- Email: [email protected]
Biogas production and saccharification of Salix pretreated at different steam explosion conditions
Bioresource Technology 102 (2011) 7932–7936
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Biogas production and saccharification of Salix pretreated at different steam explosion conditions Svein J. Horn a,⇑, Maria M. Estevez b, Henrik K. Nielsen c, Roar Linjordet d, Vincent G.H. Eijsink a a
Department of Chemistry, Biotechnology and Food Science, Norwegian University of Life Sciences, P.O. Box 5003, N-1432 Ås, Norway Department of Mathematical Sciences and Technology, Norwegian University of Life Sciences, P.O. Box 5003, N-1432 Ås, Norway c Faculty of Engineering and Science, University of Agder, P.O. Box 509, N-4898 Grimstad, Norway d Bioforsk, Norwegian Institute for Agricultural and Environmental Research, N-1432 Ås, Norway b
a r t i c l e
i n f o
Article history: Received 9 March 2011 Received in revised form 30 May 2011 Accepted 11 June 2011 Available online 17 June 2011 Keywords: Biogas Methane Enzymatic hydrolysis Cellulase Biofuel
a b s t r a c t Different steam explosion conditions were applied to Salix chips and the effect of this pretreatment was evaluated by running both enzymatic hydrolysis and biogas tests. Total enzymatic release of glucose and xylose increased with pretreatment harshness, with maximum values being obtained after pretreatment for 10 min at 210 °C. Harsher pretreatment conditions did not increase glucose release, led to degradation of xylose and to formation of furfurals. Samples pretreated at 220 and 230 °C initially showed low production of biogas, probably because of inhibitors produced during the pretreatment, but the microbial community was able to adapt and showed high final biogas production. Interestingly, final biogas yields correlated well with sugar yields after enzymatic hydrolysis, suggesting that at least in some cases a 24 h enzymatic assay may be developed as a quick method to predict the effects of pretreatment of lignocellulosic biomass on biogas yields. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Lignocellulosic materials may be hydrolyzed and fermented to ethanol, or digested anaerobically to biogas. However, due to the recalcitrant nature of lignocelluloses and despite recent improvements in enzyme technology (Harris et al., 2010; Vaaje-Kolstad et al., 2010), efficient biochemical processing to biofuels requires a pretreatment step. One of the most efficient pre-treatment methods for lignocellulosic biomass is steam explosion (Ramos, 2003; Wyman et al., 2005), which involves heating the biomass at high temperature followed by mechanical disruption of the biomass fibers by a rapid pressure drop (explosion). Steam pretreatment of lignocellulosic materials has been applied for increasing both sugar yields (Horn and Eijsink, 2010; Kumar et al., 2010; Palmarola-Adrados et al., 2005; Sassner et al., 2005) and biogas yields (Bruni et al., 2010; Teghammar et al., 2010). Anaerobic biodegradability and methane production of a biomass resource is typically evaluated by running a microbial batch experiment (Hansen et al., 2004), to determine what is often referred to as the Biochemical Methane Potential (BMP). In such a batch test the biomass is degraded anaerobically to methane by a microbial inoculum that is most often collected at a local biogas plant. The test is laborious and time consuming, typically running
for 30 or more days. There have been some attempts to correlate compositional data with biogas yields, or using aerobic microbial tests to predict biogas yields (Lesteur et al., 2010; Schievano et al., 2008). None of these methods have been adopted as a standard to predict the biogas potential of a biomass, probably because of too low correlations (usually R2 < 0.80). Less laborious and time consuming tests to evaluate biogas potential would be useful. In this study, we have analyzed the effects of a wide range of steam explosion conditions on the digestibility of non-impregnated Salix, either by enzymes (saccharification) or by anaerobic digestion (biogas production). The results show trade-offs between cellulose accessibility on the one hand and sugar degradation and inhibitor formation on the other hand during steam explosion, and reveal potentially optimal steam explosion conditions for Salix. Furthermore, the results show the effects of varying steam explosion pretreatments on biogas yields and the kinetics of biogas formation during anaerobic fermentation. Interestingly, we observed a correlation between the yields of enzymatic saccharification and biogas yields.
2. Methods 2.1. Raw material
⇑ Corresponding author. Tel.: +47 64965907; fax: +47 64965901. E-mail address: [email protected] (S.J. Horn). 0960-8524/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2011.06.042
Shoots of willow (Salix viminalis ‘Christina’) were harvested after the second growing season in November 2009 in a 7 years
S.J. Horn et al. / Bioresource Technology 102 (2011) 7932–7936
7933
old short rotation coppice plantation near Grimstad in Southern Norway. The harvest was carried out manually with pruning shears and the shoots were immediately chopped into plastic bags with a standard wood disk chipper (nominal cutting length 7 mm). The dry matter content of the fresh chips was 48.4%, and the C/N ratio was 74.0.
according to Tappi standard T222 and UM250, respectively. Total nitrogen and carbon contents of solid substrates, used to estimate C/N ratios, were determined by Dumas combustion (AOAC 990.03). Merck Spectroquant analytical kits (Merck, Darmstadt, Germany) were used for COD (Kit No: 1.14555.0001) and Total-Nitrogen (Kit No: 1.14763.0001) determination of the inoculum.
2.2. Steam explosion
2.5.2. HPLC analysis HPLC analyses of reaction mixtures were run in-house using a Dionex Ultimate 3000 system (Dionex, Sunnyvale, CA, USA) set up with combined refractive index (for glucose, xylose and acetic acid) and UV (for furfural and HMF; 280 nm) detection. Samples were identified and quantified by running standards. The HPLC samples were prepared by diluting samples from the reactions 5fold with the mobile phase followed by centrifugation and filtration (0.2 lm Sarstedt Filtropur S). Samples (10 ll) were applied to a MetaCarb 87H column (50 4.6 mm precolumn and a 250 4.6 mm analytical column; Varian Inc., Lake Forest, CA, USA) conditioned at 85 °C with 0.2 ml min 1 5 mM H2SO4.
Steam explosion was carried out in a steam explosion unit (Cambi AS, Asker, Norway) at the Norwegian University of Life Sciences. 517 g of fresh Salix chips (250 g DM) was added to the 20 l pressure vessel in each treatment. The wood chips were pretreated at temperatures from 170 to 230 °C (7.9–28.0 bar) using intervals of 10 °C. The samples were kept in the reactor for 5, 10 or 15 min. Before each pretreatment the pressure vessel was preheated for 10 min at the desired temperature. The pretreated materials were stored at 4 °C until the hydrolysis and biogas experiments were carried out. 2.3. Enzymatic hydrolysis Hydrolysis was carried out in triplicate using 30 ml reaction volumes in 50 ml screw-capped centrifuge tubes that were horizontally shaken at 130 rpm, 50 °C. The pH in the hydrolysis reactions was adjusted by adding citrate–phosphate buffer, pH 5.0, to a final concentration of 50 mM. Reaction tubes were preheated at 50 °C before the enzymes were added. The substrate concentration in the tubes was 50 g DM/l and reactions were started by adding 400 ll Celluclast and 100 ll Novozym 188 (both from Novozymes, Bagsvaerd, Denmark) yielding an enzyme load of 20 FPU g/ substrate. 2.4. Biogas production The biogas potential of the steam treated fractions was tested by running anaerobic digestion in sealed batch flasks. The total volume of the flasks was 1.1 l and they contained 8 g of steam treated Salix, 500 ml water and 200 ml inoculum containing 15 g VS/l. Due to variations in water content after pretreatment the actual amount of Salix VS added to each bottle varied and was in the range 1.3–1.9 g. The flasks were incubated at 37 °C and shaken at 90 rpm. The inoculum was sludge from the anaerobic digester at Nordre Follo Wastewater Treatment Plant (Vinterbro, Norway). Prior to the experiments the inoculum was anaerobically incubated for 9 days at 37 °C to reduce its endogenous biogas production. The inoculum had a COD and total-N content of 3.65 and 0.85 g/l, respectively. The production of biogas was followed by twice a week measuring pressure increase in the headspace of the flasks using a digital pressure transducer (GMH 3161 Greisinger Electronic, Germany) with an incorporated needle that was injected into the septum cap. After each measurement the overpressure was released by penetrating the septum with a needle. The volume of biogas production was calculated from the measured pressure by using the ideal gas law. Biogas composition was determined by gas chromatography.
2.5.3. Gas composition Biogas composition was analyzed using a gas chromatograph (Perkin Elmer Autoanalyser System GC, Waltham, MA, USA) equipped with a thermal conductivity detector (TCD) and an Alltech STR 1 column. Carbon dioxide content was determined using a 1.83 m 3.18 mm (inside diameter) stainless-steel column packed with Porapak Q (80/100 mesh). Methane content was measured using a 1.83 m 3.18 mm (inside diameter) stainless steel column packed with molecular sieve 5A. The operational temperatures of injector, detector and column were kept at 200, 250 and 60 °C, respectively. Helium was used as a carrier gas at a flow rate of 60 ml/min. 2.5.4. Ash and dry matter For the determination of the percentage of dry matter, samples were weighed before and after drying in air at 105 °C for 16 h. Ash percentages were calculated by reweighing the samples after subsequent 16 h of incubation at 550 °C. 3. Results and discussion 3.1. Steam explosion Pretreatment of Salix chips was carried out using 13 different steam explosion conditions with temperatures ranging from 170 to 230 °C and residence times ranging from 5 to 15 min. At the harsher conditions, the pretreatment led to a relative increase in the contents of cellulose and lignin, and a relative reduction in the hemicellulose content (see Table 1). This is a typical effect of steam explosion (Ballesteros et al., 2004; Horn and Eijsink, 2010) which is caused by degradation of hemicellulose sugars and loss of volatile compounds in the outlet steam. The change in relative composition can also be attributed to condensation and incorporation into lignin of non-lignin components (Chua and Wayman, 1979).
2.5. Analysis
3.2. Enzymatic hydrolysis
2.5.1. Compositional analysis Analyses of the carbohydrate and lignin contents (see Table 1) in Salix samples before and after steam explosion (whole slurry) were carried out by Innventia (Stockholm, Sweden). The sugar compositions were determined using ion-chromatography following a two-step acid hydrolysis procedure according to Tappi standard T249. Insoluble (Klason) and soluble lignin were determined
The 13 different un-washed samples of steam treated Salix were hydrolyzed using commercial cellulase preparations. Fig. 1A shows the concentration of released glucose after 0, 4 and 24 h of enzymatic hydrolysis. The data show that maximal release of glucose was obtained for material treated at 210 °C for 10 min as well as for material derived from harsher pretreatment conditions. These results are in good accordance with the results of a previous study
7934
S.J. Horn et al. / Bioresource Technology 102 (2011) 7932–7936
Table 1 Compositional analysis of the Salix before and after steam explosion. Samplea
Arabinan
Galactan
Mannan
Xylan
Glucan
Klason lignin
Acid soluble lignin
Untreated 210–15b
0.9 0.2
0.9 0.7
1.4 1.2
11.9 6.0
36.2 45.8
25.3 31.9
2.9 3.2
a
Amounts are expressed as percentage of dry matter. The amount of carbohydrates was calculated using the masses of anhydrous sugars. Composition of the whole slurry after steam explosion with 15 min residence time at 210 °C. The theoretical maximum yield of glucose starting with this material at 50 g/ l DM is 25.4 g/l. b
A
Fig. 2. Concentrations of HMF (dark gray) and furfural (light gray) in the enzymatic hydrolysates of Fig. 1 (measured after 4 h). Samples are labeled as in Fig. 1.
B
3.3. Inhibitors Furfural and HMF are degradation products of xylose and glucose, respectively. Fig. 2 shows the concentration of these inhibitors in the 50 g DM/l hydrolysates. It is readily seen that most inhibitors are produced during the harshest pretreatments, as expected (Sassner et al., 2005). Acetate is released from the xylan fraction during steam explosion and during enzymatic hydrolysis, the latter being due to deacetylase activity in the commercial cellulase preparation (Hakulinen et al., 2000; Horn and Eijsink, 2010). Fig. 3 shows that for the harshest pretreatment conditions most of the acetate is released in the pretreatment (bars marked 0 h). However, for the material that has undergone less harsh pretreatments maximum acetate release requires enzymatic activity. Acetate is a microbial inhibitor but also a substrate for methane production (Ferry, 1992).
Fig. 1. Amount of glucose (panel A) and xylose (panel B) released during enzymatic hydrolysis of the pretreated samples after 0, 4 and 24 h. The reactions were carried out at pH 5.0 and 50 °C with 50 g DM/l. The labels at the X-axis identify the samples by the pretreatment conditions (temperature in °C–residence time in minutes). The theoretical maximum concentration of glucose for the material pretreated for 15 min at 210 °C is 25.4 g/l (Table 1).
by Sassner et al., who found the optimal steam explosion pretreatment conditions for non-impregnated Salix in terms of maximum enzymatic glucose release were 14 min at 210 °C (Sassner et al., 2005). Under the optimal conditions used in the present study (210 °C, 10 min and harsher), between 70% and 80% of the glucose was already released after 4 h. The profile for xylose release versus pretreatment conditions (Fig. 1B) is clearly different from glucose release. The highest xylose levels were found for material pretreated at 210 °C for 5 min. Harsher pretreatments led to lower xylose yields, probably because of xylose degradation (Sassner et al., 2005). The total sugar release after 24 h (glucose + xylose, data not shown) was highest for the material pretreated at 210 °C for 10 min, but all pretreatments at 210, 220 and 230 °C gave similar total sugar concentrations in the range of 20–21 g/l. The only exception was the 230 °C and 15 min sample which gave a lower yield of 18.8 g/l.
3.4. Biogas production Fig. 4 shows the production of biogas after 11 and 57 days. The results show that initial biogas production increases with the
Fig. 3. Amount of acetate released during the enzymatic hydrolysis reactions of Fig. 1. Samples are labeled as in Fig. 1.
S.J. Horn et al. / Bioresource Technology 102 (2011) 7932–7936
7935
produced by Hypocrea jecorina. The correlation found in the present study may be taken to indicate that fungal and bacterial cellulase systems are similar in terms of their ability to gain access to pretreated hardwood. It remains to be investigated whether this correlation holds for other types of lignocellulosic biomass such as softwood and herbaceous residues (e.g. corn stover). Other raw materials, particularly softwood, may yield different results since efficient steam explosion pretreatment of these materials requires acid impregnation of the biomass (Ramos, 2003; Soderstrom et al., 2003). The acid used is typically sulfuric acid, the presence of which may activate sulfate-reducing bacteria in biogas processes, resulting in decreased methane yields (Kristjansson et al., 1982). Fig. 4. Accumulated biogas production (STP) after 11 (light gray bars) and 57 (dark gray bars) days. Endogenous biogas production from the inoculum, measured in a parallel control fermentation without added substrate was subtracted from the data.
harshness of the pretreatment conditions up to 210 °C, whereas initial production levels were lower after harsher pretreatments (220 and 230 °C). This is probably due to inhibition of the microorganisms by compounds such as HMF, furfural (see Fig. 2) and phenolics produced during the steam pretreatment. However, the microbial community seems to be able to adapt to the inhibitors since after 57 days the biogas yields were rather similar for all the harsher conditions. For the pretreatments at 220 and 230 °C biogas yields correlated negatively with residence time in the pretreatment. The highest yields of biogas were around 440 ml/g VS corresponding to about 240 ml methane/g VS (STP). The final methane content of the biogas was similar and in the range 52.5–54.3% for all the samples. In Fig. 5 the total amounts of released sugars after 24 h of enzymatic hydrolysis (Fig. 1) are plotted versus biogas production after 57 days (Fig. 4). The plot shows that there is a good correlation (R2 = 0.91) between sugar yield and biogas production. This suggests that tests based on enzymatic hydrolysis could offer a quick method to predict the effects of pretreatment on biogas yields from lignocellulosic materials. Since the methane concentrations in the batch tests were very similar the correlation is also good for sugar yield and methane production (R2 = 0.89; Fig. 5). The amount of sugars released from the pretreated biomass is a measure of how accessible the biomass is for enzymatic degradation. The anaerobic microbial community produces extracellular enzymes to degrade the biomass, and the initial hydrolytic step is typically the rate limiting step in anaerobic degradation of lignocellulosics (Zverlov et al., 2010). Thus, a correlation between enzymatic hydrolyzability and microbial fermentation is not surprising. The commercial enzymes used in this study are fungal enzymes
Fig. 5. Correlation between sugar concentrations (xylose + glucose) achieved after 24 h of enzymatic hydrolysis (Fig. 1) and final methane (triangles) and biogas (inverted squares) yields after 57 days (Fig. 4).
4. Conclusions This study shows that steam explosion is an efficient pretreatment method for both biogas and sugar production from hardwood. So far, relatively little has been known about the effects of thermal pretreatments on the efficiency of anaerobic digestion; our data add to the increasing recent awareness that such pretreatments are useful for lignocellulosic substrates. Interestingly, the results presented above show a good correlation between enzymatic sugar release after 24 h and biogas yields after 57 days. If such a correlation also can be found for other types of lignocellulosic biomass an enzymatic assay may be developed as a standard method to predict biogas yields from lignocellulosic substrates. Acknowledgements This work was financially supported by the Norwegian Research Council Projects Nos. 193817 and 190877, as well as by Cambi AS. We thank Novozymes for supplying enzymes and Jon Fredrik Hanssen for help with the GC analysis. References Ballesteros, M., Oliva, J.M., Negro, M.J., Manzanares, P., Ballesteros, I., 2004. Ethanol from lignocellulosic materials by a simultaneous saccharification and fermentation process (SFS) with Kluyveromyces marxianus CECT 10875. Process Biochemistry 39, 1843–1848. Bruni, E., Jensen, A.P., Angelidaki, I., 2010. Steam treatment of digested biofibers for increasing biogas production. Bioresource Technology 101, 7668–7671. Chua, M.G.S., Wayman, M., 1979. Characterization of autohydrolysis aspen (populustremuloides) lignins. Part 1. Composition and molecular weight distribution of extracted autohydrolysis lignin. Canadian Journal of Chemistry-Revue Canadienne De Chimie 57, 1141–1149. Ferry, J.G., 1992. Biochemistry of methanogenesis. Critical Reviews in Biochemistry and Molecular Biology 27, 473–503. Hakulinen, N., Tenkanen, M., Rouvinen, J., 2000. Three-dimensional structure of the catalytic core of acetylxylan esterase from Trichoderma reesei: insights into the deacetylation mechanism. Journal of Structural Biology 132, 180–190. Hansen, T.L., Schmidt, J.E., Angelidaki, I., Marca, E., Jansen, J.L., Mosbaek, H., Christensen, T.H., 2004. Method for determination of methane potentials of solid organic waste. Waste Management 24, 393–400. Harris, P.V., Welner, D., McFarland, K.C., Re, E., Poulsen, J.C.N., Brown, K., Salbo, R., Ding, H.S., Vlasenko, E., Merino, S., Xu, F., Cherry, J., Larsen, S., Lo Leggio, L., 2010. Stimulation of lignocellulosic biomass hydrolysis by proteins of glycoside hydrolase family 61: structure and function of a large, enigmatic family. Biochemistry 49, 3305–3316. Horn, S.J., Eijsink, V.G.H., 2010. Enzymatic hydrolysis of steam-exploded hardwood using short processing times. Bioscience, Biotechnology, and Biochemistry 74, 1157–1163. Kristjansson, J.K., Schonheit, P., Thauer, R.K., 1982. Different Ks values for hydrogen of methanogenic bacteria and sulfate reducing bacteria: an explanation for the apparent inhibition of methanogenesis by sulfate. Archives of Microbiology 131, 278–282. Kumar, L., Chandra, R., Chung, P.A., Saddler, J., 2010. Can the same steam pretreatment conditions be used for most softwoods to achieve good, enzymatic hydrolysis and sugar yields? Bioresource Technology 101, 7827–7833. Lesteur, M., Bellon-Maurel, V., Gonzalez, C., Latrille, E., Roger, J.M., Junqua, G., Steyer, J.P., 2010. Alternative methods for determining anaerobic biodegradability: a review. Process Biochemistry 45, 431–440.
7936
S.J. Horn et al. / Bioresource Technology 102 (2011) 7932–7936
Palmarola-Adrados, B., Galbe, M., Zacchi, G., 2005. Pretreatment of barley husk for bioethanol production. Journal of Chemical Technology and Biotechnology 80, 85–91. Ramos, L.P., 2003. The chemistry involved in the steam treatment of lignocellulosic materials. Quimica Nova 26, 863–871. Sassner, P., Galbe, M., Zacchi, G., 2005. Steam pretreatment of Salix with and without SO2 impregnation for production of bioethanol. Applied Biochemistry and Biotechnology 121, 1101–1117. Schievano, A., Pognani, M., D’Imporzano, G., Adani, F., 2008. Predicting anaerobic biogasification potential of ingestates and digestates of a full-scale biogas plant using chemical and biological parameters. Bioresource Technology 99, 8112– 8117. Soderstrom, J., Pilcher, L., Galbe, M., Zacchi, G., 2003. Two-step steam pretreatment of softwood by dilute H2SO4 impregnation for ethanol production. Biomass and Bioenergy 24, 475–486.
Teghammar, A., Yngvesson, J., Lundin, M., Taherzadeh, M.J., Horvath, I.S., 2010. Pretreatment of paper tube residuals for improved biogas production. Bioresource Technology 101, 1206–1212. Vaaje-Kolstad, G., Westereng, B., Horn, S.J., Liu, Z.L., Zhai, H., Sorlie, M., Eijsink, V.G.H., 2010. An oxidative enzyme boosting the enzymatic conversion of recalcitrant polysaccharides. Science 330, 219–222. Wyman, C.E., Dale, B.E., Elander, R.T., Holtzapple, M., Ladisch, M.R., Lee, Y.Y., 2005. Comparative sugar recovery data from laboratory scale application of leading pretreatment technologies to corn stover. Bioresource Technology 96, 2026– 2032. Zverlov, V.V., Hiegl, W., Kock, D.E., Kellermann, J., Kollmeier, T., Schwarz, W.H., 2010. Hydrolytic bacteria in mesophilic and thermophilic degradation of plant biomass. Engineering in Life Sciences 10, 528–536.
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Biogas production and saccharification of Salix pretreated at different steam explosion conditions Svein J. Horn a,⇑, Maria M. Estevez b, Henrik K. Nielsen c, Roar Linjordet d, Vincent G.H. Eijsink a a
Department of Chemistry, Biotechnology and Food Science, Norwegian University of Life Sciences, P.O. Box 5003, N-1432 Ås, Norway Department of Mathematical Sciences and Technology, Norwegian University of Life Sciences, P.O. Box 5003, N-1432 Ås, Norway c Faculty of Engineering and Science, University of Agder, P.O. Box 509, N-4898 Grimstad, Norway d Bioforsk, Norwegian Institute for Agricultural and Environmental Research, N-1432 Ås, Norway b
a r t i c l e
i n f o
Article history: Received 9 March 2011 Received in revised form 30 May 2011 Accepted 11 June 2011 Available online 17 June 2011 Keywords: Biogas Methane Enzymatic hydrolysis Cellulase Biofuel
a b s t r a c t Different steam explosion conditions were applied to Salix chips and the effect of this pretreatment was evaluated by running both enzymatic hydrolysis and biogas tests. Total enzymatic release of glucose and xylose increased with pretreatment harshness, with maximum values being obtained after pretreatment for 10 min at 210 °C. Harsher pretreatment conditions did not increase glucose release, led to degradation of xylose and to formation of furfurals. Samples pretreated at 220 and 230 °C initially showed low production of biogas, probably because of inhibitors produced during the pretreatment, but the microbial community was able to adapt and showed high final biogas production. Interestingly, final biogas yields correlated well with sugar yields after enzymatic hydrolysis, suggesting that at least in some cases a 24 h enzymatic assay may be developed as a quick method to predict the effects of pretreatment of lignocellulosic biomass on biogas yields. Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Lignocellulosic materials may be hydrolyzed and fermented to ethanol, or digested anaerobically to biogas. However, due to the recalcitrant nature of lignocelluloses and despite recent improvements in enzyme technology (Harris et al., 2010; Vaaje-Kolstad et al., 2010), efficient biochemical processing to biofuels requires a pretreatment step. One of the most efficient pre-treatment methods for lignocellulosic biomass is steam explosion (Ramos, 2003; Wyman et al., 2005), which involves heating the biomass at high temperature followed by mechanical disruption of the biomass fibers by a rapid pressure drop (explosion). Steam pretreatment of lignocellulosic materials has been applied for increasing both sugar yields (Horn and Eijsink, 2010; Kumar et al., 2010; Palmarola-Adrados et al., 2005; Sassner et al., 2005) and biogas yields (Bruni et al., 2010; Teghammar et al., 2010). Anaerobic biodegradability and methane production of a biomass resource is typically evaluated by running a microbial batch experiment (Hansen et al., 2004), to determine what is often referred to as the Biochemical Methane Potential (BMP). In such a batch test the biomass is degraded anaerobically to methane by a microbial inoculum that is most often collected at a local biogas plant. The test is laborious and time consuming, typically running
for 30 or more days. There have been some attempts to correlate compositional data with biogas yields, or using aerobic microbial tests to predict biogas yields (Lesteur et al., 2010; Schievano et al., 2008). None of these methods have been adopted as a standard to predict the biogas potential of a biomass, probably because of too low correlations (usually R2 < 0.80). Less laborious and time consuming tests to evaluate biogas potential would be useful. In this study, we have analyzed the effects of a wide range of steam explosion conditions on the digestibility of non-impregnated Salix, either by enzymes (saccharification) or by anaerobic digestion (biogas production). The results show trade-offs between cellulose accessibility on the one hand and sugar degradation and inhibitor formation on the other hand during steam explosion, and reveal potentially optimal steam explosion conditions for Salix. Furthermore, the results show the effects of varying steam explosion pretreatments on biogas yields and the kinetics of biogas formation during anaerobic fermentation. Interestingly, we observed a correlation between the yields of enzymatic saccharification and biogas yields.
2. Methods 2.1. Raw material
⇑ Corresponding author. Tel.: +47 64965907; fax: +47 64965901. E-mail address: [email protected] (S.J. Horn). 0960-8524/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2011.06.042
Shoots of willow (Salix viminalis ‘Christina’) were harvested after the second growing season in November 2009 in a 7 years
S.J. Horn et al. / Bioresource Technology 102 (2011) 7932–7936
7933
old short rotation coppice plantation near Grimstad in Southern Norway. The harvest was carried out manually with pruning shears and the shoots were immediately chopped into plastic bags with a standard wood disk chipper (nominal cutting length 7 mm). The dry matter content of the fresh chips was 48.4%, and the C/N ratio was 74.0.
according to Tappi standard T222 and UM250, respectively. Total nitrogen and carbon contents of solid substrates, used to estimate C/N ratios, were determined by Dumas combustion (AOAC 990.03). Merck Spectroquant analytical kits (Merck, Darmstadt, Germany) were used for COD (Kit No: 1.14555.0001) and Total-Nitrogen (Kit No: 1.14763.0001) determination of the inoculum.
2.2. Steam explosion
2.5.2. HPLC analysis HPLC analyses of reaction mixtures were run in-house using a Dionex Ultimate 3000 system (Dionex, Sunnyvale, CA, USA) set up with combined refractive index (for glucose, xylose and acetic acid) and UV (for furfural and HMF; 280 nm) detection. Samples were identified and quantified by running standards. The HPLC samples were prepared by diluting samples from the reactions 5fold with the mobile phase followed by centrifugation and filtration (0.2 lm Sarstedt Filtropur S). Samples (10 ll) were applied to a MetaCarb 87H column (50 4.6 mm precolumn and a 250 4.6 mm analytical column; Varian Inc., Lake Forest, CA, USA) conditioned at 85 °C with 0.2 ml min 1 5 mM H2SO4.
Steam explosion was carried out in a steam explosion unit (Cambi AS, Asker, Norway) at the Norwegian University of Life Sciences. 517 g of fresh Salix chips (250 g DM) was added to the 20 l pressure vessel in each treatment. The wood chips were pretreated at temperatures from 170 to 230 °C (7.9–28.0 bar) using intervals of 10 °C. The samples were kept in the reactor for 5, 10 or 15 min. Before each pretreatment the pressure vessel was preheated for 10 min at the desired temperature. The pretreated materials were stored at 4 °C until the hydrolysis and biogas experiments were carried out. 2.3. Enzymatic hydrolysis Hydrolysis was carried out in triplicate using 30 ml reaction volumes in 50 ml screw-capped centrifuge tubes that were horizontally shaken at 130 rpm, 50 °C. The pH in the hydrolysis reactions was adjusted by adding citrate–phosphate buffer, pH 5.0, to a final concentration of 50 mM. Reaction tubes were preheated at 50 °C before the enzymes were added. The substrate concentration in the tubes was 50 g DM/l and reactions were started by adding 400 ll Celluclast and 100 ll Novozym 188 (both from Novozymes, Bagsvaerd, Denmark) yielding an enzyme load of 20 FPU g/ substrate. 2.4. Biogas production The biogas potential of the steam treated fractions was tested by running anaerobic digestion in sealed batch flasks. The total volume of the flasks was 1.1 l and they contained 8 g of steam treated Salix, 500 ml water and 200 ml inoculum containing 15 g VS/l. Due to variations in water content after pretreatment the actual amount of Salix VS added to each bottle varied and was in the range 1.3–1.9 g. The flasks were incubated at 37 °C and shaken at 90 rpm. The inoculum was sludge from the anaerobic digester at Nordre Follo Wastewater Treatment Plant (Vinterbro, Norway). Prior to the experiments the inoculum was anaerobically incubated for 9 days at 37 °C to reduce its endogenous biogas production. The inoculum had a COD and total-N content of 3.65 and 0.85 g/l, respectively. The production of biogas was followed by twice a week measuring pressure increase in the headspace of the flasks using a digital pressure transducer (GMH 3161 Greisinger Electronic, Germany) with an incorporated needle that was injected into the septum cap. After each measurement the overpressure was released by penetrating the septum with a needle. The volume of biogas production was calculated from the measured pressure by using the ideal gas law. Biogas composition was determined by gas chromatography.
2.5.3. Gas composition Biogas composition was analyzed using a gas chromatograph (Perkin Elmer Autoanalyser System GC, Waltham, MA, USA) equipped with a thermal conductivity detector (TCD) and an Alltech STR 1 column. Carbon dioxide content was determined using a 1.83 m 3.18 mm (inside diameter) stainless-steel column packed with Porapak Q (80/100 mesh). Methane content was measured using a 1.83 m 3.18 mm (inside diameter) stainless steel column packed with molecular sieve 5A. The operational temperatures of injector, detector and column were kept at 200, 250 and 60 °C, respectively. Helium was used as a carrier gas at a flow rate of 60 ml/min. 2.5.4. Ash and dry matter For the determination of the percentage of dry matter, samples were weighed before and after drying in air at 105 °C for 16 h. Ash percentages were calculated by reweighing the samples after subsequent 16 h of incubation at 550 °C. 3. Results and discussion 3.1. Steam explosion Pretreatment of Salix chips was carried out using 13 different steam explosion conditions with temperatures ranging from 170 to 230 °C and residence times ranging from 5 to 15 min. At the harsher conditions, the pretreatment led to a relative increase in the contents of cellulose and lignin, and a relative reduction in the hemicellulose content (see Table 1). This is a typical effect of steam explosion (Ballesteros et al., 2004; Horn and Eijsink, 2010) which is caused by degradation of hemicellulose sugars and loss of volatile compounds in the outlet steam. The change in relative composition can also be attributed to condensation and incorporation into lignin of non-lignin components (Chua and Wayman, 1979).
2.5. Analysis
3.2. Enzymatic hydrolysis
2.5.1. Compositional analysis Analyses of the carbohydrate and lignin contents (see Table 1) in Salix samples before and after steam explosion (whole slurry) were carried out by Innventia (Stockholm, Sweden). The sugar compositions were determined using ion-chromatography following a two-step acid hydrolysis procedure according to Tappi standard T249. Insoluble (Klason) and soluble lignin were determined
The 13 different un-washed samples of steam treated Salix were hydrolyzed using commercial cellulase preparations. Fig. 1A shows the concentration of released glucose after 0, 4 and 24 h of enzymatic hydrolysis. The data show that maximal release of glucose was obtained for material treated at 210 °C for 10 min as well as for material derived from harsher pretreatment conditions. These results are in good accordance with the results of a previous study
7934
S.J. Horn et al. / Bioresource Technology 102 (2011) 7932–7936
Table 1 Compositional analysis of the Salix before and after steam explosion. Samplea
Arabinan
Galactan
Mannan
Xylan
Glucan
Klason lignin
Acid soluble lignin
Untreated 210–15b
0.9 0.2
0.9 0.7
1.4 1.2
11.9 6.0
36.2 45.8
25.3 31.9
2.9 3.2
a
Amounts are expressed as percentage of dry matter. The amount of carbohydrates was calculated using the masses of anhydrous sugars. Composition of the whole slurry after steam explosion with 15 min residence time at 210 °C. The theoretical maximum yield of glucose starting with this material at 50 g/ l DM is 25.4 g/l. b
A
Fig. 2. Concentrations of HMF (dark gray) and furfural (light gray) in the enzymatic hydrolysates of Fig. 1 (measured after 4 h). Samples are labeled as in Fig. 1.
B
3.3. Inhibitors Furfural and HMF are degradation products of xylose and glucose, respectively. Fig. 2 shows the concentration of these inhibitors in the 50 g DM/l hydrolysates. It is readily seen that most inhibitors are produced during the harshest pretreatments, as expected (Sassner et al., 2005). Acetate is released from the xylan fraction during steam explosion and during enzymatic hydrolysis, the latter being due to deacetylase activity in the commercial cellulase preparation (Hakulinen et al., 2000; Horn and Eijsink, 2010). Fig. 3 shows that for the harshest pretreatment conditions most of the acetate is released in the pretreatment (bars marked 0 h). However, for the material that has undergone less harsh pretreatments maximum acetate release requires enzymatic activity. Acetate is a microbial inhibitor but also a substrate for methane production (Ferry, 1992).
Fig. 1. Amount of glucose (panel A) and xylose (panel B) released during enzymatic hydrolysis of the pretreated samples after 0, 4 and 24 h. The reactions were carried out at pH 5.0 and 50 °C with 50 g DM/l. The labels at the X-axis identify the samples by the pretreatment conditions (temperature in °C–residence time in minutes). The theoretical maximum concentration of glucose for the material pretreated for 15 min at 210 °C is 25.4 g/l (Table 1).
by Sassner et al., who found the optimal steam explosion pretreatment conditions for non-impregnated Salix in terms of maximum enzymatic glucose release were 14 min at 210 °C (Sassner et al., 2005). Under the optimal conditions used in the present study (210 °C, 10 min and harsher), between 70% and 80% of the glucose was already released after 4 h. The profile for xylose release versus pretreatment conditions (Fig. 1B) is clearly different from glucose release. The highest xylose levels were found for material pretreated at 210 °C for 5 min. Harsher pretreatments led to lower xylose yields, probably because of xylose degradation (Sassner et al., 2005). The total sugar release after 24 h (glucose + xylose, data not shown) was highest for the material pretreated at 210 °C for 10 min, but all pretreatments at 210, 220 and 230 °C gave similar total sugar concentrations in the range of 20–21 g/l. The only exception was the 230 °C and 15 min sample which gave a lower yield of 18.8 g/l.
3.4. Biogas production Fig. 4 shows the production of biogas after 11 and 57 days. The results show that initial biogas production increases with the
Fig. 3. Amount of acetate released during the enzymatic hydrolysis reactions of Fig. 1. Samples are labeled as in Fig. 1.
S.J. Horn et al. / Bioresource Technology 102 (2011) 7932–7936
7935
produced by Hypocrea jecorina. The correlation found in the present study may be taken to indicate that fungal and bacterial cellulase systems are similar in terms of their ability to gain access to pretreated hardwood. It remains to be investigated whether this correlation holds for other types of lignocellulosic biomass such as softwood and herbaceous residues (e.g. corn stover). Other raw materials, particularly softwood, may yield different results since efficient steam explosion pretreatment of these materials requires acid impregnation of the biomass (Ramos, 2003; Soderstrom et al., 2003). The acid used is typically sulfuric acid, the presence of which may activate sulfate-reducing bacteria in biogas processes, resulting in decreased methane yields (Kristjansson et al., 1982). Fig. 4. Accumulated biogas production (STP) after 11 (light gray bars) and 57 (dark gray bars) days. Endogenous biogas production from the inoculum, measured in a parallel control fermentation without added substrate was subtracted from the data.
harshness of the pretreatment conditions up to 210 °C, whereas initial production levels were lower after harsher pretreatments (220 and 230 °C). This is probably due to inhibition of the microorganisms by compounds such as HMF, furfural (see Fig. 2) and phenolics produced during the steam pretreatment. However, the microbial community seems to be able to adapt to the inhibitors since after 57 days the biogas yields were rather similar for all the harsher conditions. For the pretreatments at 220 and 230 °C biogas yields correlated negatively with residence time in the pretreatment. The highest yields of biogas were around 440 ml/g VS corresponding to about 240 ml methane/g VS (STP). The final methane content of the biogas was similar and in the range 52.5–54.3% for all the samples. In Fig. 5 the total amounts of released sugars after 24 h of enzymatic hydrolysis (Fig. 1) are plotted versus biogas production after 57 days (Fig. 4). The plot shows that there is a good correlation (R2 = 0.91) between sugar yield and biogas production. This suggests that tests based on enzymatic hydrolysis could offer a quick method to predict the effects of pretreatment on biogas yields from lignocellulosic materials. Since the methane concentrations in the batch tests were very similar the correlation is also good for sugar yield and methane production (R2 = 0.89; Fig. 5). The amount of sugars released from the pretreated biomass is a measure of how accessible the biomass is for enzymatic degradation. The anaerobic microbial community produces extracellular enzymes to degrade the biomass, and the initial hydrolytic step is typically the rate limiting step in anaerobic degradation of lignocellulosics (Zverlov et al., 2010). Thus, a correlation between enzymatic hydrolyzability and microbial fermentation is not surprising. The commercial enzymes used in this study are fungal enzymes
Fig. 5. Correlation between sugar concentrations (xylose + glucose) achieved after 24 h of enzymatic hydrolysis (Fig. 1) and final methane (triangles) and biogas (inverted squares) yields after 57 days (Fig. 4).
4. Conclusions This study shows that steam explosion is an efficient pretreatment method for both biogas and sugar production from hardwood. So far, relatively little has been known about the effects of thermal pretreatments on the efficiency of anaerobic digestion; our data add to the increasing recent awareness that such pretreatments are useful for lignocellulosic substrates. Interestingly, the results presented above show a good correlation between enzymatic sugar release after 24 h and biogas yields after 57 days. If such a correlation also can be found for other types of lignocellulosic biomass an enzymatic assay may be developed as a standard method to predict biogas yields from lignocellulosic substrates. Acknowledgements This work was financially supported by the Norwegian Research Council Projects Nos. 193817 and 190877, as well as by Cambi AS. We thank Novozymes for supplying enzymes and Jon Fredrik Hanssen for help with the GC analysis. References Ballesteros, M., Oliva, J.M., Negro, M.J., Manzanares, P., Ballesteros, I., 2004. Ethanol from lignocellulosic materials by a simultaneous saccharification and fermentation process (SFS) with Kluyveromyces marxianus CECT 10875. Process Biochemistry 39, 1843–1848. Bruni, E., Jensen, A.P., Angelidaki, I., 2010. Steam treatment of digested biofibers for increasing biogas production. Bioresource Technology 101, 7668–7671. Chua, M.G.S., Wayman, M., 1979. Characterization of autohydrolysis aspen (populustremuloides) lignins. Part 1. Composition and molecular weight distribution of extracted autohydrolysis lignin. Canadian Journal of Chemistry-Revue Canadienne De Chimie 57, 1141–1149. Ferry, J.G., 1992. Biochemistry of methanogenesis. Critical Reviews in Biochemistry and Molecular Biology 27, 473–503. Hakulinen, N., Tenkanen, M., Rouvinen, J., 2000. Three-dimensional structure of the catalytic core of acetylxylan esterase from Trichoderma reesei: insights into the deacetylation mechanism. Journal of Structural Biology 132, 180–190. Hansen, T.L., Schmidt, J.E., Angelidaki, I., Marca, E., Jansen, J.L., Mosbaek, H., Christensen, T.H., 2004. Method for determination of methane potentials of solid organic waste. Waste Management 24, 393–400. Harris, P.V., Welner, D., McFarland, K.C., Re, E., Poulsen, J.C.N., Brown, K., Salbo, R., Ding, H.S., Vlasenko, E., Merino, S., Xu, F., Cherry, J., Larsen, S., Lo Leggio, L., 2010. Stimulation of lignocellulosic biomass hydrolysis by proteins of glycoside hydrolase family 61: structure and function of a large, enigmatic family. Biochemistry 49, 3305–3316. Horn, S.J., Eijsink, V.G.H., 2010. Enzymatic hydrolysis of steam-exploded hardwood using short processing times. Bioscience, Biotechnology, and Biochemistry 74, 1157–1163. Kristjansson, J.K., Schonheit, P., Thauer, R.K., 1982. Different Ks values for hydrogen of methanogenic bacteria and sulfate reducing bacteria: an explanation for the apparent inhibition of methanogenesis by sulfate. Archives of Microbiology 131, 278–282. Kumar, L., Chandra, R., Chung, P.A., Saddler, J., 2010. Can the same steam pretreatment conditions be used for most softwoods to achieve good, enzymatic hydrolysis and sugar yields? Bioresource Technology 101, 7827–7833. Lesteur, M., Bellon-Maurel, V., Gonzalez, C., Latrille, E., Roger, J.M., Junqua, G., Steyer, J.P., 2010. Alternative methods for determining anaerobic biodegradability: a review. Process Biochemistry 45, 431–440.
7936
S.J. Horn et al. / Bioresource Technology 102 (2011) 7932–7936
Palmarola-Adrados, B., Galbe, M., Zacchi, G., 2005. Pretreatment of barley husk for bioethanol production. Journal of Chemical Technology and Biotechnology 80, 85–91. Ramos, L.P., 2003. The chemistry involved in the steam treatment of lignocellulosic materials. Quimica Nova 26, 863–871. Sassner, P., Galbe, M., Zacchi, G., 2005. Steam pretreatment of Salix with and without SO2 impregnation for production of bioethanol. Applied Biochemistry and Biotechnology 121, 1101–1117. Schievano, A., Pognani, M., D’Imporzano, G., Adani, F., 2008. Predicting anaerobic biogasification potential of ingestates and digestates of a full-scale biogas plant using chemical and biological parameters. Bioresource Technology 99, 8112– 8117. Soderstrom, J., Pilcher, L., Galbe, M., Zacchi, G., 2003. Two-step steam pretreatment of softwood by dilute H2SO4 impregnation for ethanol production. Biomass and Bioenergy 24, 475–486.
Teghammar, A., Yngvesson, J., Lundin, M., Taherzadeh, M.J., Horvath, I.S., 2010. Pretreatment of paper tube residuals for improved biogas production. Bioresource Technology 101, 1206–1212. Vaaje-Kolstad, G., Westereng, B., Horn, S.J., Liu, Z.L., Zhai, H., Sorlie, M., Eijsink, V.G.H., 2010. An oxidative enzyme boosting the enzymatic conversion of recalcitrant polysaccharides. Science 330, 219–222. Wyman, C.E., Dale, B.E., Elander, R.T., Holtzapple, M., Ladisch, M.R., Lee, Y.Y., 2005. Comparative sugar recovery data from laboratory scale application of leading pretreatment technologies to corn stover. Bioresource Technology 96, 2026– 2032. Zverlov, V.V., Hiegl, W., Kock, D.E., Kellermann, J., Kollmeier, T., Schwarz, W.H., 2010. Hydrolytic bacteria in mesophilic and thermophilic degradation of plant biomass. Engineering in Life Sciences 10, 528–536.