- Email: [email protected]
Fungal pretreatment of agricultural residues for bioethanol production
Industrial Crops and Products 89 (2016) 486–492
Contents lists available at ScienceDirect
Industrial Crops and Products journal homepage: www.elsevier.com/locate/indcrop
Fungal pretreatment of agricultural residues for bioethanol production M. García-Torreiro, M. López-Abelairas, T.A. Lu-Chau ∗ , J.M. Lema Department of Chemical Engineering, Institute of Technology, Universidade de Santiago de Compostela, 15782, Santiago de Compostela, Spain
a r t i c l e
i n f o
Article history: Received 20 January 2016 Received in revised form 14 May 2016 Accepted 23 May 2016 Keywords: Bioethanol Agricultural residues Lignocellulosic biomass Biological pretreatment Irpex lacteus
a b s t r a c t The biological pretreatment of lignocellulosic biomass for the production of ethanol is an environmentally friendly process that needs to be evaluated with different feedstocks in order to avoid dependence on a single feedstock. In this study, four agricultural residues (corn stover, barley straw, corncob and wheat straw), selected in terms of their composition and geographic availability, were pretreated using the white-rot fungus Irpex lacteus. After the fungal pretreatment, the biggest reduction in lignin content (45.8 ± 3.5%), lowest sugars consumption (11.5 ± 1.4%) and highest lignin selectivity removal (2.1 ± 0.15) were achieved with corn stover. Moreover, total holocellulose digestibility was significantly increased after the biological pretreatment with all the substrates (37–103%), excepting corncob. In this study, fungal pretreatment was successfully applied to three of the most common agricultural residues available in Europe for producing ethanol, demonstrating that it is capable of handling feedstocks of variable origin. © 2016 Elsevier B.V. All rights reserved.
1. Introduction In the European Union, agricultural residues are abundant and cheap lignocellulosic resources that can be used to obtain valuable products such as bioethanol. This raw material has the advantage of not being competitive in food markets, like other substrates derived from plants such as cereals or sugarcane, used as well in the bioethanol industry. Nevertheless, the challenge to obtain the sugar fractions from these substrates, due to its complex and heterogeneous structure, is the main bottleneck of the second generation bioethanol production. Several pretreatments based on physical, chemical or physicochemical mechanisms have been extensively studied (Alvira et al., 2010). They are frequently based on the use of chemicals, extreme conditions of temperature or pressure and special equipment to deal with these extreme conditions. The sugar yield reported with these treatments is used to be very high; nevertheless inhibitory compounds are also produced in different amounts depending on the severity or the chemicals used (Jönsson et al., 2013). In this context, the biological pretreatment appears as an environmental friendly alternative. It is based on the special abilities of a small group of filamentous fungi belonging to the phylum basidiomy-
∗ Corresponding author: Rúa Constantino Candeira s/n, 15782, Santiago de Compostela, Spain. E-mail address: [email protected] (T.A. Lu-Chau). http://dx.doi.org/10.1016/j.indcrop.2016.05.036 0926-6690/© 2016 Elsevier B.V. All rights reserved.
cota, known as white-rot fungi (WRF), which have evolved the capability of degrading lignin, the most recalcitrant component of plant cell wall (Hatakka, 1983), by using a set of extracellular ligninolytic enzymes. With the partial removal of lignin, an easier degradability of the substrate is achieved, without producing inhibitors as the commonly found with the other pretreatments (Salvachúa et al., 2011). In particular, the white-rot basidiomycete Irpex lacteus presents a great biodegradation capability (Novotny´ et al., 2009; Salvachúa et al., 2011). The advantages of the biological pretreatment include low energy requirement and mild conditions. However, there is an inevitable dependence between lignin removal and sugars consumption, since the fungal strategy consists of degrading the lignin fraction in order to reach easily the cellulose and hemicellulose. The other main disadvantage, compared to physicochemical pretreatmets, is the prolonged time needed to reach similar digestibility improvements, which can be as long as 4–8 weeks (Sarkar et al., 2012). This time can be reduced to 2–3 weeks by combining the biological treatment with an alkaline wash at soft conditions and optimizing operational conditions (Salvachúa et al., 2011; López-Abelairas et al., 2013a). In order to be considered a feasible alternative, the biological pretreatment should be evaluated using different feedstocks to demonstrate its flexibility. There are several previous research works that studied the effect of using different lignocellulosic raw materials on the performance of chemical and/or physicochemical pretreatments (Martin et al., 2007; Njoku et al., 2012; Srivastava et al., 2014; Wang et al., 2014), but only one previous work
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
(Wan and Li, 2011) have studied the effect of using a fungal biological pretreatment with another WRF (Ceriporiopsis subvermispora) on different types of feedstocks. In the present study, the biological pretreatment of four lignocellulosic agricultural residues (corn stover, barley straw, corncob and wheat straw) was carried out. These agricultural residues were chosen according to their composition and their availability in Europe. Despite the fact that the main composition of the raw materials is basically the same, a different fungal behavior during the growth on the different plant species could be expected. The efficiency of the fungal process was studied in detail in order to clarify if the fungal performance presents any dependence with the substrate origin. 2. Materials and methods 2.1. Microorganisms and inocula preparation The basidiomycete I. lacteus (Fr. 238 617/93) was obtained from the culture collection of the Chemical Engineering Department of the University of Santiago de Compostela (Spain). Mycelium production in static cultures and fungal inoculum preparation was carried out as described by Salvachúa et al. (2011). The microorganism used for fermenting the lignocellulosic hydrolysates was the yeast Pachysolen tannophilus (CETC 1426). The yeast was maintained in YPD medium plates, composed of (per L): peptone, 10 g; yeast extract, 20 g; glucose 10 g and agar 15 g. For inoculum preparation, yeast from a fresh plate, was transferred to 100 mL of liquid media composed of (per L): malt extract, 20 g; glucose, 20 g and peptone 1 g. Inoculum was incubated in an orbital shaker at 150 rpm and 32.5 ◦ C for 16 h or when an optical density of approx. 1.5 was reached. 2.2. Fungal pretreatment Lignocellulosic residues procured from the local market (Santiago de Compostela, Spain) were used as the lignocellulosic feedstock source for ethanol production after fungal pretreatment. The moisture content of the raw biomasses was in the range of 7.3–8.5%, these values were taking into account for the composition calculations in dry basis. The biological pretreatment parameters and conditions were optimized in a previous work (López-Abelairas et al., 2013a). In brief, 6 g of blended (approx. 0.5 cm of particle size) lignocellulosic material were mixed with 20 mL of tap water in 250 mL Erlenmeyer flasks and sterilized by autoclave at 121 ◦ C during 20 min. After sterilization, 3 mL of fungal inoculum (as pellets) were added to each flask and maintained at 30 ◦ C during 21 d. The biologically pretreated biomass, with its liquid fraction, was freeze-dried. Total weight (TW) loss was calculated as the percentage of total solids lost after pretreatment and part of the dried solid was used for digestibility and composition measures. All the experiments were carried out in triplicate, data in tables and figures are presented with the mean and its standard deviation. 2.3. Ethanol production After the fungal pretreatment, the different substrates were subjected to an alkali recycling treatment as described by GarcíaTorreiro et al. (2016). In this protocol, the solid phases accumulated during up to three successive alkali washes with NaOH (0.36 g NaOH g−1 substrate) were mixed with the remaining liquid phase and citrate buffer 0.1 M. This buffer was added to adjust the solids percentage to 10% (w/v). Then, H2 SO4 was used to adjust the pH between 4.8–5, and 200 g mL−1 of tetracycline was finally added to avoid contamination.
487
Simultaneous saccharification and fermentation (SSF) configuration, as previously described by López-Abelairas et al. (2013b), was chosen to convert the resulting sugars into ethanol. The SSF configuration consists on a short enzymatic hydrolysis followed by the fermentation of the resulting sugars while the hydrolysing enzymes are still acting. The enzymatic doses used were previously optimized for biopretreated biomass subjected to alkali recycling treatment (García-Torreiro et al., 2016). Cellulase and beta-glucosidase were provided by the enzyme mixtures Cellic CTec2 and NS50010, respectively, while beta-xylosidase and xylanase by Cellic Htec2 (Novozymes). The cocktail doses applied were 96, 18 and 5 L g−1 raw substrate of Cellic CTec2, Cellic Htec2 and NS50010, respectively. The hydrolysis was carried out in citrate buffer 0.1 M (pH 4.8) with a solid load of 10% (w/v). After 24 h of incubation at 150 rpm and 50 ◦ C, the yeast P. tannophilus was inoculated using a ratio of 10% v/v of hydrolyzate. The fermentation and simultaneous saccharification was carried out at 37 ◦ C and 150 rpm. Samples were taken periodically during the fermentation. The concentration of total reducing sugars, glucose and ethanol was determined in these samples. 2.4. Analytical protocols 2.4.1. Composition determination An adapted protocol from NREL (National Renewable Energy Laboratory, Golden, USA) was used for the determination of the biomass composition in glucan, xylan and lignin. Acid hydrolysis allows lignin separation from sugar fraction and it can be determined by weight. The concentration of xylan and glucan, were calculated from the concentration of the corresponding monomeric sugars using an anhydro correction of 0.88 and 0.90 for C5 and C6 sugars, respectively. 2.4.2. Glucan and xylan digestibilities Glucan (Dg ) and xylan (Dx ) digestibilities were evaluated and expressed, according to Eqs. (1) and (2), respectively. Dg (%) =
Gr (g) × 100 Gs (g)
(1)
Dx (%) =
X r (g) × 100 Xs (g)
(2)
where Gr and Xr are the glucose and xylose released from the biomass after pretreatment (including the enzymatic hydrolysis step), repectively and the Gs and Xs are the theoretical maximum amount of glucose and xylose available in the raw material, respectively. Total digestibility was calculated using the total amount of reducing sugars, instead of glucose or xylose. This parameter gives an idea of how available the fermentable sugars are. 2.4.3. Determination of substrate and product concentrations Glucose content was measured using HPLC (Sluiter et al., 2008) and an enzymatic colorimetric method (GOD-PAP/Trinder, Spinreact, GI, Spain). No significant differences were found between HPLC and enzymatic method, for that reason the latter was used as routine method for glucose determination. Total reducing sugars (TRS) were determined by the dinitrosalicylic acid (DNS) method (Miller, 1959). The concentration of ethanol was determined using an HPLC system (Hewlett Packard 1100, USA) with an IR 1047 detector and an Aminex HPX-87H column. H2 SO4 (5 mM) was used as the mobile phase with a flow of 0.6 mL min−1 at 60 ◦ C. Samples were filtered through 0.22 m syringe filters prior to HPLC. 2.4.4. FTIR analysis FTIR was used to provide complementary information to understand the structural differences due to the effect of fungal
488
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
Table 1 Composition (% dry basis) of the feedstock used in the study. Wheat straw
Glucan Xylan Lignin Ash Othersa a
Corn stover
Barley straw
Corncob
Raw
Pretreated
Raw
Pretreated
Raw
Pretreated
Raw
Pretreated
33.7 ± 1.3 29.9 ± 4.1 23.4 ± 0.1 4.2 ± 0.04 8.9
27.6 ± 1.0 31.8 ± 1.4 19.5 ± 1.2 7.6 ± 0.2 13.5
35.3 ± 0.2 29.3 ± 0.5 23.2 ± 0.8 4.6 ± 0.2 7.6
33.9 ± 1.6 38.9 ± 5.9 16.4 ± 2.4 5.4 ± 0.3 5.4
34.4 ± 4.8 25.7 ± 3.5 21.1 ± 1.3 5.9 ± 0.1 12.9
28.5 ± 3.3 35.4 ± 3.7 20.2 ± 0.9 8.1 ± 0.5 9.8
30.5 ± 7.6 33.9 ± 3.2 23.6 ± 4.4 2.1 ± 0.05 9.8
24.7 ± 1.4 40.7 ± 1.2 22.1 ± 2.1 3.2 ± 0.7 9.4
Calculated as the difference between 100% and the sum of the composition of the first four components.
Fig. 1. Distribution of the main cereals produced in Europe and Spain (FAOSTAT, 2014).
pretreatment on the four assayed substrates and to relate this information with other more commonly evaluated parameters such as the increase of carbohydrate digestibilty and the reduction of lignin content. In order to carry out this analysis, IR absorption spectra in attenuated total reflectance (ATR) mode of freeze-dried raw and biopretreated substrates were collected with a Varian 670IR spectrophotometer (Varian, Santa Clara, CA) fitted with a universal ATR sampling accessory GladiATR (PIKE Technologies, Madison, WI). FTIR-ATR spectra were recorded in the 4000–400 cm−1 spectral range at 4 cm−1 resolution for 64 scans. Spectroscopic acquisition was performed using the software Agilent Resolution Pro (Agilent Technologies Inc, Santa Clara, CA). Spectral data was exported to cvs files and later analyzed using the software Spekwin32 (version 1.71.6.1). Peak height was determined as described by Pandey and Pitman (2003), by constructing a baseline connecting the lowest data points on both sides of the peak. A vertical line drawn from the top of the peak to the baseline was considered as the peak height.
2.5. Statistical analysis The statistical analysis was conducted with the software R (version 3.2.0, R Core Team, 2015). A one-way analysis of variance (ANOVA) was carried out to determine if there were statistically significant differences among the values of total digestibility, lignin and total sugar loss and composition obtained with the different substrates. Then, if the ANOVA confirmed the difference between the mean values, a post hoc analysis (Tukey’s HSD) was applied to determine between which values the difference was significant, considering a level of significance of 0.05. The ANOVA tables and other complementary statistical results (Levene’s test for homogeneity of variance, Tukey multiple comparisons of means) and
figures (box plot, Tukey’s HDS pair comparisons) that may help to verify the statistical significance of the study were included as supplementary material (Statistical analysis).
3. Results and discussion 3.1. Raw material selection Two different types of restrictions were used in order to select the lignocellulosic materials for the study: composition and geographical availability. Wheat straw (WS) was used as control material because the proposed fungal pretreatment has been previously proved and implemented using this crop residue (Salvachúa et al., 2011; López-Abelairas et al., 2013a). The comparison of the results obtained with other agricultural residues would demonstrate the flexibility of the biological pretreatment towards the utilization of other plant species with similar composition in terms of glucan, xylan and lignin. Barley straw (BS) and two different lignocellulosic materials derived from maize, corn stover (CS) and corncob (CC) presented similar composition to wheat straw (Table 1). Then, these four lignocellulosic materials were considered in the study. The second restriction was the geographic availability. According to FAO statistics (Fig. 1), wheat, barley and maize are the main cereals produced in Europe (87.7%), as well as in Spain (89.5%) (FAOSTAT, 2014). In 2009, wheat straw, barley straw and corn stover represented all together more than 79.8% of the available agricultural crop residues in the EU (Scarlat et al., 2010). Local availability is an important factor to take into account for biofuel production from an economical and environmental perspective. To produce lignocellulosic ethanol at commercial scale the eventual cost of production will be determined not only by the performance
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
489
of the conversion process but also by the performance of the entire supply-chain from feedstock production to consumption (Slade et al., 2009). Besides, a comparative environmental performance about the production of the raw materials should be also take into account, e.g using tools such as Life Cycle Assessment (GonzálezGarcía et al., 2010). 3.2. Biological pretreatment Lignin degradation by WRF is associated to variable levels of sugar consumption, obtained from the hydrolysis of holocellulose, for growth requirements. For that reason is important to reach a balanced state between both pathways. The selectivity towards lignin degradation and growth rate is an intrinsic characteristic of each WRF as shown by Salvachúa et al. (2011). When selecting a WRF for the biological pretreatment, both parameters should be considered, as they represent an important effect in the economics of the process. The best results will be that where the lignin reduction is the highest with the lowest sugar consumption in the shortest time. As expected, the behavior of the fungal treatment had different repercussions on each substrate, but in general, a reduction of the glucan fraction and an increase of xylan content were observed (Table 1). The glucan fraction was significantly reduced only with wheat straw (from 33.7 ± 1.30 to 27.6 ± 1.0), while the increase of xylan content was significant for corn stover, barley straw and corncob. Regarding lignin content, it was significantly reduced only with wheat straw and corn stover. Total weight loss and main components weight loss due to fungal growth are represented in Fig. 2. The loss of glucose, TRS and lignin takes into account both the modification of the percentage of each fraction and TW loss. As observed in Fig. 2, the highest TW loss was achieved with wheat straw (30.2 ± 0.3%), followed by corn stover and barley straw (21.1 ± 1.5 and 20.1 ± 2.3%, respectively), while the lowest TW loss was obtained with corncob (12.2 ± 2.4%). The total weight loss produced by the biological pretreatment on wheat straw was about 2.5-fold that obtained with corncob, which indicates that the fungal action was more extensive in the case of wheat straw. Total sugars consumption was the lowest in the case of maize, either with corn stover or corncob (11.5 ± 1.4 and 9.1 ± 1.0%, respectively), followed by barley straw (19.8 ± 1.4%). The highest sugar consumption was found with wheat straw (34.4 ± 1.9%). The highest lignin removal was reached with wheat straw and corn stover (42.3 ± 2.3 and 45.8 ± 3.5%, respectively), while the lower value was that of corncob (17.1 ± 5.3%). Lignin selectivity removal, defined as the ratio between lignin loss and total weight loss, for wheat straw, corn stover, barley straw and corncob were 1.38 ± 0.07, 2.10 ± 0.15, 1.54 ± 0.10 and 1.40 ± 0.38, respectively. A value less than or equal to one indicates that the fungal action is not selective with respect to the other components of the
(A)
Fig. 2. Reduction in weight of the main components of the biomass after fungal pretreatment, total weight, glucose, TRS (Total Reducing Sugars) and lignin.
residue. On the contrary, the higher the ratio the more selective the action of WRF in removing the lignin fraction. Taking this into account the most selective lignin removal was caused by I. lacteus on corn stover. The raw material with the most balanced performance between sugar consumption and lignin degradation was corn stover, which presented the greatest lignin degradation, greatest lignin selectivity removal, lowest total weigth loss and medium glucose consumption. The digestibility of the four residues was measured before and after the biopretreatment (Fig. 3). Barley straw presented the highest initial glucan digestibility (45.3 ± 1.5%). The values reached with wheat straw, corn stover and corncob (31.1 ± 0.6, 32.1 ± 1.5 and 37.6 ± 5.2%, respectively) were similar among them, but significantly lower than that obtained with barley straw. In the case of the xylan digestibility, the highest initial value was measured on wheat straw (48.1 ± 2.4%), which was significantly higher than the initial digestibilities obtained with corn stover and corncob (38.1 ± 0.4 and 31.0 ± 2.5%, respectively), but only slightly higher than that of barley straw (41.8 ± 3.4%, respectively). After fungal pretreatment, significant increases in glucan digestibility were detected with wheat straw and corn stover, while it did not change in the case of barley straw and corncob. Xylan digestibility increased after the fungal pretreatment with all the substrates, with the exception of corncob, which was reduced from 31.0 ± 2.5 to 22.4 ± 3.0%. This reduction in corncob could be attributed to the consumption of sugars by the fungus, without any effect on corncob structure. Fungus metabolized the easier
(B)
Fig. 3. Glucan (A) and xylan (B) digestibility of wheat straw (WS), corn stover (CS), barley straw (BS) and corncob (CC), before (white bar) and after (black bar) the fungal pretreatment with Irpex lacteus.
490
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
Table 2 Comparison of biological pretreatment of different lignocellulosic materials using the white-rot fungus Irpex lacteus. Substrate
I. lacteus strain
Treatment time (d)
Total weight loss (%)
Lignin removal (%)
Lignin selectivity removal
Total digestibility (%)
Reference
Wheat straw
IJFM A792
21
20.4
34
1.42
79.9
Fr.238 617/93
21
30.0
55.6
1.85
94.9
21 20 42
30.2 12.7 23
41.8 25.4 43.8
1.38 2.00 1.90
67.2 66.4 66.9
Salvachúa et al., 2011 López-Abelairas et al., 2013a This study Xu et al., 2010 Song et al., 2013
21 21 21 28 30
21 20.1 12.2 32.2 6.1
44.2 31.0 17.1 34.3 N.D.
2.10 1.54 1.40 1.07 N.D.
71.4 56.3 25.4 82 106.6 mg g−1
This study This study This study Du et al., 2011 Hwang et al., 2008
30
35.8
N.D.
N.D.
103.8 mg g−1
30
21.2
20.16
0.95
N.D.
Fr.238 617/93 CD2 FJ744594 (GenBank) Fr.238 617/93 Fr.238 617/93 Barley straw Corncob Fr.238 617/93 CD2 Cornstalks Strain isolated in Wood chips of white pine Korea Wood chips of tulip Strain isolated in tree Korea Bamboo culm SA 101 (India) Corn stover
a
a
a
Hwang et al., 2008 Albert and Pandya, 2014
Digestibility expressed as mg of glucose released per g of degraded substrate, as reported by the authors.
available sugars of the raw corncob and no other effect on the lignocellulosic structure was detected. There were not significant differences among the final glucan digestibilities of corn stover, wheat straw and barley straw (59.2 ± 3.5, 54.8 ± 4.6 and 53.9 ± 7.7%, respectively), while the lowest value (30.3 ± 4.5%) was obtained with corncob. In the case of xylan, the highest final digestibilities were obtained with wheat straw and corn stover (78.0 ± 4.7 and 82.1 ± 11.5%, respectively), while the lowest values were reached with barley straw and corncob (58.2 ± 4.3 and 22.4 ± 3.0% respectively). In terms of increase of total digestibility, corn stover presented also the highest value (103.4%), followed by wheat straw and barley straw (49.9 and 36.8%, respectively). Analysing together the results of lignin reduction, sugars consumption and total digestibility, the crop residue with the best performance under fungal pretreatment, in these terms, was corn stover. A comparison of these results with those obtained in other studies using also Irpex lacteus is presented in Table 2. It can be observed that from eight lignocellulosic biomasses, corn stover presented the highest lignin delignification and lowest total weight loss which is reflected in a higher lignin selectivity removal; while it obtained the second highest results of final holocellulosic digestibility, after wheat straw. In a previous study, in which 33 strains of WRF were evaluated on three lignocellulosic biomasses (bamboo, rice straw and corn stover) it was found that Irpex lacteus was the fungus that produced the highest saccharification rates after 15 d of fungal pretreatment (Xu et al., 2009). As it was hypothesized in the introduction a clear different fungal behaviour was observed between the different substrates. Since, despite its similar initial composition and digestibility, the structure of lignocellulose varies according to the specie, tissue, origin and growth period and these factors seem to have a repercussion on fungal growth. 3.3. FTIR comparision of raw and biological pretreated substrates The FTIR spectra of the four substrates before and after the biological pretreatment are shown in Fig. 4. The changes in wood chemistry following the decay by brown- and white-rot fungi were previously studied applying FTIR by Pandey and Pitman (2003). In this study we have applied the same FTIR analysis reported by those researchers using the ratio between the intensity of the lignin band at 1505 cm−1 (Ia ) and that of the main carbohydrate bands at 898, 1158, 1375 and 1738 cm−1 (ICH ). This ratio decreases when
the intensity of the Ia band (at 1505 cm−1 ) is reduced or when the bands of the carbohydrate fractions are increased. In Table 3, the ratio Ia /ICH for the four substrates calculated from their corresponding peak heights, as well as its reduction percentage after the biological pretreatment are presented. It was observed that the highest reduction of Ia /ICH ratios (88.5–91.5%) and the lowest final values (0.048–0.076) were obtained with corn stover, the substrate that increased more total carbohydrate digestibility and whose lignin fraction was more selectively degraded by WRF. On the contrary, the highest final Ia /ICH ratios (0.228–0.384) were obtained with corncob. The ratios Ia /I1158 and Ia /I1375 relate the aromatic skeletal vibration (C C) in lignin at 1505 cm−1 (Pandey and Pitman, 2003) with that of C O C asymetric stretching (at 1158 cm−1 ) and symetric C H deformation (at 1375 cm−1 ) of crystalline cellulose and hemicellulose (Bekiaris et al., 2015). The bands at 898 and 1375 cm−1 have been related to the C O C stretching of amorphous cellulose and the un-conjugated C O stretching of hemicellulose, respectively (Bekiaris et al., 2015). When the intensitive values at I1505 were plotted againts the final lignin content of the four substrates, both parameters were linearly correlated (R2 = 0.954). Moreover, we have calculated the mean value of the ratios Ia /I1158 and Ia /I1375 , which reduction during the pretreatment can be related to the selective removal of lignin with respect to crystalline cellulose and hemicellulose. These values were correlated with the final total digestibility of the four substrates. A linear correlation between these two parameters was obtained (R2 = 0.949, Fig. 5). When the initial digestibilities were included an exponential trend was observed. Leaving out the final digestibility of corncob, which was reduced after the fungal pretreatment, a correlation coefficient of 0.941 was obtained (Fig. 5). Previous studies have demonstrated that it is possible to predict the digestibility of lignocellulosic materials with the information obtained from FTIR-ATR and FTIR-PAS spectra applying complex algorithms (Sills and Gossett, 2012; Bekiaris et al., 2015). The structural information obtained from the FTIR analysis supports the results obtained with the other analytical techniques and demonstrates that it is useful to determine the level of effectiveness of the biological pretreatment. 3.4. Ethanol production In order to prove that the pretreated material could be fermented to ethanol without carrying out an addional detoxification step, the alkaline treatment reported by García-Torreiro et al.
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
491
Fig. 4. FTIR spectra of raw (black line) and fungal pretreated (red line) wheat straw (A), corn stover (B), barley straw (C) and corncob (D). Lignin band at 1505 cm−1 (2) and main carbohydrate bands at 898 (5), 1158 (4), 1375 (3) and 1738 cm−1 (1) are indicated. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Table 3 Ratio of the intensity of lignin and carbohydrate bands for the four substrates at initial and final pretreatment time. Substrate
Time
Relative intensities of aromatic skeletal vibration (Ia ) against typical bands for carbohydrates (ICH )a Ia /ICH Ia /I898
Wheat straw Corn stover Barley straw Corncob a
Initial Final Initial Final Initial Final Initial Final
0.2432 0.2638 0.6882 0.0712 0.2152 0.1331 1.3393 0.3842
Ia /I1158 (−8.48) (89.65) (38.12) (71.32)
0.2885 0.1071 0.4136 0.0477 0.1871 0.1110 0.5443 0.3035
Ia /I1375 (62.88) (88.46) (40.65) (44.24)
0.6549 0.1720 0.8964 0.0761 0.3336 0.1745 1.1279 0.3658
Ia /I1738 (73.74) (91.52) (47.69) (67.57)
0.3295 0.2494 0.6550 0.0619 0.1428 0.1404 0.9419 0.2278
(24.31) (89.99) (1.64) (75.84)
The percentage of the reduction of the ratio Ia /ICH between initial and final times is shown in parentheses.
(2016) was applied. It has been previously reported that the combined application of a physico-chemical step and a biological pretreatment can significantly improve the enzymatic hydrolysis of holocellulose (Shirkavand et al., 2016). The ethanol production was carried out applying a sequential SSF configuration, in which the saccharification step was conducted during the first 24 h and then, P. tannophilus was inoculated to start the fermentation. After 94 h of SSF, final ethanol concentrations of 12.5 ± 0.8, 13.5 ± 1.0, 10.8 ± 0.2 and 11.5 ± 1.1 g L−1 were reached with WS, CS, BS and CC, respectively. The final ethanol yields for wheat straw, corn stover, barley straw and corncob were 79 ± 14, 102 ± 8, 91 ± 2 and 106 ± 10 (mg ethanol g−1 dry substrate), respectively. The mass balance considered for calculating these yields are presented as supplementary material (Mass Balance). The ethanol yields obtained with corn
stover, barley straw and corncob were not significantly different among them (p = 0.1116), while the lowest value was obtained with wheat straw. When the yields of the four residues were analysed, the only significant difference was found between wheat straw and corncob (p = 0.035). These results demonstrate that the combination of the biological and alkaline pretreatment increased significantly the holocellulose digestibility, yielding similar ethanol productions with either of the four residues. 4. Conclusions Fungal pretreatment was demonstrated to be a flexible process, in terms of the substrate origin, as shown by the results of final digestibility, lignin removal and total weight loss obtained with
492
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
Fig. 5. Correlation between total digestibility and the ratio Ia /ICH obtained from FTIR analysis. Two correlations are shown: (a) one including only the final total digestibilities (diamonds) and (b) another considering both the initial and final values (circles). In the latter case, the final total digestibility of corncob was not included.
three of the most common agricultural residues available in Europe for producing ethanol. The effect of fungal pretreatment was also reflected in the biomass structural analysis carried out by applying FTIR measurements. Moreover, it was also demonstrated that combining the fungal pretreatment with an alkaline treatment, similar ethanol yields can be obtained with the different evaluated residues without requiring any additional detoxification step. Acknowledgments This work was economically supported by the BioSos Project (CDTI CEN-20091040) and by the Ministry of Economy and Competitiveness of Spain through the Local Investment Fund for Employment—Government of Spain. The authors thank Novozymes for providing the commercial enzymes used in this study. The authors belong to the GRC 2013-032 Research Group and to the strategic group CRETUS (AGRUP2015/02). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.indcrop.2016.05. 036. References Albert, S., Pandya, B., 2014. Pattern of bamboo culm degradation by Daedaleopsis confragrosa when co-cultured with selected fungi. Ann. Plant Sci. 2, 563–574. Alvira, P., Tomás-Pejó, E., Ballesteros, M., Negro, M.J., 2010. Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review. Bioresour. Technol. 101, 4851–4861. Bekiaris, G., Lindedam, J., Peltre, C., Decker, S.R., Turner, G.B., Magid, J., Bruun, S., 2015. Rapid estimation of sugar release from winter wheat straw during bioethanol production using FTIR-photoacoustic spectroscopy. Biotechnol. Biofuels 8, 1–12. Du, W., Yu, H., Song, L., Zhang, J., Weng, C., Ma, F., Zhang, X., 2011. The promoting effect of byproducts from Irpex lacteus on subsequent enzymatic hydrolysis of bio-pretreated cornstalks. Biotechnol. Biofuels 4, 8. FAOSTAT Database, 2014. http://faostat3.fao.org/download/Q/QC/E (last accessed 19.01.16).
García-Torreiro, M., Álvarez Pallín, M., López-Abelairas, M., Lu-Chau, T.A., Lema, J.M., 2016. Alkali treatment of fungal pretreated wheat straw for bioethanol production. Bioethanol 2, 32–43. González-García, S., Moreira, M.T., Feijoo, G., 2010. Comparative environmental performance of lignocellulosic ethanol from different feedstocks. J. Renew. Sustain. Energy 14, 2077–2085. Hatakka, A.I., 1983. Pretreatment of wheat straw by white-rot fungi for enzymic saccharification of cellulose. Appl. Microbiol. Biotechnol. 18, 350–357. Hwang, S.S., Lee, S.J., Kim, H.K., Ka, J.O., Kim, K.J., Song, H.G., 2008. Biodegradation and saccharification of wood chips of Pinus strobus and Liriodendron tulipifera by white rot fungi. J. Microbiol. Biotechnol. 18, 1819–1826. Jönsson, L.J., Alriksson, B., Nilvebrant, N.O., 2013. Bioconversion of lignocellulose: inhibitors and detoxification. Biotechnol. Biofuels 6, 16. López-Abelairas, M., Álvarez Pallín, M., Salvachúa, D., Lu-Chau, T.A., Martínez, M.J., Lema, J.M., 2013a. Optimisation of the biological pretreatment of wheat straw with white-rot fungi for ethanol production. Bioprocess Biosyst. Eng. 36, 1251–1260. López-Abelairas, M., Lu-Chau, T.A., Lema, J.M., 2013b. Fermentation of biologically pretreated wheat straw for ethanol production: comparison of fermentative microorganisms and process configurations. Appl. Biochem. Biotechnol. 170, 1838–1852. Martin, C., Alriksson, B., Sjöde, A., Nilvebrant, N.O., Jönsson, L.J., 2007. Dilute sulfuric acid pretreatment of agricultural and agro-industrial residues for ethanol production. Appl. Biochem. Biotechnol. 137, 339–352. Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugars. Anal. Chem. 31, 426–428. Njoku, S.I., Ahring, B.K., Uellendahl, H., 2012. Pretreatment as the crucial step for a cellulosic ethanol biorefinery: testing the efficiency of wet explosion on different types of biomass. Bioresour. Technol. 124, 105–110. ˇ Cajthaml, T., Svobodova, K., Suˇ ˇ sla, M., Saˇ ˇ sek, V., 2009. Irpex lacteus: a ´ C., Novotny, white-rot fungus with biotechnological potential − review. Folia Microbiol. 54, 375–390. Pandey, K.K., Pitman, A.J., 2003. FTIR studies of the changes in wood chemistry following decay by brown-rot and white-rot fungi. Int. Biodeterior. Biodegrad. 52, 151–160. R Core Team, 2015. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria http://www.R-project. org/. Salvachúa, D., Prieto, A., López-Abelairas, M., Lu-Chau, T.A., Martínez, A.T., Martínez, M.J., 2011. Fungal pretreatment: an alternative in second-generation ethanol from wheat straw. Bioresour. Technol. 102, 7500–7506. Sarkar, N., Ghosh, S.K., Bannerjee, S., Aikat, K., 2012. Bioethanol production from agricultural wastes: an overview. Renew. Energy 37, 19–27. Scarlat, N., Martinov, M., Dallemand, J.F., 2010. Assessment of the availability of agricultural crop residues in the European Union: potential and limitations for bioenergy use. Waste Manag. 30, 1889–1897. Shirkavand, E., Baroutian, S., Gapes, D.J., Young, B.R., 2016. Combination of fungal and physicochemical processes for lignocellulosic biomass pretreatment—a review. Renew. Sustain. Energy Rev. 29, 217–234. Sills, D.L., Gossett, J.M., 2012. Using FTIR to predict saccharification from enzymatic hydrolysis of alkali-pretreated biomasses. Biotechnol. Bioeng. 109, 353–362. Slade, R., Bauen, A., Shah, N., 2009. The commercial performance of cellulosic ethanol supply-chains in Europe. Biotechnol. Biofuels 2, 1–20. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D., 2008. Determination of structural carbohydrates and lignin in biomass. Laboratory Analytical Procedure (LAP). Tech. Rep., NREL/TP-510-42618. Song, L., Ma, F., Zeng, Y., Zhang, X., Yu, H., 2013. The promoting effects of manganese on biological pretreatment with Irpex lacteus and enzymatic hydrolysis of corn stover. Bioresour. Technol. 135, 89–92. Srivastava, A.K., Agrawal, P., Rahiman, A., 2014. Pretreatment and production of bioethanol from different lignocellulosic biomass. Int. J. Adv. Res. 2, 888–896. Wan, C., Li, Y., 2011. Effectiveness of microbial pretreatment by Ceriporiopsis subvermispora on different biomass feedstocks. Bioresour. Technol. 102, 7507–7512. Wang, Q., Wang, Z., Shen, F., Hu, J., Sun, F., Lin, L., Yang, G., Zhang, Y.Z., Deng, S., 2014. Pretreating lignocellulosic biomass by the concentrated phosphoric acid plus hydrogen peroxide (PHP) for enzymatic hydrolysis: evaluating the pretreatment flexibility on feedstocks and particle sizes. Bioresour. Technol. 166, 420–428. Xu, C., Yang, X., Ma, F., Yu, H., Zhang, X., 2009. Screening of white rot fungi for biological pretreatment of lignocellulosic biomass [J]. Food Ferment. Ind. 2, 006. Xu, C., Ma, F., Zhang, X., Chen, S., 2010. Biological pretreatment of corn stover by Irpex lacteus for enzymatic hydrolysis. J. Agric. Food Chem. 58, 10893–10898.
Contents lists available at ScienceDirect
Industrial Crops and Products journal homepage: www.elsevier.com/locate/indcrop
Fungal pretreatment of agricultural residues for bioethanol production M. García-Torreiro, M. López-Abelairas, T.A. Lu-Chau ∗ , J.M. Lema Department of Chemical Engineering, Institute of Technology, Universidade de Santiago de Compostela, 15782, Santiago de Compostela, Spain
a r t i c l e
i n f o
Article history: Received 20 January 2016 Received in revised form 14 May 2016 Accepted 23 May 2016 Keywords: Bioethanol Agricultural residues Lignocellulosic biomass Biological pretreatment Irpex lacteus
a b s t r a c t The biological pretreatment of lignocellulosic biomass for the production of ethanol is an environmentally friendly process that needs to be evaluated with different feedstocks in order to avoid dependence on a single feedstock. In this study, four agricultural residues (corn stover, barley straw, corncob and wheat straw), selected in terms of their composition and geographic availability, were pretreated using the white-rot fungus Irpex lacteus. After the fungal pretreatment, the biggest reduction in lignin content (45.8 ± 3.5%), lowest sugars consumption (11.5 ± 1.4%) and highest lignin selectivity removal (2.1 ± 0.15) were achieved with corn stover. Moreover, total holocellulose digestibility was significantly increased after the biological pretreatment with all the substrates (37–103%), excepting corncob. In this study, fungal pretreatment was successfully applied to three of the most common agricultural residues available in Europe for producing ethanol, demonstrating that it is capable of handling feedstocks of variable origin. © 2016 Elsevier B.V. All rights reserved.
1. Introduction In the European Union, agricultural residues are abundant and cheap lignocellulosic resources that can be used to obtain valuable products such as bioethanol. This raw material has the advantage of not being competitive in food markets, like other substrates derived from plants such as cereals or sugarcane, used as well in the bioethanol industry. Nevertheless, the challenge to obtain the sugar fractions from these substrates, due to its complex and heterogeneous structure, is the main bottleneck of the second generation bioethanol production. Several pretreatments based on physical, chemical or physicochemical mechanisms have been extensively studied (Alvira et al., 2010). They are frequently based on the use of chemicals, extreme conditions of temperature or pressure and special equipment to deal with these extreme conditions. The sugar yield reported with these treatments is used to be very high; nevertheless inhibitory compounds are also produced in different amounts depending on the severity or the chemicals used (Jönsson et al., 2013). In this context, the biological pretreatment appears as an environmental friendly alternative. It is based on the special abilities of a small group of filamentous fungi belonging to the phylum basidiomy-
∗ Corresponding author: Rúa Constantino Candeira s/n, 15782, Santiago de Compostela, Spain. E-mail address: [email protected] (T.A. Lu-Chau). http://dx.doi.org/10.1016/j.indcrop.2016.05.036 0926-6690/© 2016 Elsevier B.V. All rights reserved.
cota, known as white-rot fungi (WRF), which have evolved the capability of degrading lignin, the most recalcitrant component of plant cell wall (Hatakka, 1983), by using a set of extracellular ligninolytic enzymes. With the partial removal of lignin, an easier degradability of the substrate is achieved, without producing inhibitors as the commonly found with the other pretreatments (Salvachúa et al., 2011). In particular, the white-rot basidiomycete Irpex lacteus presents a great biodegradation capability (Novotny´ et al., 2009; Salvachúa et al., 2011). The advantages of the biological pretreatment include low energy requirement and mild conditions. However, there is an inevitable dependence between lignin removal and sugars consumption, since the fungal strategy consists of degrading the lignin fraction in order to reach easily the cellulose and hemicellulose. The other main disadvantage, compared to physicochemical pretreatmets, is the prolonged time needed to reach similar digestibility improvements, which can be as long as 4–8 weeks (Sarkar et al., 2012). This time can be reduced to 2–3 weeks by combining the biological treatment with an alkaline wash at soft conditions and optimizing operational conditions (Salvachúa et al., 2011; López-Abelairas et al., 2013a). In order to be considered a feasible alternative, the biological pretreatment should be evaluated using different feedstocks to demonstrate its flexibility. There are several previous research works that studied the effect of using different lignocellulosic raw materials on the performance of chemical and/or physicochemical pretreatments (Martin et al., 2007; Njoku et al., 2012; Srivastava et al., 2014; Wang et al., 2014), but only one previous work
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
(Wan and Li, 2011) have studied the effect of using a fungal biological pretreatment with another WRF (Ceriporiopsis subvermispora) on different types of feedstocks. In the present study, the biological pretreatment of four lignocellulosic agricultural residues (corn stover, barley straw, corncob and wheat straw) was carried out. These agricultural residues were chosen according to their composition and their availability in Europe. Despite the fact that the main composition of the raw materials is basically the same, a different fungal behavior during the growth on the different plant species could be expected. The efficiency of the fungal process was studied in detail in order to clarify if the fungal performance presents any dependence with the substrate origin. 2. Materials and methods 2.1. Microorganisms and inocula preparation The basidiomycete I. lacteus (Fr. 238 617/93) was obtained from the culture collection of the Chemical Engineering Department of the University of Santiago de Compostela (Spain). Mycelium production in static cultures and fungal inoculum preparation was carried out as described by Salvachúa et al. (2011). The microorganism used for fermenting the lignocellulosic hydrolysates was the yeast Pachysolen tannophilus (CETC 1426). The yeast was maintained in YPD medium plates, composed of (per L): peptone, 10 g; yeast extract, 20 g; glucose 10 g and agar 15 g. For inoculum preparation, yeast from a fresh plate, was transferred to 100 mL of liquid media composed of (per L): malt extract, 20 g; glucose, 20 g and peptone 1 g. Inoculum was incubated in an orbital shaker at 150 rpm and 32.5 ◦ C for 16 h or when an optical density of approx. 1.5 was reached. 2.2. Fungal pretreatment Lignocellulosic residues procured from the local market (Santiago de Compostela, Spain) were used as the lignocellulosic feedstock source for ethanol production after fungal pretreatment. The moisture content of the raw biomasses was in the range of 7.3–8.5%, these values were taking into account for the composition calculations in dry basis. The biological pretreatment parameters and conditions were optimized in a previous work (López-Abelairas et al., 2013a). In brief, 6 g of blended (approx. 0.5 cm of particle size) lignocellulosic material were mixed with 20 mL of tap water in 250 mL Erlenmeyer flasks and sterilized by autoclave at 121 ◦ C during 20 min. After sterilization, 3 mL of fungal inoculum (as pellets) were added to each flask and maintained at 30 ◦ C during 21 d. The biologically pretreated biomass, with its liquid fraction, was freeze-dried. Total weight (TW) loss was calculated as the percentage of total solids lost after pretreatment and part of the dried solid was used for digestibility and composition measures. All the experiments were carried out in triplicate, data in tables and figures are presented with the mean and its standard deviation. 2.3. Ethanol production After the fungal pretreatment, the different substrates were subjected to an alkali recycling treatment as described by GarcíaTorreiro et al. (2016). In this protocol, the solid phases accumulated during up to three successive alkali washes with NaOH (0.36 g NaOH g−1 substrate) were mixed with the remaining liquid phase and citrate buffer 0.1 M. This buffer was added to adjust the solids percentage to 10% (w/v). Then, H2 SO4 was used to adjust the pH between 4.8–5, and 200 g mL−1 of tetracycline was finally added to avoid contamination.
487
Simultaneous saccharification and fermentation (SSF) configuration, as previously described by López-Abelairas et al. (2013b), was chosen to convert the resulting sugars into ethanol. The SSF configuration consists on a short enzymatic hydrolysis followed by the fermentation of the resulting sugars while the hydrolysing enzymes are still acting. The enzymatic doses used were previously optimized for biopretreated biomass subjected to alkali recycling treatment (García-Torreiro et al., 2016). Cellulase and beta-glucosidase were provided by the enzyme mixtures Cellic CTec2 and NS50010, respectively, while beta-xylosidase and xylanase by Cellic Htec2 (Novozymes). The cocktail doses applied were 96, 18 and 5 L g−1 raw substrate of Cellic CTec2, Cellic Htec2 and NS50010, respectively. The hydrolysis was carried out in citrate buffer 0.1 M (pH 4.8) with a solid load of 10% (w/v). After 24 h of incubation at 150 rpm and 50 ◦ C, the yeast P. tannophilus was inoculated using a ratio of 10% v/v of hydrolyzate. The fermentation and simultaneous saccharification was carried out at 37 ◦ C and 150 rpm. Samples were taken periodically during the fermentation. The concentration of total reducing sugars, glucose and ethanol was determined in these samples. 2.4. Analytical protocols 2.4.1. Composition determination An adapted protocol from NREL (National Renewable Energy Laboratory, Golden, USA) was used for the determination of the biomass composition in glucan, xylan and lignin. Acid hydrolysis allows lignin separation from sugar fraction and it can be determined by weight. The concentration of xylan and glucan, were calculated from the concentration of the corresponding monomeric sugars using an anhydro correction of 0.88 and 0.90 for C5 and C6 sugars, respectively. 2.4.2. Glucan and xylan digestibilities Glucan (Dg ) and xylan (Dx ) digestibilities were evaluated and expressed, according to Eqs. (1) and (2), respectively. Dg (%) =
Gr (g) × 100 Gs (g)
(1)
Dx (%) =
X r (g) × 100 Xs (g)
(2)
where Gr and Xr are the glucose and xylose released from the biomass after pretreatment (including the enzymatic hydrolysis step), repectively and the Gs and Xs are the theoretical maximum amount of glucose and xylose available in the raw material, respectively. Total digestibility was calculated using the total amount of reducing sugars, instead of glucose or xylose. This parameter gives an idea of how available the fermentable sugars are. 2.4.3. Determination of substrate and product concentrations Glucose content was measured using HPLC (Sluiter et al., 2008) and an enzymatic colorimetric method (GOD-PAP/Trinder, Spinreact, GI, Spain). No significant differences were found between HPLC and enzymatic method, for that reason the latter was used as routine method for glucose determination. Total reducing sugars (TRS) were determined by the dinitrosalicylic acid (DNS) method (Miller, 1959). The concentration of ethanol was determined using an HPLC system (Hewlett Packard 1100, USA) with an IR 1047 detector and an Aminex HPX-87H column. H2 SO4 (5 mM) was used as the mobile phase with a flow of 0.6 mL min−1 at 60 ◦ C. Samples were filtered through 0.22 m syringe filters prior to HPLC. 2.4.4. FTIR analysis FTIR was used to provide complementary information to understand the structural differences due to the effect of fungal
488
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
Table 1 Composition (% dry basis) of the feedstock used in the study. Wheat straw
Glucan Xylan Lignin Ash Othersa a
Corn stover
Barley straw
Corncob
Raw
Pretreated
Raw
Pretreated
Raw
Pretreated
Raw
Pretreated
33.7 ± 1.3 29.9 ± 4.1 23.4 ± 0.1 4.2 ± 0.04 8.9
27.6 ± 1.0 31.8 ± 1.4 19.5 ± 1.2 7.6 ± 0.2 13.5
35.3 ± 0.2 29.3 ± 0.5 23.2 ± 0.8 4.6 ± 0.2 7.6
33.9 ± 1.6 38.9 ± 5.9 16.4 ± 2.4 5.4 ± 0.3 5.4
34.4 ± 4.8 25.7 ± 3.5 21.1 ± 1.3 5.9 ± 0.1 12.9
28.5 ± 3.3 35.4 ± 3.7 20.2 ± 0.9 8.1 ± 0.5 9.8
30.5 ± 7.6 33.9 ± 3.2 23.6 ± 4.4 2.1 ± 0.05 9.8
24.7 ± 1.4 40.7 ± 1.2 22.1 ± 2.1 3.2 ± 0.7 9.4
Calculated as the difference between 100% and the sum of the composition of the first four components.
Fig. 1. Distribution of the main cereals produced in Europe and Spain (FAOSTAT, 2014).
pretreatment on the four assayed substrates and to relate this information with other more commonly evaluated parameters such as the increase of carbohydrate digestibilty and the reduction of lignin content. In order to carry out this analysis, IR absorption spectra in attenuated total reflectance (ATR) mode of freeze-dried raw and biopretreated substrates were collected with a Varian 670IR spectrophotometer (Varian, Santa Clara, CA) fitted with a universal ATR sampling accessory GladiATR (PIKE Technologies, Madison, WI). FTIR-ATR spectra were recorded in the 4000–400 cm−1 spectral range at 4 cm−1 resolution for 64 scans. Spectroscopic acquisition was performed using the software Agilent Resolution Pro (Agilent Technologies Inc, Santa Clara, CA). Spectral data was exported to cvs files and later analyzed using the software Spekwin32 (version 1.71.6.1). Peak height was determined as described by Pandey and Pitman (2003), by constructing a baseline connecting the lowest data points on both sides of the peak. A vertical line drawn from the top of the peak to the baseline was considered as the peak height.
2.5. Statistical analysis The statistical analysis was conducted with the software R (version 3.2.0, R Core Team, 2015). A one-way analysis of variance (ANOVA) was carried out to determine if there were statistically significant differences among the values of total digestibility, lignin and total sugar loss and composition obtained with the different substrates. Then, if the ANOVA confirmed the difference between the mean values, a post hoc analysis (Tukey’s HSD) was applied to determine between which values the difference was significant, considering a level of significance of 0.05. The ANOVA tables and other complementary statistical results (Levene’s test for homogeneity of variance, Tukey multiple comparisons of means) and
figures (box plot, Tukey’s HDS pair comparisons) that may help to verify the statistical significance of the study were included as supplementary material (Statistical analysis).
3. Results and discussion 3.1. Raw material selection Two different types of restrictions were used in order to select the lignocellulosic materials for the study: composition and geographical availability. Wheat straw (WS) was used as control material because the proposed fungal pretreatment has been previously proved and implemented using this crop residue (Salvachúa et al., 2011; López-Abelairas et al., 2013a). The comparison of the results obtained with other agricultural residues would demonstrate the flexibility of the biological pretreatment towards the utilization of other plant species with similar composition in terms of glucan, xylan and lignin. Barley straw (BS) and two different lignocellulosic materials derived from maize, corn stover (CS) and corncob (CC) presented similar composition to wheat straw (Table 1). Then, these four lignocellulosic materials were considered in the study. The second restriction was the geographic availability. According to FAO statistics (Fig. 1), wheat, barley and maize are the main cereals produced in Europe (87.7%), as well as in Spain (89.5%) (FAOSTAT, 2014). In 2009, wheat straw, barley straw and corn stover represented all together more than 79.8% of the available agricultural crop residues in the EU (Scarlat et al., 2010). Local availability is an important factor to take into account for biofuel production from an economical and environmental perspective. To produce lignocellulosic ethanol at commercial scale the eventual cost of production will be determined not only by the performance
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
489
of the conversion process but also by the performance of the entire supply-chain from feedstock production to consumption (Slade et al., 2009). Besides, a comparative environmental performance about the production of the raw materials should be also take into account, e.g using tools such as Life Cycle Assessment (GonzálezGarcía et al., 2010). 3.2. Biological pretreatment Lignin degradation by WRF is associated to variable levels of sugar consumption, obtained from the hydrolysis of holocellulose, for growth requirements. For that reason is important to reach a balanced state between both pathways. The selectivity towards lignin degradation and growth rate is an intrinsic characteristic of each WRF as shown by Salvachúa et al. (2011). When selecting a WRF for the biological pretreatment, both parameters should be considered, as they represent an important effect in the economics of the process. The best results will be that where the lignin reduction is the highest with the lowest sugar consumption in the shortest time. As expected, the behavior of the fungal treatment had different repercussions on each substrate, but in general, a reduction of the glucan fraction and an increase of xylan content were observed (Table 1). The glucan fraction was significantly reduced only with wheat straw (from 33.7 ± 1.30 to 27.6 ± 1.0), while the increase of xylan content was significant for corn stover, barley straw and corncob. Regarding lignin content, it was significantly reduced only with wheat straw and corn stover. Total weight loss and main components weight loss due to fungal growth are represented in Fig. 2. The loss of glucose, TRS and lignin takes into account both the modification of the percentage of each fraction and TW loss. As observed in Fig. 2, the highest TW loss was achieved with wheat straw (30.2 ± 0.3%), followed by corn stover and barley straw (21.1 ± 1.5 and 20.1 ± 2.3%, respectively), while the lowest TW loss was obtained with corncob (12.2 ± 2.4%). The total weight loss produced by the biological pretreatment on wheat straw was about 2.5-fold that obtained with corncob, which indicates that the fungal action was more extensive in the case of wheat straw. Total sugars consumption was the lowest in the case of maize, either with corn stover or corncob (11.5 ± 1.4 and 9.1 ± 1.0%, respectively), followed by barley straw (19.8 ± 1.4%). The highest sugar consumption was found with wheat straw (34.4 ± 1.9%). The highest lignin removal was reached with wheat straw and corn stover (42.3 ± 2.3 and 45.8 ± 3.5%, respectively), while the lower value was that of corncob (17.1 ± 5.3%). Lignin selectivity removal, defined as the ratio between lignin loss and total weight loss, for wheat straw, corn stover, barley straw and corncob were 1.38 ± 0.07, 2.10 ± 0.15, 1.54 ± 0.10 and 1.40 ± 0.38, respectively. A value less than or equal to one indicates that the fungal action is not selective with respect to the other components of the
(A)
Fig. 2. Reduction in weight of the main components of the biomass after fungal pretreatment, total weight, glucose, TRS (Total Reducing Sugars) and lignin.
residue. On the contrary, the higher the ratio the more selective the action of WRF in removing the lignin fraction. Taking this into account the most selective lignin removal was caused by I. lacteus on corn stover. The raw material with the most balanced performance between sugar consumption and lignin degradation was corn stover, which presented the greatest lignin degradation, greatest lignin selectivity removal, lowest total weigth loss and medium glucose consumption. The digestibility of the four residues was measured before and after the biopretreatment (Fig. 3). Barley straw presented the highest initial glucan digestibility (45.3 ± 1.5%). The values reached with wheat straw, corn stover and corncob (31.1 ± 0.6, 32.1 ± 1.5 and 37.6 ± 5.2%, respectively) were similar among them, but significantly lower than that obtained with barley straw. In the case of the xylan digestibility, the highest initial value was measured on wheat straw (48.1 ± 2.4%), which was significantly higher than the initial digestibilities obtained with corn stover and corncob (38.1 ± 0.4 and 31.0 ± 2.5%, respectively), but only slightly higher than that of barley straw (41.8 ± 3.4%, respectively). After fungal pretreatment, significant increases in glucan digestibility were detected with wheat straw and corn stover, while it did not change in the case of barley straw and corncob. Xylan digestibility increased after the fungal pretreatment with all the substrates, with the exception of corncob, which was reduced from 31.0 ± 2.5 to 22.4 ± 3.0%. This reduction in corncob could be attributed to the consumption of sugars by the fungus, without any effect on corncob structure. Fungus metabolized the easier
(B)
Fig. 3. Glucan (A) and xylan (B) digestibility of wheat straw (WS), corn stover (CS), barley straw (BS) and corncob (CC), before (white bar) and after (black bar) the fungal pretreatment with Irpex lacteus.
490
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
Table 2 Comparison of biological pretreatment of different lignocellulosic materials using the white-rot fungus Irpex lacteus. Substrate
I. lacteus strain
Treatment time (d)
Total weight loss (%)
Lignin removal (%)
Lignin selectivity removal
Total digestibility (%)
Reference
Wheat straw
IJFM A792
21
20.4
34
1.42
79.9
Fr.238 617/93
21
30.0
55.6
1.85
94.9
21 20 42
30.2 12.7 23
41.8 25.4 43.8
1.38 2.00 1.90
67.2 66.4 66.9
Salvachúa et al., 2011 López-Abelairas et al., 2013a This study Xu et al., 2010 Song et al., 2013
21 21 21 28 30
21 20.1 12.2 32.2 6.1
44.2 31.0 17.1 34.3 N.D.
2.10 1.54 1.40 1.07 N.D.
71.4 56.3 25.4 82 106.6 mg g−1
This study This study This study Du et al., 2011 Hwang et al., 2008
30
35.8
N.D.
N.D.
103.8 mg g−1
30
21.2
20.16
0.95
N.D.
Fr.238 617/93 CD2 FJ744594 (GenBank) Fr.238 617/93 Fr.238 617/93 Barley straw Corncob Fr.238 617/93 CD2 Cornstalks Strain isolated in Wood chips of white pine Korea Wood chips of tulip Strain isolated in tree Korea Bamboo culm SA 101 (India) Corn stover
a
a
a
Hwang et al., 2008 Albert and Pandya, 2014
Digestibility expressed as mg of glucose released per g of degraded substrate, as reported by the authors.
available sugars of the raw corncob and no other effect on the lignocellulosic structure was detected. There were not significant differences among the final glucan digestibilities of corn stover, wheat straw and barley straw (59.2 ± 3.5, 54.8 ± 4.6 and 53.9 ± 7.7%, respectively), while the lowest value (30.3 ± 4.5%) was obtained with corncob. In the case of xylan, the highest final digestibilities were obtained with wheat straw and corn stover (78.0 ± 4.7 and 82.1 ± 11.5%, respectively), while the lowest values were reached with barley straw and corncob (58.2 ± 4.3 and 22.4 ± 3.0% respectively). In terms of increase of total digestibility, corn stover presented also the highest value (103.4%), followed by wheat straw and barley straw (49.9 and 36.8%, respectively). Analysing together the results of lignin reduction, sugars consumption and total digestibility, the crop residue with the best performance under fungal pretreatment, in these terms, was corn stover. A comparison of these results with those obtained in other studies using also Irpex lacteus is presented in Table 2. It can be observed that from eight lignocellulosic biomasses, corn stover presented the highest lignin delignification and lowest total weight loss which is reflected in a higher lignin selectivity removal; while it obtained the second highest results of final holocellulosic digestibility, after wheat straw. In a previous study, in which 33 strains of WRF were evaluated on three lignocellulosic biomasses (bamboo, rice straw and corn stover) it was found that Irpex lacteus was the fungus that produced the highest saccharification rates after 15 d of fungal pretreatment (Xu et al., 2009). As it was hypothesized in the introduction a clear different fungal behaviour was observed between the different substrates. Since, despite its similar initial composition and digestibility, the structure of lignocellulose varies according to the specie, tissue, origin and growth period and these factors seem to have a repercussion on fungal growth. 3.3. FTIR comparision of raw and biological pretreated substrates The FTIR spectra of the four substrates before and after the biological pretreatment are shown in Fig. 4. The changes in wood chemistry following the decay by brown- and white-rot fungi were previously studied applying FTIR by Pandey and Pitman (2003). In this study we have applied the same FTIR analysis reported by those researchers using the ratio between the intensity of the lignin band at 1505 cm−1 (Ia ) and that of the main carbohydrate bands at 898, 1158, 1375 and 1738 cm−1 (ICH ). This ratio decreases when
the intensity of the Ia band (at 1505 cm−1 ) is reduced or when the bands of the carbohydrate fractions are increased. In Table 3, the ratio Ia /ICH for the four substrates calculated from their corresponding peak heights, as well as its reduction percentage after the biological pretreatment are presented. It was observed that the highest reduction of Ia /ICH ratios (88.5–91.5%) and the lowest final values (0.048–0.076) were obtained with corn stover, the substrate that increased more total carbohydrate digestibility and whose lignin fraction was more selectively degraded by WRF. On the contrary, the highest final Ia /ICH ratios (0.228–0.384) were obtained with corncob. The ratios Ia /I1158 and Ia /I1375 relate the aromatic skeletal vibration (C C) in lignin at 1505 cm−1 (Pandey and Pitman, 2003) with that of C O C asymetric stretching (at 1158 cm−1 ) and symetric C H deformation (at 1375 cm−1 ) of crystalline cellulose and hemicellulose (Bekiaris et al., 2015). The bands at 898 and 1375 cm−1 have been related to the C O C stretching of amorphous cellulose and the un-conjugated C O stretching of hemicellulose, respectively (Bekiaris et al., 2015). When the intensitive values at I1505 were plotted againts the final lignin content of the four substrates, both parameters were linearly correlated (R2 = 0.954). Moreover, we have calculated the mean value of the ratios Ia /I1158 and Ia /I1375 , which reduction during the pretreatment can be related to the selective removal of lignin with respect to crystalline cellulose and hemicellulose. These values were correlated with the final total digestibility of the four substrates. A linear correlation between these two parameters was obtained (R2 = 0.949, Fig. 5). When the initial digestibilities were included an exponential trend was observed. Leaving out the final digestibility of corncob, which was reduced after the fungal pretreatment, a correlation coefficient of 0.941 was obtained (Fig. 5). Previous studies have demonstrated that it is possible to predict the digestibility of lignocellulosic materials with the information obtained from FTIR-ATR and FTIR-PAS spectra applying complex algorithms (Sills and Gossett, 2012; Bekiaris et al., 2015). The structural information obtained from the FTIR analysis supports the results obtained with the other analytical techniques and demonstrates that it is useful to determine the level of effectiveness of the biological pretreatment. 3.4. Ethanol production In order to prove that the pretreated material could be fermented to ethanol without carrying out an addional detoxification step, the alkaline treatment reported by García-Torreiro et al.
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
491
Fig. 4. FTIR spectra of raw (black line) and fungal pretreated (red line) wheat straw (A), corn stover (B), barley straw (C) and corncob (D). Lignin band at 1505 cm−1 (2) and main carbohydrate bands at 898 (5), 1158 (4), 1375 (3) and 1738 cm−1 (1) are indicated. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Table 3 Ratio of the intensity of lignin and carbohydrate bands for the four substrates at initial and final pretreatment time. Substrate
Time
Relative intensities of aromatic skeletal vibration (Ia ) against typical bands for carbohydrates (ICH )a Ia /ICH Ia /I898
Wheat straw Corn stover Barley straw Corncob a
Initial Final Initial Final Initial Final Initial Final
0.2432 0.2638 0.6882 0.0712 0.2152 0.1331 1.3393 0.3842
Ia /I1158 (−8.48) (89.65) (38.12) (71.32)
0.2885 0.1071 0.4136 0.0477 0.1871 0.1110 0.5443 0.3035
Ia /I1375 (62.88) (88.46) (40.65) (44.24)
0.6549 0.1720 0.8964 0.0761 0.3336 0.1745 1.1279 0.3658
Ia /I1738 (73.74) (91.52) (47.69) (67.57)
0.3295 0.2494 0.6550 0.0619 0.1428 0.1404 0.9419 0.2278
(24.31) (89.99) (1.64) (75.84)
The percentage of the reduction of the ratio Ia /ICH between initial and final times is shown in parentheses.
(2016) was applied. It has been previously reported that the combined application of a physico-chemical step and a biological pretreatment can significantly improve the enzymatic hydrolysis of holocellulose (Shirkavand et al., 2016). The ethanol production was carried out applying a sequential SSF configuration, in which the saccharification step was conducted during the first 24 h and then, P. tannophilus was inoculated to start the fermentation. After 94 h of SSF, final ethanol concentrations of 12.5 ± 0.8, 13.5 ± 1.0, 10.8 ± 0.2 and 11.5 ± 1.1 g L−1 were reached with WS, CS, BS and CC, respectively. The final ethanol yields for wheat straw, corn stover, barley straw and corncob were 79 ± 14, 102 ± 8, 91 ± 2 and 106 ± 10 (mg ethanol g−1 dry substrate), respectively. The mass balance considered for calculating these yields are presented as supplementary material (Mass Balance). The ethanol yields obtained with corn
stover, barley straw and corncob were not significantly different among them (p = 0.1116), while the lowest value was obtained with wheat straw. When the yields of the four residues were analysed, the only significant difference was found between wheat straw and corncob (p = 0.035). These results demonstrate that the combination of the biological and alkaline pretreatment increased significantly the holocellulose digestibility, yielding similar ethanol productions with either of the four residues. 4. Conclusions Fungal pretreatment was demonstrated to be a flexible process, in terms of the substrate origin, as shown by the results of final digestibility, lignin removal and total weight loss obtained with
492
M. García-Torreiro et al. / Industrial Crops and Products 89 (2016) 486–492
Fig. 5. Correlation between total digestibility and the ratio Ia /ICH obtained from FTIR analysis. Two correlations are shown: (a) one including only the final total digestibilities (diamonds) and (b) another considering both the initial and final values (circles). In the latter case, the final total digestibility of corncob was not included.
three of the most common agricultural residues available in Europe for producing ethanol. The effect of fungal pretreatment was also reflected in the biomass structural analysis carried out by applying FTIR measurements. Moreover, it was also demonstrated that combining the fungal pretreatment with an alkaline treatment, similar ethanol yields can be obtained with the different evaluated residues without requiring any additional detoxification step. Acknowledgments This work was economically supported by the BioSos Project (CDTI CEN-20091040) and by the Ministry of Economy and Competitiveness of Spain through the Local Investment Fund for Employment—Government of Spain. The authors thank Novozymes for providing the commercial enzymes used in this study. The authors belong to the GRC 2013-032 Research Group and to the strategic group CRETUS (AGRUP2015/02). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.indcrop.2016.05. 036. References Albert, S., Pandya, B., 2014. Pattern of bamboo culm degradation by Daedaleopsis confragrosa when co-cultured with selected fungi. Ann. Plant Sci. 2, 563–574. Alvira, P., Tomás-Pejó, E., Ballesteros, M., Negro, M.J., 2010. Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review. Bioresour. Technol. 101, 4851–4861. Bekiaris, G., Lindedam, J., Peltre, C., Decker, S.R., Turner, G.B., Magid, J., Bruun, S., 2015. Rapid estimation of sugar release from winter wheat straw during bioethanol production using FTIR-photoacoustic spectroscopy. Biotechnol. Biofuels 8, 1–12. Du, W., Yu, H., Song, L., Zhang, J., Weng, C., Ma, F., Zhang, X., 2011. The promoting effect of byproducts from Irpex lacteus on subsequent enzymatic hydrolysis of bio-pretreated cornstalks. Biotechnol. Biofuels 4, 8. FAOSTAT Database, 2014. http://faostat3.fao.org/download/Q/QC/E (last accessed 19.01.16).
García-Torreiro, M., Álvarez Pallín, M., López-Abelairas, M., Lu-Chau, T.A., Lema, J.M., 2016. Alkali treatment of fungal pretreated wheat straw for bioethanol production. Bioethanol 2, 32–43. González-García, S., Moreira, M.T., Feijoo, G., 2010. Comparative environmental performance of lignocellulosic ethanol from different feedstocks. J. Renew. Sustain. Energy 14, 2077–2085. Hatakka, A.I., 1983. Pretreatment of wheat straw by white-rot fungi for enzymic saccharification of cellulose. Appl. Microbiol. Biotechnol. 18, 350–357. Hwang, S.S., Lee, S.J., Kim, H.K., Ka, J.O., Kim, K.J., Song, H.G., 2008. Biodegradation and saccharification of wood chips of Pinus strobus and Liriodendron tulipifera by white rot fungi. J. Microbiol. Biotechnol. 18, 1819–1826. Jönsson, L.J., Alriksson, B., Nilvebrant, N.O., 2013. Bioconversion of lignocellulose: inhibitors and detoxification. Biotechnol. Biofuels 6, 16. López-Abelairas, M., Álvarez Pallín, M., Salvachúa, D., Lu-Chau, T.A., Martínez, M.J., Lema, J.M., 2013a. Optimisation of the biological pretreatment of wheat straw with white-rot fungi for ethanol production. Bioprocess Biosyst. Eng. 36, 1251–1260. López-Abelairas, M., Lu-Chau, T.A., Lema, J.M., 2013b. Fermentation of biologically pretreated wheat straw for ethanol production: comparison of fermentative microorganisms and process configurations. Appl. Biochem. Biotechnol. 170, 1838–1852. Martin, C., Alriksson, B., Sjöde, A., Nilvebrant, N.O., Jönsson, L.J., 2007. Dilute sulfuric acid pretreatment of agricultural and agro-industrial residues for ethanol production. Appl. Biochem. Biotechnol. 137, 339–352. Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugars. Anal. Chem. 31, 426–428. Njoku, S.I., Ahring, B.K., Uellendahl, H., 2012. Pretreatment as the crucial step for a cellulosic ethanol biorefinery: testing the efficiency of wet explosion on different types of biomass. Bioresour. Technol. 124, 105–110. ˇ Cajthaml, T., Svobodova, K., Suˇ ˇ sla, M., Saˇ ˇ sek, V., 2009. Irpex lacteus: a ´ C., Novotny, white-rot fungus with biotechnological potential − review. Folia Microbiol. 54, 375–390. Pandey, K.K., Pitman, A.J., 2003. FTIR studies of the changes in wood chemistry following decay by brown-rot and white-rot fungi. Int. Biodeterior. Biodegrad. 52, 151–160. R Core Team, 2015. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria http://www.R-project. org/. Salvachúa, D., Prieto, A., López-Abelairas, M., Lu-Chau, T.A., Martínez, A.T., Martínez, M.J., 2011. Fungal pretreatment: an alternative in second-generation ethanol from wheat straw. Bioresour. Technol. 102, 7500–7506. Sarkar, N., Ghosh, S.K., Bannerjee, S., Aikat, K., 2012. Bioethanol production from agricultural wastes: an overview. Renew. Energy 37, 19–27. Scarlat, N., Martinov, M., Dallemand, J.F., 2010. Assessment of the availability of agricultural crop residues in the European Union: potential and limitations for bioenergy use. Waste Manag. 30, 1889–1897. Shirkavand, E., Baroutian, S., Gapes, D.J., Young, B.R., 2016. Combination of fungal and physicochemical processes for lignocellulosic biomass pretreatment—a review. Renew. Sustain. Energy Rev. 29, 217–234. Sills, D.L., Gossett, J.M., 2012. Using FTIR to predict saccharification from enzymatic hydrolysis of alkali-pretreated biomasses. Biotechnol. Bioeng. 109, 353–362. Slade, R., Bauen, A., Shah, N., 2009. The commercial performance of cellulosic ethanol supply-chains in Europe. Biotechnol. Biofuels 2, 1–20. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D., 2008. Determination of structural carbohydrates and lignin in biomass. Laboratory Analytical Procedure (LAP). Tech. Rep., NREL/TP-510-42618. Song, L., Ma, F., Zeng, Y., Zhang, X., Yu, H., 2013. The promoting effects of manganese on biological pretreatment with Irpex lacteus and enzymatic hydrolysis of corn stover. Bioresour. Technol. 135, 89–92. Srivastava, A.K., Agrawal, P., Rahiman, A., 2014. Pretreatment and production of bioethanol from different lignocellulosic biomass. Int. J. Adv. Res. 2, 888–896. Wan, C., Li, Y., 2011. Effectiveness of microbial pretreatment by Ceriporiopsis subvermispora on different biomass feedstocks. Bioresour. Technol. 102, 7507–7512. Wang, Q., Wang, Z., Shen, F., Hu, J., Sun, F., Lin, L., Yang, G., Zhang, Y.Z., Deng, S., 2014. Pretreating lignocellulosic biomass by the concentrated phosphoric acid plus hydrogen peroxide (PHP) for enzymatic hydrolysis: evaluating the pretreatment flexibility on feedstocks and particle sizes. Bioresour. Technol. 166, 420–428. Xu, C., Yang, X., Ma, F., Yu, H., Zhang, X., 2009. Screening of white rot fungi for biological pretreatment of lignocellulosic biomass [J]. Food Ferment. Ind. 2, 006. Xu, C., Ma, F., Zhang, X., Chen, S., 2010. Biological pretreatment of corn stover by Irpex lacteus for enzymatic hydrolysis. J. Agric. Food Chem. 58, 10893–10898.