Optimization of pretreatment, enzymatic hydrolysis and fermentation for more efficient ethanol production by Jerusalem artichoke stalk

Accepted Manuscript

Optimization of pr etr eatment, enzymatic hydr olysis and fer mentation for mor e efficient ethanol pr oduction by Jer usalem ar tichoke stalk Kai Li, Jin-Cheng Qin, Chen-Guang Liu, Feng-Wu Bai PII: DOI: Reference:

S0960-8524(16)31276-7 http://dx.doi.org/10.1016/j.biortech.2016.09.021 BITE 17041

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

27 July 2016 4 September 2016 6 September 2016

Please cite this article as: Li, K., Qin, J-C., Liu, C-G., Bai, F-W., Optimization of pr etr eatment, enzymatic hydr olysis and fer mentation for mor e efficient ethanol pr oduction by Jer usalem ar tichoke stalk, Bioresource Technology (2016), doi: http://dx.doi.org/10.1016/j.biortech.2016.09.021

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Optimization of pretreatment, enzymatic hydrolysis and fermentation for more efficient ethanol production by Jerusalem artichoke stalk Kai Lia, Jin-Cheng Qinb, Chen-Guang Liua*, Feng-Wu Baia, b a

State Key Laboratory of Microbial Metabolism, and School of Life Sciences &

Biotechnology, Shanghai Jiao Tong University, Shanghai, 200240, China b

School of Life Science and Biotechnology, Dalian University of Technology, Dalian,

Liaoning, 116023, China

Email: Kai Li: [email protected] Jin-Cheng Qin: [email protected] Chen-Guang Liu: [email protected] Feng-Wu Bai: [email protected]

Abstract Jerusalem artichoke (JA) is a potential energy crop for biorefinery due to its unique agronomic traits such as resistance to environmental stresses and high biomass yield in marginal lands. Although JA tubers have been explored for inulin extraction and biofuels production, there is little concern on its stalk (JAS). In this article, the pretreatment of JAS by alkaline hydrogen peroxide was optimized using the response surface methodology to improve sugars yield and reduce chemicals usage. Scanning electron microscopy, X-ray diffraction, and thermogravimetric analysis were applied to characterize the structures of the pretreated JAS to evaluate the effectiveness of the

*

Corresponding author.Tel: +86-21-3420-5125, fax: +86-21-3420-5708 ;E-mail:[email protected] 1

pretreatment. Furthermore, the feeding of the pretreated JAS and cellulase was performed for high solid uploading (up to 30%) to increase ethanol titer, and simultaneous saccharification and fermentation with 55.6 g/L ethanol produced, 36.5% more than that produced through separate hydrolysis and fermentation, was validated to be more efficient.

Keywords：Jerusalem artichoke stalk; pretreatment; response surface methodology; enzymatic hydrolysis; ethanol fermentation

1. Introduction At present, starch- and sugar-based biomass is still major resources for the production of biofuels and bio-based chemicals, which is not sustainable, particularly in developing countries. Therefore, non-food related feedstock has to be developed. Jerusalem artichoke (JA) is tolerant to environmental stresses such as drought, salinity and plant diseases and pests, which can thus grow well in marginal lands with high biomass yield, making it an alternative energy crop (Long et al., 2016). The major biomass of JA is from its tubers (JAT) and stalk (JAS). While JAT has been explored for inulin extraction (Li et al., 2012; Li et al., 2015) and biofuel production as well (Matías et al., 2015; Sarchami et al., 2014; Gunnarsson et al., 2014), less concerns have been focused on how to utilize JAS to credit the JAT biorefinery. Kim et al. (2013; 2014) converted the mixture of JAS and JAT into ethanol through simultaneous saccharification and fermentation (SSF) by the inulinase-producing yeast Kluyveromyces marxianus, but apparently such a process compromised the advantage for JAT to be used for producing value-added inulin and other products. Most recently, Khatun et al. (2015) explored lignocellulosic ethanol production solely from JAS by an 2

engineered inulinase-producing yeast Saccharomyces cerevisiae. However, almost no research on the integration and optimization of the major unit operations has been reported. Pretreatment is the first step of lignocellulose bioconversion, which aims to destroy the rigid structure of the feedstock and separate major components cellulose, hemicelluloses and lignin from each other for more efficient hydrolysis of the cellulose component. Among various pretreatment technologies, alkaline hydrogen peroxide (AHP) pretreatment could effectively remove lignin at moderate temperature (Banerjee et al., 2012; Correia et al., 2013; Mou et al., 2013), leading to high cellulose recovery and low inhibitor formation. The second unit operation is the enzymatic hydrolysis of the cellulose component in the pretreated JAS to release glucose for microbial fermentation. Although the development of more efficient cellulase is essential, it is vital to explore the process engineering strategy to maximize the enzyme’s potential. In order to meet the industrial requirement for more than 5% ethanol produced during the fermentation, glucose released from the enzymatic hydrolysis should be higher than 12%, resulting the uploading of the pretreatment biomass at least 20% (Chu et al., 2012). But the high solids uploading significantly deteriorates mixing and mass transfer performance of the hydrolysis and fermentation system, making it a necessities for optimizing the feeding of both the feedstock and enzyme (Chu et al., 2012; Yang et al., 2010; Olofsson et al., 2010; Liu et al., 2015; Unrean et al., 2015). Fermentation can be coupled with the saccharification (simultaneous saccharification and fermentation, SSF) or separated from the unit operation (separate hydrolysis and fermentation, SHF). Compared to SHF, SSF enables the fermentation system to maintain at low sugar levels, which consequently alleviate substrate inhibition, 3

particularly the inhibition of glucose in the cellulase activity, and in the meantime decrease the contamination risk for ethanol fermentation. However, temperature for ethanol fermentation by yeast is much lower than that for cellulase to efficiently hydrolyze the cellulose component. On the other hand, hydrolysate of the pretreated biomass presents various environmental stresses on yeast growth and ethanol fermentation, and the self-flocculating yeast exhibits better tolerance to these stresses (Liu et al., 2012). In this study, AHP pretreatment was optimized by the response surface methodology (RSM) analysis, followed by the studies on uploading strategies for the pretreated biomass and enzyme and ethanol fermentation by the SSF or SHF processes using the self-flocculating S. cerevisiae SPSC01. At the end, the integrate process for ethanol production from JAS was assessed.

2. Materials and Methods

2.1 Feedstock, strain and culture medium JAS harvested from Dongying (Shandong Province), Yancheng (Jiangsu Province) and Yinchuan (Ningxia Province), China (Table 1), was dried naturally and milled to a size range of 1-10 mm and rinsed by water to remove dust and other impurities for reliable results, then dried at 50 °C for 48 h in an oven. The self-flocculating yeast Saccharomyces cerevisiae SPSC01 developed at the authors’ laboratory and deposited at China General Microbiological Culture Collection Center (CGMCC) with the reference number of CGMCC No. 1602 was employed in this work. The YPD medium for seed culture consists of (g/L): glucose 30, yeast extract 4 and peptone 3.

4

2.2 AHP pretreatment and regression optimization JAS was pretreated by the mixture of NaOH (1%, 2% or 3%, w/w) and H2O2 (2%, 3% or4%, v/v) in 5 L flasks at 10% (w/v) solid-liquid ratio and 121°C for 90 min. The treated JAS was washed to neutral pH by water and dried at 45°C for 48 h in an oven. The central composite design was chosen to evaluate the effect of the pretreatment with NaOH and H2O2 on JAS by the software Design expert 8.0 (Statease, USA, MN). The model quality was estimated by variance analysis (ANOVA). As the response values (Y), the contents of cellulose, hemicelluloses, lignin and biomass recovery were fitted in the form of a quadratic polynomial equation: k

k

k

Y = β 0 + ∑ β i xi + ∑ β ij xi x j + ∑ β ii xi 2 i =1

i< j

i =1

(1)

Where, xi and xj are independent variables. β0, βi, βij and βii are interception effect, linear effect, linear-by-linear interaction and quadratic effect, respectively (Liu et al., 2014)

2.3 Chemical composition analysis The chemical compositions of JAS including cellulose, hemicelluloses, lignin and ash were analyzed by the NREL laboratory analytical procedures with the two-step acid hydrolysis (Sluiter et al., 2011). Briefly, JAS was firstly hydrolyzed by concentrated acid (72% H2SO4 at 30oC for 1 h), and then the mixture was diluted by deionized water and hydrolyzed in autoclave by diluted acid (4% H2SO4 at 121 oC for 1 h). Finally, solid residues collected through filtration were dried to determine the acid insoluble lignin and ash, and the filtrate was collected to analyze the chemical compositions of cellulose, hemicelluloses and acid soluble lignin. The carbohydrates and metabolites in the supernatant were determined by HPLC 5

(Waters 2695, Waters, Taunton, MA) with the column (Biored Aminex HPX-87H, 300 mm × 7.8 mm, Hercules, CA) and Waters 410 refractive detector. A flow rate of 0.6 mL/min was applied with 10 mmol/L H2SO4 as the mobile phase. The temperature of column and detector was 65°C and 50°C, respectively.

2.4 Hydrolysis analysis The amount of 10 g JAS or AHP-JAS was enzymatic hydrolyzed by the commercial cellulase (Cellic CTec2, Novozyme, Bagsvaerd, Denmark) at the loading 20 FPU/g substrate in 100 mL citric acid-Na2PO4 buffer solution (0.1 mol/L sodium citrate and 0.2 mol/L Na2HPO4, pH4.8) at 50°C. In order to elevate the impact of the mass loading on the enzymatic hydrolysis, fed-batch strategy was performed with 5 g biomass and corresponding cellulase supplemented into the reaction vessel every 12 h until reaching 30% solid-liquid ratio.

2.5 Morphological analysis Scanning electron microscopy (SEM: Quanta 450, FEI, USA) was used to observe the morphology of JAS. The samples coated with gold under high vacuum conditions were fixed on the aluminum sample stubs. Images were acquired with the 20 kV acceleration voltage. The X-ray diffraction (XRD) analysis of the samples was obtained by the X-ray diffractometer (D/MAX-2400, RIGAKU, Japan). The lignocellulosic biomass was scanned in the range of 10–80° (2θ) with a step size of 0.02° and step time of 1 s at 40 kV and 100 mA under 25°C. The crystallinity index (CrI) is defined as: CrI = (I002 − Iam)/I002 × 100% (2) Where I002 is the maximal intensity of crystalline portion, and Iam is the intensity 6

attributed to the amorphous portion. (Xu et al., 2013) Thermal stability was detected by thermogravimetric analysis (TGA) with a simultaneous thermal analyzer (TGA/Q500, TA Instruments, New Castle, DE). A sample of 5 mg was heated from 30°C to 500°C at a speed of 10°C/min with nitrogen as the carrier gas. (Liu et al., 2016)

2.6 SSF and SHF Simultaneous saccharification and fermentation (SSF) was initialed with 10 g AHP-JAS in 100 ml water and 10% inoculation level supplemented with 20 FPU cellulase/g substrate, and performed at 30 oC. Separate hydrolysis and fermentation (SHF) was started also at 10% AHP-JAS loading, and the hydrolysis was performed at 50 oC for 12 h, 36 h and 60 h, respectively, for 10%, 20% and 30% AHP-JAS loading. Then, the temperature was decreased to 30 oC for ethanol fermentation by the self-flocculating yeast with 10% inoculation. The stirring rate was kept at 150 rpm for all the processes. The biomass loading was same for SSF and SHF. Every 12 h, 5 g AHP-JAS and 100 FPU cellulase were supplemented into the flasks until the total solid-liquid ratio reaching to 30%. Ethanol fermentation in SHF began after 12 h when the biomass loading increased to 20% and 30%, which was lasted for another 24 h.

3. Results and discussion

3.1 JAS composition Jerusalem artichoke fits varied environments under which food crops grow poorly, thus the component of JAS should be dependent on the conditions of soils and climates. As listed in Table 1, the lowest cellulose and the highest lignin were found in JAS 7

harvested from the saline land in Yancheng, Jiangsu province, but JAS grown in saline land in Dongying, Shandong province contained 37.06% cellulose, the highest among all the JAS, but less hemicelluloses and lignin than JAS from Yancheng. The JAS from Yinchuan showed the highest carbohydrates of 58.14% and the lowest lignin of 17.26%, because Yinchuan in Ningxia province is located near the desert with big day-night temperature difference and good sunlight irradiation, which favor sugar accumulation in the plant. Therefore, JAS harvested in Yinchuan was selected as the feedstock for ethanol production.

Table 1

3.2 Optimization of pretreatment conditions Based on the central composite design, 9 experimental runs were performed for the AHP pretreatment with contents of cellulose, hemicelluloses, lignin, and biomass recovery as the response values. As illustrated in Table 2, NaOH and H2O2 significantly removed lignin, and consequently increased the content of cellulose, but 30-40% biomass was lost. In order to identify the effect of NaOH and H2O2, the regression coefficient and ANOVA are analyzed. Second order models were suggested to describe the correlation between chemical dosage and the response values, which displayed high R2 (>0.95) for good fitting between the predicted and experimental data, and low p-value (<0.01) indicating that the reliability of the models were higher than 99%. Whereas, the intricate variation of lignin led to a simple first order regression model, the best choice among linear, quadratic, and cubic models. The model with p-value < 0.05 was accepted in respect of trend analysis, though the R2 was too low to predict the data. According to the same positive and negative coefficients of NaOH and H2O2, it 8

indicated that these two chemicals behaved consistently on the components and biomass recovery. However, the coefficients of NaOH were 2.47, 1.54, 2.59, and 5.19 times larger than that with H2O2 for cellulose, hemicelluloses, lignin and biomass recovery, respectively, thus NaOH acted as a major factor with more significant influence on the process. This conclusion could also be deduced in Fig. 1, since the color changed notably along as more NaOH was supplemented.

Table 2

In Fig. 1, red color representing the biggest values emerged in the four sub-figures for cellulose, hemicelluloses, lignin and biomass recovery, respectively, but only the cellulose change included the closed optimal conditions in the range of the experiment design. The highest cellulose content was able to reach 51.18% under the pretreatment by 2.8% NaOH and 3.2% H2O2. However, other factors should be considered together for the benefit of whole cellulosic ethanol production such as the carbohydrate content, biomass recovery, lignin, and chemical dosage. As a result, the combination of 2% NaOH and 4% H2O2 was screened from several optional conditions provided by the software Design Expert based on the sequential importance analysis.

Fig. 1

3.3 Morphological and structural changes of JAS Morphological and structural changes of JAS are important for improving the sugar yield of the enzymatic hydrolysis due to their impact on cellulase accessibility. SEM, XRD and TGA were thus employed to provide evidences for these changes before 9

and after the pretreatment. The significant changes in morphology were observed through SEM. Compared to raw JAS with compact surface and regular shape, both NaOH and H2O2 dissolved the lignin component and created many punctures in the smooth surface. The combined usage of NaOH and H2O2 further altered the microstructure of JAS, not only for forming more porous hollows, but also for destroying the native structure, which was tight, regular, and cross-link among different components, and thus restricted enzymatic hydrolysis of the cellulose component. The NaOH-H2O2 pretreatment converted JAS into nonstructural conditions due to the delignification, and consequently enhanced cellulase accessibility to the cellulose component. The XRD intensity profile collected for raw JAS presented a well resolved spectrum of cellulose I, also known as the crystalline cellulose, with the two characteristic reflections (Iam) and (I002) at 2θ values of 13−18° and near 22.6°, by which the CrI could be calculated according to the equation (2). Usually, the low CrI means a high amount of amorphous cellulose presented in materials. The CrI of all pretreated JAS was higher than that for raw JAS one, increased by 9.56%, 16.51% and 21.87% for the pretreatment by 2% NaOH, 4% H2O2, and 2% NaOH+4%H2O2, respectively. Neither NaOH nor H2O2 was capable of disrupting the crystal structure of cellulose notably in JAS, but both of them could perform efficiently in delignification. Therefore, the increased CrI of JAS after the pretreatment would attribute to the unconverted crystalline cellulose. However, high content cellulose in the pretreated JAS and its good exposure to cellulase would benefit the enzymatic hydrolysis. Thermogravimetric analysis in non-oxidative conditions was used to evaluate the thermal stability of lignocellulose. As shown in Fig. 2A, the more weight loss of raw JAS below ~240°C meant the removal of volatile components, which were removed at 10

121°C for the pretreated JAS. On the other hand, the minimal residues for AHP-JAS at ~370°C indicated that the AHP pretreatment broke down a quantity of heat stable components. These phenomena confirmed that the pretreatment benefited the increase of carbohydrate content which exhibited steep decrease profiles from 240°C to 370°C. In spite of the difference among raw and pretreated JAS observed in TGA, it was inconvenient to analyze carbohydrate properties between 240°C and 370°C due to the overlaps. Thus differential thermal gravity (DTG) transformed from TGA was adopt to unveil the correlation between the weight loss and temperature. As can be seen in Fig. 2B, the peak around 350°C for the thermal decomposition of cellulose became narrower and higher after the pretreatment. AHP-JAS showed the highest peak for the weight loss, which reflected the worst thermal stability and the highest cellulose content. These results indicated that the removal of lignin by the pretreatment loosened the compact structure of JAS and facilitated decomposition of some compounds, and AHP-JAS obtained larger cellulose content and looser construction.

Fig. 2

3.4 Enzymatic hydrolysis of the pretreated JAS Raw and AHP-JAS were enzymatically hydrolyzed to determine glucose yield from cellulose under a step-by-step increase in the solid-liquid ratio from 10% to 30%. The effectiveness of the AHP process was significant delignification and the increase in cellulose content, which benefited the high solid-liquid ratio loading for the enzymatic hydrolysis and fermentation. The optimal dosage for cellulase supplementation was 20 FPU/g, since no more glucose was released when the enzyme dosage was further increased to 25 FPU/g. On the other hand, the rate for glucose production decreased 11

drastically after 12 h. Therefore, the retention time for the enzymatic hydrolysis should be controlled within 12 h. As shown in Fig. 3A, after the enzymatic hydrolysis for 12 h, 10% raw JAS produced only 9 g/L glucose and 1.7 g/L xylose, but 22 g/L glucose and 8 g/L xylose were produced from 10% AHP-JAS. The difference in glucose production between raw and AHP-JAS was more significant when the feeding of the feedstock and enzyme was undertaken in an interval of 12 h. Continuous accumulation of glucose was observed with AHP-JAS, but not with raw JAS. The delignification by AHP led to a rapid hydrolysis of the cellulose component in AHP-JAS, and consequently improved mixing and mass transfer. As a result, more feedstock could be supplemented through fed-batch to increase glucose concentration without significant impact on glucose yield, compared to batch operation under high feedstock loading (Fig. 3B). In order to increase glucose concentration, more feedstock and cellulase need to be supplemented. However, when all feedstock and cellulase were fed at the beginning, the high viscosity would make mixing extremely difficult. Fortunately, this problem could be addressed properly by the stepwise feeding of the feedstock and cellulase. As shown in Fig. 3A, much higher glucose concentration was achieved, and in the meantime glucose yield was not significantly compromised (Fig. 3B).

Fig. 3

3.5 Ethanol production through SSF or SHF from AHP- JAS Ethanol production through SSF or SHF from AHP-JAS under different biomass uploading conditions is illustrated in Fig. 4. As can be seen, 0.43 g/L and 2.6 g/L more ethanol was produced through the SHF process compared to that produced through the 12

SSF process when 10% and 20% AHP-JAS was uploaded, respectively, because the enzymatic hydrolysis at 50 oC released glucose more efficiently. However, when AHP-JAS uploading increased to 30%, much less ethanol was produced by the SHF process due to the inhibition of glucose in the activity of cellulase, which consequently resulted in incomplete hydrolysis of the cellulose component with lower glucose yield. As a result, 55.65 g/L ethanol was produced through the SSF process during which no glucose accumulation was observed compared to only 40.78 g/L ethanol produced through the SHF process.

Fig. 4

3.6 Overall mass balance Biomass loss with single step is inevitable, thus an overall mass balance is necessary to evaluate the integrate process. As shown in Fig. 5 developed from experimental data in Table 2 and Fig. 4C at 30% AHP-JAS uploading, although 37.5% biomass was removed from raw JAS during the AHP process, cellulose loss was much lower and its recovery was as high as 96.42%, while the mass recovery of hemicelluloses and lignin were only 47.56% and 25.71%, respectively. For the SHF process, 51.44% glucose was released from AHP-JAS and converted to ethanol, and ethanol yield based on the total sugar was less than 0.270 g/g. On the other hand, the SSF process produced 55.65 g/L ethanol, and ethanol yield was increased to 0.368 g/g. Since most xylose was removed during the pretreatment, the self-flocculating yeast used in this work was good enough for ethanol production. Certainly, it would be better to develop yeast strains able to metabolize both pentose and hexose sugars to further 13

improve ethanol yield.

Fig. 5

Among limited studies on ethanol production from JAS (Table 3), as high as 55.65 g/L ethanol was produced in this work, but ethanol yield was lower due to the high mass uploading of 30%. Compared with ethanol production from the mixture of JAS and JAT, JAS contains more non-sugar components, which compromises the overall ethanol yield. However, JAT can be explored as food ingredients and other value-added products (Li et al., 2015). There, it is not reasonable to produce ethanol from JAT.

Table 3

As for AHP, different feedstocks exhibited similar overall ethanol yield except for cashew apple bagasse. The final ethanol titer thus was mainly determined by the mass loading during fermentation. In this study, 30% AHP-JAS loading enabled to achieve high ethanol titer. No significant different was observed in AHP-JAS compared to other lignocellulosic residues.

4. Conclusions The optimal conditions for JAS pretreatment by AHP were 2% NaOH and 4% H2O2, which were validated by SEM, XRD and TGA. The enzymatic hydrolysis of the AHP -JAS indicated significant improvement in glucose yield under high biomass loading up to 30%, with 93.8 g/L glucose released, and 55.6 g/L ethanol produced by the SSF process, 36.5% more than that produced through the SHF process. These results 14

provide insights for utilizing JAS for ethanol production.

Acknowledgements This work was financially supported by the National Natural Science Foundation of China (NSFC) with grant numbers of 21406030 and 51561145014, and China Postdoctoral Science Foundation Special Project (2015T80240).

Appendix A. Supplementary data Supplementary data associated with this article can be found at http://xxx

References 1. Banerjee, G., Car, S., Liu, T., Williams, D.L., Meza, S.L., Walton, J.D., Hodge, D.B., 2012. Scale-up and integration of alkaline hydrogen peroxide pretreatment, enzymatic hydrolysis, and ethanolic fermentation. Biotechnol. Bioeng. 109, 922-931. 2. Cao, W., Sun, C., Liu, R., Yin, R., Wu, X., 2012. Comparison of the effects of five pretreatment methods on enhancing the enzymatic digestibility and ethanol production from sweet sorghum bagasse. Bioresour. Technol. 111, 215-221. 3. Chu, D., Zhang, J., Bao, J., 2012. Simultaneous saccharification and ethanol fermentation of corn stover at high temperature and high solids loading by a thermo tolerant strain Saccharomyces cerevisiae DQ1. Bioenergy. Res. 5, 1020-1026. 4. Correia, J.A.D.C., Júnior, J.E.M., Gonçalves, L.R.B., Rocha, M.V.P., 2013. Alkaline hydrogen peroxide pretreatment of cashew apple bagasse for ethanol production: study of parameters. Bioresour. Technol. 139, 249-256. 15

5. da-Costa, J.A., Marques, J.E., Jr-Goncalves, L.R., Rocha, M.V., 2015. Enhanced enzymatic hydrolysis and ethanol production from cashew apple bagasse pretreated with alkaline hydrogen peroxide. Bioresour. Technol. 179, 249-259. 6. Gunnarsson, I.B., Karakashev, D., Angelidaki, I., 2014. Succinic acid production by fermentation of Jerusalem artichoke tuber hydrolysate with Actinobacillus succinogenes 130Z. Ind. Crop. Prod. 62, 125-129. 7. Khatun, M.M., Li, Y.H., Liu, C.G., Zhao, X.Q., Bai, F.W., 2015. Fed-batch saccharification and ethanol fermentation of Jerusalem artichoke stalks by an inulinase producing Saccharomyces cerevisiae MK01. RSC Adv. 5, 107112-107118. 8. Kim, S., Kim, C.H., 2014. Evaluation of whole Jerusalem artichoke (Helianthus tuberosus L.) for consolidated bioprocessing ethanol production. Renew. Energy. 65, 83-91. 9. Kim, S., Park, J.M., Kim, C.H., 2013. Ethanol production using whole plant biomass of Jerusalem artichoke by Kluyveromyces marxianus CBS1555. Appl. Biochem. Biotechnol. 169, 1531-1545. 10. Li, H.D., Zhu, H.J., Qiao, J.J., Du, J.H., Zhang, H., 2012. Optimization of the main liming process for inulin crude extract from Jerusalem artichoke tubers. Front. Chem. Sci. Eng. 6, 348-355. 11. Li, W., Zhang, J., Yu, C., Li, Q., Dong, F., Wang, G., Gu, G.D., Guo, Z.Y., 2015. Extraction, degree of polymerization determination and prebiotic effect evaluation of inulin from Jerusalem artichoke. Carbohyd. Polym. 121, 315-319. 12. Liu, C.G., Liu, L.Y., Zi, L.H., Zhao, X.Q., Xu, Y.H., Bai, F.W., 2014. Assessment and regression analysis on instant catapult steam explosion pretreatment of corn stover. Bioresour. Technol. 166, 368-372. 16

13. Liu, C.G., Qin, J.C., Liu, L.Y., Jin, B.W., Bai, F.W., 2016. Combination of ionic liquid and instant catapult steam explosion pretreatments for enhanced enzymatic digestibility of rice straw. ACS Sustain. Chem. Eng. 4, 577-582. 14. Liu, C.G., Wang, N., Lin, Y.H., Bai, F.W., 2012. Very high gravity ethanol fermentation by flocculating yeast under redox potential-controlled conditions. Biotechnol. Biofuels. 5, 61. 15. Liu, Y., Xu, J., Zhang, Y., Yuan, Z., He, M., Liang, C., Zhuang, X., Xie, J., 2015. Sequential bioethanol and biogas production from sugarcane bagasse based on high solids fed-batch SSF. Energy. 90, 1199-1205. 16. Long, X.H., Shao, H.B., Liu, L., Liu, L.P., Liu, Z.P., 2016. Jerusalem artichoke: a sustainable biomass feedstock for biorefinery. Renew. Sust. Energ. Rev. 54, 1382-1388. 17. Matías, J., Encinar, J.M., González, J., González, J.F., 2015. Optimisation of ethanol fermentation of Jerusalem artichoke tuber juice using simple technology for a decentralised and sustainable ethanol production. Energy. Sustain. Dev. 25, 34-39. 18. Mou, H.Y., Heikkilä, E., Fardim, P., 2013. Topochemistry of alkaline, alkaline-peroxide and hydrotropic pretreatments of common reed to enhance enzymatic hydrolysis efficiency. Bioresour. Technol. 150, 36-41. 19. Olofsson, K., Palmqvis,t B., Lidén, G., 2010. Improving simultaneous saccharification and co-fermentation of pretreated wheat straw using both enzyme and substrate feeding. Biotechnol. Biofuels. 3, 17. 20. Rabelo, S.C., Andrade, R.R., Filho, R.M., Costa, A.C., 2014. Alkaline hydrogen peroxide pretreatment, enzymatic hydrolysis and fermentation of sugarcane bagasse to ethanol. Fuel. 136, 349-357.

17

21. Saha, B.C., Cotta, M.A., 2014. Alkaline peroxide pretreatment of corn stover for enzymatic saccharification and ethanol production. Ind. Biotechnol. 10, 34-41. 22. Sarchami, T., Rehmann, L., 2014. Optimizing enzymatic hydrolysis of inulin from Jerusalem artichoke tubers for fermentative butanol production. Biomass. Bioenerg. 69, 175-182. 23. 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, NREL, Report No. TP-510-42618. 24. Unrean, P., Khajeeram, S., Laoteng, K., 2015. Systematic optimization of fed-batch simultaneous saccharification and fermentation at high-solid loading based on enzymatic hydrolysis and dynamic metabolic modeling of Saccharomyces cerevisiae. Appl. Microbiol. Biotechnol. 5, 2459-2470. 25. Wang, K., Yang, H., Chen, Q., Sun, R.C., 2013. Influence of delignification efficiency with alkaline peroxide on the digestibility of furfural residues for bioethanol production. Bioresour. Technol. 146, 208-214. 26. Xu, F., Shi, Y.C., Wang, D., 2013. X-ray scattering studies of lignocellulosic biomass: a review. Carbohyd. Polym. 94, 904-917. 27. Yang, M., Li, W., Liu, B, Li, Q., Xing J., 2010. High-concentration sugars production from corn stover based on combined pretreatments and fed-batch process. Bioresour. Technol. 101, 4884-4888. 28. Zhang, L., You, T., Zhang, L., Yang, H., Xu, F., 2014a. Enhanced fermentability of poplar by combination of alkaline peroxide pretreatment and semi-simultaneous saccharification and fermentation. Bioresour. Technol. 164, 292-298. 29. Zhang, X., Nghiem, N.P., Hicks, K.B., Johnston, D.B., 2014b. Hector RE. Ethanol production by high-solids simultaneous saccharification and fermentation of 18

cellulose-enriched barley straw and hull residues obtained by alkaline hydrogen peroxide pretreatment. Ind. Biotechnol. 10, 292-298.

Figures Legends

Fig. 1 Effect of the concentration of NaOH and H2O2 on the content of cellulose, hemicelluloses and lignin as well as biomass recovery.

Fig. 2 Thermogravimetric profiles of JAS under different pretreatment conditions. A TGA; B DTG

Fig. 3 Enzymatic hydrolysis of raw JAS and AHP-JAS at biomass loading from 10% to 30%. A Glucose production; B Glucose yield

Fig. 4 Ethanol production from AHP-JAS by the SSF or SHF process. For the SHF process, the hydrolysis was performed at 50 °C for 12 h, 36 h and 60 h for biomass loading of 10% (A), 20% (B) and 30% (C).

Fig. 5 Diagram for mass balance. The mass unit is gram (g).

19

Table 1 Composition of JAS from different regions Region

Yancheng

Dongying

Yinchuan

120.13W, 33.38N

118.49W, 37.46N

106.27W, 38.47N

Soil conditions

Alkali soil

Saline soil

Sierozem

Climate

Sub-tropical

Sub-tropical

Desert

Cellulose (%)

32.63±0.38

37.06 ± 0.41

35.93 ± 0.19

Hemicellulose (%)

19.03 ± 0.11

18.12 ± 0.17

22.21 ± 0.54

Lignin (%)

20.47 ± 1.19

18.10 ± 1.10

17.26 ± 0.52

Ash (%)

2.39 ± 0.15

2.17 ± 0.17

2.31 ± 0.50

Others (%)

23.65 ± 1.83

24.55 ± 1.17

22.38 ± 0.09

Geographic location

20

Table 2 Components (%) and biomass recovery (%) of pretreated JAS Biomass NaOH+H2O2

Cellulose

Hemicelluloses

Lignin

Ash

recovery* Raw JAS

100

35.93 ± 0.19

22.21 ± 0.54

17.26 ± 0.52

2.31 ± 0.50

1%+2%

74.25 ± 0.50

39.15 ± 0.76

15.52 ± 0.45

9.78 ± 0.62

5.00 ± 0.60

1%+3%

75.75 ± 2.50

41.76 ± 2.65

14.86 ± 0.06

8.08 ± 1.19

7.20 ± 0.30

1%+4%

73.00 ± 1.25

42.41 ± 2.74

14.54 ± 0.64

10.60 ± 1.50

6.40 ± 0.50

2%+2%

68.25 ± 0.75

45.65 ± 0.60

13.46 ± 3.13

11.78 ± 2.00

1.90 ± 0.20

2%+3%

64.00 ± 2.00

49.59 ± 1.64

13.27 ± 0.19

7.14 ± 0.15

1.40 ± 0.10

2%+4%

62.50 ± 2.75

50.37 ± 0.39

13.33 ± 1.28

7.10 ± 0.92

2.03 ± 0.70

3%+2%

61.25 ± 0.50

50.89 ± 2.61

12.14 ± 0.51

6.05 ± 1.32

1.09 ± 0.76

3%+3%

60.75 ± 0.25

49.72 ± 1.43

11.89 ± 0.38

5.90 ± 0.15

0.97 ± 0.60

3%+4%

61.50 ± 1.00

50.50 ± 0.52

12.52 ± 0.51

5.94 ± 1.79

1.03 ± 0.90

*Biomass recovery was defined as the ratio of pretreated biomass to raw biomass

21

Table 3 Ethanol production from JAS and other feedstocks under AHP conditions.

Feedstock

JAS

Pretreatment (solid-liquid ratio) NaOH 2%, H2O2 4%, (10%)

Ethanol yield* (g/g)

Reference

55.65

0.116

This study

45.3

0.252

Kim et al. 2014

20%

38.3

0.164

Khatuna et al. 2015

10%

30.53

0.139

Rabelo et al. 2014

Hydrolysis loading

Ethanol

30% 10% JAS

(g/L)

JAS + tubers

H2SO4 0.5%, (10%)

JAS

NaOH 2%, (10%)

Sugarcane

NaOH pH 11.5,

bagasse

H2O2 7.36%, (4%)

Cashew apple bagasse

NaOH pH11.5, H2O2 4.3%, (5%)

9%

18

0.075

da-Costa et al. 2015

Corn stover

NaOH pH 11.5, H2O2 20%, (10%)

10%

25

0.149

Saha et al. 2014

5%

9.9

0.132

Zhang et al. 2014a

3%

8.9

0.130

Zhang et al. 2014b

8%

19.3

0.135

Cao et al. 2012

5%

16.9

0.121

Wang et al. 2013

Poplar

Barley hull Sweet sorghum bagasse Furfural residue

NaOH 20%, H2O2 10%, (10%) NaOH pH 11.5, H2O2 2.1%, (5%) NaOH 2%, H2O2 5%, (10%) NaOH 2%, H2O2 2%, (10%)

8% tuber

* g ethanol/g raw feedstock

22

Fig. 1

23

Fig. 2

24

Fig. 3

25

Fig. 4

26

Fig. 5

27

28

Non-food feedstock JAS suits ethanol production. Regression analysis optimized AHP pretreatment. Stepwise addition of biomass and enzyme benefited high mass loading. SSF performed better than SHF under high mass loading.