplying estimated sugar concentration (g/g) by recovered solids after pretreatment. The % delignification was calculated using the following equation:
ð2Þ
2.6.2. Chemical characterization The total reducing sugars in enzymatic hydrolysates of SCB was determined by the DNS method (Miller, 1959) and cellulose concentration was estimated spectrophotometrically by the anthrone method (Updegroff, 1969). Anthrone and DNS reagents were obtained from SRL, India. The concentration of sugars, acetic acid and furfural was determined using high performance liquid chromatography (HPLC). The HPLC system (Shimadzu, CA, USA) consisted of a liquid pump (LC-10AD), a refractive index detector (RID-6A) and a system controller with Shimadzu EZS software. The monosaccharides (glucose, xylose and arabinose) were separated in an ion-exchange column (Aminex HPX-87P Bio-Rad Laboratories, Hercules, CA) at 85 °C. Ultra-pure water was used as mobile phase at a flow rate of 0.6 ml/min (Sluiter et al., 2008). The degradative products (acetic acid and furfural) were separated using an ion exchange column (Aminex HPX-87H) at 65 °C. The mobile phase, 5 mM H2SO4, was used at a flow rate of 0.6 ml/min (Sluiter et al., 2008). The composition of raw and pretreated SCBs was determined by strong acid hydrolysis as described previously (Velmurugan and Muthukumar, 2011).The conversion factors for hexose (glucose) and pentoses (xylose and arabinose) were 1.111 and 1.136, respectively (Demirbas, 2005). Hemicellulose, lignin and ash contents were estimated by methods described by Goering and Vansoest (1970) and APHA (2005). Cellulase activity
295
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299
was determined by standard filter paper assay and expressed as filter paper units (FPU). One FPU is defined as the enzyme that releases 1 lmol of glucose equivalents per minute from Whatman No. 1 filter paper (Ghose, 1987). Endoglucanase activity was measured using carboxymethyl cellulose as a substrate. The b-glucosidase activity was measured by cellobiase assay method (Ghose, 1987). 3. Results and discussion 3.1. Composition of raw SCB Raw SCB contained 35.6% cellulose, 26.6% hemicellulose, 17.1% lignin, 1.1% ash, 11.5% moisture and 6.4% protein (w/w). Xylan and arabinan was accounted for 17.84% and 4.42% (w/w), respectively. The compositional variability among five different particle size fractions was within the error limit of 5%. 3.2. Placket–Burman design Particle size, LSR, NaOH concentration, temperature and sonication time on delignification were identified as most significant and contributing variables and their ranges were 0.27–0.91 mm, 10–20 ml/g, 0.25–2.5% (w/v), 30–50 °C and 5–50 min, respectively (Table 1). The analysis of experimental data indicated wide variation in delignification (37.77–74.27%) and reducing sugar yield (67.08–88.94%) (Table 2). These variations indicated the necessity for optimization of the parameters. The statistically significant effects of each factor were screened by probability test and factors with a confidence interval greater than 90% (Prob > [t] < 0.1) were considered as a significant parameter. The lowest probability (<0.0001) values were obtained for sonication time and this indicates the significant influence of this factor. The LSR, NaOH concentration and temperature were also significant (Prob > [t] = 0.003). The particle size had limited significance on lignin removal (Prob > [t] = 0.053). The sequential effect of the factors was explained by a Pareto chart (Fig. 1a) and the most important factors determining delignification were found to be in the order: sonication time, temperature, NaOH concentration, LSR and particle size. For reducing sugar production, the preferred order was: sonication time, particle size, NaOH concentration, temperature and LSR (Fig. 1b). Sonication intensifies the chemical reaction by generating cavitation and shear forces (Yaldagard et al., 2008; Filson and Dawson-Andoh, 2009) but pretreatment conditions such as temperature, NaOH concentration, LSR and particle size are also important for lignin removal. The influence of particle size was more significant for reducing sugar production than for delignification due to the fact that the biomass particle size variation occurred even during pretreatment. Generally, the reduction in biomass
particle size improves the mass transfer and increases the accessible area available to the enzyme (Guo et al., 2008). 3.3. Central composite design In order to optimize the significant factors, both delignification and reducing sugar production was considered. In five factor optimization, concentrations of sugars, lignin and degradative products in solid and liquid content were used. The responses of 32 sets of experiments were analyzed statistically to obtain the optimum value of each factor and the results are summarized in Tables 3 and 4. 3.3.1. Effect of pretreatment on composition of liquor Xylose was the major product in all experiments (Table 4). Furfural concentrations were ranged from 0.21 to 0.71 g/L. Acetic acid release was from 1.79 to 9.28 g/L, which is below the level of 6 g/L that inhibits cellulase activity and ethanol production (Larsson et al., 1999). 3.3.2. Effect of pretreatment on composition of solids The recovery of various important components achieved in the solid phase is shown in Table 4 and the % recovery was in the range of 72.14–89.12. The recovery in the solid phase was expressed as the ratio of dry weight of pretreated solid to the raw SCB. The maximum recovery (89.12%) was obtained with 0.75 mm SCB particles at a LSR of 10, 1% NaOH, 30 °C and 20 min reaction time. The pretreatment conditions were chosen to produce a solid substrate with the lowest possible lignin content and the least amount of sugar degradation. The glucan recovery was in the range of 97.71–99.43% (Table 4) which indicates that the pretreatment conditions did not affect cellulose content. In contrast, the pretreatment solubilized more pentosan than glucan, which resulted in a higher glucan recovery in the solid content. The recovery of xylan and arabinan was in the range of 69.64–93.10% and 61.19–86.22%, respectively. The higher solid content (run 26) showed higher pentose recovery but greater solid losses and less delignification. Therefore, while selecting pretreatment based on delignification, polysaccharides recovery in the solid phase and saccharification of solids after pretreatment should also be considered. 3.3.3. Effect of pretreatment on delignification An analysis of variance (ANOVA) was performed and the results are presented in Table 5a. The results indicated that the fitting model and lack of fit (p < 0.001) were highly significant. The coefficient of determination (R2) was 0.9954, which indicated that the model could explain 99.54% variability of the response variable. The adjusted coefficient of determination (R2 = 98.69) was satisfactory and confirmed the significance of the model. Analysis of the
Table 1 Plackett–Burman design variables (in coded levels) with delignification and reducing sugar concentration as responsea.
a
Run order
A
B
C
D
E
Delignification (%)
Reducing sugar yield (%)
1 2 3 4 5 6 7 8 9 10 11 12
+ + + + + +
+ + + + + +
+ + + + + +
+ + + + + +
+ + + + + +
65.57 39.71 37.77 46.47 29.24 40.59 73.39 71.63 64.75 74.27 38.77 62.24
81.00 70.76 67.08 69.88 67.50 65.62 88.94 88.48 81.08 83.80 70.22 84.76
A: Particle size (mm), B: liquid to solid ratio (v/w), C: NaOH (%, w/v), D: temperature (°C), E: time (min).
296
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299
Table 2 Levels of the variables and statistical analysis of Plackett–Burman design. Factors
Particle size (mm) LSR (v/w)a NaOH (%, w/v) Temperature (°C) Time (min)
Code
A B C D E
Low level (1)
0.27 10 0.25 30 5
High level (+1)
0.91 20 2.5 50 50
Delignification (%) b
Effect
Coeff
6.255 12.027 12.459 12.585 21.225
3.128 6.013 6.230 6.292 10.612
Reducing sugar yield (%) t-Value
p-Value
Effect
Coeffb
t-Value
p-Value
2.42 4.66 4.83 4.88 8.23
0.052 0.003 0.003 0.003 0.000
76.593 7.320 3.880 5.633 4.687
3.660 1.940 2.817 2.343 6.273
0.00 6.09 3.23 4.69 3.90
0.000 0.001 0.018 0.003 0.008
a
LSR: Liquid to solid ratio. Coeff: Coefficient. R for delignification = 95.96% (adjusted R2 = 92.59%), R2 for reducing sugar yield = 97.00% (adjusted R2 = 94.50%).
b
2
value for delignification was 99.64% at a particle size of 0.27 mm, LSR of 25 ml/g, NaOH concentration of 3.25%, temperature of 75 °C and sonication time of 61.97 min.
Fig. 1. Pareto charts (a) standardized effects of process terms on delignification; and (b) standardized effect of process terms on reducing sugar yield (a = 0.1).
3.3.4. Effect of pretreatment on enzymatic saccharification An ANOVA was performed for the reducing sugar yield and the results are presented in Table 5a. The values of p were less than 0.001, indicating that the model was highly significant. The coefficient of determination (R2) value was 0.9884 and this indicates that the model could explain 98.84% variability of the response variable. The adjusted value of coefficient of determination (R2 = 0.9674) was also satisfactory, indicating the significance of the model. Analysis of the response trends showed that the model could explain the effect of particle size, LSR, NaOH concentration, temperature and sonication time on reducing sugar yield satisfactorily. The coefficients calculated by regression analysis for each variable are presented in Table 5b and the regression coefficients of linear and squared terms were highly significant (p < 0.1). The interactive terms were not significant for reducing sugar yield, indicating a negligible interactive effect on reducing sugar yield. The positive coefficients of LSR, NaOH concentration, temperature and sonication time indicate that such factors had a positive effect on reducing sugar yield. The particle size had a negative coefficient (8.607) on reducing sugar yield and this indicates that the increase in particle size had a negative effect on the reducing sugar yield. The optimum values of variables were obtained by varying the factors within the experimental range and the other factors were maintained constant at their middle level. The maximum response value for reducing sugar yield was estimated as 96.27% at a particle size of 0.27 mm, LSR of 25 ml/g, NaOH concentration of 2.89%, temperature of 70.15 °C and sonication time of 47.42 min. 3.4. Comparative analysis
response trends was reasonable and the model explained the effect of particle size, LSR, NaOH concentration, temperature and sonication time on delignification satisfactorily. Model coefficients were evaluated by regression analysis and their significance was tested. The values for linear, square and interaction effects are presented in Table 5b. The interactive effects of particle size and NaOH concentration, and LSR and temperature on delignification were highly significant; however, the interactive effect for eight terms on delignification was insignificant (p > 0.1), indicating that these interactions had little impact on delignification. Positive coefficients indicate a linear increase in delignification while negative coefficients indicate a linear decrease in delignification. Among the various factors, particle size had a negative effect on delignification during ultrasound-assisted alkaline pretreatment. The optimum values of different variables were obtained by varying the selected factor within the experimental range and maintaining the other factors constant at their middle level. The maximum response
Ultrasound-assisted alkaline pretreatment solubilized the lignin matrix more effectively than autoclave pretreatment and hence, the performance of saccharification was improved considerably. SEM analysis (Fig. S1) (Supplementary material) demonstrated that raw SCB displayed a regular and compact surface structure while the pretreated SCB showed histological changes. The outer surface was destroyed and perforated by the pretreatment processes and pores were approximately 2.3 lm in size. These observations may be due to the oxidation of lignin matrix by hydroxyl radicals formed during sonication (Moiser et al., 2005). The pores formed may increase enzyme accessibility, which in turn increases the enzyme digestibility of the SCB and shorten the enzymatic hydrolysis time (Moiser et al., 2005). The diffraction patterns of raw, ultrasound-assisted alkaline pretreated and alkaline pretreated SCB (Fig. S2) showed broad peak for raw SCB and pretreated (both ultrasound and autoclave) SCB showed a narrow peak at 2h = 21.9°. The crystallinity index of pretreated SCB (sono-assisted)
297
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299 Table 3 Central composite design for the optimization of factors identified by Plackett–Burman designa.
a
Run order
A
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
0.75 0.59 0.59 0.59 0.59 0.59 0.43 0.59 0.59 0.43 0.75 0.43 0.43 0.75 0.59 0.43 0.75 0.59 0.43 0.43 0.59 0.27 0.75 0.59 0.75 0.59 0.59 0.59 0.75 0.75 0.91 0.43
(1) (0) (0) (0) (0) (0) (1) (0) (0) (1) (1) (1) (1) (1) (0) (1) (1) (0) (1) (1) (0) (2) (1) (0) (1) (0) (0) (0) (1) (1) (2) (1)
B
C
20 (1) 15 (0) 15 (0) 15 (0) 15 (0) 15 (0) 20 (1) 15 (0) 15 (0) 20 (1) 10 (1) 10 (1) 10 (1) 20 (1) 25 (2) 10 (1) 10 (1) 15 (0) 20 (1) 10 (1) 15 (0) 15 (0) 20 (1) 15 (0) 20 (1) 5 (2) 15 (0) 15 (0) 10 (1) 10 (1) 15 (0) 20 (1)
2.50 3.25 0.25 1.75 1.75 1.75 2.50 1.75 1.75 1.00 2.50 1.00 2.50 1.00 1.75 2.50 2.50 1.75 1.00 1.00 1.75 1.75 2.50 1.75 1.00 1.75 1.75 1.75 1.00 1.00 1.75 2.50
(1) (2) (2) (0) (0) (0) (1) (0) (0) (1) (1) (1) (1) (1) (0) (1) (1) (0) (1) (1) (0) (0) (1) (0) (1) (0) (0) (0) (1) (1) (0) (1)
D
E
60 (1) 45 (0) 45 (0) 45 (0) 45 (0) 45 (0) 30 (1) 45 (0) 45 (0) 30 (1) 60 (1) 30 (1) 60 (1) 60 (1) 45 (0) 30 (1) 30 (1) 75 (2) 60 (1) 60 (1) 45 (0) 45 (0) 30 (1) 15 (2) 30 (1) 45 (0) 45 (0) 45 (0) 60 (1) 30 (1) 45 (0) 60 (1)
50 (1) 35 (0) 35 (0) 35 (0) 35 (0) 35 (0) 50 (1) 5 (2) 35 (0) 20 (1) 20 (1) 50 (1) 50 (1) 20 (1) 35 (0) 20 (1) 50 (1) 35 (0) 50 (1) 20 (1) 35 (0) 35 (0) 20 (1) 35 (0) 50 (1) 35 (0) 65 (2) 35 (0) 50 (1) 20 (1) 35 (0) 20 (1)
Delignification(%)
Reducing sugar yield (%)
Actual
Predicted
Actual
Predicted
77.71 74.93 58.29 68.71 68.71 68.71 77.24 59.10 68.71 61.14 62.84 58.96 74.61 63.98 73.86 65.22 65.01 73.79 73.40 59.23 68.71 75.05 67.39 61.46 63.12 58.24 70.89 68.71 61.51 52.54 63.57 78.31
79.50 76.51 58.47 70.92 70.92 70.92 74.96 60.04 70.92 60.32 64.89 57.88 73.88 66.47 75.38 64.41 64.89 73.42 74.76 59.98 70.92 74.16 68.28 59.57 63.52 58.84 69.81 70.92 62.97 50.84 68.44 78.06
86.13 85.22 74.74 83.70 83.70 83.70 86.98 76.51 83.70 79.07 79.19 75.84 86.04 79.84 86.90 81.34 78.45 84.45 85.51 77.93 83.70 88.19 81.96 77.12 77.05 76.46 83.92 83.70 76.08 71.49 80.09 89.36
86.47 85.61 74.34 83.70 83.70 83.70 87.28 77.98 83.70 78.83 78.70 76.25 86.30 79.63 86.71 80.81 78.64 84.33 86.07 77.65 83.70 88.14 81.50 77.23 77.53 76.64 82.44 83.70 76.52 71.13 80.13 88.98
A: Particle size (mm), B: liquid to solid ratio (v/w), C: NaOH (%, w/v), D: temperature (°C), E: time (min).
Table 4 Recovery of sugars and composition of liquor obtained after the pretreatment of sugarcane bagasse. Run order
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Solids recovered (%)
74.53 77.05 85.63 82.59 80.94 82.29 79.04 84.23 82.29 86.66 81.69 85.62 79.35 82.48 79.49 82.80 80.50 77.38 78.68 85.37 80.94 80.25 79.92 82.86 82.78 72.14 80.26 82.14 82.46 89.12 82.00 79.04
Sugars recovered (%)
Composition of liquor obtained after pretreatment (g/L)
Glucan
Xylan
Arabinan
Glucose
Xylose
Arabinose
Acetic acid
Furfural
97.71 98.43 99.06 98.88 98.88 98.88 98.21 99.12 98.88 99.33 98.58 98.95 98.40 98.49 98.44 99.00 98.23 98.35 98.41 99.27 98.88 99.02 98.39 98.94 98.24 99.05 98.46 98.88 98.19 99.43 98.48 98.58
69.64 75.44 84.14 80.32 80.32 80.32 77.65 84.45 80.32 87.01 77.63 86.56 77.54 76.81 76.17 82.24 77.52 74.64 77.95 85.77 80.32 83.10 77.25 81.88 78.13 93.10 78.50 80.32 77.29 89.48 76.56 77.63
61.19 68.13 79.64 74.84 74.84 74.84 70.80 79.94 74.84 84.37 72.15 82.15 71.20 71.40 70.99 77.75 70.47 68.37 72.10 82.82 74.84 78.26 70.93 76.75 71.54 86.22 71.77 74.84 70.90 86.86 70.15 72.15
0.46 0.41 0.24 0.29 0.30 0.29 0.35 0.22 0.29 0.13 0.55 0.41 0.62 0.28 0.24 0.40 0.70 0.44 0.31 0.28 0.29 0.26 0.31 0.28 0.34 0.49 0.40 0.30 0.71 0.23 0.39 0.29
6.08 6.46 4.43 5.15 5.20 5.44 4.64 4.30 5.06 2.80 8.78 5.65 9.00 4.47 4.05 6.96 8.86 6.62 4.20 5.65 5.06 4.35 4.68 4.77 4.26 3.80 5.65 5.15 9.00 4.30 6.24 4.77
0.95 1.04 0.66 0.82 0.82 0.81 0.70 0.61 0.82 0.37 1.36 0.87 1.41 0.68 0.55 1.09 1.46 1.03 0.68 0.85 0.82 0.72 0.72 0.75 0.69 1.36 0.91 0.82 1.42 0.62 0.97 0.67
1.91 2.86 4.75 3.56 3.56 3.56 1.95 4.66 3.56 3.32 6.35 7.02 4.34 3.08 1.79 5.95 5.98 2.98 2.27 6.97 3.56 2.84 2.79 4.39 3.15 9.28 3.32 3.56 6.58 8.12 4.15 1.85
0.71 0.58 0.41 0.48 0.48 0.48 0.52 0.39 0.49 0.28 0.51 0.31 0.52 0.55 0.56 0.40 0.52 0.64 0.49 0.29 0.47 0.41 0.51 0.44 0.50 0.40 0.50 0.47 0.54 0.21 0.57 0.51
298
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299
Table 5a ANOVA for selected central composite design for delignification and reducing sugar yield. Source
Regression Linear Square Interaction Residual Error Lack-of-fit Pure error Total
Delignification (%)
Reducing sugar yield (%)
DF
Sum of squares
Mean value
F value
Prob > F
DF
Sum of squares
Mean value
F value
Prob > F
20 5 5 10 11 6 5 31
1398.21 1336.62 45.22 16.36 6.51 6.51 0.00 1404.72
69.910 267.325 9.045 1.636 0.592 1.085 0.000
118.07 451.50 15.28 2.76
0.00 0.00 0.00 0.06
30.46 108.83 12.13 0.43 0.65 1.19 0.00
0.000 0.000 0.000 0.735
0.00
609.127 544.146 60.660 4.321 7.139 7.139 0.000 616.266
46.93 167.69 18.69 0.67
0.00
20 5 5 10 11 6 5 31
0.00 0.00
0.000
R2 for delinification = 99.54% (adjusted R2 = 98.69%); R2 for reducing sugar yield = 98.84% (adjusted R2 = 96.74%).
Table 5b Regression analysis for selected central composite design for delignification and reducing sugar yield. Terma
Constant A B C D E AA BB CC DD EE AB AC AD AE BC BD BE CD CE DE a b
Delignification
Reducing sugar yield b
Coeff
SE coeff
t
p
Coeffb
SE coeffb
t
p
22.477 3.478 1.262 12.599 0.262 0.397 5.250 0.027 0.962 0.001 0.004 0.138 5.658 0.130 0.008 0.059 0.007 0.000 0.021 0.004 0.001
0.3069 0.1571 0.1571 0.1571 0.1571 0.1571 0.1421 0.1421 0.1421 0.1421 0.1421 0.1924 0.1924 0.1924 0.1924 0.1924 0.1924 0.1924 0.1924 0.1924 0.1924
223.936 15.119 24.828 28.576 17.412 17.108 0.946 4.791 3.810 2.014 6.651 0.573 3.529 1.627 0.099 1.158 2.615 0.119 1.214 0.226 1.571
0.000 0.000 0.000 0.000 0.000 0.000 0.364 0.001 0.003 0.069 0.000 0.578 0.005 0.132 0.923 0.271 0.024 0.908 0.250 0.825 0.145
45.037 8.607 1.026 12.050 0.425 0.347 4.246 0.020 1.658 0.003 0.004 0.001 2.124 0.086 0.037 0.012 0.003 0.001 0.022 0.002 0.001
0.3213 0.1644 0.1644 0.1644 0.1644 0.1644 0.1487 0.1487 0.1487 0.1487 0.1487 0.2014 0.2014 0.2014 0.2014 0.2014 0.2014 0.2014 0.2014 0.2014 0.2014
260.491 12.186 15.313 17.127 10.781 6.774 0.731 3.411 6.268 4.912 5.866 0.002 1.266 1.031 0.446 0.224 1.137 0.317 1.217 0.126 0.924
0.000 0.000 0.000 0.000 0.000 0.000 0.480 0.006 0.000 0.000 0.000 0.998 0.232 0.325 0.665 0.827 0.280 0.757 0.249 0.902 0.375
A: Particle size (mm), B: liquid to solid ratio (v/w), C: NaOH (%, w/v), D: temperature (°C), E: time (min). Coeff: Coefficient.
Table 6 Comparative analysis of different pretreatment methods. Factors
Raw SCB NaOH, room temperature Ultrasound NaOH, ultrasoundb NaOH, autoclave
Composition of liquor obtained after pre-treatment (g/L)
Percentage recovered (%)
Glucose
Xylose
Arabinose
Acetic acid
Furfural
Glucan
Xylan
Arabinan
ND 0.01 0.01 0.13 0.28
ND 0.20 0.85 1.60 4.16
ND 0.28 0.37 0.47 0.64
ND 0.23 0.18 0.36 0.53
ND ND ND ND 0.12
ND 99.5 99.5 98.32 96
ND 97.8 90.47 77.95 34.98
ND 87.3 83.26 71.16 23.1
Lignin removal (%)
Reducing sugar yield (%)a
ND 2.50 7.30 82.32 46.50
28.90 30.50 31.20 92.11 72.00
ND – not detectable. a Theoretical yield of reducing sugar obtained after saccharification. b RSM optimized conditions.
was 67%, whereas that for raw SCB was 48.50% due to the removal of hemicellulose and lignin fractions which increased the relative content of crystalline cellulose. The crystallinity index for alkaline pretreatment at 121 °C was 65.40%, which confirmed that the removal of amorphous domains was more or less similar to that of the ultrasound-assisted alkaline pretreatment. However, after the pretreatment, the recovery of domains containing sugars should be high. The sugar recovery and reducing sugar yield for ultrasound-assisted pretreatment and other pretreatment methods are presented in Table 6. The lignin removal and cellulose recovery
achieved with ultrasound assisted pretreatment were 82.32% and 98.32%, respectively. The autoclave pretreatment reduced the lignin content by 46.50% and observed cellulose recovery was 96%. The ultrasound assisted pretreated solids produced 92.11% of theoretical reducing sugar yield, which is comparatively more or less similar to RSM predicted yield (96.27%) and higher than the yield obtained from autoclave pretreatment (72%). The formation of degradative product (acetic acid) was observed in ultrasound assisted pretreatment but its concentration was very less (0.36 g/L). The improved lignin removal and the reducing sugar yield clearly
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299
demonstrated that the application of ultrasound improved the alkaline pretreatment effectively. 4. Conclusions SCB pretreated with ultrasound-assisted alkaline pretreatment showed better reducing sugar yield than commercial alkaline pretreatment. The substantial reduction in pretreatment time and temperature with improved efficiency is the most attractive feature of the ultrasound-assisted alkaline pretreatment. However, to implement this process at a larger scale, design of suitable reactor and the optimization of energy required need to be done. Acknowledgements The authors express their sincere thanks to the Editor and the Reviewers for their constructive suggestions. The authors gratefully acknowledge Department of Science and Technology, New Delhi for providing financial support to carry out this research work under PURSE scheme. One of the authors Mr. R. Velmurugan is grateful to DST, New Delhi for the award of DST-PURSE fellowship. The authors also express their sincere thanks to Dr. A. Peer Mohamed, Department of Textile Technology, Anna University, Chennai for his valuable suggestions. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.biortech.2012.01.168. References APHA, 2005. Standard methods for the examination of water and wastewater. In: Eaton, A.D., Clesceri, L.S., Rice, E.W., Greenberg, A.E., Franson, M.A.H. (Eds.), 21st ed. American Public Health Association, American Water Works Association, Water Environment Federation, USA (centennial edition). Aswathy, U.S., Sukumaran, R.K., LalithaDevi, G., Rajasree, K.P., Singhania, R.R., Pandey, A., 2010. Bio-ethanol from water hyacinth biomass: an evaluation of enzymatic saccharification strategy. Bioresour. Technol. 101, 925–930. Chuanyun, D., Bochu, W., Huan, Z., Conglin, H., Chuanren, D., Wangqian, L., Toyama, Y., Sakanishi, A., 2004. Effect of low frequency ultrasonic stimulation on the secretion of siboflavin produced by Ecemothecium ashbyii. Colloids Surf. B 34 (1), 7–11. Dasari, R.K., Berson, R.E., 2007. The effect of particle size on hydrolysis reaction rates and rheological properties in cellulosic slurries. Appl. Biochem. Biotechnol., 136–140. Demirbas, A., 2005. Bioethanol from cellulosic materials: a renewable motor fuel from biomass. Energy Sources 21, 327–337. Fengel, D., Wegener, G., 1984. Wood – Chemistry, Ultrastructure, Reactions. Walter de Gruyter, Berlin, New York.
299
Filson, P.B., Dawson-Andoh, B.E., 2009. Sono-chemical preparation of cellulose nanocrystals from lignocellulose derived materials. Bioresour. Technol. 100, 2259–2264. Ghose, T.K., 1987. Measurement of cellulase activities. Pure Appl. Chem. 59, 257– 268. Goering, H.K., Vansoest, P.J., 1970. Forage fibre analysis. Agriculture Handbook. Agricultural Research Services, United States Department of Agriculture. Gonzalez, G., Lopes-Santin, J., Caminal, G., Sola, C., 1986. Dilute acid hydrolysis of wheat straw hemicellulose at moderate temperature: a simplified kinetic model. Biotechnol. Bioeng. 28, 288–293. Guo, G.L., Chen, W.H., Chen, W.H., Men, L.C., Hwang, W.S., 2008. Characterization of dilute acid pretreatment of silvergrass for ethanol production. Bioresour. Technol. 99, 6046–6053. Hendriks, A.T.W.M., Zeeman, G., 2009. Pretreatments to enhance the digestibility of lignocellulosic biomass. Bioresour. Technol. 100, 10–18. Hernandez-Salas, J.M., Villa-Ramirez, M.S., Veloz-Rendon, J.S., Rivera-Hernandez, K.N., Gonzalez-Cesar, R.A., Plascencia-Espinosa, M.A., Trejo-Estrada, S.R., 2009. Comparative hydrolysis and fermentation of sugarcane and agave bagasse. Bioresour. Technol. 100, 1238–1245. Larsson, S., Palmqvist, E., Hahn-Hagerdal, B., Tengborg, C., Stenberg, K., Zacchi, G., Nilvebrant, N., 1999. The generation of fermentation inhibitors during dilute acid hydrolysis of softwood. Enzyme Microb. Technol. 24, 151–159. Lu, X., Zhang, Y., Angelidaki, I., 2009. Optimization of H2SO4-catalyzed hydrothermal pretreatment of rapeseed straw for bioconversion to ethanol: focusing on pretreatment at high solids content. Bioresour. Technol. 100, 3048–3053. Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Annl. Chem. 31, 426–428. Moiser, N., Wyman, C., Dale, B., Elander, R., Lee, Y.Y., Holtzapple, M., Ladisch, M., 2005. Features of promising technologies for pretreatment of lignocellulosic biomass. Bioresour. Technol. 96, 673–686. Pan, X., 2008. Role of functional groups in lignin inhibition of enzymatic hydrolysis of cellulose to glucose. J. Biobased Mater. Bioenergy 2 (1), 25–32. Rosgaard, L., Pedersen, S., Meyer, A.S., 2007. Comparison of different pretreatment strategies for enzymatic hydrolysis of wheat and barley straw. Appl. Biochem. Biotechnol. 143, 284–296. Segal, L., Creely, J.J., Martin, J.A.E., Conrad, C.M., 1959. An empirical method for estimating the degree of crystallinity of native cellulose using the X-ray diffractometer. Text. Res. J., 786–794. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., 2008. Determination of sugars, byproducts, and degradation products in liquid fraction Process Samples. Laboratory Analytical Procedure (LAP). Technical, Report NREL/TP-510-42623. Sun, J.X., Sun, R., Sun, X.F., Su, Y., 2004. Fractional and physico-chemical characterization of hemicelluloses from ultrasonic irradiated sugarcane bagasse. Carbohydr. Res. 339, 291–300. Updegroff, D.M., 1969. Semi micro determination of cellulose in biological materials. Anal. Biochem. 32, 420. Velmurugan, R., Muthukumar, K., 2011. Utilization of sugarcane bagasse for bioethanol production: sonoassisted acid hydrolysis approach. Bioresour. Technol.. doi:10.1016/j.biortech.2011.04.045. Wooley, R., Ruth, M., Sheehan, J., Ibsen, K., Majdeski, H., Galvez, A., 1999. Lignocellulosic Biomass to Ethanol – Process Design and Economics Utilizing Co-current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis – Current and Futuristic Scenarios. NREL, Golden. Wright, J.D., 1998. Economics of Enzymatic Hydrolysis Processes. Solar Energy Research Institute (SERI), Presented at the American Institute of Chemical Engineers Spring National Meeting, 6–10. New Orleans, LA. Yaldagard, M., Mortazavi, S.A., Tabatabaie, F., 2008. The effect of ultrasound in combination with thermal treatment on the germinated barley’s alpha-amylase activity. Korean J. Chem. Eng. 25 (3), 517–523. Zhang, Q., Cai, W., 2008. Enzymatic hydrolysis of alkali-pretreated rice straw by Trichoderma reesei ZM4-F3. Biomass Bioenergy 32, 1130–1135.
Contents lists available at SciVerse ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Ultrasound-assisted alkaline pretreatment of sugarcane bagasse for fermentable sugar production: Optimization through response surface methodology Rajendran Velmurugan, Karuppan Muthukumar ⇑ Department of Chemical Engineering, Alagappa College of Technology Campus, Anna University Chennai, Chennai 600 025, India
a r t i c l e
i n f o
Article history: Received 29 April 2011 Received in revised form 26 January 2012 Accepted 30 January 2012 Available online 8 February 2012 Keywords: Sugarcane bagasse Ultrasound Alkaline pretreatment Lignin removal RSM
a b s t r a c t Ultrasound-assisted alkaline pretreatment of sugarcane bagasse (SCB) for fermentable sugar production was carried out and the influence of particle size, liquid to solid ratio (LSR), NaOH concentration, temperature and sonication time on delignification and reducing sugar production was ascertained with Placket–Burman design. The best combination of each significant factor was determined by a central composite design (CCD) and optimum pretreatment conditions for maximum reducing sugar yield (96.27%) were particle size of 0.27 mm, LSR of 25 ml/g, NaOH concentration of 2.89% (w/v), temperature of 70.15 °C and pretreatment time of 47.42 min. Under these conditions, 92.11% of theoretical reducing sugar yield was observed experimentally. The substantial reduction in pretreatment time and temperature with improved efficiency is the most attractive features of the ultrasound-assisted alkaline pretreatment. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Enzymatic hydrolysis is regarded as the most promising technique for converting lignocellulosic compounds into fermentable sugars (Wooley et al., 1999; Wright, 1998); however, this approach requires biomass in small particles and removal of lignin (Dasari and Berson, 2007; Pan, 2008). Pretreatments for lignocellulosic materials include steam explosion (Hernandez-Salas et al., 2009), dilute acid hydrolysis (Hendriks and Zeeman, 2009), alkaline pretreatment (Zhang and Cai, 2008; Hendriks and Zeeman, 2009) and wet oxidation (Rosgaard et al., 2007). Alkaline pretreatments show less sugar degradation and furan derivatives formation (Gonzalez et al., 1986) than thermal and acid pretreatments (Fengel and Wegener, 1984). The alkaline pretreatment process can be improved further by the application of ultrasound (Filson and Dawson-Andoh, 2009). The ultrasonic treatment of aqueous media produces cavitation, which generates high temperature, pressure and extreme shear forces. The decomposition of water molecules into free radicals by cavitation aids in cleaving the linkages in lignin and xylan networks (Chuanyun et al., 2004; Yaldagard et al., 2008). Sun et al. (2004) treated sugarcane bagasse (SCB) with alkali/alkaline peroxide and ultrasound for the extraction of hemicellulose from SCB and reported 90% hemicellulose and lignin removal, but more detailed analyses of factors influencing delignification or cellulose recovery are needed. Therefore, the aim of the present study was to optimize particle size, liquid to solid ratio (LSR),
NaOH concentration and sonication time on ultrasound-assisted alkaline pretreatment by response surface methodology (RSM) with the intention to improve the delignification and reducing sugar yield. 2. Methods 2.1. Materials Sugarcane bagasse was obtained from a sugarcane juice shop in Chennai, India. The SCB was air dried, milled using a laboratory blender (Remi Anupam Mixie Ltd., Mumbai) and screened to obtain particles of different size ranges. The particle size ranges are average values of upper and lower sieves openings of ASTM and BS sieves. Determination of the main fractions (cellulose, hemicelluloses, and lignin) was carried out by detergent extraction methods (Goering and Vansoest, 1970). 2.2. Sonochemical reactor The ultrasonic treatment was carried out using a titanium probe-type sonolyzer (Hielscher UP 400S, Germany). The operating frequency and power of the sonolyzer were 24 kHz and 400 W, respectively. The amplitude was maintained at 100% and the temperature was controlled using a water bath. 2.3. Design of experiments
⇑ Corresponding author. Tel.: +91 44 22359153; fax: +91 44 22352642. E-mail address: [email protected] (K. Muthukumar). 0960-8524/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2012.01.168
The entire design was carried out in two stages. Initially, five variables were screened to identify significant factors using a
294
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299
two-level Plackett–Burman design (PBD) and in the second stage the level of these factors were optimized using a central composite design (CCD). 2.3.1. Screening of parameters The effect of particle size, LSR, NaOH concentration, temperature and sonication time on reducing sugar production was evaluated using PBD in twelve experimental runs. The software used for PBD was MINITAB 16. The variables with a confidence level higher than 90% were considered significant for both delignification and reducing sugar production. All experiments were carried out in duplicate and the averages were taken as responses. 2.3.2. Optimization of significant parameters Effects of the variables such as particle size, LSR, NaOH concentration, temperature and sonication time on response were fit with the second-order polynomial model given in Eq. (1).
Y ¼ b0 þ b1 A þ b2 B þ b3 C þ b4 D þ b5 E þ b11 A21 þ b22 A22 þ b33 A23 þ b44 A24 þ b55 A25 þ b12 AB þ b13 AC þ b14 AD þ b15 AE þ b23 BC þ b24 BD þ b25 BE þ b34 CD þ b35 CE þ b45 DE
ð1Þ
where Y is the predicted response; b0 is the constant; b1, b2, b3, b4, and b5 are linear coefficients; b11, b22, b33, b44, and b55 are quadratic coefficients; b12, b13, b14, b15, b23, b24, b25, b34, b35, and b45 are interaction coefficients; A, B, C, D, and E are factors representing particle size, liquid to solid ratio, NaOH concentration, temperature and sonication time, respectively. Five significant variables were considered for the optimization using a five-level CCD. ANOVA and regression analyses were carried out to evaluate the effects of variables and their interactive effects. Coefficients of the full model were analyzed for their significance and the insignificant ones were eliminated from the model. The predicted response and optimal levels of the variables were found by solving the second-order polynomial model for maximization of response using MINITAB 16 software (Lu et al., 2009).
SCB biomass was dispersed in 100 ml of desired concentration of NaOH solution (0.25%, 1.00%, 1.75%, 2.50% and 3.25%) in an Erlenmeyer flask and treated with ultrasound at different operating conditions based on the experimental design. The performance of ultrasound-assisted alkaline pretreatment was compared with those of alkaline pretreatment carried out at 30 ± 1 °C, ultrasound treatment in the absence of NaOH and commercial pretreatment (2% NaOH in an autoclave at121 °C for 1 h) (Zhang and Cai, 2008; Hendriks and Zeeman, 2009). After the treatment, the contents were filtered through Whatman No. 1 filter paper and the supernatant was subjected to sugar, acetic acid and furfural analyses. The solid residue was repeatedly washed with water until the filtrate was neutral. The residue was dried at 50 °C until constant weight was observed and used for saccharification experiments. Experiments were performed in triplicate and mean values were employed in the analyses. The sugar concentration in terms of gram per gram was calculated by dividing the sugar concentration in pretreated solids by the total solids recovered after pretreatment. The polysaccharide (glucan, xylan and arabinan) recovery in the solid content was calculated using the following equation:
PPT-SCB 100 PSCB
% delignification ¼
LSCB LPT-SCB 100 LSCB
ð3Þ
where LSCB is the concentration of lignin in SCB, LPT-SCB is the concentration of lignin in pretreated SCB, which was calculated by multiplying estimated lignin concentration (g/g) by recovered solids after the pretreatment. 2.5. Enzymatic convertibility Pretreated SCBs were weighed and suspended in 100 ml of 50 mM sodium citrate buffer (pH 4.8) supplemented with 0.02% sodium azide under aseptic conditions. The contents were mixed with 25 FPU/g dry matter and 0.46 CBU/g cellulose of commercially available cellulase and b-glucosidase (Sisco Research Laboratory (SRL), India), respectively. The reaction mixture was incubated in a shaking incubator at 45 °C and 150 rpm for 26 h (Aswathy et al., 2010). The release of soluble reducing sugars was periodically measured with the dinitrosalicylic acid (DNS) method (Miller, 1959). 2.6. Analytical methods 2.6.1. Physical characterization The morphological differences of SCB pretreated by various methods were examined using scanning electron microscopy (SEM, JEOL Ltd., Tokyo, Japan). The crystalline nature of the raw and treated SCB was analyzed using a Rigaku RINT-TTR3 X-ray diffractometer (Rigaku Co., Tokyo, Japan). The nickel-filtered Cu Ka radiation (k = 0.1542 nm) was applied at 50 kV and 30 mA. Samples were scanned over the range of 2h = 5–50° and the crystallinity index (CrI) was determined using Eq. (4) (Segal et al., 1959):
CrI ¼
ICrystalline IAmorphous 100 % ICrystalline
ð4Þ
where, Icrystaline = intensity at 21° and Iamorphous = intensity at 18.8°.
2.4. Experimental
% recovered ¼
where PSCB is the concentration of polysaccharide in SCB, PPT-SCB is the concentration of polysaccharide in pretreated SCB, which is calculated by multiplying estimated sugar concentration (g/g) by recovered solids after pretreatment. The % delignification was calculated using the following equation:
ð2Þ
2.6.2. Chemical characterization The total reducing sugars in enzymatic hydrolysates of SCB was determined by the DNS method (Miller, 1959) and cellulose concentration was estimated spectrophotometrically by the anthrone method (Updegroff, 1969). Anthrone and DNS reagents were obtained from SRL, India. The concentration of sugars, acetic acid and furfural was determined using high performance liquid chromatography (HPLC). The HPLC system (Shimadzu, CA, USA) consisted of a liquid pump (LC-10AD), a refractive index detector (RID-6A) and a system controller with Shimadzu EZS software. The monosaccharides (glucose, xylose and arabinose) were separated in an ion-exchange column (Aminex HPX-87P Bio-Rad Laboratories, Hercules, CA) at 85 °C. Ultra-pure water was used as mobile phase at a flow rate of 0.6 ml/min (Sluiter et al., 2008). The degradative products (acetic acid and furfural) were separated using an ion exchange column (Aminex HPX-87H) at 65 °C. The mobile phase, 5 mM H2SO4, was used at a flow rate of 0.6 ml/min (Sluiter et al., 2008). The composition of raw and pretreated SCBs was determined by strong acid hydrolysis as described previously (Velmurugan and Muthukumar, 2011).The conversion factors for hexose (glucose) and pentoses (xylose and arabinose) were 1.111 and 1.136, respectively (Demirbas, 2005). Hemicellulose, lignin and ash contents were estimated by methods described by Goering and Vansoest (1970) and APHA (2005). Cellulase activity
295
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299
was determined by standard filter paper assay and expressed as filter paper units (FPU). One FPU is defined as the enzyme that releases 1 lmol of glucose equivalents per minute from Whatman No. 1 filter paper (Ghose, 1987). Endoglucanase activity was measured using carboxymethyl cellulose as a substrate. The b-glucosidase activity was measured by cellobiase assay method (Ghose, 1987). 3. Results and discussion 3.1. Composition of raw SCB Raw SCB contained 35.6% cellulose, 26.6% hemicellulose, 17.1% lignin, 1.1% ash, 11.5% moisture and 6.4% protein (w/w). Xylan and arabinan was accounted for 17.84% and 4.42% (w/w), respectively. The compositional variability among five different particle size fractions was within the error limit of 5%. 3.2. Placket–Burman design Particle size, LSR, NaOH concentration, temperature and sonication time on delignification were identified as most significant and contributing variables and their ranges were 0.27–0.91 mm, 10–20 ml/g, 0.25–2.5% (w/v), 30–50 °C and 5–50 min, respectively (Table 1). The analysis of experimental data indicated wide variation in delignification (37.77–74.27%) and reducing sugar yield (67.08–88.94%) (Table 2). These variations indicated the necessity for optimization of the parameters. The statistically significant effects of each factor were screened by probability test and factors with a confidence interval greater than 90% (Prob > [t] < 0.1) were considered as a significant parameter. The lowest probability (<0.0001) values were obtained for sonication time and this indicates the significant influence of this factor. The LSR, NaOH concentration and temperature were also significant (Prob > [t] = 0.003). The particle size had limited significance on lignin removal (Prob > [t] = 0.053). The sequential effect of the factors was explained by a Pareto chart (Fig. 1a) and the most important factors determining delignification were found to be in the order: sonication time, temperature, NaOH concentration, LSR and particle size. For reducing sugar production, the preferred order was: sonication time, particle size, NaOH concentration, temperature and LSR (Fig. 1b). Sonication intensifies the chemical reaction by generating cavitation and shear forces (Yaldagard et al., 2008; Filson and Dawson-Andoh, 2009) but pretreatment conditions such as temperature, NaOH concentration, LSR and particle size are also important for lignin removal. The influence of particle size was more significant for reducing sugar production than for delignification due to the fact that the biomass particle size variation occurred even during pretreatment. Generally, the reduction in biomass
particle size improves the mass transfer and increases the accessible area available to the enzyme (Guo et al., 2008). 3.3. Central composite design In order to optimize the significant factors, both delignification and reducing sugar production was considered. In five factor optimization, concentrations of sugars, lignin and degradative products in solid and liquid content were used. The responses of 32 sets of experiments were analyzed statistically to obtain the optimum value of each factor and the results are summarized in Tables 3 and 4. 3.3.1. Effect of pretreatment on composition of liquor Xylose was the major product in all experiments (Table 4). Furfural concentrations were ranged from 0.21 to 0.71 g/L. Acetic acid release was from 1.79 to 9.28 g/L, which is below the level of 6 g/L that inhibits cellulase activity and ethanol production (Larsson et al., 1999). 3.3.2. Effect of pretreatment on composition of solids The recovery of various important components achieved in the solid phase is shown in Table 4 and the % recovery was in the range of 72.14–89.12. The recovery in the solid phase was expressed as the ratio of dry weight of pretreated solid to the raw SCB. The maximum recovery (89.12%) was obtained with 0.75 mm SCB particles at a LSR of 10, 1% NaOH, 30 °C and 20 min reaction time. The pretreatment conditions were chosen to produce a solid substrate with the lowest possible lignin content and the least amount of sugar degradation. The glucan recovery was in the range of 97.71–99.43% (Table 4) which indicates that the pretreatment conditions did not affect cellulose content. In contrast, the pretreatment solubilized more pentosan than glucan, which resulted in a higher glucan recovery in the solid content. The recovery of xylan and arabinan was in the range of 69.64–93.10% and 61.19–86.22%, respectively. The higher solid content (run 26) showed higher pentose recovery but greater solid losses and less delignification. Therefore, while selecting pretreatment based on delignification, polysaccharides recovery in the solid phase and saccharification of solids after pretreatment should also be considered. 3.3.3. Effect of pretreatment on delignification An analysis of variance (ANOVA) was performed and the results are presented in Table 5a. The results indicated that the fitting model and lack of fit (p < 0.001) were highly significant. The coefficient of determination (R2) was 0.9954, which indicated that the model could explain 99.54% variability of the response variable. The adjusted coefficient of determination (R2 = 98.69) was satisfactory and confirmed the significance of the model. Analysis of the
Table 1 Plackett–Burman design variables (in coded levels) with delignification and reducing sugar concentration as responsea.
a
Run order
A
B
C
D
E
Delignification (%)
Reducing sugar yield (%)
1 2 3 4 5 6 7 8 9 10 11 12
+ + + + + +
+ + + + + +
+ + + + + +
+ + + + + +
+ + + + + +
65.57 39.71 37.77 46.47 29.24 40.59 73.39 71.63 64.75 74.27 38.77 62.24
81.00 70.76 67.08 69.88 67.50 65.62 88.94 88.48 81.08 83.80 70.22 84.76
A: Particle size (mm), B: liquid to solid ratio (v/w), C: NaOH (%, w/v), D: temperature (°C), E: time (min).
296
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299
Table 2 Levels of the variables and statistical analysis of Plackett–Burman design. Factors
Particle size (mm) LSR (v/w)a NaOH (%, w/v) Temperature (°C) Time (min)
Code
A B C D E
Low level (1)
0.27 10 0.25 30 5
High level (+1)
0.91 20 2.5 50 50
Delignification (%) b
Effect
Coeff
6.255 12.027 12.459 12.585 21.225
3.128 6.013 6.230 6.292 10.612
Reducing sugar yield (%) t-Value
p-Value
Effect
Coeffb
t-Value
p-Value
2.42 4.66 4.83 4.88 8.23
0.052 0.003 0.003 0.003 0.000
76.593 7.320 3.880 5.633 4.687
3.660 1.940 2.817 2.343 6.273
0.00 6.09 3.23 4.69 3.90
0.000 0.001 0.018 0.003 0.008
a
LSR: Liquid to solid ratio. Coeff: Coefficient. R for delignification = 95.96% (adjusted R2 = 92.59%), R2 for reducing sugar yield = 97.00% (adjusted R2 = 94.50%).
b
2
value for delignification was 99.64% at a particle size of 0.27 mm, LSR of 25 ml/g, NaOH concentration of 3.25%, temperature of 75 °C and sonication time of 61.97 min.
Fig. 1. Pareto charts (a) standardized effects of process terms on delignification; and (b) standardized effect of process terms on reducing sugar yield (a = 0.1).
3.3.4. Effect of pretreatment on enzymatic saccharification An ANOVA was performed for the reducing sugar yield and the results are presented in Table 5a. The values of p were less than 0.001, indicating that the model was highly significant. The coefficient of determination (R2) value was 0.9884 and this indicates that the model could explain 98.84% variability of the response variable. The adjusted value of coefficient of determination (R2 = 0.9674) was also satisfactory, indicating the significance of the model. Analysis of the response trends showed that the model could explain the effect of particle size, LSR, NaOH concentration, temperature and sonication time on reducing sugar yield satisfactorily. The coefficients calculated by regression analysis for each variable are presented in Table 5b and the regression coefficients of linear and squared terms were highly significant (p < 0.1). The interactive terms were not significant for reducing sugar yield, indicating a negligible interactive effect on reducing sugar yield. The positive coefficients of LSR, NaOH concentration, temperature and sonication time indicate that such factors had a positive effect on reducing sugar yield. The particle size had a negative coefficient (8.607) on reducing sugar yield and this indicates that the increase in particle size had a negative effect on the reducing sugar yield. The optimum values of variables were obtained by varying the factors within the experimental range and the other factors were maintained constant at their middle level. The maximum response value for reducing sugar yield was estimated as 96.27% at a particle size of 0.27 mm, LSR of 25 ml/g, NaOH concentration of 2.89%, temperature of 70.15 °C and sonication time of 47.42 min. 3.4. Comparative analysis
response trends was reasonable and the model explained the effect of particle size, LSR, NaOH concentration, temperature and sonication time on delignification satisfactorily. Model coefficients were evaluated by regression analysis and their significance was tested. The values for linear, square and interaction effects are presented in Table 5b. The interactive effects of particle size and NaOH concentration, and LSR and temperature on delignification were highly significant; however, the interactive effect for eight terms on delignification was insignificant (p > 0.1), indicating that these interactions had little impact on delignification. Positive coefficients indicate a linear increase in delignification while negative coefficients indicate a linear decrease in delignification. Among the various factors, particle size had a negative effect on delignification during ultrasound-assisted alkaline pretreatment. The optimum values of different variables were obtained by varying the selected factor within the experimental range and maintaining the other factors constant at their middle level. The maximum response
Ultrasound-assisted alkaline pretreatment solubilized the lignin matrix more effectively than autoclave pretreatment and hence, the performance of saccharification was improved considerably. SEM analysis (Fig. S1) (Supplementary material) demonstrated that raw SCB displayed a regular and compact surface structure while the pretreated SCB showed histological changes. The outer surface was destroyed and perforated by the pretreatment processes and pores were approximately 2.3 lm in size. These observations may be due to the oxidation of lignin matrix by hydroxyl radicals formed during sonication (Moiser et al., 2005). The pores formed may increase enzyme accessibility, which in turn increases the enzyme digestibility of the SCB and shorten the enzymatic hydrolysis time (Moiser et al., 2005). The diffraction patterns of raw, ultrasound-assisted alkaline pretreated and alkaline pretreated SCB (Fig. S2) showed broad peak for raw SCB and pretreated (both ultrasound and autoclave) SCB showed a narrow peak at 2h = 21.9°. The crystallinity index of pretreated SCB (sono-assisted)
297
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299 Table 3 Central composite design for the optimization of factors identified by Plackett–Burman designa.
a
Run order
A
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
0.75 0.59 0.59 0.59 0.59 0.59 0.43 0.59 0.59 0.43 0.75 0.43 0.43 0.75 0.59 0.43 0.75 0.59 0.43 0.43 0.59 0.27 0.75 0.59 0.75 0.59 0.59 0.59 0.75 0.75 0.91 0.43
(1) (0) (0) (0) (0) (0) (1) (0) (0) (1) (1) (1) (1) (1) (0) (1) (1) (0) (1) (1) (0) (2) (1) (0) (1) (0) (0) (0) (1) (1) (2) (1)
B
C
20 (1) 15 (0) 15 (0) 15 (0) 15 (0) 15 (0) 20 (1) 15 (0) 15 (0) 20 (1) 10 (1) 10 (1) 10 (1) 20 (1) 25 (2) 10 (1) 10 (1) 15 (0) 20 (1) 10 (1) 15 (0) 15 (0) 20 (1) 15 (0) 20 (1) 5 (2) 15 (0) 15 (0) 10 (1) 10 (1) 15 (0) 20 (1)
2.50 3.25 0.25 1.75 1.75 1.75 2.50 1.75 1.75 1.00 2.50 1.00 2.50 1.00 1.75 2.50 2.50 1.75 1.00 1.00 1.75 1.75 2.50 1.75 1.00 1.75 1.75 1.75 1.00 1.00 1.75 2.50
(1) (2) (2) (0) (0) (0) (1) (0) (0) (1) (1) (1) (1) (1) (0) (1) (1) (0) (1) (1) (0) (0) (1) (0) (1) (0) (0) (0) (1) (1) (0) (1)
D
E
60 (1) 45 (0) 45 (0) 45 (0) 45 (0) 45 (0) 30 (1) 45 (0) 45 (0) 30 (1) 60 (1) 30 (1) 60 (1) 60 (1) 45 (0) 30 (1) 30 (1) 75 (2) 60 (1) 60 (1) 45 (0) 45 (0) 30 (1) 15 (2) 30 (1) 45 (0) 45 (0) 45 (0) 60 (1) 30 (1) 45 (0) 60 (1)
50 (1) 35 (0) 35 (0) 35 (0) 35 (0) 35 (0) 50 (1) 5 (2) 35 (0) 20 (1) 20 (1) 50 (1) 50 (1) 20 (1) 35 (0) 20 (1) 50 (1) 35 (0) 50 (1) 20 (1) 35 (0) 35 (0) 20 (1) 35 (0) 50 (1) 35 (0) 65 (2) 35 (0) 50 (1) 20 (1) 35 (0) 20 (1)
Delignification(%)
Reducing sugar yield (%)
Actual
Predicted
Actual
Predicted
77.71 74.93 58.29 68.71 68.71 68.71 77.24 59.10 68.71 61.14 62.84 58.96 74.61 63.98 73.86 65.22 65.01 73.79 73.40 59.23 68.71 75.05 67.39 61.46 63.12 58.24 70.89 68.71 61.51 52.54 63.57 78.31
79.50 76.51 58.47 70.92 70.92 70.92 74.96 60.04 70.92 60.32 64.89 57.88 73.88 66.47 75.38 64.41 64.89 73.42 74.76 59.98 70.92 74.16 68.28 59.57 63.52 58.84 69.81 70.92 62.97 50.84 68.44 78.06
86.13 85.22 74.74 83.70 83.70 83.70 86.98 76.51 83.70 79.07 79.19 75.84 86.04 79.84 86.90 81.34 78.45 84.45 85.51 77.93 83.70 88.19 81.96 77.12 77.05 76.46 83.92 83.70 76.08 71.49 80.09 89.36
86.47 85.61 74.34 83.70 83.70 83.70 87.28 77.98 83.70 78.83 78.70 76.25 86.30 79.63 86.71 80.81 78.64 84.33 86.07 77.65 83.70 88.14 81.50 77.23 77.53 76.64 82.44 83.70 76.52 71.13 80.13 88.98
A: Particle size (mm), B: liquid to solid ratio (v/w), C: NaOH (%, w/v), D: temperature (°C), E: time (min).
Table 4 Recovery of sugars and composition of liquor obtained after the pretreatment of sugarcane bagasse. Run order
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Solids recovered (%)
74.53 77.05 85.63 82.59 80.94 82.29 79.04 84.23 82.29 86.66 81.69 85.62 79.35 82.48 79.49 82.80 80.50 77.38 78.68 85.37 80.94 80.25 79.92 82.86 82.78 72.14 80.26 82.14 82.46 89.12 82.00 79.04
Sugars recovered (%)
Composition of liquor obtained after pretreatment (g/L)
Glucan
Xylan
Arabinan
Glucose
Xylose
Arabinose
Acetic acid
Furfural
97.71 98.43 99.06 98.88 98.88 98.88 98.21 99.12 98.88 99.33 98.58 98.95 98.40 98.49 98.44 99.00 98.23 98.35 98.41 99.27 98.88 99.02 98.39 98.94 98.24 99.05 98.46 98.88 98.19 99.43 98.48 98.58
69.64 75.44 84.14 80.32 80.32 80.32 77.65 84.45 80.32 87.01 77.63 86.56 77.54 76.81 76.17 82.24 77.52 74.64 77.95 85.77 80.32 83.10 77.25 81.88 78.13 93.10 78.50 80.32 77.29 89.48 76.56 77.63
61.19 68.13 79.64 74.84 74.84 74.84 70.80 79.94 74.84 84.37 72.15 82.15 71.20 71.40 70.99 77.75 70.47 68.37 72.10 82.82 74.84 78.26 70.93 76.75 71.54 86.22 71.77 74.84 70.90 86.86 70.15 72.15
0.46 0.41 0.24 0.29 0.30 0.29 0.35 0.22 0.29 0.13 0.55 0.41 0.62 0.28 0.24 0.40 0.70 0.44 0.31 0.28 0.29 0.26 0.31 0.28 0.34 0.49 0.40 0.30 0.71 0.23 0.39 0.29
6.08 6.46 4.43 5.15 5.20 5.44 4.64 4.30 5.06 2.80 8.78 5.65 9.00 4.47 4.05 6.96 8.86 6.62 4.20 5.65 5.06 4.35 4.68 4.77 4.26 3.80 5.65 5.15 9.00 4.30 6.24 4.77
0.95 1.04 0.66 0.82 0.82 0.81 0.70 0.61 0.82 0.37 1.36 0.87 1.41 0.68 0.55 1.09 1.46 1.03 0.68 0.85 0.82 0.72 0.72 0.75 0.69 1.36 0.91 0.82 1.42 0.62 0.97 0.67
1.91 2.86 4.75 3.56 3.56 3.56 1.95 4.66 3.56 3.32 6.35 7.02 4.34 3.08 1.79 5.95 5.98 2.98 2.27 6.97 3.56 2.84 2.79 4.39 3.15 9.28 3.32 3.56 6.58 8.12 4.15 1.85
0.71 0.58 0.41 0.48 0.48 0.48 0.52 0.39 0.49 0.28 0.51 0.31 0.52 0.55 0.56 0.40 0.52 0.64 0.49 0.29 0.47 0.41 0.51 0.44 0.50 0.40 0.50 0.47 0.54 0.21 0.57 0.51
298
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299
Table 5a ANOVA for selected central composite design for delignification and reducing sugar yield. Source
Regression Linear Square Interaction Residual Error Lack-of-fit Pure error Total
Delignification (%)
Reducing sugar yield (%)
DF
Sum of squares
Mean value
F value
Prob > F
DF
Sum of squares
Mean value
F value
Prob > F
20 5 5 10 11 6 5 31
1398.21 1336.62 45.22 16.36 6.51 6.51 0.00 1404.72
69.910 267.325 9.045 1.636 0.592 1.085 0.000
118.07 451.50 15.28 2.76
0.00 0.00 0.00 0.06
30.46 108.83 12.13 0.43 0.65 1.19 0.00
0.000 0.000 0.000 0.735
0.00
609.127 544.146 60.660 4.321 7.139 7.139 0.000 616.266
46.93 167.69 18.69 0.67
0.00
20 5 5 10 11 6 5 31
0.00 0.00
0.000
R2 for delinification = 99.54% (adjusted R2 = 98.69%); R2 for reducing sugar yield = 98.84% (adjusted R2 = 96.74%).
Table 5b Regression analysis for selected central composite design for delignification and reducing sugar yield. Terma
Constant A B C D E AA BB CC DD EE AB AC AD AE BC BD BE CD CE DE a b
Delignification
Reducing sugar yield b
Coeff
SE coeff
t
p
Coeffb
SE coeffb
t
p
22.477 3.478 1.262 12.599 0.262 0.397 5.250 0.027 0.962 0.001 0.004 0.138 5.658 0.130 0.008 0.059 0.007 0.000 0.021 0.004 0.001
0.3069 0.1571 0.1571 0.1571 0.1571 0.1571 0.1421 0.1421 0.1421 0.1421 0.1421 0.1924 0.1924 0.1924 0.1924 0.1924 0.1924 0.1924 0.1924 0.1924 0.1924
223.936 15.119 24.828 28.576 17.412 17.108 0.946 4.791 3.810 2.014 6.651 0.573 3.529 1.627 0.099 1.158 2.615 0.119 1.214 0.226 1.571
0.000 0.000 0.000 0.000 0.000 0.000 0.364 0.001 0.003 0.069 0.000 0.578 0.005 0.132 0.923 0.271 0.024 0.908 0.250 0.825 0.145
45.037 8.607 1.026 12.050 0.425 0.347 4.246 0.020 1.658 0.003 0.004 0.001 2.124 0.086 0.037 0.012 0.003 0.001 0.022 0.002 0.001
0.3213 0.1644 0.1644 0.1644 0.1644 0.1644 0.1487 0.1487 0.1487 0.1487 0.1487 0.2014 0.2014 0.2014 0.2014 0.2014 0.2014 0.2014 0.2014 0.2014 0.2014
260.491 12.186 15.313 17.127 10.781 6.774 0.731 3.411 6.268 4.912 5.866 0.002 1.266 1.031 0.446 0.224 1.137 0.317 1.217 0.126 0.924
0.000 0.000 0.000 0.000 0.000 0.000 0.480 0.006 0.000 0.000 0.000 0.998 0.232 0.325 0.665 0.827 0.280 0.757 0.249 0.902 0.375
A: Particle size (mm), B: liquid to solid ratio (v/w), C: NaOH (%, w/v), D: temperature (°C), E: time (min). Coeff: Coefficient.
Table 6 Comparative analysis of different pretreatment methods. Factors
Raw SCB NaOH, room temperature Ultrasound NaOH, ultrasoundb NaOH, autoclave
Composition of liquor obtained after pre-treatment (g/L)
Percentage recovered (%)
Glucose
Xylose
Arabinose
Acetic acid
Furfural
Glucan
Xylan
Arabinan
ND 0.01 0.01 0.13 0.28
ND 0.20 0.85 1.60 4.16
ND 0.28 0.37 0.47 0.64
ND 0.23 0.18 0.36 0.53
ND ND ND ND 0.12
ND 99.5 99.5 98.32 96
ND 97.8 90.47 77.95 34.98
ND 87.3 83.26 71.16 23.1
Lignin removal (%)
Reducing sugar yield (%)a
ND 2.50 7.30 82.32 46.50
28.90 30.50 31.20 92.11 72.00
ND – not detectable. a Theoretical yield of reducing sugar obtained after saccharification. b RSM optimized conditions.
was 67%, whereas that for raw SCB was 48.50% due to the removal of hemicellulose and lignin fractions which increased the relative content of crystalline cellulose. The crystallinity index for alkaline pretreatment at 121 °C was 65.40%, which confirmed that the removal of amorphous domains was more or less similar to that of the ultrasound-assisted alkaline pretreatment. However, after the pretreatment, the recovery of domains containing sugars should be high. The sugar recovery and reducing sugar yield for ultrasound-assisted pretreatment and other pretreatment methods are presented in Table 6. The lignin removal and cellulose recovery
achieved with ultrasound assisted pretreatment were 82.32% and 98.32%, respectively. The autoclave pretreatment reduced the lignin content by 46.50% and observed cellulose recovery was 96%. The ultrasound assisted pretreated solids produced 92.11% of theoretical reducing sugar yield, which is comparatively more or less similar to RSM predicted yield (96.27%) and higher than the yield obtained from autoclave pretreatment (72%). The formation of degradative product (acetic acid) was observed in ultrasound assisted pretreatment but its concentration was very less (0.36 g/L). The improved lignin removal and the reducing sugar yield clearly
R. Velmurugan, K. Muthukumar / Bioresource Technology 112 (2012) 293–299
demonstrated that the application of ultrasound improved the alkaline pretreatment effectively. 4. Conclusions SCB pretreated with ultrasound-assisted alkaline pretreatment showed better reducing sugar yield than commercial alkaline pretreatment. The substantial reduction in pretreatment time and temperature with improved efficiency is the most attractive feature of the ultrasound-assisted alkaline pretreatment. However, to implement this process at a larger scale, design of suitable reactor and the optimization of energy required need to be done. Acknowledgements The authors express their sincere thanks to the Editor and the Reviewers for their constructive suggestions. The authors gratefully acknowledge Department of Science and Technology, New Delhi for providing financial support to carry out this research work under PURSE scheme. One of the authors Mr. R. Velmurugan is grateful to DST, New Delhi for the award of DST-PURSE fellowship. The authors also express their sincere thanks to Dr. A. Peer Mohamed, Department of Textile Technology, Anna University, Chennai for his valuable suggestions. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.biortech.2012.01.168. References APHA, 2005. Standard methods for the examination of water and wastewater. In: Eaton, A.D., Clesceri, L.S., Rice, E.W., Greenberg, A.E., Franson, M.A.H. (Eds.), 21st ed. American Public Health Association, American Water Works Association, Water Environment Federation, USA (centennial edition). Aswathy, U.S., Sukumaran, R.K., LalithaDevi, G., Rajasree, K.P., Singhania, R.R., Pandey, A., 2010. Bio-ethanol from water hyacinth biomass: an evaluation of enzymatic saccharification strategy. Bioresour. Technol. 101, 925–930. Chuanyun, D., Bochu, W., Huan, Z., Conglin, H., Chuanren, D., Wangqian, L., Toyama, Y., Sakanishi, A., 2004. Effect of low frequency ultrasonic stimulation on the secretion of siboflavin produced by Ecemothecium ashbyii. Colloids Surf. B 34 (1), 7–11. Dasari, R.K., Berson, R.E., 2007. The effect of particle size on hydrolysis reaction rates and rheological properties in cellulosic slurries. Appl. Biochem. Biotechnol., 136–140. Demirbas, A., 2005. Bioethanol from cellulosic materials: a renewable motor fuel from biomass. Energy Sources 21, 327–337. Fengel, D., Wegener, G., 1984. Wood – Chemistry, Ultrastructure, Reactions. Walter de Gruyter, Berlin, New York.
299
Filson, P.B., Dawson-Andoh, B.E., 2009. Sono-chemical preparation of cellulose nanocrystals from lignocellulose derived materials. Bioresour. Technol. 100, 2259–2264. Ghose, T.K., 1987. Measurement of cellulase activities. Pure Appl. Chem. 59, 257– 268. Goering, H.K., Vansoest, P.J., 1970. Forage fibre analysis. Agriculture Handbook. Agricultural Research Services, United States Department of Agriculture. Gonzalez, G., Lopes-Santin, J., Caminal, G., Sola, C., 1986. Dilute acid hydrolysis of wheat straw hemicellulose at moderate temperature: a simplified kinetic model. Biotechnol. Bioeng. 28, 288–293. Guo, G.L., Chen, W.H., Chen, W.H., Men, L.C., Hwang, W.S., 2008. Characterization of dilute acid pretreatment of silvergrass for ethanol production. Bioresour. Technol. 99, 6046–6053. Hendriks, A.T.W.M., Zeeman, G., 2009. Pretreatments to enhance the digestibility of lignocellulosic biomass. Bioresour. Technol. 100, 10–18. Hernandez-Salas, J.M., Villa-Ramirez, M.S., Veloz-Rendon, J.S., Rivera-Hernandez, K.N., Gonzalez-Cesar, R.A., Plascencia-Espinosa, M.A., Trejo-Estrada, S.R., 2009. Comparative hydrolysis and fermentation of sugarcane and agave bagasse. Bioresour. Technol. 100, 1238–1245. Larsson, S., Palmqvist, E., Hahn-Hagerdal, B., Tengborg, C., Stenberg, K., Zacchi, G., Nilvebrant, N., 1999. The generation of fermentation inhibitors during dilute acid hydrolysis of softwood. Enzyme Microb. Technol. 24, 151–159. Lu, X., Zhang, Y., Angelidaki, I., 2009. Optimization of H2SO4-catalyzed hydrothermal pretreatment of rapeseed straw for bioconversion to ethanol: focusing on pretreatment at high solids content. Bioresour. Technol. 100, 3048–3053. Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Annl. Chem. 31, 426–428. Moiser, N., Wyman, C., Dale, B., Elander, R., Lee, Y.Y., Holtzapple, M., Ladisch, M., 2005. Features of promising technologies for pretreatment of lignocellulosic biomass. Bioresour. Technol. 96, 673–686. Pan, X., 2008. Role of functional groups in lignin inhibition of enzymatic hydrolysis of cellulose to glucose. J. Biobased Mater. Bioenergy 2 (1), 25–32. Rosgaard, L., Pedersen, S., Meyer, A.S., 2007. Comparison of different pretreatment strategies for enzymatic hydrolysis of wheat and barley straw. Appl. Biochem. Biotechnol. 143, 284–296. Segal, L., Creely, J.J., Martin, J.A.E., Conrad, C.M., 1959. An empirical method for estimating the degree of crystallinity of native cellulose using the X-ray diffractometer. Text. Res. J., 786–794. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., 2008. Determination of sugars, byproducts, and degradation products in liquid fraction Process Samples. Laboratory Analytical Procedure (LAP). Technical, Report NREL/TP-510-42623. Sun, J.X., Sun, R., Sun, X.F., Su, Y., 2004. Fractional and physico-chemical characterization of hemicelluloses from ultrasonic irradiated sugarcane bagasse. Carbohydr. Res. 339, 291–300. Updegroff, D.M., 1969. Semi micro determination of cellulose in biological materials. Anal. Biochem. 32, 420. Velmurugan, R., Muthukumar, K., 2011. Utilization of sugarcane bagasse for bioethanol production: sonoassisted acid hydrolysis approach. Bioresour. Technol.. doi:10.1016/j.biortech.2011.04.045. Wooley, R., Ruth, M., Sheehan, J., Ibsen, K., Majdeski, H., Galvez, A., 1999. Lignocellulosic Biomass to Ethanol – Process Design and Economics Utilizing Co-current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis – Current and Futuristic Scenarios. NREL, Golden. Wright, J.D., 1998. Economics of Enzymatic Hydrolysis Processes. Solar Energy Research Institute (SERI), Presented at the American Institute of Chemical Engineers Spring National Meeting, 6–10. New Orleans, LA. Yaldagard, M., Mortazavi, S.A., Tabatabaie, F., 2008. The effect of ultrasound in combination with thermal treatment on the germinated barley’s alpha-amylase activity. Korean J. Chem. Eng. 25 (3), 517–523. Zhang, Q., Cai, W., 2008. Enzymatic hydrolysis of alkali-pretreated rice straw by Trichoderma reesei ZM4-F3. Biomass Bioenergy 32, 1130–1135.