Comparison of ultrasound-assisted Fenton reaction and dilute acid-catalysed steam explosion pretreatment of corncobs: cellulose characteristics and enzymatic saccharification

Abstract

As an emerging method for lignocellulose pretreatment, the ultrasound-assisted Fenton reaction is not well developed in comparison to the dilute acid-catalysed steam explosion. In this work, ultrasound-assisted Fenton reaction and dilute acid-catalysed steam explosion were investigated to evaluate their different effects as the pretreatment methods. The changes of cellulose characteristics were analysed by X-ray diffraction, ¹³C CP/MAS solid state NMR, water retention value, Simons' staining and scanning electron microscopy. It was indicated that ultrasound-assisted Fenton reaction removed more lignin and created slightly more accessible area and pores on the surface of the substrate compared with dilute acid-catalysed steam explosion. After 72 h of enzymatic hydrolysis, the conversion of cellulose from ultrasound-assisted Fenton reaction and dilute acid-catalysed steam explosion pretreatment reached 90.34% and 86.80% of the theoretical value, respectively, which indicated ultrasound-assisted Fenton reaction had the same capacity of pretreatment as dilute acid-catalysed steam explosion.

---

Introduction

The global corn production for 2014 was more than nine hundred million metric tons according to the report from United States Department of Agriculture and approximately forty–fifty million metric tons of corncobs could be collected.¹ So the corncob is an important lignocellulosic biomass resource, most of which has not been efficiently utilized in China.² Because of the recalcitrance of lignocellulose, pretreatment of lignocellulosic biomass plays a critical role in the process of conversion of biomass to energy, fuels and chemicals. Several types of pretreatments have been applied to biomass for improving the enzymatic hydrolysis, such as physical, physicochemical, chemical and biological methods. An effective pretreatment can disrupt and remove the cross-linked matrix of lignin and hemicellulose, change crystallinity, and increase accessibility of cellulose.³

Among a considerable amount of pretreatment technologies, dilute sulfuric acid pretreatment has been shown as a leading pretreatment process that is currently under commercial development.⁴ Steam explosion is one of the most commonly used methods for pretreatment of lignocellulosic materials.⁵ The combination of dilute acid with steam explosion is an effective and environmental-friendly method, which applies few acid to hydrolyze hemicellulose, leaving most of the cellulose intact for cellulase based hydrolysis.^6,7

Recently, a newly emerging pretreatment method, Fenton reaction has been applied to degrading lignocellulosic biomass.⁸ Fenton reaction is the main source of free radicals in biological systems.⁹ It catalyzes ferrous ions and hydrogen peroxide to generate hydroxyl radicals, which could destroy lignin by electrophilic addition to π systems in aqueous solution without additional energy input.¹⁰ Kato et al. reported that mimicing white-rot fungi lignin degradation via in vivo Fenton chemistry was applied to miscanthus, switchgrass, corn stover and wheat straw.¹¹ The enzymatic saccharification of Fenton pretreated biomass showed an average 212% increase relative to untreated control across all four feedstocks.¹¹ Using Fenton's reagent to pretreat rice straw for 24 h, the enzymatic digestibility achieved 93.2% of the theoretical glucose yield.¹² As an effective oxidation technology, Fenton reaction is famous for the environmental-friendly, economic and recyclable characteristics.¹³

However, single Fenton reaction consumed longer reaction time (24 h) in pretreating lignocellulose.^12,14 So it is necessary to combine Fenton reaction with other pretreatment methods for reducing the time and improving efficiency. Among the methods, ultrasound is a good assistant method, which is environmental-friendly and simple to manipulate. It has been already used to enhance the Fenton reaction by regeneration of hydroxyl radicals from water and oxygen; in addition, it also could accelerate the regeneration of ferrous ions.¹⁵ Several studies have indicated that the combination Fenton reaction with ultrasound can degrade lignin in the lignocellulose.^16–18 However, to the best of our knowledge, few reports are available for ultrasound-assisted Fenton reaction pretreatment of corncob for cellulose hydrolysis.

To investigate the chemical and physical changes characteristics of lignocellulosic biomass after pretreatment, Xin et al. have characterized and compared the bamboo fractions resulting from two leading pretreatments: dilute acid and aqueous ammonia technologies.¹⁹ To date, however, no comprehensive comparative analysis has been conducted in order to evaluate dilute acid catalysed steam explosion (DASE) technique with the novel ultrasound-assisted Fenton reaction (UFR) pretreatment. In addition, performances of different pretreatments may vary with feedstock types, the mechanisms of how the UFR and DASE techniques act on corncob, cause the changes of substrate, disrupt the lignin-carbohydrate linkages, and influence on subsequent enzymatic hydrolysis have not been illustrated clearly. Thus, here we compared the products after UFR and DASE pretreatments of corncob, especially the conversion of cellulose to monosaccharides.

Results and discussion

Chemical composition variation after UFR and DASE pretreatments

From the point of view of biofuel, an ideal pretreatment should remove lignin and hemicellulose as much as possible, and retain the most cellulose, which are important for subsequent enzymatic hydrolysis. As shown in Table 1, the untreated corncob (UC) was composed of 33.25% cellulose, 28.97% hemicellulose, 12.33% acid soluble lignin (ASL) and 18.98% acid insoluble lignin (AIL). Furthermore, after UFR pretreatment for 2 h, the content of lignin was reduced from 31.31% to 16.72% and hemicellulose was reduced from 28.97% to 23.27%, indicating the stronger delignification and poorer removal of hemicellulose capacity of UFR pretreatment. It was the hydroxy radicals generated by UFR that removed amorphous lignin and hemicellulose from lignocellulose.¹² However, the single Fenton reaction is time-consuming, which is similar to the wood decay process by fungi. It was reported that the lignin content of rice straw was reduced from 17.5% to 9.3% after 24 h of Fenton treatment,¹² while in the present paper similar result was obtained with the assistance of ultrasound only for 2 h, which demonstrated that ultrasound could improve the efficiency of Fenton reaction for degrading lignin. Under the ultrasound condition, acoustic cavitation could generate millions of small bubbles, creating partial high temperature and high pressure in acid environment to facilitate the degradation of lignin.²⁰ The combinational effect of hydroxy radicals and acoustic cavitation prompted the delignifying process and reduced the reaction time.

Table 1 Chemical composition and crystal index of UC, DASE and UFR substrates

Substrate	Cellulose	Hemicellulose	ASL^a	AIL^b	Ash	CrI (%)
a ASL: acid soluble lignin.b AIL: acid insoluble lignin.c UC: untreated corncob.d DASE: dilute acid-catalytic steam explosion.e UFR: ultrasound-assisted Fenton reaction; , where Itotal is the scattered intensity at 2θ = 22.5°, and Iam is the scattered intensity at 2θ = 16.4°.
UC^c	33.25 ± 1.20	28.97 ± 0.62	12.33 ± 0.36	18.98 ± 0.56	2.36 ± 0.18	26.85 ± 1.12
DASE^d	58.22 ± 1.61	1.15 ± 0.34	1.6 ± 0.12	27.07 ± 1.12	1.4 ± 0.08	39.40 ± 1.24
UFR^e	44.79 ± 1.52	23.27 ± 0.52	1.39 ± 0.08	15.33 ± 0.38	5.87 ± 0.21	33.17 ± 1.08

---

In order to evaluate the efficiency of UFR and DASE, DASE was carried out to pretreat the corncob. As indicated in Table 1, after DASE pretreatment the content of hemicellulose was reduced from 28.97% to 1.15%, indicating that DASE could degrade hemicellulose effectively. On the basis of the variation of chemical composition after UFR and DASE pretreatments in Table 1, it was demonstrated that UFR pretreatment could remove lignin effectively, while DASE could remove hemicellulose effectively.

Enzymatic hydrolysis of UFR and DASE substrates

The composition and structure of lignocellulose play critical roles during cellulose hydrolysis. Effective pretreatment can make conversion in UC, DASE and UFR substrates were as low as 5.41%, 19.2% and 26.05% after enzymatic hydrolysis for 4 h. When the hydrolysis time reached 72 h, the cellulose conversion of DASE and UFR substrates were 86.80% and 90.34%, respectively. However, the cellulose conversion of UC was still as low as 21.40%, which indicated that the DASE and UFR pretreatments could promote enzymatic hydrolysis of cellulose effectively.

Under the investigated conditions, the cellulose conversion of UFR substrate was higher than that of DASE substrate, which indicated that the UFR pretreatment possessed the hydrolyzing advantage over DASE pretreatment. The reason could be explained by lignin content difference of UFR and DASE substrates. It was evident that the Langmuir absorption and electrostatic interaction between lignin and cellulase could reduce the activity of enzymes, consequently the lignin content has a negative correlation with cellulose enzymatic activity.²¹ As shown in Table 1, the total content of lignin of UC, DASE and UFR substrates were 31.31%, 28.67% and 16.72%, respectively, the higher lignin content showed the stronger inhibitory effect, resulting in the lower cellulose digestibility. In addition, the hemicellulose content of UC, DASE and UFR substrates were 28.97%, 1.15% and 23.27%, respectively. The removal of hemicellulose also made cellulose more accessible to cellulase and hence improved cellulose digestibility. Furthermore, the results also indicated that the removal of lignin might be more effective than the removal of hemicellulose in the enzymatic hydrolysis of cellulose. The similar results were also reported by others.²² Besides lignin and hemicellulose contents many other factors also could affect cellulose enzymatic hydrolysis, such as cellulose structure, morphology, accessible surface area, pore distribution, etc.

X-ray diffraction analysis of UFR and DASE substrates

Cellulose enzymatic hydrolysis was also affected significantly by crystallinity of cellulose. Cellulose could be divided into cellulose I and cellulose II according to crystal structure, which showed different peaks in X-ray diffraction (XRD) patterns.²³ Untreated corncob apparently was dominated by cellulose I at peaks of 16.4°, 22.5° and 34.5°, representing the 110, 200 and 004 crystallographic planes.^23,24 The main peak at 22.5° is indicative of the instance between hydrogen-bonded sheets in cellulose I lattice.²⁵ The small peak at 34.5° is due to 1/4 of the length of one cellobiose unit and caused by ordering along the fiber direction.²⁶ However, after DASE and UFR pretreatments, a small shoulder was emerging at peak 21.8° (cellulose II), arising from the expansion of the cellulose matrix structure and the removal of hemicellulose and lignin during the pretreatment.²⁷ All in all, there were no significant changes in XRD spectrum of both DASE and UFR substrates in comparision to that of UC substrate, but they showed a trend to form cellulose II.

The cellulose crystalline index (CrI) was calculated with the XRD spectra's peak height and width between the crystalline peaks (101–002). After DASE and UFR pretreatment, the crystal index (CrI) increased to 39.40% and 33.17%, respectively from 26.85%. The results here were in good accordance with previous research that the CrI of lignocellulose increased after being pretreated by Fenton reaction, dilute acid and steam explosion.^28–30 The increase of CrI after DASE and UFR pretreatments was caused by the removal of amorphous hemicellulose and lignin (Table 1). These results indicated that more amorphous contents of corncob during DASE pretreatment were removed than that during UFR pretreatment.

¹³C CP/MAS solid state NMR spectroscopy of UFR and DASE substrates

¹³C cross polarization/magic-angle spinning solid state NMR spectroscopy (¹³C CP/MAS solid-state NMR) was chosen to provide further information about the composition and crystallinity of corncob substrates. ¹³C CP/MAS solid-state NMR spectra of the corncob substrate was shown in Fig. 3. There are two key regions in the NMR spectra: one is from 165 to 110 ppm, which is characterized by low intensity and broad resonances (aromatic structures in lignin); the other one is from 110 to 60 ppm assigning to carbohydrates (cellulose and hemicellulose), which includes sharp intense peaks. Peak at 104.8 ppm represents cellulose's C1, and the most intense peaks at 74.8 and 72.2 ppm are assigned to C2, C3 and C5 carbons in pyranoid and furan ring of carbohydrates. The two peaks at 88.7 and 64.8 ppm are due to C4 and C6 in crystalline cellulose, which indicated that the crystalline region enlarged after DASE, UFR pretreatments. Moreover, the crystalline region of DASE substrate was larger than that of UFR substrate, which were well agreement with XRD spectrum analysis (Fig. 2). Two other peaks (83.8 and 63 ppm) are related to C4 and C6 of hemicellulose and amorphous cellulose.³¹ Peak at 55.4 ppm is assigned to the methoxyl group of lignin. Only two peaks could undoubtedly belong to hemicellulose: the methyl carbon (21 ppm) and the carboxylic carbon (175 ppm). Peaks from 185 to 165 ppm are assigned to carbonyl groups derived from acetyl groups in hemicellulose. Due to the removal of hemicellulose during DASE pretreatment, the content of carbonyl groups in the DASE substrate was significantly reduced, which was corresponding to chemical composition analysis in Table 1. Peak at 148 ppm pertaining to phenolic S3/5 increased in DASE substrate, indicating that β-O-4 ether linkages in lignin were broken during DASE pretreatment and free phenolic hydroxyl groups were generated. In contrast, UFR substrate had less phenolic S3/5 (148 ppm) and a little more non-phenolic (154 ppm) groups in lignin. Significant variation in the ¹³C spectra of DASE and UFR samples was caused by the removal of hemicellulose and lignin. From the ¹³C spectra, DASE substrate had less hemicellulose than UFR substrate, which indicated that DASE pretreatment removed almost all hemicellulose, while UFR substrate contained less lignin since about 50% was degraded by hydroxy radicals. Those results were consistent with the analysis of chemical composition in Table 1.

Fig. 1 The enzymatic hydrolysis of different substrates: UC (red), DASE (blue) and URF (green) substrates.

Fig. 2 XRD profiles of UC (red), DASE (blue) and UFR (green) substrates.

Fig. 3 ¹³C cross polarization/magic-angle spinning (CP/MAS) solid-state nuclear magnetic resonance of UC (red), DASE (blue) and UFR (green) substrates. Spectra were normalized to the cellulose peak at 72 ppm.

The porosity and accessibility of cellulose after UFR and DASE pretreatments

The measurement of water retention value (WRV) could provide information on cellulose swelling and expansion caused by different pretreatments.³² Swelling of cellulose, resulting in the increase of cellulose accessibility, is essential to achieve efficient cellulose hydrolysis. Table 2 showed that the WRV of UC, DASE and UFR substrates were 279.5, 325.9, 344.9%, respectively. The DASE pretreatment disrupted the network structure among carbohydrate polymers and broke the hydrogen bonds between hemicellulose and cellulose, which eventually increased the enzymatic accessible area of cellulose.³³ The WRV of UFR substrate was also increased after pretreatment, which was caused by the removal of hemicellulose and lignin, and the increased accessibility of cellulose was indicated by the water diffusion inside the channels of the UFR substrate.⁶ WRV in Table 2 demonstrated that the accessible surface area of UFR substrate was larger than that of DASE substrate, which led to the higher rate of enzymatic hydrolysis. This was consistent with the curve of enzymatic hydrolysis of three substrates (Fig. 1).

Table 2 Water retention value and dyes adsorption of UC, DASE and UFR substrates

Substrate	Water retention value (%)	Maximum adsorbed DO^a (mg per g substrate)	Maximum adsorbed DB (mg per g substrate)	Total adsorbed dye (mg per g substrate)	DO/DB ratio
a The maximum amount of DO dye (>100 kDa) absorbed by UC, DASE and UFR substrates was described as mg dye per g substrate during Simons' staining, the same to DB dye.
UC	279.5 ± 2.1	28.5 ± 0.8	20.4 ± 0.3	49.9 ± 0.9	1.40 ± 0.04
DASE	325.9 ± 1.8	32.9 ± 0.6	22.5 ± 0.2	55.4 ± 0.6	1.46 ± 0.02
UFR	344.9 ± 2.2	35.4 ± 0.7	23.1 ± 0.3	58.5 ± 0.8	1.53 ± 0.02

---

The accessibility to cellulose was evaluated further by Simons' staining technique. Pontamine Sky Blue 6BX (DB) is a dye which has the chemical formula of C₃₄H₂₈N₆O₁₆S₄ with 992.8 g mol⁻¹ molar mass and 1 nm molecular diameter. Direct Orange 15 (DO) is a dye which has high molecular weight (>100 K) and 5–36 nm molecular diameter. Both the direct dyes could be used to evaluate the accessibility of cellulase to cellulose because of their linear structures and outstanding substantivity toward cellulose.^34,35 The adsorption of DO and DB dyes by cellulose can provide information about the distribution of large and small pores in porous substrates. Therefore, the adsorption of DO/DB ratio can be used to analyse the amount of large pores and small pores. The increase of maximum adsorbed DO dye indicated that DASE and UFR pretreatments could improve accessible area of cellulose. In other words, there was a positive relationship between porosity and hydrolyzing efficiency. The more large pores created by pretreatment, the higher cellulose enzymatic conversion and faster enzymatic rate achieved. However, the maximum adsorption of DB also increased slightly, which was a sign of the generation of small pores. This might be caused by the collapse of large pores during pretreatments or the hornification of substrate during drying process.³² The porosity of biomass depended on the size and distribution of interfibrillar spaces created by the association of cellulose with hemicellulose and lignin. The significant increase adsorption of DO and DO/DB ration indicated that both pretreatments could generate more pores or accessible surface area by removal of hemicellulose and lignin, which leading to the improvement of enzymatic hydrolysing rate. From Table 2, the DO adsorption and DO/DB ration of UFR substrate were higher than that of DASE substrate, which indicated that more pores and accessible surface area were created by UFR pretreatment. Those results might be the reasons that the cellulose conversion of UFR substrate was higher than that of DASE substrate (Fig. 1). The result of Simons' staining was in good agreement with that of WRV (Table 2).

The scanning electron microscopy of UFR and DASE substrates

The removal of hemicellulose and lignin by UFR and DASE pretreatments would result in physical and surface morphological changes in the corncob, consequently SEM analysis was conducted to observe the changes of corncob substrates. For the UC substrate (Fig. 4A), the surface was rigid and highly ordered. However, after DASE pretreatment the cell wall was broken into pieces and the surface structure became disorder (Fig. 4B), which was the result of thermal explosion decomposition.⁶ In addition, there were some particles attaching to the surface, which might be the redeposited lignin.³⁶ Those re-deposited lignin would inhibit the cellulase activity by unproductive absorption. From Fig. 4C, it was observed that the UFR substrate possessed a lot of penetrations on the surface, possibly created by the hydroxy radicals in combination with acoustic cavitation.³⁷ The significant morphological changes of corncob substrates indicated that more accessible areas and pores were formed after DASE and UFR pretreatments.

Fig. 4 Scanning electron micrographs of corncob substrates: (A) UC, (B) DASE, (C) UFR substrate.

In the case of industrial application for Fenton reaction system, the hydrogen peroxide is green and feasible in industry, ferrous ions solution could be substituted by solid catalyst to avoid the production of too much iron sludge;^38,39 in addition, automated Fenton reactor is available to simplify manipulation. Consequently, UFR treatment has been applied extensively in wastewater treatment. All above provide evidences for industrial feasibility of UFR for lignocellulose pretreatment.

Conclusions

It is important that employing pretreatments reduce the effects of lignocellulosic biomass recalcitrance, which could enhance the accessibility of cellulose to enzymes. DASE pretreatment could remove most of hemicellulose. However, large amount of lignin is still retained in substrate disturbing cellulose enzymatic hydrolysis after DASE pretreatment. Compared with DASE, UFR pretreatment removed more lignin and created more accessible area on the surface of substrate by the strong oxidation of OH˙, which might be the reasons of the enzymatic hydrolysis difference between DASE (86.8%) and UFR (90.34%) substrates. In comparison to the single Fenton pretreatment, ultrasound-assisted Fenton reaction reduced the pretreatment time significantly. As an environment-friendly and economic pretreatment, UFR is promising in biorefinery but further work still need to be done to explore its effects on different types of biomass and apply it to the biomass energy industry.

Experimental

Materials and pretreatments

Corncob used in all the experiments was collected from Shandong province and ground using a high-speed rotary cutting mill to produce a powder with particle sizes less than 1 mm, then stored at room temperature. For the optimum UFR pretreatment, FeSO₄·7H₂O (25 Mm) was added to the milled corncob sample suspension and the pH was adjusted to 4.0. The sample was at a solids loading of 5% (w/v) in a 500 mL glass conical flask at 25 °C under ultrasonic irradiation (35 W, 24 kHz). Finally, 30% H₂O₂ (5 M) was added to the reaction system every 20 min. After incubation for 120 min, the sample was vacuum filtered by Buchner funnel and washed several times to remove ferri and ferrous ion, which were adsorbed on the UFR sample. The corncob was pretreated by DASE with the optimum conditions H₂SO₄ (0.2% w/w), time (10 min) and pressure (1.0 MPa). The specific experimental steps were according to Zhang et al. (2014).⁴⁰

Enzymatic digestibility analysis

The effects of DASE and UFR pretreatments were evaluated by performing the enzymatic digestibility analysis. The UC, DASE and UFR substrates were enzymatic hydrolysis with 5% (w/v) substrates loading. The cellulase (Novozyme Cellic CTec2) activity was 206 FPU mL⁻¹ and the enzyme loading was 18 FPU g⁻¹ cellulose. The enzymatic reaction mixture of substrate in 10 mL of citrate buffer (50 mM, pH 4.8) were shaking in an air incubator at 50 °C and 150 rpm. Substrates were taken periodically and analyzed by HPLC. Cellulose conversion was defined as the percentage of glucose released compared to the maximum theoretical yield.

Analytical methods

The composition analysis of samples was conducted according to the National Renewable Energy Laboratory (NREL; Golden, CO).⁴¹ Crystallinity of the samples was determined by X-ray diffraction.⁴² The high-resolution ¹³C CP/MAS solid-state NMR spectra was performed on a Bruker Avance 300 MHz spectrometer at 4.7 T with a 7 mm BL probe (HP WB 73A MAS 7 BL CP VTN), operating at 50.13 MHz for ¹³C at room temperature.³¹ WRV and Simons' stain methods were used to determine porosity and accessibility of cellulose. The WRV was measured following the Scandinavian test method SCAN-C 62:00. DB and DO (MP Biomedical products, Santa Ana, USA) were used for Simons' staining, which was performed according to Chandra & Saddler (2012).⁴³ The SEM analysis of the corncob samples was carried out with the Philips XL30 scanning electron microscope.

Abbreviations

UC	Untreated corncob
DASE	Dilute acid-catalysed steam explosion
UFR	Ultrasound-assisted Fenton reaction
ASL	Acid soluble lignin
AIL	Acid insoluble lignin
DO	Direct orange
DB	Direct blue

Acknowledgements

We are indebted to the National High Technology Research and Development Program of China (2014AA021906; 2015AA021001), Beijing Natural Science Foundation (5162019) and National Key Research Programme (2016YFD0400601) for the generous financial supports.

Notes and references

1. W. H. Qu, Y. Y. Xu, A. H. Lu, X. Q. Zhang and W. C. Li, Bioresour. Technol., 2015, 189, 285.
2. L. Xie, J. Zhao, J. Wu, M. Gao, Z. Zhao, X. Lei, Y. Zhao, W. Yang, X. Gao, C. Ma, H. Liu, F. Wu, X. Wang, F. Zhang, P. Guo and G. Dai, Bioresour. Technol., 2015, 193, 331.
3. R. Kumar and C. E. Wyman, Biotechnol. Bioeng., 2009, 103, 252.
4. Z. Li, Z. Jiang, B. Fei, Z. Cai and X. Pan, Bioresour. Technol., 2014, 151, 91.
5. J. D. McMillan, ACS Symp. Ser., 1994, 566, 292.
6. W. H. Chen, B. L. Pen, C. T. Yu and W. S. Hwang, Bioresour. Technol., 2011, 102, 2916.
7. J. Sumphanwanichi, N. Leepipatpiboon, T. Srinorakutara and A. Akaracharanya, Ann. Microbiol., 2008, 58, 219.
8. V. Arantes, A. F. Milagres, T. Filley and B. Goodell, J. Ind. Microbiol. Biotechnol., 2011, 38, 541.
9. V. Melin, A. Henríquez, J. Freer and D. Contreras, Redox Rep., 2015, 20, 89.
10. U. Jurva, H. V. WlkstrÖm and A. P. Brulns, Rapid Commun. Mass Spectrom., 2002, 16, 1934.
11. D. M. Kato, N. Elía, M. Flythe and B. C. Lynn, Bioresour. Technol., 2014, 162, 273.
12. Y. H. Jung, H. K. Kim, H. M. Park, Y. C. Park, K. Park, J. H. Seo and K. H. Kim, Bioresour. Technol., 2015, 179, 467.
13. C. Oliveira, A. Alves and L. M. Madeira, Chem. Eng. J., 2014, 241, 190.
14. Y. C. He, Y. Ding, Y. F. Xue, B. Yang, F. Liu, C. Wang, Z. Z. Zhu, Q. Qing, H. Wu, C. Zhu, Z. C. Tao and D. P. Zhang, Bioresour. Technol., 2015, 193, 324.
15. P. V. Nidheesh, RSC Adv., 2015, 5, 40552.
16. K. Ninomiya, H. Takamatsu, A. Onishi, K. Takahashi and N. Shimizu, Ultrason. Sonochem., 2013, 20, 1092.
17. G. Ramadoss and K. Muthukumar, Ultrason. Sonochem., 2016, 28, 207.
18. A. Kawee-ai, A. Srisuwun, N. Tantiwa, W. Nontaman, P. Boonchuay, A. Kuntiya, T. Chaiyaso and P. Seesuriyachan, Ultrason. Sonochem., 2016, 31, 184.
19. D. Xin, Z. Yang, F. Liu, X. Xu and J. Zhang, Bioresour. Technol., 2015, 175, 529.
20. M. J. Bussemaker and D. Zhang, Ind. Eng. Chem. Res., 2013, 52, 3563.
21. S. Nakagame, R. P. Chandra, J. F. Kadla and J. N. Saddler, Bioresour. Technol., 2011, 102, 4507.
22. K. Li, J. Wan, X. Wang, J. Wang and J. Zhang, Ind. Crops Prod., 2016, 83, 414.
23. G. Cheng, X. Zhang, B. Simmons and S. Singh, Energy Environ. Sci., 2015, 8, 436.
24. G. Sèbe, F. Ham-Pichavant, E. Ibarboure, A. L. Koffi and P. Tingaut, Biomacromolecules, 2012, 13, 570.
25. X. Fan, G. Cheng, H. Zhang, M. Li, S. Wang and Q. Yuan, Carbohydr. Polym., 2014, 114, 21.
26. G. Cheng, P. Varanasi, C. Li, H. Liu, Y. B. Melnichenko, B. A. Simmons, M. S. Kent and S. Singh, Biomacromolecules, 2011, 12, 933.
27. J. Singh, M. Suhag and A. Dhaka, Carbohydr. Polym., 2015, 117, 624.
28. P. Jain and N. Vigneshwaran, Bioresour. Technol., 2012, 103, 219.
29. R. Kumar and C. E. Wyman, Bioresour. Technol., 2009, 100, 4203.
30. F. Pang, S. Xue, S. Yu, C. Zhang, B. Li and Y. Kang, Bioresour. Technol., 2013, 42, 402.
31. N. Baccile, C. Falco and M. M. Titirici, Green Chem., 2014, 16, 4839.
32. X. Luo and J. Y. Zhu, Enzyme Microb. Technol., 2011, 48, 92.
33. N. Mosier, C. Wyman, B. Dale, R. Elander, Y. Y. Lee, M. Holtzapple and M. Ladisch, Bioresour. Technol., 2005, 96, 673.
34. R. P. Chandra, V. Arantes and J. Saddler, Bioresour. Technol., 2015, 185, 302.
35. A. R. Esteghlalian, M. Bilodeau, S. D. Mansfield and J. N. Saddler, Biotechnol. Prog., 2001, 17, 1049.
36. M. J. Selig, S. Viamajala, S. R. Decker, M. P. Tucker, M. E. Himmel and T. B. Vinzant, Biotechnol. Prog., 2007, 23, 1333.
37. G. Ramadoss and K. Muthukumar, Ultrason. Sonochem., 2016, 28, 207.
38. E. G. Garrido-Ramírez, B. K. G. Theng and M. L. Mora, Appl. Clay Sci., 2010, 47, 182.
39. P. V. Nidheesh, RSC Adv., 2015, 5, 40552.
40. H. J. Zhang, X. G. Fan, X. L. Qiu, Q. X. Zhang, W. Y. Wang, S. X. Li, L. H. Deng, M. A. Koffas, D. S. Wei and Q. P. Yuan, Bioprocess Biosyst. Eng., 2014, 37, 2425.
41. A. Sluiter, B. Hames, R. Ruiz, C. Scarlata, J. Sluiter, D. Templeton and D. Crocker, Determination of structural carbohydrates and lignin in biomass, NREL, 2008.
42. X. Ouyang, W. Wang, Q. Yuan, S. Li, Q. Zhang and P. Zhao, RSC Adv., 2015, 5, 61650.
43. R. P. Chandra and J. Saddler, Ind. Biotechnol., 2012, 8, 230.

---