- Email: [email protected]
Pretreatment of switchgrass for sugar production with the combination of sodium hydroxide and lime
Bioresource Technology 102 (2011) 3861–3868
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Pretreatment of switchgrass for sugar production with the combination of sodium hydroxide and lime Jiele Xu, Jay J. Cheng ⇑ Department of Biological and Agricultural Engineering, Campus Box 7625, North Carolina State University, Raleigh, NC 27695-7625, USA
a r t i c l e
i n f o
Article history: Received 5 October 2010 Received in revised form 7 December 2010 Accepted 8 December 2010 Available online 15 December 2010 Keywords: Biofuels Biomass conversion Lignocellulose Lime Sodium hydroxide
a b s t r a c t Sodium hydroxide (NaOH) and lime (Ca(OH)2) were innovatively used together in this study to improve the cost-effectiveness of alkaline pretreatment of switchgrass at ambient temperature. Based on the sugar production in enzymatic hydrolysis, the best pretreatment conditions were determined as: residence time of 6 h, NaOH loading of 0.10 g/g raw biomass, NaOH addition at the beginning, Ca(OH)2 loading of 0.02 g/g raw biomass, and biomass wash intensity of 100 ml water/g raw biomass, at which the glucose and xylose yields were respectively 59.4% and 57.3% of the theoretical yields. The sugar yield of the biomass pretreated using the combination of 0.10 g NaOH/g raw biomass and 0.02 g Ca(OH)2/g raw biomass was found comparable with that of the biomass pretreated using 0.20 g NaOH/g raw biomass at the same conditions, while the chemical expense was remarkably reduced due to the low cost of lime and the reduced loading of NaOH. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction Switchgrass (Panicum virgatum L.) is considered as one of the best feedstocks for ethanol production because of its high annual biomass yield and high carbohydrate content (McLaughlin et al., 1999; Parrish and Fike, 2005). However, a pretreatment step is required before its enzymatic hydrolysis to disrupt recalcitrant lignocellulosic matrix. Alkaline pretreatment makes the lignocellulose swollen through solvation and saponification reactions, resulting in a more porous structure for enzyme access (Hendriks and Zeeman, 2009), and has attracted much attention due to its high efficiency and potential low cost (Belkacemi et al., 1998; Chang et al., 2001; Chen et al., 2007; Xu et al., 2010a,b). Sodium hydroxide (NaOH) and lime (Ca(OH)2) based alkaline pretreatments have been intensively investigated and proven efficient at high temperatures (MacDonald et al., 1983; Chang et al., 1997; Kaar and Holtzapple, 2000; Silverstein et al.,2007). However, explorations at mild temperatures were less encouraging from the perspective of economic promise. A previous study on NaOH pretreatment of switchgrass shows that, at ambient temperature, the chemical loading had to be doubled based on the amount required at 121 °C to achieve a comparable sugar production within a reasonable time frame of 24 h (Xu et al., 2010a). Since NaOH is costly, substantially increasing its usage to offset the impact of the reduced chemical reaction rates at mild temperatures may not be an economically justifiable solution. Lime is much cheaper than NaOH (Chang ⇑ Corresponding author. Tel.: +1 919 515 6733; fax: +1 919 515 7760. E-mail address: [email protected] (J.J. Cheng). 0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2010.12.038
et al., 1998, 2001; Kaar and Holtzapple, 2000). However, it is considered as a weak base due to its poor solubility in water (1.73 g/L at 20 °C). At ambient temperature, residence times of several days or even weeks are normally required for effective lime pretreatments. Kim and Holtzapple (2005) reported that, even if the residence time was extended to two weeks, the overall glucan conversion of corn stover pretreated using lime at 25 °C barely exceeded 50%. The study on lime pretreatment of switchgrass also demonstrated the poor performance of lime at ambient temperature (Xu et al., 2010b). However, based on the properties of NaOH and lime, it is possible to manipulate the usage of these two alkali reagents to make them work together to achieve a cost-effective pretreatment. First, lime is much cheaper than NaOH, thus being able to replace part of the NaOH alkalinity at a very low cost. Second, due to its poor solubility, a significant part of lime exists as solid and would gradually dissolve to supplement the alkalinity consumed by the biomass, thus stabilizing the pH at a high level throughout the pretreatment. Third, calcium ions, each carrying two positive charges, is expected to provide linkages within the biomass which are negatively charged at alkaline conditions due to the ionization of some functional groups including carboxyl, methoxy, and hydroxyl, thus preventing serious solid loss which is commonly observed in NaOH pretreatment (Torre et al., 1992; Xu et al., 2010b). Therefore, lime, although not strong enough by itself, can be used as a supplementary reagent to strong but expensive NaOH to improve the economic promise of alkaline pretreatment at ambient temperature. In this research, pretreatment of switchgrass using the combination of NaOH and lime was explored at ambient temperature,
3862
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
and the effects of residence time, NaOH loading, time point for NaOH addition, lime loading, and the volume of water for biomass washing on pretreatment efficiency were studied. Time point for NaOH addition was investigated because of its potential impact on sugar recovery. Since NaOH is a strong base and could cause serious solid loss, postponing its addition would probably improve biomass preservation by giving calcium ions more time to dissociate from lime and form linkages within biomass before harsh NaOH attack occurs. However, overly delaying NaOH addition will have negative impact on biomass digestibility improvement by excessively reducing the exposure time of biomass to the higher pH. Therefore, determining the best time point for NaOH addition is necessary. The optimal pretreatment conditions were determined based on the sugar production in enzymatic hydrolysis and the result was compared with that of NaOH pretreatment. Material balances were performed to investigate the compositional changes of biomass caused by pretreatments, and enzyme loadings in hydrolysis were optimized.
slurry at the beginning of experiment while NaOH of desired loading was added at a specific time point during the experiment. The pretreated biomass was washed using DI water of desired volume at ambient temperature and recovered by vacuum filtration using Fisherbrand P8 qualitative filter paper. About 1 g (dry basis) of the biomass was dried at 45 °C in a Fisher Scientific Isotemp Oven to constant weight for composition analysis, and the rest was stored in a sealed plastic bag at 4 °C for enzymatic hydrolysis. The pH of pretreatment liquor was determined using a Denver Instrument Model 250 pH meter. 2.3. Enzymatic hydrolysis
The biomass feedstock was ‘Performer’ switchgrass, a lowland cultivar recently released by Burns et al. (2008) as an animal feed of improved quality. Switchgrass plants, harvested in late July, 2007, were obtained from the Central Crops Research Station near Clayton, North Carolina. A harvest strip was taken randomly from each quarter of the field, and four strips were combined to form one bulk sample. The biomass sample was oven dried at 50 °C for 72 h, ground to pass a 2 mm sieve in a Thomas Wiley Laboratory Mill (Model No. 4), and stored in sealed plastic bags at ambient temperature. The chemical composition of the raw biomass was as follows: glucan 32.0%, xylan 17.9%, galactan 1.73%, arabinan 1.87%, lignin 21.4%, ash 3.77%, and others 21.4%.
One gram of pretreated biomass (dry basis) was immersed in 30 ml of 50 mM sodium citrate buffer (pH 4.8) in a 250 ml Erlenmeyer flask. Cellulase from Trichoderma reesei (E.C. 3.2.1.4) was added at an enzyme loading of 35 FPU (filter paper unit)/g dry biomass. To prevent cellobiose inhibition to the enzymatic hydrolysis of the lignocellulose, cellobiase from Aspergillus niger (E.C. 3.2.1.21) at an enzyme loading of 61.5 CBU (cellobiase unit)/g dry biomass was supplemented in the flask. FPU is defined as the amount of enzyme that produces 1 lmol of glucose from filter paper in a minute, and CBU is defined as the amount of enzyme that produces 2 lmol of glucose from cellobiose in one minute. Enzymes were obtained from Novozymes North America, Inc. (Franklinton, North Carolina, USA), and the activities of cellulase and cellobiase determined using the methods adapted from Ghose (1987) were respectively 80 FPU/ml and 277 CBU/ml. To prevent the mixed liquor in the flask from microbial contamination, sodium azide (0.3%, w/v) was added into the mixture. The enzymatic hydrolysis was carried out in an air bath shaker at 55 °C and 150 rpm for 72 h. After the hydrolysis time elapsed, the flasks were immediately chilled in an ice bath to stop the reactions. The hydrolysate was then collected by centrifugation at 8 103g for 5 min, and the supernatant was stored at 80 °C for sugar analysis at a later time.
2.2. Pretreatment
2.4. Analytical methods
The pretreatment study was divided into two parts. In Study 1, three major factors which were residence time, NaOH loading, and time point for NaOH addition were investigated (Table 1). In Study 2, the other two factors which were lime loading and wash intensity were investigated. Lime loading and wash intensity were studied as a separate group because they don’t have much interaction with the three major factors. Since the solubility of lime is poor, to provide as much cheap alkalinity as possible, an excessive lime loading is required anyway to maintain a saturated lime solution throughout the pretreatment. Wash intensity and lime loading were related because lime loading might affect the amount of water required to wash away the excess lime after the pretreatment. In the 3 3 factorial study on the three major factors, lime loading and wash intensity were held constant respectively at 0.10 g Ca(OH)2/g raw biomass and 200 ml water/g raw biomass, while the subsequent studies on lime loading and wash intensity were based on the best combination of the three major factors. Four grams of biomass sample was mixed with deionized (DI) water in a serum bottle, forming a slurry at a solid concentration of 0.1 g/ml. Lime was added into the bottle and mixed with the
Total solids, structural carbohydrates, lignin, and ash in raw and pretreated biomass were measured according to Laboratory Analytical Procedures (LAP) established by National Renewable Energy Laboratory (NREL) (Sluiter et al., 2005a,b, 2008). Total reducing sugars in hydrolysate were measured using DNS method adapted from Miller (1959) and Ghose (1987). Monomeric sugars (glucose, xylose, galactose, arabinose, and mannose) derived from cellulose and hemicellulose in hydrolysate and liquor from acid hydrolysis of biomass for composition analysis were measured with a Shimadzu high performance liquid chromatography (HPLC). The HPLC system consisted of a Bio-Rad Aminex HPX-87P column (300 7.8 mm) tailored for analysis of lignocellulose-derived sugars, a Bio-Rad Micro-Guard column (30 4.6 mm), a thermostatted autosampler, a quaternary pump, and a refractive index detector. The analytical column was operated at 80 °C with HPLC grade water as the mobile phase at a flow rate of 0.6 ml/min. The samples were injected at 10 ll and the acquisition time was 35 min. A post-run time of 25 min was included between injections to allow for late-eluting compounds to come off the column. The standards used were glucose, xylose, galactose, arabinose, and
2. Methods 2.1. Biomass preparation
Table 1 Factors studied in the pretreatment of swithgrass using the combination of NaOH and lime at ambient temperature. Study 1
*
Study 2
Residence time (h)
Time point for NaOH addition
NaOH loading (g/g raw biomass)
Lime loading (g/g raw biomass)
Wash intensity (ml water/g raw biomass)
3, 6, 9
0, 1/3, 2/3*
0.05, 0.10, 0.20
0, 0.02, 0.04, 0.06, 0.08, 0.10
100, 200
0, 1/3, 2/3 represent adding NaOH at the beginning of pretreatment, after 1/3 of the residence time elapses, or after 2/3 of the residence time elapses.
3863
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
mannose at concentrations of 0.5, 2.0, 5.0, 7.5, 10.0 g/L. Mass balance was conducted based on the composition analysis of raw and pretreated biomass. After pretreatment and enzymatic hydrolysis, the percentage sugar yield was calculated as follows:
% Sugar yield ¼
SH 100 PR CF
where SH is the sugar released from per gram of raw biomass in hydrolysis, PR is the content of sugar polymer in per gram of raw biomass, and CF is the correction factor (1.11 and 1.14 for glucan and xylan respectively). 2.5. Statistical analysis All treatments in this study were conducted in triplicate and a 95% confidence level was applied for data analysis. The GLM procedure in SAS 9.1 software (SAS Institute Inc., Cary, NC) was used for all statistical analysis. Analysis of variance (ANOVA) was used to determine the effects of various factors on pretreatment and Tukey simultaneous tests were conducted to determine the statistical differences between treatments. 3. Results and discussion 3.1. Solid loss Solid loss associated with alkaline attack, which substantially compromises sugar recovery in the subsequent enzymatic hydrolysis, is of great concern especially when a strong base or severe pretreatment conditions are applied. At ambient temperature, however, only 13.9–28.4% of the total solids were lost during the pretreatment (Fig. 1), which was considerably lower than those of alkaline pretreatments at elevated temperatures (Silverstein et al., 2007; Xu et al., 2010a). NaOH loading had significant (P < 0.05) effect on solid loss at all the combinations of residence time and time point for NaOH addition except 3 h and 2/3 point NaOH addition, at which the biomass might be exposed to NaOH attack for too short a time, while residence time had significant (P < 0.05) effect on solid loss at all the combinations of NaOH loading and time point for NaOH addition. Postponing NaOH addition was crucial for the reduction of solid loss especially at higher NaOH loadings. Adding NaOH after 1/3 of the residence time elapsed substantially reduced solid loss. However, further postponing NaOH addition till 2/3 of the residence time elapsed did not substantially reduce the solid loss, which probably indicated that the majority of calcium linkages formed at the initial stage of pretreatment.
3.2. Sugar production in Study 1 After pretreatment and enzymatic hydrolysis, sugars in hydrolysate were analyzed. The yields of glucose and xylose were reported to understand the impact of pretreatments on cellulose and hemicellulose conversions. Galactose and arabinose, the two minor hemicellulose constituents, were measured but not reported in this paper due to their very low concentrations in the hydrolysate. The overall pretreatment effectiveness was evaluated using the total reducing sugar yield based on raw biomass. Glucose yields were significantly (P < 0.05) increased with the increase of residence time and NaOH loading (Fig. 2a). Postponing NaOH addition did not improve the glucose yield. This is probably because of the semicrystalline structure of cellulose. The alteration from crystalline to amorphous cellulose for improved digestibility needs time under a relatively high pH. The maximum glucose yield of 263.0 mg/g raw biomass was obtained at the most severe pretreatment conditions of 9 h, 0.20 g NaOH/g raw biomass, and NaOH addition at the very beginning (0 point), at which the overall glucan conversion reached 74.0%, 3.46 times that of untreated biomass. Hemicellulose has amorphous, heterogeneous and branched structure with little strength, which makes it more susceptible to solubilization than semicrystalline cellulose at alkaline conditions. At 3 h, increasing NaOH loading significantly (P < 0.05) improved xylose yield while postponing NaOH addition did not result in better yields (Fig. 2b). The correlations between xylose yield and pretreatment conditions were more complicated at longer residence times. Because of the enhanced solubilization of hemicellulose during pretreatment, increasing NaOH loading did not necessarily lead to higher xylose yields at 6 and 9 h, especially when NaOH was added at the earlier stages (0 and 1/3 point) of the experiment (Fig. 2b). At the NaOH loading of 0.20 g/g raw biomass, unlike glucose yield, there was no significant (P < 0.05) difference between xylose yields at 0 and 1/3 point NaOH addition, which indicated that postponing NaOH addition was more favorable for hemicellulose preservation. The maximum xylose yield of 135.3 mg/g raw biomass was obtained at 3 h, 0.20 g NaOH/g raw biomass, and 0 point NaOH addition, at which the xylan conversion reached 66.5%, 5.15 times that of untreated biomass. The effects of pretreatment conditions on total sugar yield are clearly shown in Fig. 2c. NaOH loading had significant (P < 0.05) effect on total reducing sugar yield at all the combinations of residence time and time point for NaOH addition, while extending residence time from 6 to 9 h did not improve the sugar production at 0 and 1/3 point NaOH addition due to more serious carbohydrate
Solid loss following wash (%)
30
Time for NaOH dosing 0 1/3 2/3
25
20
15
10
5
0 3h/0.05g 3h/0.10g 3h/0.20g 6h/0.05g 6h/0.10g 6h/0.20g 9h/0.05g 9h/0.10g 9h/0.20g 3 hr 3 hr 3 hr 6 hr 6 hr 6 hr 9 hr 9 hr 9 hr 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g
Treatment Fig. 1. Solid losses of swithgrass biomass pretreated at different combinations of residence time, NaOH loading, and time point for NaOH addition.
3864
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
Glucose yield (g/g raw biomass)
300
Time for NaOH dosing
a
0 1/3 2/3
250 200 150 100 50
Xylose yield (g/g raw biomass)
0 140
b3h/0.05g 3h/0.10g 3h/0.20g 6h/0.05g 6h/0.10g 6h/0.20g 12h/0.05g12h/0.10g12h/0.20g Treatment
120 100 80 60 40 20
Reducing sugar yield (g eq. glucose/g raw biomass)
0
c3h/0.05g 3h/0.10g 3h/0.20g 6h/0.05g 6h/0.10g 6h/0.20g 12h/0.05g12h/0.10g12h/0.20g 400
Treatment
300
200
100
0 3h/0.05g 3h/0.10g 3h/0.20g 6h/0.05g 6h/0.10g 6h/0.20g 3 hr 3 hr 3 hr 6 hr 6 hr 6 hr 12h/0.05g12h/0.10g12h/0.20g 9 hr 9 hr 9 hr 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g
Treatment Fig. 2. Yields of (a) glucose, (b) xylose, and (c) total reducing sugars in enzymatic hydrolysis of switchgrass biomass pretreated at different combinations of residence time, NaOH loading, and time point for NaOH addition.
loss at the extended exposure time to NaOH attack. Similar to glucose and xylose yields, although postponing NaOH addition could result in better carbohydrate preservation, reducing the exposure time to the high pH substantially compromise the biomass digestibility improvement, resulting in lower sugar productions. The maximum total reducing sugar yield of 435.8 mg/g raw biomass was obtained at 9 h, 0.20 g NaOH/g raw biomass, and 0 point NaOH addition, at which the conversion of total available carbohydrate reached 68.1%, 3.63 times that of untreated biomass. However, cutting the residence time down to 6 h and applying half the NaOH loading only resulted in a 5% reduction in sugar yield. Therefore, based on cost-benefit considerations, 6 h, 0.10 g NaOH/g raw biomass, and 0 point NaOH addition were applied in the following lime loading and biomass washing studies. 3.3. Sugar production in Study 2 Using the best combination of residence time, NaOH loading, and time point for NaOH addition, lime loadings of 0, 0.02, 0.04, 0.06, 0.08 and 0.10 g Ca(OH)2/g raw biomass and wash intensities of 100 ml and 200 ml water/g raw biomass were further studied.
The wash intensity of 200 ml water/g raw biomass was sufficient to remove residual lime and the sugar production was maximized at the lime loading of 0.04 g Ca(OH)2/g raw biomass (Fig. 3). Reducing wash intensity to 100 ml water/g raw biomass resulted in significant (P < 0.05) decreases of sugar production at lime loadings of 0.04–0.10 g Ca(OH)2/g raw biomass. This is because that the reduced wash intensity was not sufficient to remove all the residual lime mixed with biomass. Although sodium citrate buffer (pH 4.8) was used according to the standard procedure to maintain the hydrolysis pH at the optimum level, the dissolution of residual lime particles still substantially raised the pH (pH exceeded 6.5 at 0.10 g Ca(OH)2/g raw biomass). However, there was no significant (P < 0.05) decrease in sugar production at the lowest lime loading of 0.02 g Ca(OH)2/g raw biomass, at which the total reducing sugar yield, 30% higher than that of the biomass treated without lime supplementation, whereas the total chemical expense was barely increased considering the low cost of lime and the small amount required. Although using the combination of 0.04 g Ca(OH)2/g raw biomass and 200 ml water/g raw biomass resulted in a 6% higher total reducing sugar yield than using the combination of 0.02 g Ca(OH)2/g raw biomass and 100 ml water/g raw bio-
3865
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
Total reducing sugar yield (g eq. glucose/g raw biomass)
500
Wash volume 100 ml 200 ml
400
300
200
100
0 0.00
0.02
0.04
0.06
0.08
0.10
Lime loading (g Ca(OH)2/g raw biomass) Fig. 3. Total reducing sugar yields in enzymatic hydrolysis of switchgrass biomass pretreated at different combinations of lime loading and wash intensity.
mass, doubling both lime loading and wash intensity could potentially compromise the cost-effectiveness of the process. Preliminary economic analysis, therefore, is necessary before any scale-up studies. Based on total sugar production and processing convenience, 6 h, 0.10 g NaOH/g raw biomass, 0 point NaOH addition, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass were determined as the best conditions for the ambient temperature pretreatment of switchgrass in this study. 3.4. Lignin reduction Lignin is a three dimensional complex aromatic polymer which forms a sheath surrounding cellulose and hemicellulose, stiffening and holding together the fibers of polysaccharides (Fan et al., 1987). A number of studies show that NaOH can effectively remove lignin barriers, thus exposing the carbohydrates to enzymes (MacDonald et al., 1983; Chen et al., 2007; Silverstein et al., 2007). However, the delignification capability of lime is quite limited. Wang (2009) reported that, at 121 °C, the total lignin reduction of lime pretreatment of coastal Bermuda grass was only 10–20% which was much lower than that of NaOH pretreatment. This is probably because of the interaction between lignin and calcium ions within the biomass structure. Many studies show that divalent calcium
ions had high affinity for lignin and could effectively crosslink lignin molecules especially at alkaline conditions (Torre et al., 1992; Sundin and Harlter, 2000a,b; Duong et al., 2005). In this study, all of the three parameters (residence time, NaOH loading, and time point for NaOH addition) had significant (P < 0.05) effect on lignin reduction (Fig. 4). Increasing residence time, NaOH loading, and the exposure time of biomass to NaOH attack favored the removal of lignin barrier. It was also found that improving biomass preservation by postponing NaOH addition till 1/3 of the residence time elapsed caused a remarkable reduction in lignin removal, which is in agreement with the previous reports that calcium ions provide linkages between lignin molecules, thus providing better resistance against solubilization during alkaline pretreatment (Wang, 2009; Xu et al., 2010b). The lignin reduction ranged from 23.8% at 3 h, 0.05 g NaOH/g raw biomass, 2/3 point NaOH addition, to 45.1% at 9 h, 0.20 g NaOH/g raw biomass, 0 point NaOH addition, and was positively correlated to NaOH loading and residence time. The high lignin content of pretreated biomass did not affect enzymatic hydrolysis. This is in correspondence with our previous research (Xu et al., 2010b) which shows that as long as the chemical bonds stiffening the lignocellulose were disrupted and the biomass porosities substantially increased, enzymatic hydrolysis of high efficiency can still be achieved even in the presence of relatively high lignin contents.
50
Time for NaOH dosing 0 1/3 2/3
Lignin reduction (%)
40
30
20
10
0 3h/0.05g 3h/0.20g 6h/0.05g 6h/0.10g 6h/0.20g 9h/0.05g 9h/0.10g 9h/0.20g 3 hr 3h/0.10g 3 hr 3 hr 6 hr 6 hr 6 hr 9 hr 9 hr 9 hr 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g
Treatment Fig. 4. Lignin reductions of switchgrass biomass pretreated at different combinations of residence time, NaOH loading, and time point for NaOH addition.
3866
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
3.5. Comparison with NaOH pretreatment Since lime was too weak to effectively improve biomass digestibility within a reasonable time frame at ambient temperature (Kim and Holtzapple, 2005; Xu et al., 2010b), the pretreatment using the combination of NaOH and lime at the best conditions (0.10 g NaOH/g raw biomass, 0.02 g Ca(OH)2/g raw biomass, 6 h) was further evaluated based on the comparison with NaOH pretreatment. Two NaOH loadings of 0.10 and 0.20 g NaOH/g raw biomass were explored with all the other pretreatment conditions kept the same. The carbohydrate conversions, lignin reductions, pH changes of the three pretreatment strategies are shown in Table 2. After pretreatment and enzymatic hydrolysis, the overall glucan conversion of the biomass treated with lime supplementation was 23.5% higher than that of the biomass treated just using 0.10 g NaOH/g raw biomass, but 14.5% lower than that of the biomass treated using 0.20 g NaOH/g raw biomass. However, the xylan conversion of the biomass treated with lime supplementation was not only 47.3% higher than that of the biomass treated using 0.10 g NaOH/g raw biomass, but also 4.6% higher than that of the biomass treated using 0.20 g NaOH/g raw biomass, indicating better hemicellulose preservations by calcium bonding. As a result, the total carbohydrate conversion of the biomass treated with lime supplementation was 34.5% higher than that of the biomass treated using 0.10 g NaOH/g raw biomass, and comparable with that of the biomass treated using 0.20 g NaOH/g raw biomass. Since the cost of supplementing 0.02 g Ca(OH)2/g raw biomass was remarkably lower than that of doubling the loading of expensive NaOH, using the combination of NaOH and lime shows great economic promise. Assuming that the market price of NaOH is 3.5 times as much as that of lime (Lerner, 2007; USGS, 2010), to produce the same amount of fermentable sugars from biomass feedstock, supple-
menting 0.02 g Ca(OH)2 to 0.1 g NaOH/g raw biomass instead of applying 0.20 g NaOH/g raw biomass in the pretreatment lead to a 43.4% reduction in chemical cost. The pH changes indicated that lime did help to avoid substantial pH drop during pretreatment. After 6 h, the decline of pH at 0.10 g NaOH/g raw biomass without lime supplementation was significantly (P < 0.05) greater than that with lime supplementation. To make the differences more outstanding, the pretreatment time was extended to 24 h. The final pH of the pretreatment with lime supplementation was not significantly (P < 0.05) different from that of the pretreatment using 0.20 g NaOH/g raw biomass, while considerably higher than that of the pretreatment using 0.10 g NaOH/g raw biomass. 3.6. Mass balances Mass balances were performed on the biomass pretreated at the best conditions (6 h, 0.10 g NaOH/g raw biomass, and 0 point NaOH addition, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass). The biomass pretreated using 0.10 and 0.20 g NaOH/g raw biomass was also analyzed for comparison. The total dry weight of the sample was measured after pretreatment, and the compositions of the pretreated biomass (glucan, xylan, galactan, arabinan, lignin, ash, and other) were determined and compared with that of raw biomass. The solid loss significantly (P < 0.05) increased with the increase of NaOH loading (Fig. 5). However, lime supplementation, although contributed to a greater alkalinity than that of just using 0.10 g NaOH/g raw biomass, did not cause higher solid loss due to the better biomass preservation in the presence of calcium linkages. Glucan (cellulose) was the best preserved carbohydrate at all pretreatment conditions because of its semicrystilline structure. The impacts of pretreatment on xylan and lignin were similar to that on total solids. Better xylan and lig-
Table 2 Comparisons of carbohydrate conversions, lignin reductions, and pH changes between NaOH pretreatment and the pretreatment using the combination of NaOH and lime.
*
Chemical loading (g raw biomass1)
Glucan conversion (%)
Xylan conversion (%)
Total carbohydrate conversion (%)
Lignin reduction (%)
Initial pH
0.20 g NaOH 0.10 g NaOH 0.10 g NaOH + 0.02 g Ca(OH)2
69.5 (1.09)* 48.1 (3.12) 59.4 (0.79)
54.8 (0.87) 38.9 (2.90) 57.3 (0.79)
63.5 (1.01) 44.1 (1.73) 59.3 (1.04)
48.6 (1.03) 34.6 (1.51) 37.8 (0.55)
12.92 (0.02) 12.84 (0.02) 12.86 (0.01)
Final pH 6h
24 h
12.81 (0.02) 12.59 (0.02) 12.77 (0.03)
12.66 (0.04) 12.32 (0.03) 12.65 (0.04)
The number in parentheses is standard deviation of triplicate samples.
Content (g)
100 90
Raw biomass 0.10 g NaOH+0.02 g Ca(OH)2
80
0.10 g NaOH 0.20 g NaOH
70
30 20 10 0 Total
Glucan
Xylan
Galactan Arabinan Lignin
Ash
Others
Components Fig. 5. Material balances for raw switchgrass, the biomass pretreated using the combination of NaOH and lime at the best conditions (6 h, 0.10 g NaOH/g raw biomass, 0 point NaOH addition, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass), and the biomass pretreated using 0.10 and 0.20 g NaOH/g raw biomass at the same conditions.
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
nin preservations were achieved with lime supplementation. The biomass pretreated using the combination of 0.10 g NaOH and 0.02 g Ca(OH)2/g raw biomass was comparable with that pretreated using 0.20 g NaOH/g raw biomass in terms of sugar production, however, its composition was more like that of the biomass pretreated using 0.10 g NaOH/g raw biomass. 3.7. Study on enzyme loading To determine the best enzyme loadings, cellulase loadings of 0– 35 FPU/g dry biomass and cellobiase loadings of 0–50 CBU/g dry biomass were investigated. The glucose, xylose, and total reducing sugar yields of switchgrass pretreated at the best conditions (6 h, 0.10 g NaOH/g raw biomass, 0 point NaOH addition, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass) were determined at different cellulase loadings (Fig. 6a). The cellobiase loading was kept constant at an excess level of 61.5 CBU/g biomass in the cellulase loading test to eliminate the impact of cellobiase limitation on sugar production. With the increase of cellulase loading from 0 to 20 FPU/g dry biomass, the total reducing sugar yield increased by 6.52 times, while further increasing cellulase loading did not improve the sugar yield. Similar trends were observed for both glucose and xylose yields. Cellobiase loadings were studied based on the best cellulase loading of 20 FPU/g dry biomass. The results show that supplementing cellobiase significantly (P < 0.05) increased sugar yields and a cellobiase loading of 10 CBU/g dry biomass was sufficient (Fig. 6b). The optimal enzyme loadings were comparable with those obtained in the study on switchgrass pretreatment using NaOH (Xu et al., 2010a).
3867
4. Conclusions Lime, which is very inexpensive but too weak to work effectively at ambient temperature, helped to reduce the requirement for strong but expensive NaOH in the alkaline pretreatment of switchgrass by maintaining a high pH level and mitigating carbohydrate solubilization during pretreatment, thus contributing to a novel pretreatment technology of high cost-effectiveness. The best pretreatment conditions were determined as: 6 h, 0.10 g NaOH/g raw biomass, NaOH addition at the beginning, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass, and the enzyme loadings of 20 FPU cellulase/g raw biomass and 10 CBU cellobiase/g raw biomass were recommended for hydrolysis. A pretreatment at the best conditions resulted in a total reducing sugar yield of 59.3%, 34.5% higher than that of NaOH pretreatment at the same conditions without lime supplementation, while the chemical cost was barely increased considering the low cost of lime and the minor loading required.
Acknowledgements The authors would like to acknowledge the financial support of this research from North Carolina Agricultural Research Service (NCARS) and North Carolina Agricultural Foundation (NCAF). We would also like to thank Dr. Joseph Burns from Crop Science Department at North Carolina State University for providing the switchgrass material and Novozymes North America, Inc. for donating the enzymes for this study.
Sugar yields (mg/g raw biomass)
References
a
400
300
200
100 Cellobiase loading: 61.5 CBU/g dry biomass 0 0
10
20
30
40
Sugar yields (mg/g raw biomass)
Cellulase loading (FPU/g dry biomass)
b
400
300
200
100 Cellulase loading: 20 FPU/g dry biomass 0 0
10
20
30
40
50
60
Cellobiase loading (CBU/g dry biomass) Fig. 6. Effects of cellulase and cellobiase loadings on (d) glucose, (s) xylose, and (.) total reducing sugar yields in enzymatic hydrolysis of switchgrass pretreated at the best conditions (6 h, 0.10 g NaOH/g raw biomass, 0 point NaOH addition, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass).
Belkacemi, K., Turcotte, G., de Halleux, D., Savoie, P., 1998. Ethanol production from AFEX-treated forages and agricultural residues. Appl. Biochem. Biotechnol. 70– 72, 441–462. Burns, J.C., Godshalk, E.B., Timothy, D.H., 2008. Registration of ‘Performer’ switchgrass. J. Plant Reg. 2 (1), 29–30. Chang, V.S., Burr, B., Holtzapple, M.T., 1997. Lime pretreatment of switchgrass. Appl. Biochem. Biotechnol. 63, 3–19. Chang, V.S., Nagwani, M., Holtzapple, M.T., 1998. Lime pretreatment of crop residues bagasse and wheat straw. Appl. Biochem. Biotechnol. 74, 135–159. Chang, V.S., Nagwani, M., Kim, C.-H., Holtzapple, M.T., 2001. Oxidative lime pretreatment of high-lignin biomass: poplar wood and newspaper. Appl. Biochem. Biotechnol. 94, 1–28. Chen, Y., Sharma-Shivappa, R.R., Keshwani, D., Chen, C., 2007. Potential of agricultural residues and hay for bioethanol production. Appl. Biochem. Biotechnol. 142, 276–290. Duong, T.D., Hoang, M., Nguyen, K.L., 2005. Sorption of Na+, Ca2+ ions from aqueous solution onto unbleached kraft fibers-kinetics and equilibrium studies. J. Colloid Interf. Sci. 287, 438–443. Fan, L.T., Gharpuray, M.M., Lee, Y.-H., 1987. Cellulose hydrolysis. Biotechnology Monographs, Springer, Berlin. pp. 8. Ghose, T.K., 1987. Measurement of cellulase activities. Pure Appl. Chem. 59 (2), 257–268. Hendriks, A.T.W.M., Zeeman, G., 2009. Pretreatments to enhance the digestibility of lignocellulosic biomass. Biotechnol. Prog. 100, 10–18. Kaar, W.E., Holtzapple, M.T., 2000. Using lime pretreatment to facilitate the enzymic hydrolysis of corn stover. Biomass Bioenergy 18, 189–199. Kim, S., Holtzapple, M.T., 2005. Lime pretreatment and enzymatic hydrolysis of corn stover. Bioresour. Technol. 96, 1994–2006. Lerner, I., 2007. Caustic soda prices bubbling. ICIS Chem. Business Am. (1936-458X) 271 (12), 24. MacDonald, D.G., Bakhshi, N.N., Mathews, J.F., Roychowdhury, A., 1983. Alkali treatment of corn stover to improve sugar production by enzymatic hydrolysis. Biotechnol. Bioeng. 25, 2067–2076. McLaughlin, S.B., Bouton, J., Bransby, D., Conger, B.V., Ocumpaugh, W.R., Parrish, D.J., Taliaferro, C., Vogel, K.P., Wullschleger, S.D., 1999. Developing switchgrass as a bioenergy crop. In: Janick, J. (Ed.), Perspectives on New Crops and New Uses. ASHS Press, Alexandria, VA, pp. 282–299. Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal. Chem. 31, 426–428. Parrish, D.J., Fike, J.H., 2005. The biology and agronomy of switchgrass for biofuels. Crit. Rev. Plant Sci. 24 (5–6), 423–459. Silverstein, R.A., Chen, Y., Sharma-Shivappa, R.R., Boyette, M.D., Osborne, J., 2007. A comparison of chemical pretreatment methods for improving saccharification of cotton stalks. Bioresour. Technol. 98, 3000–3011.
3868
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
Sluiter, A., Hames, B., Hyman, D., Payne, C., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Wolfe, J., 2005a. Determination of total solids in biomass and total dissolved solids in liquid process samples. Laboratory Analytical Procedure (LAP). NREL/TP-510–42621. National Renewable Energy Laboratory, Golden, Colorado. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., 2005b. Determination of ash in biomass. Laboratory Analytical Procedure (LAP). NREL/TP-510–42622. National Renewable Energy Laboratory, Golden, Colorado. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D., 2008. Determination of structural carbohydrates and lignin in biomass. Laboratory Analytical Procedure (LAP). NREL/TP-510–42618. National Renewable Energy Laboratory, Golden, Colorado. Sundin, J., Hartler, N., 2000a. Precipitation of kraft lignin by metal cations in alkaline solutions. Nord. Pulp Pap. Res. J. 15 (4), 306–312.
Sundin, J., Hartler, N., 2000b. Precipitation of kraft lignin by metal during pulp washing. Nord. Pulp Pap. Res. J. 15 (4), 313–318. Torre, M., Rodriguez, A.R., Saura-Calixto, F., 1992. Study of the interactions of calcium ions with lignin, cellulose, and pectin. J. Agric. Food. Chem. 40, 1762–1766. USGS, 2010. USGS Minerals Information: Lime. Retrieved Nov 27, 2010 from
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Pretreatment of switchgrass for sugar production with the combination of sodium hydroxide and lime Jiele Xu, Jay J. Cheng ⇑ Department of Biological and Agricultural Engineering, Campus Box 7625, North Carolina State University, Raleigh, NC 27695-7625, USA
a r t i c l e
i n f o
Article history: Received 5 October 2010 Received in revised form 7 December 2010 Accepted 8 December 2010 Available online 15 December 2010 Keywords: Biofuels Biomass conversion Lignocellulose Lime Sodium hydroxide
a b s t r a c t Sodium hydroxide (NaOH) and lime (Ca(OH)2) were innovatively used together in this study to improve the cost-effectiveness of alkaline pretreatment of switchgrass at ambient temperature. Based on the sugar production in enzymatic hydrolysis, the best pretreatment conditions were determined as: residence time of 6 h, NaOH loading of 0.10 g/g raw biomass, NaOH addition at the beginning, Ca(OH)2 loading of 0.02 g/g raw biomass, and biomass wash intensity of 100 ml water/g raw biomass, at which the glucose and xylose yields were respectively 59.4% and 57.3% of the theoretical yields. The sugar yield of the biomass pretreated using the combination of 0.10 g NaOH/g raw biomass and 0.02 g Ca(OH)2/g raw biomass was found comparable with that of the biomass pretreated using 0.20 g NaOH/g raw biomass at the same conditions, while the chemical expense was remarkably reduced due to the low cost of lime and the reduced loading of NaOH. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction Switchgrass (Panicum virgatum L.) is considered as one of the best feedstocks for ethanol production because of its high annual biomass yield and high carbohydrate content (McLaughlin et al., 1999; Parrish and Fike, 2005). However, a pretreatment step is required before its enzymatic hydrolysis to disrupt recalcitrant lignocellulosic matrix. Alkaline pretreatment makes the lignocellulose swollen through solvation and saponification reactions, resulting in a more porous structure for enzyme access (Hendriks and Zeeman, 2009), and has attracted much attention due to its high efficiency and potential low cost (Belkacemi et al., 1998; Chang et al., 2001; Chen et al., 2007; Xu et al., 2010a,b). Sodium hydroxide (NaOH) and lime (Ca(OH)2) based alkaline pretreatments have been intensively investigated and proven efficient at high temperatures (MacDonald et al., 1983; Chang et al., 1997; Kaar and Holtzapple, 2000; Silverstein et al.,2007). However, explorations at mild temperatures were less encouraging from the perspective of economic promise. A previous study on NaOH pretreatment of switchgrass shows that, at ambient temperature, the chemical loading had to be doubled based on the amount required at 121 °C to achieve a comparable sugar production within a reasonable time frame of 24 h (Xu et al., 2010a). Since NaOH is costly, substantially increasing its usage to offset the impact of the reduced chemical reaction rates at mild temperatures may not be an economically justifiable solution. Lime is much cheaper than NaOH (Chang ⇑ Corresponding author. Tel.: +1 919 515 6733; fax: +1 919 515 7760. E-mail address: [email protected] (J.J. Cheng). 0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2010.12.038
et al., 1998, 2001; Kaar and Holtzapple, 2000). However, it is considered as a weak base due to its poor solubility in water (1.73 g/L at 20 °C). At ambient temperature, residence times of several days or even weeks are normally required for effective lime pretreatments. Kim and Holtzapple (2005) reported that, even if the residence time was extended to two weeks, the overall glucan conversion of corn stover pretreated using lime at 25 °C barely exceeded 50%. The study on lime pretreatment of switchgrass also demonstrated the poor performance of lime at ambient temperature (Xu et al., 2010b). However, based on the properties of NaOH and lime, it is possible to manipulate the usage of these two alkali reagents to make them work together to achieve a cost-effective pretreatment. First, lime is much cheaper than NaOH, thus being able to replace part of the NaOH alkalinity at a very low cost. Second, due to its poor solubility, a significant part of lime exists as solid and would gradually dissolve to supplement the alkalinity consumed by the biomass, thus stabilizing the pH at a high level throughout the pretreatment. Third, calcium ions, each carrying two positive charges, is expected to provide linkages within the biomass which are negatively charged at alkaline conditions due to the ionization of some functional groups including carboxyl, methoxy, and hydroxyl, thus preventing serious solid loss which is commonly observed in NaOH pretreatment (Torre et al., 1992; Xu et al., 2010b). Therefore, lime, although not strong enough by itself, can be used as a supplementary reagent to strong but expensive NaOH to improve the economic promise of alkaline pretreatment at ambient temperature. In this research, pretreatment of switchgrass using the combination of NaOH and lime was explored at ambient temperature,
3862
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
and the effects of residence time, NaOH loading, time point for NaOH addition, lime loading, and the volume of water for biomass washing on pretreatment efficiency were studied. Time point for NaOH addition was investigated because of its potential impact on sugar recovery. Since NaOH is a strong base and could cause serious solid loss, postponing its addition would probably improve biomass preservation by giving calcium ions more time to dissociate from lime and form linkages within biomass before harsh NaOH attack occurs. However, overly delaying NaOH addition will have negative impact on biomass digestibility improvement by excessively reducing the exposure time of biomass to the higher pH. Therefore, determining the best time point for NaOH addition is necessary. The optimal pretreatment conditions were determined based on the sugar production in enzymatic hydrolysis and the result was compared with that of NaOH pretreatment. Material balances were performed to investigate the compositional changes of biomass caused by pretreatments, and enzyme loadings in hydrolysis were optimized.
slurry at the beginning of experiment while NaOH of desired loading was added at a specific time point during the experiment. The pretreated biomass was washed using DI water of desired volume at ambient temperature and recovered by vacuum filtration using Fisherbrand P8 qualitative filter paper. About 1 g (dry basis) of the biomass was dried at 45 °C in a Fisher Scientific Isotemp Oven to constant weight for composition analysis, and the rest was stored in a sealed plastic bag at 4 °C for enzymatic hydrolysis. The pH of pretreatment liquor was determined using a Denver Instrument Model 250 pH meter. 2.3. Enzymatic hydrolysis
The biomass feedstock was ‘Performer’ switchgrass, a lowland cultivar recently released by Burns et al. (2008) as an animal feed of improved quality. Switchgrass plants, harvested in late July, 2007, were obtained from the Central Crops Research Station near Clayton, North Carolina. A harvest strip was taken randomly from each quarter of the field, and four strips were combined to form one bulk sample. The biomass sample was oven dried at 50 °C for 72 h, ground to pass a 2 mm sieve in a Thomas Wiley Laboratory Mill (Model No. 4), and stored in sealed plastic bags at ambient temperature. The chemical composition of the raw biomass was as follows: glucan 32.0%, xylan 17.9%, galactan 1.73%, arabinan 1.87%, lignin 21.4%, ash 3.77%, and others 21.4%.
One gram of pretreated biomass (dry basis) was immersed in 30 ml of 50 mM sodium citrate buffer (pH 4.8) in a 250 ml Erlenmeyer flask. Cellulase from Trichoderma reesei (E.C. 3.2.1.4) was added at an enzyme loading of 35 FPU (filter paper unit)/g dry biomass. To prevent cellobiose inhibition to the enzymatic hydrolysis of the lignocellulose, cellobiase from Aspergillus niger (E.C. 3.2.1.21) at an enzyme loading of 61.5 CBU (cellobiase unit)/g dry biomass was supplemented in the flask. FPU is defined as the amount of enzyme that produces 1 lmol of glucose from filter paper in a minute, and CBU is defined as the amount of enzyme that produces 2 lmol of glucose from cellobiose in one minute. Enzymes were obtained from Novozymes North America, Inc. (Franklinton, North Carolina, USA), and the activities of cellulase and cellobiase determined using the methods adapted from Ghose (1987) were respectively 80 FPU/ml and 277 CBU/ml. To prevent the mixed liquor in the flask from microbial contamination, sodium azide (0.3%, w/v) was added into the mixture. The enzymatic hydrolysis was carried out in an air bath shaker at 55 °C and 150 rpm for 72 h. After the hydrolysis time elapsed, the flasks were immediately chilled in an ice bath to stop the reactions. The hydrolysate was then collected by centrifugation at 8 103g for 5 min, and the supernatant was stored at 80 °C for sugar analysis at a later time.
2.2. Pretreatment
2.4. Analytical methods
The pretreatment study was divided into two parts. In Study 1, three major factors which were residence time, NaOH loading, and time point for NaOH addition were investigated (Table 1). In Study 2, the other two factors which were lime loading and wash intensity were investigated. Lime loading and wash intensity were studied as a separate group because they don’t have much interaction with the three major factors. Since the solubility of lime is poor, to provide as much cheap alkalinity as possible, an excessive lime loading is required anyway to maintain a saturated lime solution throughout the pretreatment. Wash intensity and lime loading were related because lime loading might affect the amount of water required to wash away the excess lime after the pretreatment. In the 3 3 factorial study on the three major factors, lime loading and wash intensity were held constant respectively at 0.10 g Ca(OH)2/g raw biomass and 200 ml water/g raw biomass, while the subsequent studies on lime loading and wash intensity were based on the best combination of the three major factors. Four grams of biomass sample was mixed with deionized (DI) water in a serum bottle, forming a slurry at a solid concentration of 0.1 g/ml. Lime was added into the bottle and mixed with the
Total solids, structural carbohydrates, lignin, and ash in raw and pretreated biomass were measured according to Laboratory Analytical Procedures (LAP) established by National Renewable Energy Laboratory (NREL) (Sluiter et al., 2005a,b, 2008). Total reducing sugars in hydrolysate were measured using DNS method adapted from Miller (1959) and Ghose (1987). Monomeric sugars (glucose, xylose, galactose, arabinose, and mannose) derived from cellulose and hemicellulose in hydrolysate and liquor from acid hydrolysis of biomass for composition analysis were measured with a Shimadzu high performance liquid chromatography (HPLC). The HPLC system consisted of a Bio-Rad Aminex HPX-87P column (300 7.8 mm) tailored for analysis of lignocellulose-derived sugars, a Bio-Rad Micro-Guard column (30 4.6 mm), a thermostatted autosampler, a quaternary pump, and a refractive index detector. The analytical column was operated at 80 °C with HPLC grade water as the mobile phase at a flow rate of 0.6 ml/min. The samples were injected at 10 ll and the acquisition time was 35 min. A post-run time of 25 min was included between injections to allow for late-eluting compounds to come off the column. The standards used were glucose, xylose, galactose, arabinose, and
2. Methods 2.1. Biomass preparation
Table 1 Factors studied in the pretreatment of swithgrass using the combination of NaOH and lime at ambient temperature. Study 1
*
Study 2
Residence time (h)
Time point for NaOH addition
NaOH loading (g/g raw biomass)
Lime loading (g/g raw biomass)
Wash intensity (ml water/g raw biomass)
3, 6, 9
0, 1/3, 2/3*
0.05, 0.10, 0.20
0, 0.02, 0.04, 0.06, 0.08, 0.10
100, 200
0, 1/3, 2/3 represent adding NaOH at the beginning of pretreatment, after 1/3 of the residence time elapses, or after 2/3 of the residence time elapses.
3863
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
mannose at concentrations of 0.5, 2.0, 5.0, 7.5, 10.0 g/L. Mass balance was conducted based on the composition analysis of raw and pretreated biomass. After pretreatment and enzymatic hydrolysis, the percentage sugar yield was calculated as follows:
% Sugar yield ¼
SH 100 PR CF
where SH is the sugar released from per gram of raw biomass in hydrolysis, PR is the content of sugar polymer in per gram of raw biomass, and CF is the correction factor (1.11 and 1.14 for glucan and xylan respectively). 2.5. Statistical analysis All treatments in this study were conducted in triplicate and a 95% confidence level was applied for data analysis. The GLM procedure in SAS 9.1 software (SAS Institute Inc., Cary, NC) was used for all statistical analysis. Analysis of variance (ANOVA) was used to determine the effects of various factors on pretreatment and Tukey simultaneous tests were conducted to determine the statistical differences between treatments. 3. Results and discussion 3.1. Solid loss Solid loss associated with alkaline attack, which substantially compromises sugar recovery in the subsequent enzymatic hydrolysis, is of great concern especially when a strong base or severe pretreatment conditions are applied. At ambient temperature, however, only 13.9–28.4% of the total solids were lost during the pretreatment (Fig. 1), which was considerably lower than those of alkaline pretreatments at elevated temperatures (Silverstein et al., 2007; Xu et al., 2010a). NaOH loading had significant (P < 0.05) effect on solid loss at all the combinations of residence time and time point for NaOH addition except 3 h and 2/3 point NaOH addition, at which the biomass might be exposed to NaOH attack for too short a time, while residence time had significant (P < 0.05) effect on solid loss at all the combinations of NaOH loading and time point for NaOH addition. Postponing NaOH addition was crucial for the reduction of solid loss especially at higher NaOH loadings. Adding NaOH after 1/3 of the residence time elapsed substantially reduced solid loss. However, further postponing NaOH addition till 2/3 of the residence time elapsed did not substantially reduce the solid loss, which probably indicated that the majority of calcium linkages formed at the initial stage of pretreatment.
3.2. Sugar production in Study 1 After pretreatment and enzymatic hydrolysis, sugars in hydrolysate were analyzed. The yields of glucose and xylose were reported to understand the impact of pretreatments on cellulose and hemicellulose conversions. Galactose and arabinose, the two minor hemicellulose constituents, were measured but not reported in this paper due to their very low concentrations in the hydrolysate. The overall pretreatment effectiveness was evaluated using the total reducing sugar yield based on raw biomass. Glucose yields were significantly (P < 0.05) increased with the increase of residence time and NaOH loading (Fig. 2a). Postponing NaOH addition did not improve the glucose yield. This is probably because of the semicrystalline structure of cellulose. The alteration from crystalline to amorphous cellulose for improved digestibility needs time under a relatively high pH. The maximum glucose yield of 263.0 mg/g raw biomass was obtained at the most severe pretreatment conditions of 9 h, 0.20 g NaOH/g raw biomass, and NaOH addition at the very beginning (0 point), at which the overall glucan conversion reached 74.0%, 3.46 times that of untreated biomass. Hemicellulose has amorphous, heterogeneous and branched structure with little strength, which makes it more susceptible to solubilization than semicrystalline cellulose at alkaline conditions. At 3 h, increasing NaOH loading significantly (P < 0.05) improved xylose yield while postponing NaOH addition did not result in better yields (Fig. 2b). The correlations between xylose yield and pretreatment conditions were more complicated at longer residence times. Because of the enhanced solubilization of hemicellulose during pretreatment, increasing NaOH loading did not necessarily lead to higher xylose yields at 6 and 9 h, especially when NaOH was added at the earlier stages (0 and 1/3 point) of the experiment (Fig. 2b). At the NaOH loading of 0.20 g/g raw biomass, unlike glucose yield, there was no significant (P < 0.05) difference between xylose yields at 0 and 1/3 point NaOH addition, which indicated that postponing NaOH addition was more favorable for hemicellulose preservation. The maximum xylose yield of 135.3 mg/g raw biomass was obtained at 3 h, 0.20 g NaOH/g raw biomass, and 0 point NaOH addition, at which the xylan conversion reached 66.5%, 5.15 times that of untreated biomass. The effects of pretreatment conditions on total sugar yield are clearly shown in Fig. 2c. NaOH loading had significant (P < 0.05) effect on total reducing sugar yield at all the combinations of residence time and time point for NaOH addition, while extending residence time from 6 to 9 h did not improve the sugar production at 0 and 1/3 point NaOH addition due to more serious carbohydrate
Solid loss following wash (%)
30
Time for NaOH dosing 0 1/3 2/3
25
20
15
10
5
0 3h/0.05g 3h/0.10g 3h/0.20g 6h/0.05g 6h/0.10g 6h/0.20g 9h/0.05g 9h/0.10g 9h/0.20g 3 hr 3 hr 3 hr 6 hr 6 hr 6 hr 9 hr 9 hr 9 hr 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g
Treatment Fig. 1. Solid losses of swithgrass biomass pretreated at different combinations of residence time, NaOH loading, and time point for NaOH addition.
3864
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
Glucose yield (g/g raw biomass)
300
Time for NaOH dosing
a
0 1/3 2/3
250 200 150 100 50
Xylose yield (g/g raw biomass)
0 140
b3h/0.05g 3h/0.10g 3h/0.20g 6h/0.05g 6h/0.10g 6h/0.20g 12h/0.05g12h/0.10g12h/0.20g Treatment
120 100 80 60 40 20
Reducing sugar yield (g eq. glucose/g raw biomass)
0
c3h/0.05g 3h/0.10g 3h/0.20g 6h/0.05g 6h/0.10g 6h/0.20g 12h/0.05g12h/0.10g12h/0.20g 400
Treatment
300
200
100
0 3h/0.05g 3h/0.10g 3h/0.20g 6h/0.05g 6h/0.10g 6h/0.20g 3 hr 3 hr 3 hr 6 hr 6 hr 6 hr 12h/0.05g12h/0.10g12h/0.20g 9 hr 9 hr 9 hr 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g
Treatment Fig. 2. Yields of (a) glucose, (b) xylose, and (c) total reducing sugars in enzymatic hydrolysis of switchgrass biomass pretreated at different combinations of residence time, NaOH loading, and time point for NaOH addition.
loss at the extended exposure time to NaOH attack. Similar to glucose and xylose yields, although postponing NaOH addition could result in better carbohydrate preservation, reducing the exposure time to the high pH substantially compromise the biomass digestibility improvement, resulting in lower sugar productions. The maximum total reducing sugar yield of 435.8 mg/g raw biomass was obtained at 9 h, 0.20 g NaOH/g raw biomass, and 0 point NaOH addition, at which the conversion of total available carbohydrate reached 68.1%, 3.63 times that of untreated biomass. However, cutting the residence time down to 6 h and applying half the NaOH loading only resulted in a 5% reduction in sugar yield. Therefore, based on cost-benefit considerations, 6 h, 0.10 g NaOH/g raw biomass, and 0 point NaOH addition were applied in the following lime loading and biomass washing studies. 3.3. Sugar production in Study 2 Using the best combination of residence time, NaOH loading, and time point for NaOH addition, lime loadings of 0, 0.02, 0.04, 0.06, 0.08 and 0.10 g Ca(OH)2/g raw biomass and wash intensities of 100 ml and 200 ml water/g raw biomass were further studied.
The wash intensity of 200 ml water/g raw biomass was sufficient to remove residual lime and the sugar production was maximized at the lime loading of 0.04 g Ca(OH)2/g raw biomass (Fig. 3). Reducing wash intensity to 100 ml water/g raw biomass resulted in significant (P < 0.05) decreases of sugar production at lime loadings of 0.04–0.10 g Ca(OH)2/g raw biomass. This is because that the reduced wash intensity was not sufficient to remove all the residual lime mixed with biomass. Although sodium citrate buffer (pH 4.8) was used according to the standard procedure to maintain the hydrolysis pH at the optimum level, the dissolution of residual lime particles still substantially raised the pH (pH exceeded 6.5 at 0.10 g Ca(OH)2/g raw biomass). However, there was no significant (P < 0.05) decrease in sugar production at the lowest lime loading of 0.02 g Ca(OH)2/g raw biomass, at which the total reducing sugar yield, 30% higher than that of the biomass treated without lime supplementation, whereas the total chemical expense was barely increased considering the low cost of lime and the small amount required. Although using the combination of 0.04 g Ca(OH)2/g raw biomass and 200 ml water/g raw biomass resulted in a 6% higher total reducing sugar yield than using the combination of 0.02 g Ca(OH)2/g raw biomass and 100 ml water/g raw bio-
3865
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
Total reducing sugar yield (g eq. glucose/g raw biomass)
500
Wash volume 100 ml 200 ml
400
300
200
100
0 0.00
0.02
0.04
0.06
0.08
0.10
Lime loading (g Ca(OH)2/g raw biomass) Fig. 3. Total reducing sugar yields in enzymatic hydrolysis of switchgrass biomass pretreated at different combinations of lime loading and wash intensity.
mass, doubling both lime loading and wash intensity could potentially compromise the cost-effectiveness of the process. Preliminary economic analysis, therefore, is necessary before any scale-up studies. Based on total sugar production and processing convenience, 6 h, 0.10 g NaOH/g raw biomass, 0 point NaOH addition, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass were determined as the best conditions for the ambient temperature pretreatment of switchgrass in this study. 3.4. Lignin reduction Lignin is a three dimensional complex aromatic polymer which forms a sheath surrounding cellulose and hemicellulose, stiffening and holding together the fibers of polysaccharides (Fan et al., 1987). A number of studies show that NaOH can effectively remove lignin barriers, thus exposing the carbohydrates to enzymes (MacDonald et al., 1983; Chen et al., 2007; Silverstein et al., 2007). However, the delignification capability of lime is quite limited. Wang (2009) reported that, at 121 °C, the total lignin reduction of lime pretreatment of coastal Bermuda grass was only 10–20% which was much lower than that of NaOH pretreatment. This is probably because of the interaction between lignin and calcium ions within the biomass structure. Many studies show that divalent calcium
ions had high affinity for lignin and could effectively crosslink lignin molecules especially at alkaline conditions (Torre et al., 1992; Sundin and Harlter, 2000a,b; Duong et al., 2005). In this study, all of the three parameters (residence time, NaOH loading, and time point for NaOH addition) had significant (P < 0.05) effect on lignin reduction (Fig. 4). Increasing residence time, NaOH loading, and the exposure time of biomass to NaOH attack favored the removal of lignin barrier. It was also found that improving biomass preservation by postponing NaOH addition till 1/3 of the residence time elapsed caused a remarkable reduction in lignin removal, which is in agreement with the previous reports that calcium ions provide linkages between lignin molecules, thus providing better resistance against solubilization during alkaline pretreatment (Wang, 2009; Xu et al., 2010b). The lignin reduction ranged from 23.8% at 3 h, 0.05 g NaOH/g raw biomass, 2/3 point NaOH addition, to 45.1% at 9 h, 0.20 g NaOH/g raw biomass, 0 point NaOH addition, and was positively correlated to NaOH loading and residence time. The high lignin content of pretreated biomass did not affect enzymatic hydrolysis. This is in correspondence with our previous research (Xu et al., 2010b) which shows that as long as the chemical bonds stiffening the lignocellulose were disrupted and the biomass porosities substantially increased, enzymatic hydrolysis of high efficiency can still be achieved even in the presence of relatively high lignin contents.
50
Time for NaOH dosing 0 1/3 2/3
Lignin reduction (%)
40
30
20
10
0 3h/0.05g 3h/0.20g 6h/0.05g 6h/0.10g 6h/0.20g 9h/0.05g 9h/0.10g 9h/0.20g 3 hr 3h/0.10g 3 hr 3 hr 6 hr 6 hr 6 hr 9 hr 9 hr 9 hr 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g 0.05g 0.10g 0.20g
Treatment Fig. 4. Lignin reductions of switchgrass biomass pretreated at different combinations of residence time, NaOH loading, and time point for NaOH addition.
3866
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
3.5. Comparison with NaOH pretreatment Since lime was too weak to effectively improve biomass digestibility within a reasonable time frame at ambient temperature (Kim and Holtzapple, 2005; Xu et al., 2010b), the pretreatment using the combination of NaOH and lime at the best conditions (0.10 g NaOH/g raw biomass, 0.02 g Ca(OH)2/g raw biomass, 6 h) was further evaluated based on the comparison with NaOH pretreatment. Two NaOH loadings of 0.10 and 0.20 g NaOH/g raw biomass were explored with all the other pretreatment conditions kept the same. The carbohydrate conversions, lignin reductions, pH changes of the three pretreatment strategies are shown in Table 2. After pretreatment and enzymatic hydrolysis, the overall glucan conversion of the biomass treated with lime supplementation was 23.5% higher than that of the biomass treated just using 0.10 g NaOH/g raw biomass, but 14.5% lower than that of the biomass treated using 0.20 g NaOH/g raw biomass. However, the xylan conversion of the biomass treated with lime supplementation was not only 47.3% higher than that of the biomass treated using 0.10 g NaOH/g raw biomass, but also 4.6% higher than that of the biomass treated using 0.20 g NaOH/g raw biomass, indicating better hemicellulose preservations by calcium bonding. As a result, the total carbohydrate conversion of the biomass treated with lime supplementation was 34.5% higher than that of the biomass treated using 0.10 g NaOH/g raw biomass, and comparable with that of the biomass treated using 0.20 g NaOH/g raw biomass. Since the cost of supplementing 0.02 g Ca(OH)2/g raw biomass was remarkably lower than that of doubling the loading of expensive NaOH, using the combination of NaOH and lime shows great economic promise. Assuming that the market price of NaOH is 3.5 times as much as that of lime (Lerner, 2007; USGS, 2010), to produce the same amount of fermentable sugars from biomass feedstock, supple-
menting 0.02 g Ca(OH)2 to 0.1 g NaOH/g raw biomass instead of applying 0.20 g NaOH/g raw biomass in the pretreatment lead to a 43.4% reduction in chemical cost. The pH changes indicated that lime did help to avoid substantial pH drop during pretreatment. After 6 h, the decline of pH at 0.10 g NaOH/g raw biomass without lime supplementation was significantly (P < 0.05) greater than that with lime supplementation. To make the differences more outstanding, the pretreatment time was extended to 24 h. The final pH of the pretreatment with lime supplementation was not significantly (P < 0.05) different from that of the pretreatment using 0.20 g NaOH/g raw biomass, while considerably higher than that of the pretreatment using 0.10 g NaOH/g raw biomass. 3.6. Mass balances Mass balances were performed on the biomass pretreated at the best conditions (6 h, 0.10 g NaOH/g raw biomass, and 0 point NaOH addition, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass). The biomass pretreated using 0.10 and 0.20 g NaOH/g raw biomass was also analyzed for comparison. The total dry weight of the sample was measured after pretreatment, and the compositions of the pretreated biomass (glucan, xylan, galactan, arabinan, lignin, ash, and other) were determined and compared with that of raw biomass. The solid loss significantly (P < 0.05) increased with the increase of NaOH loading (Fig. 5). However, lime supplementation, although contributed to a greater alkalinity than that of just using 0.10 g NaOH/g raw biomass, did not cause higher solid loss due to the better biomass preservation in the presence of calcium linkages. Glucan (cellulose) was the best preserved carbohydrate at all pretreatment conditions because of its semicrystilline structure. The impacts of pretreatment on xylan and lignin were similar to that on total solids. Better xylan and lig-
Table 2 Comparisons of carbohydrate conversions, lignin reductions, and pH changes between NaOH pretreatment and the pretreatment using the combination of NaOH and lime.
*
Chemical loading (g raw biomass1)
Glucan conversion (%)
Xylan conversion (%)
Total carbohydrate conversion (%)
Lignin reduction (%)
Initial pH
0.20 g NaOH 0.10 g NaOH 0.10 g NaOH + 0.02 g Ca(OH)2
69.5 (1.09)* 48.1 (3.12) 59.4 (0.79)
54.8 (0.87) 38.9 (2.90) 57.3 (0.79)
63.5 (1.01) 44.1 (1.73) 59.3 (1.04)
48.6 (1.03) 34.6 (1.51) 37.8 (0.55)
12.92 (0.02) 12.84 (0.02) 12.86 (0.01)
Final pH 6h
24 h
12.81 (0.02) 12.59 (0.02) 12.77 (0.03)
12.66 (0.04) 12.32 (0.03) 12.65 (0.04)
The number in parentheses is standard deviation of triplicate samples.
Content (g)
100 90
Raw biomass 0.10 g NaOH+0.02 g Ca(OH)2
80
0.10 g NaOH 0.20 g NaOH
70
30 20 10 0 Total
Glucan
Xylan
Galactan Arabinan Lignin
Ash
Others
Components Fig. 5. Material balances for raw switchgrass, the biomass pretreated using the combination of NaOH and lime at the best conditions (6 h, 0.10 g NaOH/g raw biomass, 0 point NaOH addition, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass), and the biomass pretreated using 0.10 and 0.20 g NaOH/g raw biomass at the same conditions.
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
nin preservations were achieved with lime supplementation. The biomass pretreated using the combination of 0.10 g NaOH and 0.02 g Ca(OH)2/g raw biomass was comparable with that pretreated using 0.20 g NaOH/g raw biomass in terms of sugar production, however, its composition was more like that of the biomass pretreated using 0.10 g NaOH/g raw biomass. 3.7. Study on enzyme loading To determine the best enzyme loadings, cellulase loadings of 0– 35 FPU/g dry biomass and cellobiase loadings of 0–50 CBU/g dry biomass were investigated. The glucose, xylose, and total reducing sugar yields of switchgrass pretreated at the best conditions (6 h, 0.10 g NaOH/g raw biomass, 0 point NaOH addition, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass) were determined at different cellulase loadings (Fig. 6a). The cellobiase loading was kept constant at an excess level of 61.5 CBU/g biomass in the cellulase loading test to eliminate the impact of cellobiase limitation on sugar production. With the increase of cellulase loading from 0 to 20 FPU/g dry biomass, the total reducing sugar yield increased by 6.52 times, while further increasing cellulase loading did not improve the sugar yield. Similar trends were observed for both glucose and xylose yields. Cellobiase loadings were studied based on the best cellulase loading of 20 FPU/g dry biomass. The results show that supplementing cellobiase significantly (P < 0.05) increased sugar yields and a cellobiase loading of 10 CBU/g dry biomass was sufficient (Fig. 6b). The optimal enzyme loadings were comparable with those obtained in the study on switchgrass pretreatment using NaOH (Xu et al., 2010a).
3867
4. Conclusions Lime, which is very inexpensive but too weak to work effectively at ambient temperature, helped to reduce the requirement for strong but expensive NaOH in the alkaline pretreatment of switchgrass by maintaining a high pH level and mitigating carbohydrate solubilization during pretreatment, thus contributing to a novel pretreatment technology of high cost-effectiveness. The best pretreatment conditions were determined as: 6 h, 0.10 g NaOH/g raw biomass, NaOH addition at the beginning, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass, and the enzyme loadings of 20 FPU cellulase/g raw biomass and 10 CBU cellobiase/g raw biomass were recommended for hydrolysis. A pretreatment at the best conditions resulted in a total reducing sugar yield of 59.3%, 34.5% higher than that of NaOH pretreatment at the same conditions without lime supplementation, while the chemical cost was barely increased considering the low cost of lime and the minor loading required.
Acknowledgements The authors would like to acknowledge the financial support of this research from North Carolina Agricultural Research Service (NCARS) and North Carolina Agricultural Foundation (NCAF). We would also like to thank Dr. Joseph Burns from Crop Science Department at North Carolina State University for providing the switchgrass material and Novozymes North America, Inc. for donating the enzymes for this study.
Sugar yields (mg/g raw biomass)
References
a
400
300
200
100 Cellobiase loading: 61.5 CBU/g dry biomass 0 0
10
20
30
40
Sugar yields (mg/g raw biomass)
Cellulase loading (FPU/g dry biomass)
b
400
300
200
100 Cellulase loading: 20 FPU/g dry biomass 0 0
10
20
30
40
50
60
Cellobiase loading (CBU/g dry biomass) Fig. 6. Effects of cellulase and cellobiase loadings on (d) glucose, (s) xylose, and (.) total reducing sugar yields in enzymatic hydrolysis of switchgrass pretreated at the best conditions (6 h, 0.10 g NaOH/g raw biomass, 0 point NaOH addition, 0.02 g Ca(OH)2/g raw biomass, and wash intensity of 100 ml water/g raw biomass).
Belkacemi, K., Turcotte, G., de Halleux, D., Savoie, P., 1998. Ethanol production from AFEX-treated forages and agricultural residues. Appl. Biochem. Biotechnol. 70– 72, 441–462. Burns, J.C., Godshalk, E.B., Timothy, D.H., 2008. Registration of ‘Performer’ switchgrass. J. Plant Reg. 2 (1), 29–30. Chang, V.S., Burr, B., Holtzapple, M.T., 1997. Lime pretreatment of switchgrass. Appl. Biochem. Biotechnol. 63, 3–19. Chang, V.S., Nagwani, M., Holtzapple, M.T., 1998. Lime pretreatment of crop residues bagasse and wheat straw. Appl. Biochem. Biotechnol. 74, 135–159. Chang, V.S., Nagwani, M., Kim, C.-H., Holtzapple, M.T., 2001. Oxidative lime pretreatment of high-lignin biomass: poplar wood and newspaper. Appl. Biochem. Biotechnol. 94, 1–28. Chen, Y., Sharma-Shivappa, R.R., Keshwani, D., Chen, C., 2007. Potential of agricultural residues and hay for bioethanol production. Appl. Biochem. Biotechnol. 142, 276–290. Duong, T.D., Hoang, M., Nguyen, K.L., 2005. Sorption of Na+, Ca2+ ions from aqueous solution onto unbleached kraft fibers-kinetics and equilibrium studies. J. Colloid Interf. Sci. 287, 438–443. Fan, L.T., Gharpuray, M.M., Lee, Y.-H., 1987. Cellulose hydrolysis. Biotechnology Monographs, Springer, Berlin. pp. 8. Ghose, T.K., 1987. Measurement of cellulase activities. Pure Appl. Chem. 59 (2), 257–268. Hendriks, A.T.W.M., Zeeman, G., 2009. Pretreatments to enhance the digestibility of lignocellulosic biomass. Biotechnol. Prog. 100, 10–18. Kaar, W.E., Holtzapple, M.T., 2000. Using lime pretreatment to facilitate the enzymic hydrolysis of corn stover. Biomass Bioenergy 18, 189–199. Kim, S., Holtzapple, M.T., 2005. Lime pretreatment and enzymatic hydrolysis of corn stover. Bioresour. Technol. 96, 1994–2006. Lerner, I., 2007. Caustic soda prices bubbling. ICIS Chem. Business Am. (1936-458X) 271 (12), 24. MacDonald, D.G., Bakhshi, N.N., Mathews, J.F., Roychowdhury, A., 1983. Alkali treatment of corn stover to improve sugar production by enzymatic hydrolysis. Biotechnol. Bioeng. 25, 2067–2076. McLaughlin, S.B., Bouton, J., Bransby, D., Conger, B.V., Ocumpaugh, W.R., Parrish, D.J., Taliaferro, C., Vogel, K.P., Wullschleger, S.D., 1999. Developing switchgrass as a bioenergy crop. In: Janick, J. (Ed.), Perspectives on New Crops and New Uses. ASHS Press, Alexandria, VA, pp. 282–299. Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal. Chem. 31, 426–428. Parrish, D.J., Fike, J.H., 2005. The biology and agronomy of switchgrass for biofuels. Crit. Rev. Plant Sci. 24 (5–6), 423–459. Silverstein, R.A., Chen, Y., Sharma-Shivappa, R.R., Boyette, M.D., Osborne, J., 2007. A comparison of chemical pretreatment methods for improving saccharification of cotton stalks. Bioresour. Technol. 98, 3000–3011.
3868
J. Xu, J.J. Cheng / Bioresource Technology 102 (2011) 3861–3868
Sluiter, A., Hames, B., Hyman, D., Payne, C., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Wolfe, J., 2005a. Determination of total solids in biomass and total dissolved solids in liquid process samples. Laboratory Analytical Procedure (LAP). NREL/TP-510–42621. National Renewable Energy Laboratory, Golden, Colorado. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., 2005b. Determination of ash in biomass. Laboratory Analytical Procedure (LAP). NREL/TP-510–42622. National Renewable Energy Laboratory, Golden, Colorado. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D., 2008. Determination of structural carbohydrates and lignin in biomass. Laboratory Analytical Procedure (LAP). NREL/TP-510–42618. National Renewable Energy Laboratory, Golden, Colorado. Sundin, J., Hartler, N., 2000a. Precipitation of kraft lignin by metal cations in alkaline solutions. Nord. Pulp Pap. Res. J. 15 (4), 306–312.
Sundin, J., Hartler, N., 2000b. Precipitation of kraft lignin by metal during pulp washing. Nord. Pulp Pap. Res. J. 15 (4), 313–318. Torre, M., Rodriguez, A.R., Saura-Calixto, F., 1992. Study of the interactions of calcium ions with lignin, cellulose, and pectin. J. Agric. Food. Chem. 40, 1762–1766. USGS, 2010. USGS Minerals Information: Lime. Retrieved Nov 27, 2010 from