Bioresource Technology 182 (2015) 245–250
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Chemical characteristics and enzymatic saccharification of lignocellulosic biomass treated using high-temperature saturated steam: Comparison of softwood and hardwood Chikako Asada, Chizuru Sasaki, Takeshi Hirano, Yoshitoshi Nakamura ⇑ Department of Life System, Institute of Technology and Science, The University of Tokushima, 2-1 Minamijosanjima-cho, Tokushima 770-8506, Japan
h i g h l i g h t s High temperature saturated steam treatment of cedar and beech was evaluated. Steam treatment followed by milling treatment improved enzymatic saccharification. Saccharification rate of steam-treated beech with milling treatment was 94%. Necessity of milling treatment after steam treatment is dependent on wood species.
a r t i c l e
i n f o
Article history: Received 20 November 2014 Received in revised form 1 February 2015 Accepted 2 February 2015 Available online 11 February 2015 Keywords: High-temperature saturated steam Chemical characteristics Enzymatic saccharification Softwood Hardwood
a b s t r a c t This study investigated the effect of high-temperature saturated steam treatments on the chemical characteristics and enzymatic saccharification of softwood and hardwood. The weight loss and chemical modification of cedar and beech wood pieces treated at 25, 35, and 45 atm for 5 min were determined. Fourier transform infrared and X-ray diffraction analyses indicated that solubilization and removal of hemicellulose and lignin occurred by the steam treatment. The milling treatment of steam-treated wood enhanced its enzymatic saccharification. Maximum enzymatic saccharification (i.e., 94% saccharification rate of cellulose) was obtained using steam-treated beech at 35 atm for 5 min followed by milling treatment for 1 min. However, the necessity of the milling treatment for efficient enzymatic saccharification is dependent on the wood species. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Primary energy self-sufficiency in Japan is only 4%, and this value is much lower than other developed countries. Even when nuclear power generation is considered as a domestically produced energy, the rate slightly increases to 17%, implying that Japan is hugely dependent on imported energy (approximately 80%). Therefore, the use of renewable energy sources such as biomass has been considered as an alternative to fossil fuels in Japan. For the significant production of domestic biofuels, the development of a method to convert lignocellulosic biomass such as wood, bamboo, and agricultural waste (such as rice straw) into ethanol and useful materials is desirable because it would not compete with food production. However, because cellulosic components (i.e., substrates for ethanol production) are strongly covered with lignin in the lignocellulosic biomass, pretreatment (delignification) ⇑ Corresponding author. Tel.: +81 88 656 7518; fax: +81 88 656 9071. E-mail address: [email protected] (Y. Nakamura). http://dx.doi.org/10.1016/j.biortech.2015.02.005 0960-8524/Ó 2015 Elsevier Ltd. All rights reserved.
is necessary for efficient enzymatic saccharification followed by alcohol fermentation to increase the accessibility of cellulose to cellulases (Karp et al., 2013). Various pretreatment methods have been developed and applied to target the conversion of biomass to biofuel including: acid treatment (Torget et al., 1988), alkaline treatment (Ucar, 1990), ammonia fiber expansion (Dale et al., 1996), organosolv treatment (Bonn et al., 1987), liquid hot water treatment (Sreenath et al., 1999), microwave irradiation (Ooshima et al., 1984), superheated steam treatment (Bahrin et al., 2012), and steam explosion (Ramos et al., 1992). Steam explosion, introduced as a biomass pretreatment process by Mason (1926), has become one of the simplest and environmentally friendly techniques and is a novel pretreatment method with a high efficiency at breaking lignin structures to release cellulose and hemicellulose from lignocellulosic biomass. It has attracted significant attention by researchers in the field of both bioethanol and biomethane production (Bungay, 1982; Hooper and Li, 1996; Lipinsky, 1981; Sanchez and Cardona, 2008). Steam explosion involves exposing the biomass sample to saturated steam under
246
C. Asada et al. / Bioresource Technology 182 (2015) 245–250
pressure, which penetrates the cell wall by diffusion and achieves mechanical separation by the sudden pressure release, creating a shear force around the surrounding structure and resulting in the mechanical breakdown of lignocellulosic biomass cell walls. For the sudden steam explosion process, the deflation time should be short to achieve the full effect of steam explosion for destroying the lignocellulosic complex composed mainly of cellulose and lignin, increasing the accessibility of cellulose component to cellulase attack (Hendriks and Zeeman, 2009). From that point of view, the technical features of the steam explosion device itself limits the scaling-up application because deflation time is basically determined by the intrinsic structure of the equipment. In the scalingup application the traditional valve blow mode operating equipment with a large valve requires a long deflation time, which in turn decreases the shearing effect on the biomass (Yu et al., 2012). Therefore, it is necessary to examine whether a high-temperature saturated steam treatment without the sudden release of steam pressure can provide an efficient pretreatment effect of biomass or not using various wood species. Furthermore, it is more desirable and practical if the large-scale process uses the hightemperature saturated steam treatment. In this investigation, the feasibility of enzymatic saccharification of lignocellulosic biomass using a high-temperature saturated steam treatment was examined. The weight loss, chemical characteristics, and enzymatic saccharification of lignocellulosic biomass treated using high-temperature saturated steam with and without milling treatments were examined using softwood and hardwood biomass. Furthermore, the pretreatment effects of high-temperature saturated steam treatment on softwood and hardwood biomass were compared using cedar and beech, respectively. 2. Methods 2.1. Lignocellulosic biomass samples Softwood biomass, cedar (Cryptomeria japonica), and hardwood biomass, beech (Fagus japonica), were cut into pieces approximately 10-cm-long, 3-cm-wide, and 1.5-cm-thick, and then treated by high-temperature saturated steam with and without milling as described below. 2.2. High temperature saturated steam treatment The wood pieces were treated in a batch steam treatment apparatus (steam explosion apparatus NK-2L; Japan Chemical Engineering and Machinery Co. Ltd., Osaka, Japan) (Asada et al., 2012). The apparatus consisted of a steam generator, a pressurized digester, a receiver, and a condenser with a silencing action. The digester was insulated to maintain a constant temperature. The capacity of the digester was 2 L, the highest pressure was 6.7 MPa, and the highest temperature was 281 °C. Three wood pieces were introduced into the digester and exposed to high-temperature saturated steam at a pressure of 25 atm (225 °C), 35 atm (243 °C), and 45 atm (258 °C), and steaming time of 5 min. The prescribed temperature was reached in a few seconds. After saturated steam exposure, a ball valve at the top of the reactor was slowly opened to bring the digester to atmospheric pressure without applying shear force at a reduction speed of approximately 10 atm/min, and then the steam-treated wood pieces were collected from the digester for measuring the weight loss of steam-treated wood correctly. 2.3. Milling treatment The steam-treated wood pieces were ground by a crusher mill (Wonder Crush Mill D3V-10, Osaka Chemical Co. Ltd., Osaka, Japan)
at 25,000 rpm for 1 min. The particle sizes after milling were approximately 0.01–1 mm. 2.4. Fourier transform infrared (FTIR) analysis Changes in the functional groups of the pretreated wood were recorded by FTIR spectrometry (FT/IR-670 Plus; JASCO, Japan). First, the samples were ground and dried at 105 °C. The sample (1.5 mg) was mixed with 200 mg potassium bromide (KBr). The role of KBr was to hold the fiber flour during the test. Transparent pellets were prepared from the blend and analyzed from 500 to 4000 cm1. 2.5. X-ray diffraction (XRD) analysis The XRD patterns of pretreated wood were obtained by the X-ray diffractometer (RINT2000; RIGAKU, Japan). First, the samples were ground and dried at 105 °C. The sample was scanned in the range of 5–35° (2h) with a scanning speed of 2° min1 at 40 kV and 30 mA under 25 °C. The crystallinity index (CrI) of the sample was estimated according to the method proposed by Isogai and Usuda (1989).
CrI ð%Þ ¼ ðI IB Þ=I
ð1Þ
where I and IB are the peak intensity at 2h = 16° and that of slope line at the same peak, respectively. 2.6. Extraction and component analysis The amounts of the components, i.e., water soluble material, acetone soluble lignin, acid soluble lignin, acid insoluble material (a high-molecular weight lignin and ash), cellulose, and hemicellulose, in the steam-treated wood pieces with milling treatment were measured by the following extraction and separation procedure. One gram of each sample (in triplicate) was extracted with 60 mL of distilled water (DW) for 24 h at room temperature with stirring at 500 rpm. The solid and liquid materials were separated by filtration (ADVANTEC No. 131), and the filtrate (i.e., water soluble material) was recovered from the liquid, and then concentrated, dried, and weighed. The monomeric sugars, organic acids, 5-hydroxymethylfurfural (5-HMF), and furfural concentrations in the water-soluble material were determined by high performance liquid chromatography (HPLC) (LC-20AT HPLC system; Shimadzu, Kyoto, Japan) using an ion-exchange column (Aminex HPX-87H; Bio-Rad, Hercules, CA, USA: 300 7.8 mm; mobile phase, 5 mM H2SO4; temperature, 65 °C; flow rate, 0.6 mL/min; and injection volume, 10 lL) (Davis et al., 2006). The solid precipitate was extracted with 30 mL of acetone at room temperature for 24 h with stirring at 500 rpm to dissolve the acetone soluble lignin (a lowmolecular weight lignin). After filtration (ADVANTEC No. 131), concentration, and drying of the extract, the acetone soluble lignin was weighed. The residue from the acetone extraction consisted of acid soluble lignin, acid insoluble material, cellulose, and hemicellulose. Thereafter, 0.2 g of this residue was added with 3 mL of 72%(w/w) sulfuric acid and kept at room temperature for 4 h. After 4 h, the mixture was transferred to a 100 mL conical flask and diluted with 75 mL DW, and then autoclaved for 1 h at 121 °C. After the sulfuric acid insoluble residue was washed with hot water, it was oven dried at 105 °C to the constant weight, and its mass was recorded (acid insoluble material). Acid soluble lignin in the hydrolyzed liquid was determined by UV spectrophotometry at 205 nm (Sluiter et al., 2012). Furthermore, the residue from the acetone extraction was hydrolyzed with 10 mL of 72%(w/w) sulfuric acid at 30 °C for 60 min. Then, the reaction mixture was diluted to 4%(w/w) sulfuric and autoclaved at 121 °C for 60 min. The amount of cellulose was determined on the basis of the monomer
247
C. Asada et al. / Bioresource Technology 182 (2015) 245–250
2.7. Enzyme saccharification
120
(a)
Volatile component
100
Weight [%]
content (glucose) using HPLC as stated above. The amount of hemicellulose was calculated by subtraction of cellulose content from the holocellulose content. The holocellulose content was determined as the NaClO2-delignified residue (Wise et al., 1946). All analytical data shown in this work are average values of three independent experiments.
Solid component
80 60 40 20
Saccharification rate ð%Þ ¼ ðAmount of glucose produced 0:9Þ=
Untreated
25 atm
35 atm
45 atm
25 atm
35 atm
45 atm
120
(b) 100 80 60 40 20 0
Untreated
Fig. 1. Weight loss of (a) cedar and (b) beech after high temperature saturated steam treatment.
ðAmount of cellulose contained in sampleÞ 100
3. Results and discussions 3.1. Weight loss of steam-treated wood During the steam treatment, some of the volatile wood materials were lost. Fig. 1 shows the weight loss of cedar and beech wood piece treated at a pressure (temperature) of 25 (225 °C), 35 (243 °C), and 45 atm (258 °C) for a steaming time of 5 min. Before the experiment the moisture contents of cedar and beech wood pieces were 14.3% and 11.1%, respectively. The severity factor, S, of the steam treatment is calculated by Eq. (1) (Overend and Chornet, 1987). The volume of the steam treatment reactor (2 L) was small and the steam temperature could reach the target value in a very short time. Thus, the time required to heat from atmospheric temperature to the target temperature was negligible.
S ¼ log ½t expfðT 100Þ=14:75g
0
Weight [%]
Enzymatic saccharification of the untreated and treated wood pieces with and without milling treatment was conducted using a cellulolytic enzyme. The reaction was performed in 100 mL Erlenmeyer flasks at an initial substrate concentration of 2%(w/v) in 50 mL of 0.1 M sodium acetate buffer (pH 5), using enzyme (Meicelase; Meiji Seika Co. Ltd.: Trichoderma viride, 224 FPU/g; b-glucosidase activity, 264 IU/g) loading of 0.1 g of enzyme per 1 g of substrate. The reaction was performed in a reciprocating waterbath shaker at 140 strokes/min for 120 h at 50 °C. The supernatant was centrifuged, and the solid residue was removed to determine the glucose content. The glucose concentration was measured by HPLC as stated above. All trials were performed in triplicate and the means were calculated. The saccharification rate of cellulose is calculated using following equation:
ð3Þ
where T = steam temperature (°C), t = steaming time (min), and 14.75 = activation energy value under conditions where the process kinetics are first order and they obey the Arrhenius law. The overall weight loss of cedar and beech increased with the increase of steam pressure and reached 16.7% and 24.8%, respectively, at a steam pressure of 45 atm (S = 5.35). In case of cedar, the dry weight loss defined as the percentage of weight loss to dry weight of untreated wood piece was almost the same value (12%), when steam pressure was increased from 25 (S = 4.35) to 45 atm. On the other hand, in case of beech, the dry weight loss increased from 14% to 26% when steam pressure was increased from 25 to 45 atm. This suggests that a large amount of volatile matter, i.e., furfural, 5-HMF, and other degradation products (formic acid and levulinic acid), were produced from wood by high temperature saturated steam treatment (Rocha et al., 2012) and were evaporated during the decompression process by venting steam gradually. A major part of the volatile matter seems to be produced by the degradation of hemicellulose, which makes the cellulose more accessible to cellulase (Bahrin et al., 2012). The
reason why the weight loss in beech was much higher than that in cedar is dependent on the fact that the hardwood in general contains much more acetyl groups than the softwood (Boonstra and Tjeerdsma, 2006; Zhu et al., 2008), which improves the hydrolysis of hemicellulose by steam treatment, and thus, steam treatment with a higher severity factor could degrade hemicellulose in the beech significantly resulting in higher weight loss.
3.2. FTIR analysis of steam-treated wood Changes in functional groups of untreated and treated wood were examined using FT-IR spectroscopy analysis. The FT-IR spectra of untreated and treated wood at 25, 35, and 45 atm for 5 min are illustrated in Fig. S1 (Supplementary data). The peaks at 1368 and 2850 cm1 were ascribed to CAH stretching of cellulose and hemicellulose, respectively, and these seem to correspond to the crystallinity of cellulose (Xiao et al., 2014). The stretching vibration band at approximately 3340 cm1 was attributed to hydroxyl groups that indicate hydrogen bonds between molecules. Because the intensity of this band was increased by the steam treatment, it suggests that hydrogen bonds were formed between cellulose chains because of the removal of other components by the steam treatment. In both samples of cedar and beech, with an increase in steam pressure there was a disappearance or a decrease in the intensity of peaks at 1740, 1635, and 1250 cm1, which relates to C@O stretching, OAH bending, and the stretching bands CAO and CAH in the hemicelluloses, respectively. With an increase in steam pressure, there was an increase in the intensity of peaks at 1590 and 1510 cm1 related to aromatic rings, and at 1450 and 1420 cm1 related to CAH bending bands in the lignin. Furthermore, the intensity of peaks at 1160, 1100, and 1030 cm1 due to CAOAC stretching of b-1,4-glycosidic bond between sugars, CAH stretching of sugars, and C@O, CAH stretching of sugars in the cellulose also increased. As the steam pressure increased, the intensity of peaks related to cellulose and lignin increased, whereas the intensity of peaks related to hemicellulose decreased, which
248
C. Asada et al. / Bioresource Technology 182 (2015) 245–250
implies that the high-temperature saturated steam treatment caused the hydrolysis and removal of hemicellulose and a significant increase in the cellulose and lignin content in the steam-treated sample, as shown in Table 1. 3.3. XRD analysis of steam-treated wood The XRD patterns of untreated and treated wood at 25, 35, and 45 atm for 5 min are shown in Fig. S2 (Supplementary data). The sharp peak at 2h = 28.5° due to SiO2 as a marker reinforces the success of this XRD. The crystalline values of the untreated and treated cedar at 25, 35, and 45 atm were 39.1%, 49.3%, 49.7%, and 48.5%, respectively. On the other hand, the crystalline values of the untreated and treated beech at 25, 35, and 45 atm were 37.7%, 52.8%, 57.6%, and 56.4%, respectively. Higher crystalline values for treated samples were obtained compared with that of the untreated samples. This implies that the efficient removal of hemicellulose and amorphous cellulose from the sample occurred during high-temperature saturated steam treatment because the decrease in peak intensity of hemicellulose in FTIR spectra and the rearrangement of the disordered fraction of cellulose fibers (Jonoobi et al., 2011) could be clearly observed as shown in Fig. S1 (Supplementary data). The high-temperature saturated steam treatment was useful for promoting the crystallinity of cellulose; however, the crystalline value decreased slightly at >35 atm. With higher steam pressure, pseudolignin is formed from lignin and sugars liberated from hemicellulose and cellulose (e.g., glucose), thereby resulting in a low crystalline value (Sannigrahi et al., 2011). Furthermore, it was found that the crystalline value of treated beech was higher than treated cedar, which implies that the hemicellulose and lignin in the hardwood are easier to degrade than those in the softwood because the softwood contains a comparatively large amount of condensed-type lignin (Ehara et al., 2002). 3.4. Composition changes of steam-treated wood Table 1 shows the composition of untreated and treated wood at 25, 35, and 45 atm for 5 min followed by milling treatment for 1 min. In the case of untreated samples, the ratios of hemicellulose and acid insoluble material (high-molecular weight lignin and ash) components were comparatively high, whereas those of water soluble material (low-molecular weight monosaccharides, oligosaccharides, 5-HMF, furfural and etc.) and acetone soluble lignin (low-molecular weight lignin) were comparatively low. The increase in ratio of water soluble material seems to be due to the hydrolysis and conversion of some part of hemicellulose into water soluble sugars by the steam treatment. The ratio of cellulose increased by steam treatment and it decreased with an increase in steam pressure. On the other hand, the ratio of hemicellulose decreased significantly in the steam-treated wood. In general, hemicellulose hydrolyzes under less severe conditions than
cellulose (Ando et al., 2000); therefore, the hydrolysis of hemicellulose occurred under the steam treatment at 25–45 atm and degraded products volatized and dissolved as water soluble material (Dunlop, 1948). This also reinforces that the ratio of cellulose increased by the steam treatment, as the weight loss of steamtreated wood occurred due to the removal of hemicellulose as shown in Fig. 1. In the steam-treated cedar, the ratio of acetone soluble lignin increased with an increase in steam pressure and reached its maximum value (35.73%) at 45 atm, whereas the ratio of acid insoluble lignin decreased with an increase in steam pressure and reached its minimum value (23.71%) at 45 atm. This implies that high-weight molecular lignin contained in the cedar was hydrolyzed into lowweight molecular lignin by the steam treatment. The reason why the total ratios (43.92–59.44%) of acetone soluble lignin and acid insoluble lignin at 25–45 atm significantly exceeded the ratio of untreated cedar (33.41%) seems to be because not only did the relative ratio of lignin increase because of the removal of hemicellulose content but also pseudolignin formed from lignin and sugars liberated from hemicellulose and cellulose (e.g., glucose), resulting in a high ratio of acid insoluble material as described above. Furthermore, the acetone soluble lignin with a low-molecular weight (weight-average molecular weight is approximately 2000; data not shown) could be utilized as an alternative raw material for epoxy polymer synthesis (Sasaki et al., 2013). In the steam-treated beech, the ratio of acetone soluble lignin increased by 35 atm reaching its maximum value (37.02%) and then decreased. Conversely, the ratio of acid insoluble lignin decreased by 35 atm reaching its minimum value (8.25%) and then increased. This means that most of the lignin in beech was converted into a low-molecular weight lignin. Because the minimum ratio of acid insoluble material in the steam-treated beech was much lower than that in the steam-treated cedar, it means that hardwood lignin is more easily hydrolyzed than softwood lignin as described above. As a result, it was found that the effect of the steam treatment not only depended on the severity but also on the chemical composition of the wood, which varies significantly with wood species. 3.5. Enzymatic saccharification of steam-treated wood Fig. 2 shows the results of the enzymatic saccharification of the treated wood at 25, 35, and 45 atm for 5 min. In this experiment, the initial substrate concentration was 20 g/L. The ratio of the amount of glucose produced per amount of steam-treated wood with and without milling treatment for 1 min increased rapidly with an increase in incubation time and then slowly reached its maximum value. Using untreated wood with the milling treatment, little glucose was produced even after an incubation time of >120 h. When treated cedar was used as a substrate, the amount of glucose produced with milling treatment was much higher than that without milling treatment. This suggests that steam treatment only would not provide a sufficient pretreatment effect for
Table 1 Composition of untreated and treated wood at 25, 35, and 45 atm for 5 min followed by milling treatment for 1 min. Sample
Composition of ratio [%] Water soluble material
Acetone soluble lignin
Acid soluble lignin
Acid insoluble material
Cellulose
Hemicellulose
Total
Crystallinity index [%]
Cedar
Untreated 25 atm 35 atm 45 atm
2.10 10.20 7.15 5.84
1.99 10.75 27.11 35.73
0.42 0.26 0.17 0.14
31.42 33.17 25.89 23.71
37.99 45.87 42.30 33.82
26.64 3.45 0.64 0.03
100.56 103.70 103.26 99.27
39.08 49.30 49.67 48.46
Beech
Untreated 25 atm 35 atm 45 atm
1.65 10.61 7.37 7.55
1.09 24.20 37.02 36.57
2.12 0.41 0.21 0.16
21.72 11.45 8.25 14.18
36.79 49.92 45.79 38.88
37.79 2.46 2.96 0.37
101.16 99.05 101.60 97.71
37.68 52.78 57.59 56.36
249
C. Asada et al. / Bioresource Technology 182 (2015) 245–250
Ratio of amount of glucose produced from unit per amount of sample [-]
0.5
(a) 0.4
0.3
0.2
0.1
0 0
50
100
150
Incubation time [h]
Ratio of amount of glucose produced from unit per amount of sample [-]
0.5
(b) 0.4
0.3
0.2
0.1
0 0
50
100
150
Incubation time [h] Fig. 2. Enzymatic saccharification of 20 g/L of untreated and steam-treated (a) cedar and (b) beech at 25, 35, and 45 atm for 5 min with and without milling treatment for 1 min. Symbols: }, milling treatment only; s, 25 atm without milling treatment; 4, 35 atm without milling treatment; h, 45 atm without milling treatment; d, 25 atm with milling treatment; N, 35 atm with milling treatment; j, 45 atm with milling treatment.
increasing enzyme accessibility to cellulose because of the residual lignin that strongly covers the cellulose. Therefore, the milling treatment was necessary to achieve the expected saccharification from steam-treated cedar. It seems that smaller particle size improves the accessibility to the cellulose, thus improving enzyme saccharification. Furthermore, the maximum ratio (i.e., 0.35) that corresponds to 93% saccharification rate of cellulose, was obtained at 45 atm for 5 min (S = 5.35) with milling treatment. At this steam treatment condition the maximum ratio of acetone soluble lignin and the minimum ratio of acid insoluble material were obtained as shown in Table 1. This suggests that the depolymerization of lignin is desirable for efficient saccharification of cellulose. On the other hand, when the treated beech was used as a substrate, the amount of glucose produced with milling treatment was also higher than that shown without milling treatment. However, the ratio of the amount of glucose obtained at 35 atm for 5 min (S = 4.90) with milling treatment (i.e., 0.48), which corresponds to 94% saccharification rate of cellulose, was nearly the same value as that without milling treatment (i.e., 0.45), corresponding to 88% saccharification rate of cellulose. At this steam treatment condition the minimum ratio of acid insoluble material was obtained as shown in Table 1. This implies that the steam treatment at 35 atm for 5 min could depolymerize the lignin resulting in a high saccharification rate. Because the hardwood biomass of beech contains a lower amount of condensed-type lignin than the softwood biomass of cedar, which is resistant to hydrolysis, it seems that the beech sample could be delignified by steam treatment only.
Furthermore, even if the beech was treated at 25 atm for 5 min (S = 4.35), the milling treatment provided a high ratio of the amount of glucose obtained (i.e., 0.42). Hence, the necessity of the milling treatment is also dependent on the wood species. Although the milling treatment has been known as one of the most widely used mechanical activation processes to increase the surface area of wood and then obtain a high saccharification rate, it requires significant energy costs (Taherzadeh and Karimi, 2008; Zakaria et al., 2014). Therefore, the comparatively short time for steam treatment (i.e.,
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Chemical characteristics and enzymatic saccharification of lignocellulosic biomass treated using high-temperature saturated steam: Comparison of softwood and hardwood Chikako Asada, Chizuru Sasaki, Takeshi Hirano, Yoshitoshi Nakamura ⇑ Department of Life System, Institute of Technology and Science, The University of Tokushima, 2-1 Minamijosanjima-cho, Tokushima 770-8506, Japan
h i g h l i g h t s High temperature saturated steam treatment of cedar and beech was evaluated. Steam treatment followed by milling treatment improved enzymatic saccharification. Saccharification rate of steam-treated beech with milling treatment was 94%. Necessity of milling treatment after steam treatment is dependent on wood species.
a r t i c l e
i n f o
Article history: Received 20 November 2014 Received in revised form 1 February 2015 Accepted 2 February 2015 Available online 11 February 2015 Keywords: High-temperature saturated steam Chemical characteristics Enzymatic saccharification Softwood Hardwood
a b s t r a c t This study investigated the effect of high-temperature saturated steam treatments on the chemical characteristics and enzymatic saccharification of softwood and hardwood. The weight loss and chemical modification of cedar and beech wood pieces treated at 25, 35, and 45 atm for 5 min were determined. Fourier transform infrared and X-ray diffraction analyses indicated that solubilization and removal of hemicellulose and lignin occurred by the steam treatment. The milling treatment of steam-treated wood enhanced its enzymatic saccharification. Maximum enzymatic saccharification (i.e., 94% saccharification rate of cellulose) was obtained using steam-treated beech at 35 atm for 5 min followed by milling treatment for 1 min. However, the necessity of the milling treatment for efficient enzymatic saccharification is dependent on the wood species. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Primary energy self-sufficiency in Japan is only 4%, and this value is much lower than other developed countries. Even when nuclear power generation is considered as a domestically produced energy, the rate slightly increases to 17%, implying that Japan is hugely dependent on imported energy (approximately 80%). Therefore, the use of renewable energy sources such as biomass has been considered as an alternative to fossil fuels in Japan. For the significant production of domestic biofuels, the development of a method to convert lignocellulosic biomass such as wood, bamboo, and agricultural waste (such as rice straw) into ethanol and useful materials is desirable because it would not compete with food production. However, because cellulosic components (i.e., substrates for ethanol production) are strongly covered with lignin in the lignocellulosic biomass, pretreatment (delignification) ⇑ Corresponding author. Tel.: +81 88 656 7518; fax: +81 88 656 9071. E-mail address: [email protected] (Y. Nakamura). http://dx.doi.org/10.1016/j.biortech.2015.02.005 0960-8524/Ó 2015 Elsevier Ltd. All rights reserved.
is necessary for efficient enzymatic saccharification followed by alcohol fermentation to increase the accessibility of cellulose to cellulases (Karp et al., 2013). Various pretreatment methods have been developed and applied to target the conversion of biomass to biofuel including: acid treatment (Torget et al., 1988), alkaline treatment (Ucar, 1990), ammonia fiber expansion (Dale et al., 1996), organosolv treatment (Bonn et al., 1987), liquid hot water treatment (Sreenath et al., 1999), microwave irradiation (Ooshima et al., 1984), superheated steam treatment (Bahrin et al., 2012), and steam explosion (Ramos et al., 1992). Steam explosion, introduced as a biomass pretreatment process by Mason (1926), has become one of the simplest and environmentally friendly techniques and is a novel pretreatment method with a high efficiency at breaking lignin structures to release cellulose and hemicellulose from lignocellulosic biomass. It has attracted significant attention by researchers in the field of both bioethanol and biomethane production (Bungay, 1982; Hooper and Li, 1996; Lipinsky, 1981; Sanchez and Cardona, 2008). Steam explosion involves exposing the biomass sample to saturated steam under
246
C. Asada et al. / Bioresource Technology 182 (2015) 245–250
pressure, which penetrates the cell wall by diffusion and achieves mechanical separation by the sudden pressure release, creating a shear force around the surrounding structure and resulting in the mechanical breakdown of lignocellulosic biomass cell walls. For the sudden steam explosion process, the deflation time should be short to achieve the full effect of steam explosion for destroying the lignocellulosic complex composed mainly of cellulose and lignin, increasing the accessibility of cellulose component to cellulase attack (Hendriks and Zeeman, 2009). From that point of view, the technical features of the steam explosion device itself limits the scaling-up application because deflation time is basically determined by the intrinsic structure of the equipment. In the scalingup application the traditional valve blow mode operating equipment with a large valve requires a long deflation time, which in turn decreases the shearing effect on the biomass (Yu et al., 2012). Therefore, it is necessary to examine whether a high-temperature saturated steam treatment without the sudden release of steam pressure can provide an efficient pretreatment effect of biomass or not using various wood species. Furthermore, it is more desirable and practical if the large-scale process uses the hightemperature saturated steam treatment. In this investigation, the feasibility of enzymatic saccharification of lignocellulosic biomass using a high-temperature saturated steam treatment was examined. The weight loss, chemical characteristics, and enzymatic saccharification of lignocellulosic biomass treated using high-temperature saturated steam with and without milling treatments were examined using softwood and hardwood biomass. Furthermore, the pretreatment effects of high-temperature saturated steam treatment on softwood and hardwood biomass were compared using cedar and beech, respectively. 2. Methods 2.1. Lignocellulosic biomass samples Softwood biomass, cedar (Cryptomeria japonica), and hardwood biomass, beech (Fagus japonica), were cut into pieces approximately 10-cm-long, 3-cm-wide, and 1.5-cm-thick, and then treated by high-temperature saturated steam with and without milling as described below. 2.2. High temperature saturated steam treatment The wood pieces were treated in a batch steam treatment apparatus (steam explosion apparatus NK-2L; Japan Chemical Engineering and Machinery Co. Ltd., Osaka, Japan) (Asada et al., 2012). The apparatus consisted of a steam generator, a pressurized digester, a receiver, and a condenser with a silencing action. The digester was insulated to maintain a constant temperature. The capacity of the digester was 2 L, the highest pressure was 6.7 MPa, and the highest temperature was 281 °C. Three wood pieces were introduced into the digester and exposed to high-temperature saturated steam at a pressure of 25 atm (225 °C), 35 atm (243 °C), and 45 atm (258 °C), and steaming time of 5 min. The prescribed temperature was reached in a few seconds. After saturated steam exposure, a ball valve at the top of the reactor was slowly opened to bring the digester to atmospheric pressure without applying shear force at a reduction speed of approximately 10 atm/min, and then the steam-treated wood pieces were collected from the digester for measuring the weight loss of steam-treated wood correctly. 2.3. Milling treatment The steam-treated wood pieces were ground by a crusher mill (Wonder Crush Mill D3V-10, Osaka Chemical Co. Ltd., Osaka, Japan)
at 25,000 rpm for 1 min. The particle sizes after milling were approximately 0.01–1 mm. 2.4. Fourier transform infrared (FTIR) analysis Changes in the functional groups of the pretreated wood were recorded by FTIR spectrometry (FT/IR-670 Plus; JASCO, Japan). First, the samples were ground and dried at 105 °C. The sample (1.5 mg) was mixed with 200 mg potassium bromide (KBr). The role of KBr was to hold the fiber flour during the test. Transparent pellets were prepared from the blend and analyzed from 500 to 4000 cm1. 2.5. X-ray diffraction (XRD) analysis The XRD patterns of pretreated wood were obtained by the X-ray diffractometer (RINT2000; RIGAKU, Japan). First, the samples were ground and dried at 105 °C. The sample was scanned in the range of 5–35° (2h) with a scanning speed of 2° min1 at 40 kV and 30 mA under 25 °C. The crystallinity index (CrI) of the sample was estimated according to the method proposed by Isogai and Usuda (1989).
CrI ð%Þ ¼ ðI IB Þ=I
ð1Þ
where I and IB are the peak intensity at 2h = 16° and that of slope line at the same peak, respectively. 2.6. Extraction and component analysis The amounts of the components, i.e., water soluble material, acetone soluble lignin, acid soluble lignin, acid insoluble material (a high-molecular weight lignin and ash), cellulose, and hemicellulose, in the steam-treated wood pieces with milling treatment were measured by the following extraction and separation procedure. One gram of each sample (in triplicate) was extracted with 60 mL of distilled water (DW) for 24 h at room temperature with stirring at 500 rpm. The solid and liquid materials were separated by filtration (ADVANTEC No. 131), and the filtrate (i.e., water soluble material) was recovered from the liquid, and then concentrated, dried, and weighed. The monomeric sugars, organic acids, 5-hydroxymethylfurfural (5-HMF), and furfural concentrations in the water-soluble material were determined by high performance liquid chromatography (HPLC) (LC-20AT HPLC system; Shimadzu, Kyoto, Japan) using an ion-exchange column (Aminex HPX-87H; Bio-Rad, Hercules, CA, USA: 300 7.8 mm; mobile phase, 5 mM H2SO4; temperature, 65 °C; flow rate, 0.6 mL/min; and injection volume, 10 lL) (Davis et al., 2006). The solid precipitate was extracted with 30 mL of acetone at room temperature for 24 h with stirring at 500 rpm to dissolve the acetone soluble lignin (a lowmolecular weight lignin). After filtration (ADVANTEC No. 131), concentration, and drying of the extract, the acetone soluble lignin was weighed. The residue from the acetone extraction consisted of acid soluble lignin, acid insoluble material, cellulose, and hemicellulose. Thereafter, 0.2 g of this residue was added with 3 mL of 72%(w/w) sulfuric acid and kept at room temperature for 4 h. After 4 h, the mixture was transferred to a 100 mL conical flask and diluted with 75 mL DW, and then autoclaved for 1 h at 121 °C. After the sulfuric acid insoluble residue was washed with hot water, it was oven dried at 105 °C to the constant weight, and its mass was recorded (acid insoluble material). Acid soluble lignin in the hydrolyzed liquid was determined by UV spectrophotometry at 205 nm (Sluiter et al., 2012). Furthermore, the residue from the acetone extraction was hydrolyzed with 10 mL of 72%(w/w) sulfuric acid at 30 °C for 60 min. Then, the reaction mixture was diluted to 4%(w/w) sulfuric and autoclaved at 121 °C for 60 min. The amount of cellulose was determined on the basis of the monomer
247
C. Asada et al. / Bioresource Technology 182 (2015) 245–250
2.7. Enzyme saccharification
120
(a)
Volatile component
100
Weight [%]
content (glucose) using HPLC as stated above. The amount of hemicellulose was calculated by subtraction of cellulose content from the holocellulose content. The holocellulose content was determined as the NaClO2-delignified residue (Wise et al., 1946). All analytical data shown in this work are average values of three independent experiments.
Solid component
80 60 40 20
Saccharification rate ð%Þ ¼ ðAmount of glucose produced 0:9Þ=
Untreated
25 atm
35 atm
45 atm
25 atm
35 atm
45 atm
120
(b) 100 80 60 40 20 0
Untreated
Fig. 1. Weight loss of (a) cedar and (b) beech after high temperature saturated steam treatment.
ðAmount of cellulose contained in sampleÞ 100
3. Results and discussions 3.1. Weight loss of steam-treated wood During the steam treatment, some of the volatile wood materials were lost. Fig. 1 shows the weight loss of cedar and beech wood piece treated at a pressure (temperature) of 25 (225 °C), 35 (243 °C), and 45 atm (258 °C) for a steaming time of 5 min. Before the experiment the moisture contents of cedar and beech wood pieces were 14.3% and 11.1%, respectively. The severity factor, S, of the steam treatment is calculated by Eq. (1) (Overend and Chornet, 1987). The volume of the steam treatment reactor (2 L) was small and the steam temperature could reach the target value in a very short time. Thus, the time required to heat from atmospheric temperature to the target temperature was negligible.
S ¼ log ½t expfðT 100Þ=14:75g
0
Weight [%]
Enzymatic saccharification of the untreated and treated wood pieces with and without milling treatment was conducted using a cellulolytic enzyme. The reaction was performed in 100 mL Erlenmeyer flasks at an initial substrate concentration of 2%(w/v) in 50 mL of 0.1 M sodium acetate buffer (pH 5), using enzyme (Meicelase; Meiji Seika Co. Ltd.: Trichoderma viride, 224 FPU/g; b-glucosidase activity, 264 IU/g) loading of 0.1 g of enzyme per 1 g of substrate. The reaction was performed in a reciprocating waterbath shaker at 140 strokes/min for 120 h at 50 °C. The supernatant was centrifuged, and the solid residue was removed to determine the glucose content. The glucose concentration was measured by HPLC as stated above. All trials were performed in triplicate and the means were calculated. The saccharification rate of cellulose is calculated using following equation:
ð3Þ
where T = steam temperature (°C), t = steaming time (min), and 14.75 = activation energy value under conditions where the process kinetics are first order and they obey the Arrhenius law. The overall weight loss of cedar and beech increased with the increase of steam pressure and reached 16.7% and 24.8%, respectively, at a steam pressure of 45 atm (S = 5.35). In case of cedar, the dry weight loss defined as the percentage of weight loss to dry weight of untreated wood piece was almost the same value (12%), when steam pressure was increased from 25 (S = 4.35) to 45 atm. On the other hand, in case of beech, the dry weight loss increased from 14% to 26% when steam pressure was increased from 25 to 45 atm. This suggests that a large amount of volatile matter, i.e., furfural, 5-HMF, and other degradation products (formic acid and levulinic acid), were produced from wood by high temperature saturated steam treatment (Rocha et al., 2012) and were evaporated during the decompression process by venting steam gradually. A major part of the volatile matter seems to be produced by the degradation of hemicellulose, which makes the cellulose more accessible to cellulase (Bahrin et al., 2012). The
reason why the weight loss in beech was much higher than that in cedar is dependent on the fact that the hardwood in general contains much more acetyl groups than the softwood (Boonstra and Tjeerdsma, 2006; Zhu et al., 2008), which improves the hydrolysis of hemicellulose by steam treatment, and thus, steam treatment with a higher severity factor could degrade hemicellulose in the beech significantly resulting in higher weight loss.
3.2. FTIR analysis of steam-treated wood Changes in functional groups of untreated and treated wood were examined using FT-IR spectroscopy analysis. The FT-IR spectra of untreated and treated wood at 25, 35, and 45 atm for 5 min are illustrated in Fig. S1 (Supplementary data). The peaks at 1368 and 2850 cm1 were ascribed to CAH stretching of cellulose and hemicellulose, respectively, and these seem to correspond to the crystallinity of cellulose (Xiao et al., 2014). The stretching vibration band at approximately 3340 cm1 was attributed to hydroxyl groups that indicate hydrogen bonds between molecules. Because the intensity of this band was increased by the steam treatment, it suggests that hydrogen bonds were formed between cellulose chains because of the removal of other components by the steam treatment. In both samples of cedar and beech, with an increase in steam pressure there was a disappearance or a decrease in the intensity of peaks at 1740, 1635, and 1250 cm1, which relates to C@O stretching, OAH bending, and the stretching bands CAO and CAH in the hemicelluloses, respectively. With an increase in steam pressure, there was an increase in the intensity of peaks at 1590 and 1510 cm1 related to aromatic rings, and at 1450 and 1420 cm1 related to CAH bending bands in the lignin. Furthermore, the intensity of peaks at 1160, 1100, and 1030 cm1 due to CAOAC stretching of b-1,4-glycosidic bond between sugars, CAH stretching of sugars, and C@O, CAH stretching of sugars in the cellulose also increased. As the steam pressure increased, the intensity of peaks related to cellulose and lignin increased, whereas the intensity of peaks related to hemicellulose decreased, which
248
C. Asada et al. / Bioresource Technology 182 (2015) 245–250
implies that the high-temperature saturated steam treatment caused the hydrolysis and removal of hemicellulose and a significant increase in the cellulose and lignin content in the steam-treated sample, as shown in Table 1. 3.3. XRD analysis of steam-treated wood The XRD patterns of untreated and treated wood at 25, 35, and 45 atm for 5 min are shown in Fig. S2 (Supplementary data). The sharp peak at 2h = 28.5° due to SiO2 as a marker reinforces the success of this XRD. The crystalline values of the untreated and treated cedar at 25, 35, and 45 atm were 39.1%, 49.3%, 49.7%, and 48.5%, respectively. On the other hand, the crystalline values of the untreated and treated beech at 25, 35, and 45 atm were 37.7%, 52.8%, 57.6%, and 56.4%, respectively. Higher crystalline values for treated samples were obtained compared with that of the untreated samples. This implies that the efficient removal of hemicellulose and amorphous cellulose from the sample occurred during high-temperature saturated steam treatment because the decrease in peak intensity of hemicellulose in FTIR spectra and the rearrangement of the disordered fraction of cellulose fibers (Jonoobi et al., 2011) could be clearly observed as shown in Fig. S1 (Supplementary data). The high-temperature saturated steam treatment was useful for promoting the crystallinity of cellulose; however, the crystalline value decreased slightly at >35 atm. With higher steam pressure, pseudolignin is formed from lignin and sugars liberated from hemicellulose and cellulose (e.g., glucose), thereby resulting in a low crystalline value (Sannigrahi et al., 2011). Furthermore, it was found that the crystalline value of treated beech was higher than treated cedar, which implies that the hemicellulose and lignin in the hardwood are easier to degrade than those in the softwood because the softwood contains a comparatively large amount of condensed-type lignin (Ehara et al., 2002). 3.4. Composition changes of steam-treated wood Table 1 shows the composition of untreated and treated wood at 25, 35, and 45 atm for 5 min followed by milling treatment for 1 min. In the case of untreated samples, the ratios of hemicellulose and acid insoluble material (high-molecular weight lignin and ash) components were comparatively high, whereas those of water soluble material (low-molecular weight monosaccharides, oligosaccharides, 5-HMF, furfural and etc.) and acetone soluble lignin (low-molecular weight lignin) were comparatively low. The increase in ratio of water soluble material seems to be due to the hydrolysis and conversion of some part of hemicellulose into water soluble sugars by the steam treatment. The ratio of cellulose increased by steam treatment and it decreased with an increase in steam pressure. On the other hand, the ratio of hemicellulose decreased significantly in the steam-treated wood. In general, hemicellulose hydrolyzes under less severe conditions than
cellulose (Ando et al., 2000); therefore, the hydrolysis of hemicellulose occurred under the steam treatment at 25–45 atm and degraded products volatized and dissolved as water soluble material (Dunlop, 1948). This also reinforces that the ratio of cellulose increased by the steam treatment, as the weight loss of steamtreated wood occurred due to the removal of hemicellulose as shown in Fig. 1. In the steam-treated cedar, the ratio of acetone soluble lignin increased with an increase in steam pressure and reached its maximum value (35.73%) at 45 atm, whereas the ratio of acid insoluble lignin decreased with an increase in steam pressure and reached its minimum value (23.71%) at 45 atm. This implies that high-weight molecular lignin contained in the cedar was hydrolyzed into lowweight molecular lignin by the steam treatment. The reason why the total ratios (43.92–59.44%) of acetone soluble lignin and acid insoluble lignin at 25–45 atm significantly exceeded the ratio of untreated cedar (33.41%) seems to be because not only did the relative ratio of lignin increase because of the removal of hemicellulose content but also pseudolignin formed from lignin and sugars liberated from hemicellulose and cellulose (e.g., glucose), resulting in a high ratio of acid insoluble material as described above. Furthermore, the acetone soluble lignin with a low-molecular weight (weight-average molecular weight is approximately 2000; data not shown) could be utilized as an alternative raw material for epoxy polymer synthesis (Sasaki et al., 2013). In the steam-treated beech, the ratio of acetone soluble lignin increased by 35 atm reaching its maximum value (37.02%) and then decreased. Conversely, the ratio of acid insoluble lignin decreased by 35 atm reaching its minimum value (8.25%) and then increased. This means that most of the lignin in beech was converted into a low-molecular weight lignin. Because the minimum ratio of acid insoluble material in the steam-treated beech was much lower than that in the steam-treated cedar, it means that hardwood lignin is more easily hydrolyzed than softwood lignin as described above. As a result, it was found that the effect of the steam treatment not only depended on the severity but also on the chemical composition of the wood, which varies significantly with wood species. 3.5. Enzymatic saccharification of steam-treated wood Fig. 2 shows the results of the enzymatic saccharification of the treated wood at 25, 35, and 45 atm for 5 min. In this experiment, the initial substrate concentration was 20 g/L. The ratio of the amount of glucose produced per amount of steam-treated wood with and without milling treatment for 1 min increased rapidly with an increase in incubation time and then slowly reached its maximum value. Using untreated wood with the milling treatment, little glucose was produced even after an incubation time of >120 h. When treated cedar was used as a substrate, the amount of glucose produced with milling treatment was much higher than that without milling treatment. This suggests that steam treatment only would not provide a sufficient pretreatment effect for
Table 1 Composition of untreated and treated wood at 25, 35, and 45 atm for 5 min followed by milling treatment for 1 min. Sample
Composition of ratio [%] Water soluble material
Acetone soluble lignin
Acid soluble lignin
Acid insoluble material
Cellulose
Hemicellulose
Total
Crystallinity index [%]
Cedar
Untreated 25 atm 35 atm 45 atm
2.10 10.20 7.15 5.84
1.99 10.75 27.11 35.73
0.42 0.26 0.17 0.14
31.42 33.17 25.89 23.71
37.99 45.87 42.30 33.82
26.64 3.45 0.64 0.03
100.56 103.70 103.26 99.27
39.08 49.30 49.67 48.46
Beech
Untreated 25 atm 35 atm 45 atm
1.65 10.61 7.37 7.55
1.09 24.20 37.02 36.57
2.12 0.41 0.21 0.16
21.72 11.45 8.25 14.18
36.79 49.92 45.79 38.88
37.79 2.46 2.96 0.37
101.16 99.05 101.60 97.71
37.68 52.78 57.59 56.36
249
C. Asada et al. / Bioresource Technology 182 (2015) 245–250
Ratio of amount of glucose produced from unit per amount of sample [-]
0.5
(a) 0.4
0.3
0.2
0.1
0 0
50
100
150
Incubation time [h]
Ratio of amount of glucose produced from unit per amount of sample [-]
0.5
(b) 0.4
0.3
0.2
0.1
0 0
50
100
150
Incubation time [h] Fig. 2. Enzymatic saccharification of 20 g/L of untreated and steam-treated (a) cedar and (b) beech at 25, 35, and 45 atm for 5 min with and without milling treatment for 1 min. Symbols: }, milling treatment only; s, 25 atm without milling treatment; 4, 35 atm without milling treatment; h, 45 atm without milling treatment; d, 25 atm with milling treatment; N, 35 atm with milling treatment; j, 45 atm with milling treatment.
increasing enzyme accessibility to cellulose because of the residual lignin that strongly covers the cellulose. Therefore, the milling treatment was necessary to achieve the expected saccharification from steam-treated cedar. It seems that smaller particle size improves the accessibility to the cellulose, thus improving enzyme saccharification. Furthermore, the maximum ratio (i.e., 0.35) that corresponds to 93% saccharification rate of cellulose, was obtained at 45 atm for 5 min (S = 5.35) with milling treatment. At this steam treatment condition the maximum ratio of acetone soluble lignin and the minimum ratio of acid insoluble material were obtained as shown in Table 1. This suggests that the depolymerization of lignin is desirable for efficient saccharification of cellulose. On the other hand, when the treated beech was used as a substrate, the amount of glucose produced with milling treatment was also higher than that shown without milling treatment. However, the ratio of the amount of glucose obtained at 35 atm for 5 min (S = 4.90) with milling treatment (i.e., 0.48), which corresponds to 94% saccharification rate of cellulose, was nearly the same value as that without milling treatment (i.e., 0.45), corresponding to 88% saccharification rate of cellulose. At this steam treatment condition the minimum ratio of acid insoluble material was obtained as shown in Table 1. This implies that the steam treatment at 35 atm for 5 min could depolymerize the lignin resulting in a high saccharification rate. Because the hardwood biomass of beech contains a lower amount of condensed-type lignin than the softwood biomass of cedar, which is resistant to hydrolysis, it seems that the beech sample could be delignified by steam treatment only.
Furthermore, even if the beech was treated at 25 atm for 5 min (S = 4.35), the milling treatment provided a high ratio of the amount of glucose obtained (i.e., 0.42). Hence, the necessity of the milling treatment is also dependent on the wood species. Although the milling treatment has been known as one of the most widely used mechanical activation processes to increase the surface area of wood and then obtain a high saccharification rate, it requires significant energy costs (Taherzadeh and Karimi, 2008; Zakaria et al., 2014). Therefore, the comparatively short time for steam treatment (i.e.,
