- Email: [email protected]
Effects of organosolv pretreatment and enzymatic hydrolysis on cellulose structure and crystallinity in Loblolly pine
Carbohydrate Research 345 (2010) 965–970
Contents lists available at ScienceDirect
Carbohydrate Research journal homepage: www.elsevier.com/locate/carres
Note
Effects of organosolv pretreatment and enzymatic hydrolysis on cellulose structure and crystallinity in Loblolly pine Poulomi Sannigrahi a, Stephen J. Miller b, Arthur J. Ragauskas a,* a b
School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332-0400, USA Chevron Energy Technology Company, 100 Chevron Way, Richmond, CA 94802, USA
a r t i c l e
i n f o
Article history: Received 6 November 2009 Received in revised form 6 February 2010 Accepted 11 February 2010 Available online 17 February 2010 Keywords: Loblolly pine Organosolv pretreatment Cellulose structure 13 C NMR spectroscopy Bioethanol
a b s t r a c t Ethanol organosolv pretreatment was performed on Loblolly pine to enhance the efficiency of enzymatic hydrolysis of cellulose to glucose. Solid-state 13C NMR spectroscopy coupled with line shape analysis was used to determine the structure and crystallinity of cellulose isolated from pretreated and enzymehydrolyzed Loblolly pine. The results indicate reduced crystallinity of the cellulose following the organosolv pretreatment, which renders the substrate easily hydrolyzable by cellulase. The degree of crystallinity increases and the relative proportion of para-crystalline and amorphous cellulose decreases after enzymatic hydrolysis, indicating preferential hydrolysis of these regions by cellulase. The structural and compositional changes in this material resulting from the organosolv pretreatment and cellulase enzyme hydrolysis of the pretreated wood were studied with solid-state CP/MAS 13C NMR spectroscopy. NMR spectra of the solid material before and after the treatments show that hemicelluloses and lignin are degraded during the organosolv pretreatment. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction Production of bioethanol from renewable lignocellulosic biomass can be accomplished by hydrolyzing plant polysaccharides to soluble hexose and pentose sugars, which are then microbiologically fermented.1 Cellulose, lignin and hemicellulose, the key components of lignocellulosic biomass, are closely associated with each other at the plant cell level. This close association, together with the partly crystalline nature of cellulose, reduces cellulose reactivity towards enzymatic hydrolysis in native biomass.2 Thus, some type of pretreatment of the native biomass is necessary to render the carbohydrate fraction amenable to enzymatic and microbial action.3 The ethanol organosolv process is among the pretreatment strategies currently being studied for the conversion of lignocellulosic biomass feedstocks.4–6 In this pretreatment, biomass is treated at high temperature in an aqueous ethanol solution with sulfuric acid as a delignification catalyst. Lignin removed during the pretreatment is solubilized by washing in ethanol and is recovered by precipitating in water. It was originally developed as the AlcellÒ pulping process for hardwoods.7,8 Recent studies focused on evaluating and optimizing this pretreatment for the delignification of softwood, hardwood and agricultural residues,5,9–11 have shown that the organosolv pretreatment yields readily hydrolyzable substrates with good glucose recovery and low lignin content.
* Corresponding author. Tel.: +1 404 894 9701; fax: +1 404 894 4778. E-mail address: [email protected] (A.J. Ragauskas). 0008-6215/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.carres.2010.02.010
In addition, the recovered lignin remains an interesting biomaterial in its own right.12 The ordered region of native cellulose is a mixture of two crystalline allomorphs, cellulose Ia and cellulose Ib. Cellulose Ia is the dominant form in bacterial and algal celluloses and cellulose Ib is the dominant form in higher plants.13 Solid-state 13C NMR of cellulose has been shown to be a suitable analytical technique to characterize these and several other forms of cellulose including para-crystalline cellulose, cellulose at accessible fibril surfaces (i.e., surfaces in contact with water) and amorphous cellulose at inaccessible fibril surfaces (i.e., fibril–fibril contact surfaces and surfaces resulting from distortions in the fibril interior).14–18 The effects of chemical pretreatment on the structure and crystallinity of cellulose in lignocellulosic biomass are a function of the pretreatment conditions. The two-step dilute acid pretreatment often employed for softwood feedstocks, has been reported to bring about an increase in cellulose crystallinity of Loblolly pine, by preferential degradation of the less ordered amorphous cellulose.19 Steam explosion pretreatment of aspen wood also exhibited an increase in the relative proportion of crystalline cellulose.20 In contrast, the ammonia fiber explosion (AFEX) pretreatment enhances enzymatic digestibility by decrystallization of cellulose.21 The effects of the ethanol organosolv pretreatment on cellulose structure and crystallinity are not fully explored. However, it has been shown that the degree of crystallinity increases with increasing severity of the organosolv pretreatment conditions.11 This has been attributed to the preferential degradation of the less structured forms of cellulose under harsher pretreatment conditions.
966
P. Sannigrahi et al. / Carbohydrate Research 345 (2010) 965–970
The relationship between cellulose structure and rates of enzymatic hydrolysis has been extensively studied,22,23 but is still a topic of debate in the literature. Cellulose hydrolysis rates mediated by fungal cellulases are typically 3–30 times faster for amorphous cellulose as compared to highly crystalline cellulose.22 Thus it is expected that crystallinity should increase over the course of cellulose hydrolysis as a result of preferential degradation of amorphous cellulose. While some studies support this hypothesis,17,24 others have reported that crystallinity does not increase during enzymatic hydrolysis.22,23 Pu et al.17 have shown that during cellulase treatment of fully bleached softwood kraft pulp, the cellulose Ia, para-crystalline and non-crystalline regions are rapidly hydrolyzed in the initial phase of hydrolysis leading to an increase in the cellulose crystallinity index. After this rapid initial phase, all cellulose polymorphs including the more energetically stable cellulose Ib, were found to be equally susceptible to enzymatic hydrolysis. In this study, the structure and crystallinity of cellulose extracted from untreated Loblolly pine has been compared to that obtained following ethanol organosolv pretreatment of this feedstock and its subsequent enzymatic hydrolysis by cellulase enzymes. The degree of crystallinity and ultrastructure of cellulose was determined using solid-state 13C CP/MAS NMR spectroscopy coupled with advanced line fitting analysis. Changes in the bulk chemical composition of the wood resulting from these treatments have been evaluated with solid-state 13C CP/MAS NMR spectroscopy. The results provide a view of the structural changes taking place in Loblolly pine, especially in the cellulose fraction, with organosolv pretreatment and enzymatic hydrolysis. 2. Results and discussion 2.1. Composition of untreated, pretreated and enzymehydrolyzed Loblolly pine The analysis of carbohydrates and lignin in the pre- and posttreated organosolv pretreated pine and the pretreatment liquids was accomplished following literature methods,25,26 and these results are summarized in Table 1. About 79% of the glucose in the untreated biomass is recovered in the solid phase after organosolv pretreatment. The loss of glucan was attributed, in part, to the hydrolysis of glucomannans and to some extent cellulose. The hydrolysis of polysaccharides, especially hemicelluloses has been well reported in the literature.10,11 As recently reported, the organosolv process results in the hydrolysis of lignin which solubilizes and releases lignin into the aqueous ethanolic phase.11 Solubilization of lignin in the pretreatment effluent results in extensive delignification. The solid residue after enzymatic hydrolysis is composed primarily of Klason lignin (Table 1). The liquid phase recovered from the pretreatment reactor and after washing the pretreated material, contains hemicellulose sugars, furans and soluble lignin (Table 1). Furans are produced by the degradation of carbohydrates under high temperature acidic conditions.
About 3% of the glucose in the starting material is also recovered in the pretreatment liquid. The carbohydrates not detected in the pretreated solid or liquid fractions, are inferred to have degraded to furans and other degradation products such as acetic acid.5,10 The formation of acetic acid and other minor acids including levulinic and succinic acids can also lower the pH of the cooking liquor during the pretreatment. 19% of the lignin in the untreated biomass is recovered as ethanol organosolv lignin (Table 1). Hydrolysis with cellulase results in the hydrolysis of 70% of the cellulose in the pretreated biomass to soluble glucose (Fig. 1). 2.2. Solid-state NMR of untreated, pretreated and enzymehydrolyzed wood Solid-state 13C NMR spectra of extractive free Loblolly pine before and after the organosolv and enzymatic treatments are presented in Figure 2. Peaks have been assigned based on the literature27,28 and are shown on the spectra. The peak at d 21 ppm, which can be attributed to the acetyl group of hemicelluloses, is seen only in the spectra from untreated Loblolly pine. This correlates to the solubilization of hemicelluloses in the effluent fraction during organosolv pretreatment. The peaks in the region between 60 and 105 ppm are predominantly due to the different carbons of cellulose. There is some spectral overlap with hemicelluloses in this region, thus pure cellulose needs to be isolated for determination of cellulose structure and crystallinity as presented in the next section. The signal at d 56 ppm corresponds to the aryl methoxyl carbons of lignin and a reduced intensity of this peak is also seen in the sample after the organosolv pretreatment. The signal in the region from 140 to 160 ppm is diagnostic of lignin aromatic carbons. As seen in Figure 2, the lignin peaks are less intense in the wood samples after the organosolv pretreatment. Since the solid residue remaining the cellulose enzymatic hydrolysis has a relatively large proportion of lignin, the peaks at 56 and 140–160 ppm are more pronounced in this spectrum. 2.3. Cellulose structure and crystallinity The solid-state 13C NMR spectra of cellulose from untreated, pretreated and enzyme-hydrolyzed Loblolly pine (Fig. 2) have been analyzed using line shape analysis based on the method outlined by Larsson et al.15,29 Signals from d 86 to 92 correspond to the C4 carbons from the crystalline forms of cellulose, as well as paracrystalline domains, whereas the broader upfield resonance from d 80 to 86 is assigned to less ordered or non-crystalline domains of cellulose (Fig. 3). The ratio of the area in the d 86–92 region to the total peak area from d 80–92 is designated as the crystallinity index (CrI). Typical errors for this analysis are 2% (standard deviation). The crystallinity index for the untreated Loblolly pine cellulose is 63%.19 This is comparable to cellulose crystallinity values obtained by NMR, from hybrid poplar (63%),30 but is higher than spruce (48%).31 Cellulose isolated from Loblolly pine wood chips (without milling to a smaller size) has a crystallinity index of
Table 1 Composition of the Loblolly pine feedstock and the solid material following organosolv pretreatment and enzymatic hydrolysisa (% Dry weight starting material)
Untreated Organosolv-treated wood Organosolv lignin Pretreatment liquid Enzyme hydrolysis solid residue a
Carbohydrates
Lignin
Arabinan
Galactan
Glucan
Mannan
Xylan
1.6 0.0
2.6 0.0
42.0 33.3
9.6 0.2
7.8 0.3
1.2 0.1
1.3 0.0
3.0 10.7
7.2 0.9
5.6 1.5
The pretreatment liquid is the combined effluent and wash solutions.
Furans
Acid insoluble
Acid soluble
Organosolv lignin
29.4 11.4
0.5 0.2
— — 5.4
16.3
12.4 0.0
2.2
—
967
P. Sannigrahi et al. / Carbohydrate Research 345 (2010) 965–970
80.0
Cellulose to glucose converson (%)
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0 0
10
20
30
40
50
60
70
80
90
Time (hours) Figure 1. Cellulose to glucose conversion (%) during enzymatic hydrolysis of organosolv-treated Loblolly pine.
Cellulose Lignin –OCH3 Lignin Aryl C
Organosolvenzyme
Organosolv
Hemicellulose R-COCH3 Untreated
190
170
150
130
110
90
80
70
60
50
40
30
20
10
0
-20
f1 (ppm) Figure 2. CP/MAS
13
C NMR spectrum of untreated, organosolv-treated and enzyme-hydrolyzed Loblolly pine.
52%, which shows that mechanical milling removes some of the less ordered or amorphous regions of cellulose. While effects of mechanical milling on cellulose crystallinity have not been reported, Larsson et al.29 observed a greater proportion of ordered forms of cellulose in spruce Kraft pulp as compared to chips from the same wood sample. This has been attributed to a combination of the preferential loss of disordered cellulose and the loss of hemicelluloses during pulping.29 It should be noted that in this study, the biomass used for cellulose isolation (from the untreated mate-
rial) and that used for the organosolv pretreatment were milled to the same extent, which should account for the effects of mechanical milling. The cellulose crystallinity index decreases to 53% in the cellulose from organosolv-treated wood (Table 2), which indicates that the pretreatment conditions used in this study are capable of decreasing the ordering of cellulosic fibers. These results are in contrast to those obtained from two-step dilute sulfuric acid pretreatment of this biomass feedstock.19 The lower crystallinity
968
P. Sannigrahi et al. / Carbohydrate Research 345 (2010) 965–970
6
HOH2C 4
O
5
O HO
2
1
OH
3
OH 1
HO O
O
HOH2C 90.0
ppm
85.0
Organosolv-enzyme
Organosolv C 2, 3, 5
90.0
ppm
85.0
C1
C4
C6
Untreated
110
100
90
80
70
60
50
ppm (t1) Figure 3. CP/MAS
13
C NMR spectrum of cellulose isolated from untreated, organosolv-treated and enzyme-hydrolyzed Loblolly pine.
Table 2 Peak assignments and results from the spectral fitting of cellulose 13C NMR spectra for untreated, organosolv-treated and enzyme-hydrolyzed Loblolly pine Assignments
Cellulose Ia Cellulose Ia+b para-Crystalline cellulose Cellulose Ib Accessible fibril surface Inaccessible fibril surface Accessible fibril surface Crystallinity indexa (%) a b c
Chemical shift (ppm)
Proportion (%)
Line type
Untreated
OSb
OSenzymec
90.0 88.8 88.5
0.1 30.7 24.8
0.5 11.1 12.5
5.7 46.7 0.0
Lorentz Lorentz Gauss
87.9 84.9
6.9 13.0
23.4 15.6
28.3 8.8
Lorentz Gauss
83.9
15.6
14.6
8.7
Gauss
83.4
8.9
22.4
1.7
Gauss
63
53
81
Crystallinity index = d86–92 ppm/d80–92 ppm. OS: Organosolv treated OS-enzyme: Organosolv treated and enzyme hydrolyzed.
index, coupled with delignification renders the organosolv-treated wood highly amenable to enzymatic hydrolysis. This is reflected in the high (70%) cellulose to glucose conversion during enzymatic hydrolysis of pretreated Loblolly pine (Fig. 1) despite the low enzyme dosage employed. Cellulose hydrolysis proceeds at a higher
rate during the initial phases of the experiment and becomes slower as the experiment proceeds. This can be attributed in part to gradual changes in ultrastructure,14 a decrease in enzyme activity by non-productive binding to lignin and product inhibition. The crystallinity index increases to 81% in cellulose isolated from the substrate after the enzymatic hydrolysis for 80 h. Thus the cellulase treatment appears to be selectively degrading the less ordered, amorphous and para-crystalline forms of cellulose. These results are in agreement with those of Pu et al.17 and support the hypothesis that cellulose crystallinity increases during enzymatic hydrolysis. The fraction of cellulose not hydrolyzed to glucose by cellulase is highly crystalline and recalcitrant to these enzymes. Crystalline regions of cellulose are considered to be more difficult to degrade than non-crystalline regions due to strong intermolecular hydrogen bonding between the cellulose chains. As the demand for lignocellulosic feedstock grows, enzymes capable of hydrolyzing crystalline cellulose need to be developed in order to achieve optimal utilization of available resources. Line shape analysis of the cellulose C-4 region (d 82–90) was performed by using Lorentzian lines for the crystalline region and Gaussian lines for the non-crystalline region.29 This analysis is based on a non-linear least squares fitting of 13C NMR spectra enabling calculation of the relative amounts of cellulose Ia, cellulose Ib, para-crystalline cellulose and cellulose at accessible and inaccessible fibril surfaces.15,18,29 The results of spectral fitting for the cellulose isolated from untreated, organosolv-treated and enzyme-hydrolyzed Loblolly pine are compared in Table 2. Typical
P. Sannigrahi et al. / Carbohydrate Research 345 (2010) 965–970
errors for this analysis with similar materials are ±7%. The main effects of the organosolv pretreatment on cellulose structure in Loblolly pine are: increase in accessible fibril surface (16.1%); increase in cellulose Ib (16.5%) and decrease in cellulose Ia+b and para-crystalline cellulose (Table 2). An increase in the relative proportion of cellulose Ib is among the most significant modifications observed in the crystalline region of cellulose. There is a concurrent decrease in the proportion of cellulose Ia+b. It may be inferred that a fraction of the thermodynamically meta-stable cellulose Ia is being converted to cellulose Ib under the thermal and chemical conditions of the organosolv pretreatment. An increase in cellulose Ib was also reported in cellulose after dilute acid pretreatment of Loblolly pine.19 In the non-crystalline region, there is an increase in the accessible fibril surface at d 83.4 ppm indicating possible improved enzyme accessibility of the pretreated cellulose. The cellulose isolated from the solid residue from enzymatic hydrolysis is mostly crystalline. As seen in Table 2, the relative proportions of ordered cellulose regions (cellulose Ia, cellulose Ib and cellulose Ia+b) increase by 45.7%, which is accompanied by a decrease in the para-crystalline and non-crystalline region of cellulose. These results corroborate previous hypotheses,22 that fungal-derived cellulases can hydrolyze the less-ordered regions of cellulose more readily than the crystalline regions with long-range order. 3. Experimental 3.1. Materials All chemical reagents used in this study were purchased from VWR International and used as received. The enzymes were obtained from Sigma–Aldrich (St. Louis, MO) and the D-glucose assay kit was purchased from Megazyme International (Ireland). Wood from a mature Loblolly pine (Pinus taeda), was obtained from the University of Georgia research plot in Baldwin County, GA. The wood was manually debarked and chipped using a mechanical chipper. A homogenized sample has been used for all analyses. The wood chips were stored at 5 °C during the course of this study. Prior to the organosolv treatment, the wet wood chips were milled using a Wiley mill to pass a 5 mm screen.
3.2. Organosolv pretreatment Loblolly pine sawdust was treated with aqueous ethanol, along with sulfuric acid as catalyst, using the methodology and conditions optimized for Lodgepole pine.11 Briefly, 100.0 g (by dry weight) of Loblolly pine sawdust were immersed in 65% ethanol/ water solution (700.0 mL), containing 1.1% sulfuric acid (by weight of wood) and treated in a Parr pressure reactor (3.8 L) at 170 °C for 1 h. The effluents were collected after condensation. The treated wood was washed three times with 300 mL, 65% ethanol at 60 °C and the washes combined with the effluent. The wood was then washed three times with de-ionized (DI) water at 60 °C and the washes discarded. The combined effluent and ethanol wash solution was mixed with three volumes of DI water to precipitate the dissolved lignin (ethanol organosolv lignin). Carbohydrate and Klason lignin contents of the untreated and organosolv pretreated Loblolly pine were determined by acid hydrolysis following the NREL standard procedure.25 Lignin and carbohydrate contents in the pretreatment effluent were determined using the NREL standard procedure after adjusting the pH with the addition of 72% sulfuric acid.26 Total furans content of the pretreatment liquid was quantified with UV–vis spectrophotometry using the method given in Martinez et al.32 The carbohydrate and acid soluble and insoluble (Klason) lignin contents for the untreated and organosolv-pretreated Loblolly pine, the pretreatment effluent and the solid
969
residue after enzymatic hydrolysis, are presented in Table 1. The higher Klason lignin content of the solid residue after enzymatic hydrolysis as compared to the pretreated wood may be due to the presence of greater amounts of extractives-like material in this sample. Analytical errors for Klason lignin and carbohydrate analysis were calculated from analysis of duplicate samples. Error values of 1% (standard deviation) and 10% (standard deviation) were determined for the Klason lignin and carbohydrate analysis, respectively. 3.3. Enzymatic hydrolysis of pretreated wood Commercial cellulase and b-glucosidase preparations (Celluclast 1.5 L and Novozym 188) were used as received. Along with their intended use, these commercial enzyme preparations also exhibit low levels of other enzymatic activity such as xylanse, mannase, and pectinase.33 The cellulase and b-glucosidase used in this study had xylanase activities of 1.16 birch xylan units/mg enzyme preparation and 0.09 birch xylan units/mg enzyme preparation, respectively. Xylanase activity was measured using birch xylan as the substrate.34 Both these enzymes showed no laccase activity (measured using the ABTS assay35). Organosolv-treated pine (weight equivalent to 2.0 g cellulose) was suspended in 100 mL of pH 4.8, 50 mM acetate buffer solution. Celluclast was used at a loading of 8 FPU/g cellulose and Novozym 188 at a loading of 16 IU/g cellulose. The reaction mixtures were incubated at 150 rpm, 50 °C, in a rotary shaker for 80 h and sampled periodically for glucose determination. Glucose concentrations in the aqueous phase samples were determined using the Megazyme D-glucose (glucose oxidase/peroxidase: GOPOD) assay kit. The errors in these measurements are ±4.8% (std. dev.) based on triplicate analysis. The cellulose to glucose conversion (%) at different times during the experiment is shown in Figure 1. The solid residue was thoroughly washed with DI water and air dried prior to analysis. 3.4. Cellulose extraction and NMR analysis Holocellulose was isolated from the untreated, organosolv treated and enzyme hydrolysis residue by repeatedly treating extractive-free wood with a mixture of acetic acid and sodium chlorite until the sample has very low (<2%) measured Klason lignin content. The holocellulose sample was further treated with 2.50 M HCl at 100 °C for 4 h in order to remove the hemicelluloses and facilitate NMR analyses following literature procedures.15 Cellulose was also extracted from the Loblolly pine chips without any milling to smaller sizes, in order to delineate the effects of mechanical milling on cellulose crystallinity. The cellulose samples were hydrated with DI water and a sample with 40% moisture content was used for NMR analyses. Hydrating the samples prior to NMR analyses has been shown to result in significant narrowing of signals in the spectra.15 The hydrated cellulose samples were packed in 4 mm diameter zirconium oxide rotors fitted with Kel-F caps. Solid-state CP/MAS 13C NMR experiments were performed on a Bruker Avance-400 spectrometer operating at a 13C frequency of 100.59 MHz. Glycine was used for the Hartman–Hahn matching procedure and as an external standard for the calibration of the chemical shift scale relative to tetramethylsilane (TMS). All the experiments were carried out at ambient temperature using a Bruker 4-mm MAS probe. Data processing was performed offline using the NUTS software (Acorn NMR Inc.). 3.5. Solid-state NMR of wood Extractive free (extracted with dichloromethane in a Soxhlet apparatus for 8 h), ground wood samples were packed in 4 mm diameter zirconium oxide rotors fitted with Kel-F caps and
970
P. Sannigrahi et al. / Carbohydrate Research 345 (2010) 965–970
solid-state CPMAS 13C NMR analysis was carried out as described above for cellulose except that the wood samples were not hydrated. 4. Conclusions The effects of organosolv pretreatment and enzymatic hydrolysis on Loblolly pine, a softwood biomass feedstock for bioethanol, have been studied using solid-state 13C CP/MAS NMR spectroscopy of the wood and cellulose isolated from before and after the pretreatments. NMR spectra of the wood reveal that hemicellulose and lignin are degraded during the organosolv pretreatment. This is accompanied by a decrease in cellulose crystallinity, as calculated from spectral analysis of cellulose NMR spectra. Enzymatic hydrolysis results in good cellulose to glucose conversion and an increase in the crystallinity of the residual cellulose, which is inferred to arise from the selective degradation of non-crystalline and para-crystalline cellulose. Acknowledgements The authors wish to acknowledge financial support from Chevron Technology Ventures for these studies. References 1. Ragauskas, A. J.; Williams, C. K.; Davison, B. H.; Britovsek, G.; Cairney, J.; Eckert, C. A.; Frederick, W. J., Jr.; Hallett, J. P.; Leak, D. J.; Liotta, C. L.; Mielenz, J. R.; Murphy, R.; Templer, R.; Tschaplinski, T. Science 2006, 311, 484–489. 2. Zhang, Y.-H. P.; Lynd, L. R. Biotechnol. Bioeng. 2004, 88, 797–824. 3. Bing, Y.; Wyman, C. Biofuels, Bioprod. Biorefin. 2008, 2, 26–40. 4. Mabee, W. A.; Gregg, D. J.; Arato, C.; Berlin, A.; Bura, R.; Gilkes, N.; Mirochnik, O.; Pan, X.; Pye, E. K.; Saddler, J. N. Appl. Biochem. Biotechnol. 2006, 129–132, 55– 70. 5. Pan, X.; Gilkes, N.; Kadla, J. F.; Pye, E. K.; Gregg, D. J.; Ehara, K.; Xie, D.; Lam, D.; Saddler, J. N. Biotechnol. Bioeng. 2006, 94, 851–861. 6. Pan, X.; Kadla, J. F.; Ehara, K.; Gilkes, N.; Saddler, J. N. J. Agric. Food Chem. 2006, 54, 5806–5813.
7. Pye, E. K.; Lora, J. H. TAPPI J. 1991, 74, 113–118. 8. Stockburger, P. TAPPI J. 1993, 76, 71–74. 9. Araque, E.; Parra, C.; Freer, J.; Contreras, D.; Rodriguez, J.; Mendonca, R.; Baeza, J. Enzyme Microb. Technol. 2008, 43, 214–219. 10. Pan, X.; Arato, C.; Gilkes, N.; Gregg, D. J.; Mabee, W. A.; Pye, E. K.; Xiao, Z. Z.; Zhang, X.; Saddler, J. N. Biotechnol. Bioeng. 2005, 90, 473–481. 11. Pan, X.; Xie, D.; Yu, R. W.; Saddler, J. N. Biotechnol. Bioeng. 2008, 101, 39–48. 12. Arato, C.; Pye, E. K.; Gjennestad, G. Appl. Biochem. Biotechnol. 2005, 121–124, 871–882. 13. Klemm, D.; Heublein, B.; Fink, H. P.; Bohn, A. Angew. Chem., Int. Ed. 2005, 44, 3358–3393. 14. Larsson, P. T.; Westermark, U.; Iverson, T. Carbohydr. Res. 1995, 278, 339–343. 15. Larsson, P. T.; Wickholm, K.; Iverson, T. Carbohydr. Res. 1997, 302, 19–25. 16. Lennholm, H.; Larsson, P. T.; Iverson, T. Carbohydr. Res. 1994, 261, 119–131. 17. Pu, Y.; Ziemer, C.; Ragauskas, A. J. Carbohydr. Res. 2006, 341, 591–597. 18. Wickholm, K.; Larsson, P. T.; Iverson, T. Carbohydr. Res. 1998, 312, 123–129. 19. Sannigrahi, P.; Ragauskas, A. J.; Miller, S. J. Bioenerg. Res. 2008, 1, 205–214. 20. Josefsson, T.; Lennholm, H.; Gellerstedt, G. Cellulose 2001, 8, 289–296. 21. Gollapalli, L. E.; Dale, B.; Rivers, D. M. Appl. Biochem. Biotechnol. 2002, 98–100, 23–25. 22. Lynd, L. R.; Weimer, P. J.; van Zyl, W. H.; Pretorius, I. S. Microbiol. Mol. Biol. Rev. 2002, 66, 506–577. 23. Mansfield, S. D.; Mooney, C.; Saddler, J. N. Biotechnol. Prog. 1999, 15, 804–816. 24. Cao, Y.; Tan, H. Enzyme Microb. Technol. 2005, 36, 314–317. 25. Sluiter, A.; Hames, B.; Ruiz, R.; Scarlata, C.; Sluiter, J.; Templeton, D.; Crocker, D. NREL/TP-510-42618, 2008. 26. Sluiter, A.; Hames, B.; Ruiz, R.; Scarlata, C.; Sluiter, J.; Templeton, D. NREL/TP510-42623, 2006. 27. Alesiani, M.; Proietti, F.; Capuani, S.; Maraviglia, B. Appl. Magn. Reson. 2005, 28, 1–8. 28. Maunu, S. L. Prog. Nucl. Magn. Reson. Spectrosc. 2002, 40, 151–174. 29. Larsson, P. T.; Hult, E.-L.; Wickholm, K.; Pettersson, E.; Iverson, T. Solid State Nucl. Magn. Reson. 1999, 15, 31–40. 30. Foston, M.; Hubbell, C. A.; Davis, M.; Ragauskas, A. J. Bioenerg. Res. 2009, 2, 193– 197. 31. Maunu, S. L.; Littia, T.; Kaulimaki, S.; Hortling, B.; Sundquist, J. Cellulose 2000, 7, 147–159. 32. Martinez, A.; Rodriguez, M. E.; York, S. W.; Preston, J. F.; Ingram, L. O. Biotechnol. Prog. 2000, 16, 637–641. 33. Berlin, A.; Maximenko, V.; Gilkes, N.; Saddler, J. Biotechnol. Bioeng. 2007, 97, 287–296. 34. Bailey, M. J.; Biely, P.; Poutanen, K. J. Biotechnol. 1992, 23, 257–270. 35. Bourbonnais, R.; Leech, D.; Paice, G. M. Biochim. Biophys. Acta 1998, 1379, 381– 390.
Contents lists available at ScienceDirect
Carbohydrate Research journal homepage: www.elsevier.com/locate/carres
Note
Effects of organosolv pretreatment and enzymatic hydrolysis on cellulose structure and crystallinity in Loblolly pine Poulomi Sannigrahi a, Stephen J. Miller b, Arthur J. Ragauskas a,* a b
School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332-0400, USA Chevron Energy Technology Company, 100 Chevron Way, Richmond, CA 94802, USA
a r t i c l e
i n f o
Article history: Received 6 November 2009 Received in revised form 6 February 2010 Accepted 11 February 2010 Available online 17 February 2010 Keywords: Loblolly pine Organosolv pretreatment Cellulose structure 13 C NMR spectroscopy Bioethanol
a b s t r a c t Ethanol organosolv pretreatment was performed on Loblolly pine to enhance the efficiency of enzymatic hydrolysis of cellulose to glucose. Solid-state 13C NMR spectroscopy coupled with line shape analysis was used to determine the structure and crystallinity of cellulose isolated from pretreated and enzymehydrolyzed Loblolly pine. The results indicate reduced crystallinity of the cellulose following the organosolv pretreatment, which renders the substrate easily hydrolyzable by cellulase. The degree of crystallinity increases and the relative proportion of para-crystalline and amorphous cellulose decreases after enzymatic hydrolysis, indicating preferential hydrolysis of these regions by cellulase. The structural and compositional changes in this material resulting from the organosolv pretreatment and cellulase enzyme hydrolysis of the pretreated wood were studied with solid-state CP/MAS 13C NMR spectroscopy. NMR spectra of the solid material before and after the treatments show that hemicelluloses and lignin are degraded during the organosolv pretreatment. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction Production of bioethanol from renewable lignocellulosic biomass can be accomplished by hydrolyzing plant polysaccharides to soluble hexose and pentose sugars, which are then microbiologically fermented.1 Cellulose, lignin and hemicellulose, the key components of lignocellulosic biomass, are closely associated with each other at the plant cell level. This close association, together with the partly crystalline nature of cellulose, reduces cellulose reactivity towards enzymatic hydrolysis in native biomass.2 Thus, some type of pretreatment of the native biomass is necessary to render the carbohydrate fraction amenable to enzymatic and microbial action.3 The ethanol organosolv process is among the pretreatment strategies currently being studied for the conversion of lignocellulosic biomass feedstocks.4–6 In this pretreatment, biomass is treated at high temperature in an aqueous ethanol solution with sulfuric acid as a delignification catalyst. Lignin removed during the pretreatment is solubilized by washing in ethanol and is recovered by precipitating in water. It was originally developed as the AlcellÒ pulping process for hardwoods.7,8 Recent studies focused on evaluating and optimizing this pretreatment for the delignification of softwood, hardwood and agricultural residues,5,9–11 have shown that the organosolv pretreatment yields readily hydrolyzable substrates with good glucose recovery and low lignin content.
* Corresponding author. Tel.: +1 404 894 9701; fax: +1 404 894 4778. E-mail address: [email protected] (A.J. Ragauskas). 0008-6215/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.carres.2010.02.010
In addition, the recovered lignin remains an interesting biomaterial in its own right.12 The ordered region of native cellulose is a mixture of two crystalline allomorphs, cellulose Ia and cellulose Ib. Cellulose Ia is the dominant form in bacterial and algal celluloses and cellulose Ib is the dominant form in higher plants.13 Solid-state 13C NMR of cellulose has been shown to be a suitable analytical technique to characterize these and several other forms of cellulose including para-crystalline cellulose, cellulose at accessible fibril surfaces (i.e., surfaces in contact with water) and amorphous cellulose at inaccessible fibril surfaces (i.e., fibril–fibril contact surfaces and surfaces resulting from distortions in the fibril interior).14–18 The effects of chemical pretreatment on the structure and crystallinity of cellulose in lignocellulosic biomass are a function of the pretreatment conditions. The two-step dilute acid pretreatment often employed for softwood feedstocks, has been reported to bring about an increase in cellulose crystallinity of Loblolly pine, by preferential degradation of the less ordered amorphous cellulose.19 Steam explosion pretreatment of aspen wood also exhibited an increase in the relative proportion of crystalline cellulose.20 In contrast, the ammonia fiber explosion (AFEX) pretreatment enhances enzymatic digestibility by decrystallization of cellulose.21 The effects of the ethanol organosolv pretreatment on cellulose structure and crystallinity are not fully explored. However, it has been shown that the degree of crystallinity increases with increasing severity of the organosolv pretreatment conditions.11 This has been attributed to the preferential degradation of the less structured forms of cellulose under harsher pretreatment conditions.
966
P. Sannigrahi et al. / Carbohydrate Research 345 (2010) 965–970
The relationship between cellulose structure and rates of enzymatic hydrolysis has been extensively studied,22,23 but is still a topic of debate in the literature. Cellulose hydrolysis rates mediated by fungal cellulases are typically 3–30 times faster for amorphous cellulose as compared to highly crystalline cellulose.22 Thus it is expected that crystallinity should increase over the course of cellulose hydrolysis as a result of preferential degradation of amorphous cellulose. While some studies support this hypothesis,17,24 others have reported that crystallinity does not increase during enzymatic hydrolysis.22,23 Pu et al.17 have shown that during cellulase treatment of fully bleached softwood kraft pulp, the cellulose Ia, para-crystalline and non-crystalline regions are rapidly hydrolyzed in the initial phase of hydrolysis leading to an increase in the cellulose crystallinity index. After this rapid initial phase, all cellulose polymorphs including the more energetically stable cellulose Ib, were found to be equally susceptible to enzymatic hydrolysis. In this study, the structure and crystallinity of cellulose extracted from untreated Loblolly pine has been compared to that obtained following ethanol organosolv pretreatment of this feedstock and its subsequent enzymatic hydrolysis by cellulase enzymes. The degree of crystallinity and ultrastructure of cellulose was determined using solid-state 13C CP/MAS NMR spectroscopy coupled with advanced line fitting analysis. Changes in the bulk chemical composition of the wood resulting from these treatments have been evaluated with solid-state 13C CP/MAS NMR spectroscopy. The results provide a view of the structural changes taking place in Loblolly pine, especially in the cellulose fraction, with organosolv pretreatment and enzymatic hydrolysis. 2. Results and discussion 2.1. Composition of untreated, pretreated and enzymehydrolyzed Loblolly pine The analysis of carbohydrates and lignin in the pre- and posttreated organosolv pretreated pine and the pretreatment liquids was accomplished following literature methods,25,26 and these results are summarized in Table 1. About 79% of the glucose in the untreated biomass is recovered in the solid phase after organosolv pretreatment. The loss of glucan was attributed, in part, to the hydrolysis of glucomannans and to some extent cellulose. The hydrolysis of polysaccharides, especially hemicelluloses has been well reported in the literature.10,11 As recently reported, the organosolv process results in the hydrolysis of lignin which solubilizes and releases lignin into the aqueous ethanolic phase.11 Solubilization of lignin in the pretreatment effluent results in extensive delignification. The solid residue after enzymatic hydrolysis is composed primarily of Klason lignin (Table 1). The liquid phase recovered from the pretreatment reactor and after washing the pretreated material, contains hemicellulose sugars, furans and soluble lignin (Table 1). Furans are produced by the degradation of carbohydrates under high temperature acidic conditions.
About 3% of the glucose in the starting material is also recovered in the pretreatment liquid. The carbohydrates not detected in the pretreated solid or liquid fractions, are inferred to have degraded to furans and other degradation products such as acetic acid.5,10 The formation of acetic acid and other minor acids including levulinic and succinic acids can also lower the pH of the cooking liquor during the pretreatment. 19% of the lignin in the untreated biomass is recovered as ethanol organosolv lignin (Table 1). Hydrolysis with cellulase results in the hydrolysis of 70% of the cellulose in the pretreated biomass to soluble glucose (Fig. 1). 2.2. Solid-state NMR of untreated, pretreated and enzymehydrolyzed wood Solid-state 13C NMR spectra of extractive free Loblolly pine before and after the organosolv and enzymatic treatments are presented in Figure 2. Peaks have been assigned based on the literature27,28 and are shown on the spectra. The peak at d 21 ppm, which can be attributed to the acetyl group of hemicelluloses, is seen only in the spectra from untreated Loblolly pine. This correlates to the solubilization of hemicelluloses in the effluent fraction during organosolv pretreatment. The peaks in the region between 60 and 105 ppm are predominantly due to the different carbons of cellulose. There is some spectral overlap with hemicelluloses in this region, thus pure cellulose needs to be isolated for determination of cellulose structure and crystallinity as presented in the next section. The signal at d 56 ppm corresponds to the aryl methoxyl carbons of lignin and a reduced intensity of this peak is also seen in the sample after the organosolv pretreatment. The signal in the region from 140 to 160 ppm is diagnostic of lignin aromatic carbons. As seen in Figure 2, the lignin peaks are less intense in the wood samples after the organosolv pretreatment. Since the solid residue remaining the cellulose enzymatic hydrolysis has a relatively large proportion of lignin, the peaks at 56 and 140–160 ppm are more pronounced in this spectrum. 2.3. Cellulose structure and crystallinity The solid-state 13C NMR spectra of cellulose from untreated, pretreated and enzyme-hydrolyzed Loblolly pine (Fig. 2) have been analyzed using line shape analysis based on the method outlined by Larsson et al.15,29 Signals from d 86 to 92 correspond to the C4 carbons from the crystalline forms of cellulose, as well as paracrystalline domains, whereas the broader upfield resonance from d 80 to 86 is assigned to less ordered or non-crystalline domains of cellulose (Fig. 3). The ratio of the area in the d 86–92 region to the total peak area from d 80–92 is designated as the crystallinity index (CrI). Typical errors for this analysis are 2% (standard deviation). The crystallinity index for the untreated Loblolly pine cellulose is 63%.19 This is comparable to cellulose crystallinity values obtained by NMR, from hybrid poplar (63%),30 but is higher than spruce (48%).31 Cellulose isolated from Loblolly pine wood chips (without milling to a smaller size) has a crystallinity index of
Table 1 Composition of the Loblolly pine feedstock and the solid material following organosolv pretreatment and enzymatic hydrolysisa (% Dry weight starting material)
Untreated Organosolv-treated wood Organosolv lignin Pretreatment liquid Enzyme hydrolysis solid residue a
Carbohydrates
Lignin
Arabinan
Galactan
Glucan
Mannan
Xylan
1.6 0.0
2.6 0.0
42.0 33.3
9.6 0.2
7.8 0.3
1.2 0.1
1.3 0.0
3.0 10.7
7.2 0.9
5.6 1.5
The pretreatment liquid is the combined effluent and wash solutions.
Furans
Acid insoluble
Acid soluble
Organosolv lignin
29.4 11.4
0.5 0.2
— — 5.4
16.3
12.4 0.0
2.2
—
967
P. Sannigrahi et al. / Carbohydrate Research 345 (2010) 965–970
80.0
Cellulose to glucose converson (%)
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0 0
10
20
30
40
50
60
70
80
90
Time (hours) Figure 1. Cellulose to glucose conversion (%) during enzymatic hydrolysis of organosolv-treated Loblolly pine.
Cellulose Lignin –OCH3 Lignin Aryl C
Organosolvenzyme
Organosolv
Hemicellulose R-COCH3 Untreated
190
170
150
130
110
90
80
70
60
50
40
30
20
10
0
-20
f1 (ppm) Figure 2. CP/MAS
13
C NMR spectrum of untreated, organosolv-treated and enzyme-hydrolyzed Loblolly pine.
52%, which shows that mechanical milling removes some of the less ordered or amorphous regions of cellulose. While effects of mechanical milling on cellulose crystallinity have not been reported, Larsson et al.29 observed a greater proportion of ordered forms of cellulose in spruce Kraft pulp as compared to chips from the same wood sample. This has been attributed to a combination of the preferential loss of disordered cellulose and the loss of hemicelluloses during pulping.29 It should be noted that in this study, the biomass used for cellulose isolation (from the untreated mate-
rial) and that used for the organosolv pretreatment were milled to the same extent, which should account for the effects of mechanical milling. The cellulose crystallinity index decreases to 53% in the cellulose from organosolv-treated wood (Table 2), which indicates that the pretreatment conditions used in this study are capable of decreasing the ordering of cellulosic fibers. These results are in contrast to those obtained from two-step dilute sulfuric acid pretreatment of this biomass feedstock.19 The lower crystallinity
968
P. Sannigrahi et al. / Carbohydrate Research 345 (2010) 965–970
6
HOH2C 4
O
5
O HO
2
1
OH
3
OH 1
HO O
O
HOH2C 90.0
ppm
85.0
Organosolv-enzyme
Organosolv C 2, 3, 5
90.0
ppm
85.0
C1
C4
C6
Untreated
110
100
90
80
70
60
50
ppm (t1) Figure 3. CP/MAS
13
C NMR spectrum of cellulose isolated from untreated, organosolv-treated and enzyme-hydrolyzed Loblolly pine.
Table 2 Peak assignments and results from the spectral fitting of cellulose 13C NMR spectra for untreated, organosolv-treated and enzyme-hydrolyzed Loblolly pine Assignments
Cellulose Ia Cellulose Ia+b para-Crystalline cellulose Cellulose Ib Accessible fibril surface Inaccessible fibril surface Accessible fibril surface Crystallinity indexa (%) a b c
Chemical shift (ppm)
Proportion (%)
Line type
Untreated
OSb
OSenzymec
90.0 88.8 88.5
0.1 30.7 24.8
0.5 11.1 12.5
5.7 46.7 0.0
Lorentz Lorentz Gauss
87.9 84.9
6.9 13.0
23.4 15.6
28.3 8.8
Lorentz Gauss
83.9
15.6
14.6
8.7
Gauss
83.4
8.9
22.4
1.7
Gauss
63
53
81
Crystallinity index = d86–92 ppm/d80–92 ppm. OS: Organosolv treated OS-enzyme: Organosolv treated and enzyme hydrolyzed.
index, coupled with delignification renders the organosolv-treated wood highly amenable to enzymatic hydrolysis. This is reflected in the high (70%) cellulose to glucose conversion during enzymatic hydrolysis of pretreated Loblolly pine (Fig. 1) despite the low enzyme dosage employed. Cellulose hydrolysis proceeds at a higher
rate during the initial phases of the experiment and becomes slower as the experiment proceeds. This can be attributed in part to gradual changes in ultrastructure,14 a decrease in enzyme activity by non-productive binding to lignin and product inhibition. The crystallinity index increases to 81% in cellulose isolated from the substrate after the enzymatic hydrolysis for 80 h. Thus the cellulase treatment appears to be selectively degrading the less ordered, amorphous and para-crystalline forms of cellulose. These results are in agreement with those of Pu et al.17 and support the hypothesis that cellulose crystallinity increases during enzymatic hydrolysis. The fraction of cellulose not hydrolyzed to glucose by cellulase is highly crystalline and recalcitrant to these enzymes. Crystalline regions of cellulose are considered to be more difficult to degrade than non-crystalline regions due to strong intermolecular hydrogen bonding between the cellulose chains. As the demand for lignocellulosic feedstock grows, enzymes capable of hydrolyzing crystalline cellulose need to be developed in order to achieve optimal utilization of available resources. Line shape analysis of the cellulose C-4 region (d 82–90) was performed by using Lorentzian lines for the crystalline region and Gaussian lines for the non-crystalline region.29 This analysis is based on a non-linear least squares fitting of 13C NMR spectra enabling calculation of the relative amounts of cellulose Ia, cellulose Ib, para-crystalline cellulose and cellulose at accessible and inaccessible fibril surfaces.15,18,29 The results of spectral fitting for the cellulose isolated from untreated, organosolv-treated and enzyme-hydrolyzed Loblolly pine are compared in Table 2. Typical
P. Sannigrahi et al. / Carbohydrate Research 345 (2010) 965–970
errors for this analysis with similar materials are ±7%. The main effects of the organosolv pretreatment on cellulose structure in Loblolly pine are: increase in accessible fibril surface (16.1%); increase in cellulose Ib (16.5%) and decrease in cellulose Ia+b and para-crystalline cellulose (Table 2). An increase in the relative proportion of cellulose Ib is among the most significant modifications observed in the crystalline region of cellulose. There is a concurrent decrease in the proportion of cellulose Ia+b. It may be inferred that a fraction of the thermodynamically meta-stable cellulose Ia is being converted to cellulose Ib under the thermal and chemical conditions of the organosolv pretreatment. An increase in cellulose Ib was also reported in cellulose after dilute acid pretreatment of Loblolly pine.19 In the non-crystalline region, there is an increase in the accessible fibril surface at d 83.4 ppm indicating possible improved enzyme accessibility of the pretreated cellulose. The cellulose isolated from the solid residue from enzymatic hydrolysis is mostly crystalline. As seen in Table 2, the relative proportions of ordered cellulose regions (cellulose Ia, cellulose Ib and cellulose Ia+b) increase by 45.7%, which is accompanied by a decrease in the para-crystalline and non-crystalline region of cellulose. These results corroborate previous hypotheses,22 that fungal-derived cellulases can hydrolyze the less-ordered regions of cellulose more readily than the crystalline regions with long-range order. 3. Experimental 3.1. Materials All chemical reagents used in this study were purchased from VWR International and used as received. The enzymes were obtained from Sigma–Aldrich (St. Louis, MO) and the D-glucose assay kit was purchased from Megazyme International (Ireland). Wood from a mature Loblolly pine (Pinus taeda), was obtained from the University of Georgia research plot in Baldwin County, GA. The wood was manually debarked and chipped using a mechanical chipper. A homogenized sample has been used for all analyses. The wood chips were stored at 5 °C during the course of this study. Prior to the organosolv treatment, the wet wood chips were milled using a Wiley mill to pass a 5 mm screen.
3.2. Organosolv pretreatment Loblolly pine sawdust was treated with aqueous ethanol, along with sulfuric acid as catalyst, using the methodology and conditions optimized for Lodgepole pine.11 Briefly, 100.0 g (by dry weight) of Loblolly pine sawdust were immersed in 65% ethanol/ water solution (700.0 mL), containing 1.1% sulfuric acid (by weight of wood) and treated in a Parr pressure reactor (3.8 L) at 170 °C for 1 h. The effluents were collected after condensation. The treated wood was washed three times with 300 mL, 65% ethanol at 60 °C and the washes combined with the effluent. The wood was then washed three times with de-ionized (DI) water at 60 °C and the washes discarded. The combined effluent and ethanol wash solution was mixed with three volumes of DI water to precipitate the dissolved lignin (ethanol organosolv lignin). Carbohydrate and Klason lignin contents of the untreated and organosolv pretreated Loblolly pine were determined by acid hydrolysis following the NREL standard procedure.25 Lignin and carbohydrate contents in the pretreatment effluent were determined using the NREL standard procedure after adjusting the pH with the addition of 72% sulfuric acid.26 Total furans content of the pretreatment liquid was quantified with UV–vis spectrophotometry using the method given in Martinez et al.32 The carbohydrate and acid soluble and insoluble (Klason) lignin contents for the untreated and organosolv-pretreated Loblolly pine, the pretreatment effluent and the solid
969
residue after enzymatic hydrolysis, are presented in Table 1. The higher Klason lignin content of the solid residue after enzymatic hydrolysis as compared to the pretreated wood may be due to the presence of greater amounts of extractives-like material in this sample. Analytical errors for Klason lignin and carbohydrate analysis were calculated from analysis of duplicate samples. Error values of 1% (standard deviation) and 10% (standard deviation) were determined for the Klason lignin and carbohydrate analysis, respectively. 3.3. Enzymatic hydrolysis of pretreated wood Commercial cellulase and b-glucosidase preparations (Celluclast 1.5 L and Novozym 188) were used as received. Along with their intended use, these commercial enzyme preparations also exhibit low levels of other enzymatic activity such as xylanse, mannase, and pectinase.33 The cellulase and b-glucosidase used in this study had xylanase activities of 1.16 birch xylan units/mg enzyme preparation and 0.09 birch xylan units/mg enzyme preparation, respectively. Xylanase activity was measured using birch xylan as the substrate.34 Both these enzymes showed no laccase activity (measured using the ABTS assay35). Organosolv-treated pine (weight equivalent to 2.0 g cellulose) was suspended in 100 mL of pH 4.8, 50 mM acetate buffer solution. Celluclast was used at a loading of 8 FPU/g cellulose and Novozym 188 at a loading of 16 IU/g cellulose. The reaction mixtures were incubated at 150 rpm, 50 °C, in a rotary shaker for 80 h and sampled periodically for glucose determination. Glucose concentrations in the aqueous phase samples were determined using the Megazyme D-glucose (glucose oxidase/peroxidase: GOPOD) assay kit. The errors in these measurements are ±4.8% (std. dev.) based on triplicate analysis. The cellulose to glucose conversion (%) at different times during the experiment is shown in Figure 1. The solid residue was thoroughly washed with DI water and air dried prior to analysis. 3.4. Cellulose extraction and NMR analysis Holocellulose was isolated from the untreated, organosolv treated and enzyme hydrolysis residue by repeatedly treating extractive-free wood with a mixture of acetic acid and sodium chlorite until the sample has very low (<2%) measured Klason lignin content. The holocellulose sample was further treated with 2.50 M HCl at 100 °C for 4 h in order to remove the hemicelluloses and facilitate NMR analyses following literature procedures.15 Cellulose was also extracted from the Loblolly pine chips without any milling to smaller sizes, in order to delineate the effects of mechanical milling on cellulose crystallinity. The cellulose samples were hydrated with DI water and a sample with 40% moisture content was used for NMR analyses. Hydrating the samples prior to NMR analyses has been shown to result in significant narrowing of signals in the spectra.15 The hydrated cellulose samples were packed in 4 mm diameter zirconium oxide rotors fitted with Kel-F caps. Solid-state CP/MAS 13C NMR experiments were performed on a Bruker Avance-400 spectrometer operating at a 13C frequency of 100.59 MHz. Glycine was used for the Hartman–Hahn matching procedure and as an external standard for the calibration of the chemical shift scale relative to tetramethylsilane (TMS). All the experiments were carried out at ambient temperature using a Bruker 4-mm MAS probe. Data processing was performed offline using the NUTS software (Acorn NMR Inc.). 3.5. Solid-state NMR of wood Extractive free (extracted with dichloromethane in a Soxhlet apparatus for 8 h), ground wood samples were packed in 4 mm diameter zirconium oxide rotors fitted with Kel-F caps and
970
P. Sannigrahi et al. / Carbohydrate Research 345 (2010) 965–970
solid-state CPMAS 13C NMR analysis was carried out as described above for cellulose except that the wood samples were not hydrated. 4. Conclusions The effects of organosolv pretreatment and enzymatic hydrolysis on Loblolly pine, a softwood biomass feedstock for bioethanol, have been studied using solid-state 13C CP/MAS NMR spectroscopy of the wood and cellulose isolated from before and after the pretreatments. NMR spectra of the wood reveal that hemicellulose and lignin are degraded during the organosolv pretreatment. This is accompanied by a decrease in cellulose crystallinity, as calculated from spectral analysis of cellulose NMR spectra. Enzymatic hydrolysis results in good cellulose to glucose conversion and an increase in the crystallinity of the residual cellulose, which is inferred to arise from the selective degradation of non-crystalline and para-crystalline cellulose. Acknowledgements The authors wish to acknowledge financial support from Chevron Technology Ventures for these studies. References 1. Ragauskas, A. J.; Williams, C. K.; Davison, B. H.; Britovsek, G.; Cairney, J.; Eckert, C. A.; Frederick, W. J., Jr.; Hallett, J. P.; Leak, D. J.; Liotta, C. L.; Mielenz, J. R.; Murphy, R.; Templer, R.; Tschaplinski, T. Science 2006, 311, 484–489. 2. Zhang, Y.-H. P.; Lynd, L. R. Biotechnol. Bioeng. 2004, 88, 797–824. 3. Bing, Y.; Wyman, C. Biofuels, Bioprod. Biorefin. 2008, 2, 26–40. 4. Mabee, W. A.; Gregg, D. J.; Arato, C.; Berlin, A.; Bura, R.; Gilkes, N.; Mirochnik, O.; Pan, X.; Pye, E. K.; Saddler, J. N. Appl. Biochem. Biotechnol. 2006, 129–132, 55– 70. 5. Pan, X.; Gilkes, N.; Kadla, J. F.; Pye, E. K.; Gregg, D. J.; Ehara, K.; Xie, D.; Lam, D.; Saddler, J. N. Biotechnol. Bioeng. 2006, 94, 851–861. 6. Pan, X.; Kadla, J. F.; Ehara, K.; Gilkes, N.; Saddler, J. N. J. Agric. Food Chem. 2006, 54, 5806–5813.
7. Pye, E. K.; Lora, J. H. TAPPI J. 1991, 74, 113–118. 8. Stockburger, P. TAPPI J. 1993, 76, 71–74. 9. Araque, E.; Parra, C.; Freer, J.; Contreras, D.; Rodriguez, J.; Mendonca, R.; Baeza, J. Enzyme Microb. Technol. 2008, 43, 214–219. 10. Pan, X.; Arato, C.; Gilkes, N.; Gregg, D. J.; Mabee, W. A.; Pye, E. K.; Xiao, Z. Z.; Zhang, X.; Saddler, J. N. Biotechnol. Bioeng. 2005, 90, 473–481. 11. Pan, X.; Xie, D.; Yu, R. W.; Saddler, J. N. Biotechnol. Bioeng. 2008, 101, 39–48. 12. Arato, C.; Pye, E. K.; Gjennestad, G. Appl. Biochem. Biotechnol. 2005, 121–124, 871–882. 13. Klemm, D.; Heublein, B.; Fink, H. P.; Bohn, A. Angew. Chem., Int. Ed. 2005, 44, 3358–3393. 14. Larsson, P. T.; Westermark, U.; Iverson, T. Carbohydr. Res. 1995, 278, 339–343. 15. Larsson, P. T.; Wickholm, K.; Iverson, T. Carbohydr. Res. 1997, 302, 19–25. 16. Lennholm, H.; Larsson, P. T.; Iverson, T. Carbohydr. Res. 1994, 261, 119–131. 17. Pu, Y.; Ziemer, C.; Ragauskas, A. J. Carbohydr. Res. 2006, 341, 591–597. 18. Wickholm, K.; Larsson, P. T.; Iverson, T. Carbohydr. Res. 1998, 312, 123–129. 19. Sannigrahi, P.; Ragauskas, A. J.; Miller, S. J. Bioenerg. Res. 2008, 1, 205–214. 20. Josefsson, T.; Lennholm, H.; Gellerstedt, G. Cellulose 2001, 8, 289–296. 21. Gollapalli, L. E.; Dale, B.; Rivers, D. M. Appl. Biochem. Biotechnol. 2002, 98–100, 23–25. 22. Lynd, L. R.; Weimer, P. J.; van Zyl, W. H.; Pretorius, I. S. Microbiol. Mol. Biol. Rev. 2002, 66, 506–577. 23. Mansfield, S. D.; Mooney, C.; Saddler, J. N. Biotechnol. Prog. 1999, 15, 804–816. 24. Cao, Y.; Tan, H. Enzyme Microb. Technol. 2005, 36, 314–317. 25. Sluiter, A.; Hames, B.; Ruiz, R.; Scarlata, C.; Sluiter, J.; Templeton, D.; Crocker, D. NREL/TP-510-42618, 2008. 26. Sluiter, A.; Hames, B.; Ruiz, R.; Scarlata, C.; Sluiter, J.; Templeton, D. NREL/TP510-42623, 2006. 27. Alesiani, M.; Proietti, F.; Capuani, S.; Maraviglia, B. Appl. Magn. Reson. 2005, 28, 1–8. 28. Maunu, S. L. Prog. Nucl. Magn. Reson. Spectrosc. 2002, 40, 151–174. 29. Larsson, P. T.; Hult, E.-L.; Wickholm, K.; Pettersson, E.; Iverson, T. Solid State Nucl. Magn. Reson. 1999, 15, 31–40. 30. Foston, M.; Hubbell, C. A.; Davis, M.; Ragauskas, A. J. Bioenerg. Res. 2009, 2, 193– 197. 31. Maunu, S. L.; Littia, T.; Kaulimaki, S.; Hortling, B.; Sundquist, J. Cellulose 2000, 7, 147–159. 32. Martinez, A.; Rodriguez, M. E.; York, S. W.; Preston, J. F.; Ingram, L. O. Biotechnol. Prog. 2000, 16, 637–641. 33. Berlin, A.; Maximenko, V.; Gilkes, N.; Saddler, J. Biotechnol. Bioeng. 2007, 97, 287–296. 34. Bailey, M. J.; Biely, P.; Poutanen, K. J. Biotechnol. 1992, 23, 257–270. 35. Bourbonnais, R.; Leech, D.; Paice, G. M. Biochim. Biophys. Acta 1998, 1379, 381– 390.