Phosphoric acid doped polybenzimidazole as an heterogeneous catalyst for selective and efficient dehydration of saccharides to 5-hydroxymethylfurfural

Abstract

5-Hydroxymethylfurfural (HMF) is considered as a valuable platform chemical for fuels and chemical intermediates. Tremendous efforts have been made on the development of catalysts for the dehydration of saccharides to HMF with high conversion and selectivity. Here we reported a new solid acid, phosphoric acid doped polybenzimidazole (PA–PBI), for the dehydration of fructose to HMF with high selectivity (∼94%) at high fructose concentration (50 wt%) in a water–methyl isobutyl ketone (MIBK) biphasic system. PA–PBI was directly obtained by the one-pot polycondensation reaction and was very stable under harsh reaction conditions due to the strong acid–base interaction between PA and imidazole in PBI. PA–PBI was recycled to catalyze the dehydration of fructose four times without the loss of activity and could also catalyze a variety of disaccharides and polysaccharides to produce HMF. The high selectivity and the ease of recycling make PA–PBI a superior catalyst for the dehydration of saccharides to HMF.

---

1. Introduction

With the diminishing of the fossil fuel reserves and growing concerns about global warming, much attention has been paid to sustainable sources for the production of chemicals and fuels in recent decades. Biorefining provides an alternative approach to solving the current dependence on fossil fuels for energy and chemical building blocks.¹ 5-Hydroxymethylfurfural (HMF) is one of the versatile and key biorenewable platform chemicals for the production of valuable chemicals, high quality fuels and plastics.^2,3

HMF can be obtained by acid-catalyzed dehydration of renewable biomass-derived carbohydrates.^4,5 This dehydration reaction has been extensively studied using either heterogeneous or homogeneous catalysts, such as mineral acids,^6,7 organic acids,⁸ metal salts and metal oxides,^9,10 zeolites,¹¹ transition-metal ions,^12,13 phosphates,^14,15 ion exchange resins,^16,17 and ionic liquids.^18,19 Asghari et al. compared the performance of various homogeneous acid catalysts, e.g. HCl, H₂SO₄, H₃PO₄ and organic acids, and found that not only the pH, but also the acid type had great influence on the decomposition pathway.⁸ Under the same pH condition, phosphoric acid showed the best balance between activity, yield, and selectivity. Although homogeneous acid catalysts are effective and inexpensive, they have serious drawbacks, such as severe corrosion of equipments and difficulties in product separation and solvent or catalyst recycling. In comparison, heterogeneous acid catalysts can be easily separated and regenerated for multiple cycles. Sulfonated resins and silica are widely used heterogeneous catalysts for efficient dehydration of saccharides to HMF,^20–23 however, they also promote severe side reactions such as humin formation and decomposition of HMF to form levulinic acid (LA) and formic acid due to the strong acidicity of sulfonic acid.⁶ Moreover, the reaction temperature using these types of catalysts is generally limited to below 130 °C due to the thermal stability of the resin.²⁴ Solid heteropolyacids could also sufficiently catalyze the dehydration of fructose to form HMF with moderate to high HMF selectivity, but the initial fructose concentration in water phase was typically lower than 30 wt%.

It is necessary to develop an efficient catalyst which combines the advantages of both homogeneous and heterogeneous catalysts to efficiently catalyze the dehydration of high concentration saccharides to produce HMF with high selectivity. Polybenzimidazole (PBI) is an aromatic heterocyclic polymer and compatible with most organic solvents.^25,26 Phosphoric acid, a superior homogeneous dehydration catalyst, can strongly interact with PBI through acid–base interaction between OH and imidazole groups. The resulting phosphoric acid doped PBI (PA–PBI) will be an efficient heterogeneous catalyst for HMF production.

In this study, PA–PBI was prepared by the one-pot polycondensation reaction and used as an heterogeneous catalyst for the dehydration of saccharides (Scheme 1). Poly(phosphoric acid) (PPA) was used as both the polymerization solvent and phosphorylation agent, and after polymerization, PA–PBI was directly obtained by pouring the reaction solution into the water. The performance of PA–PBI was evaluated by investigating the effects of the reaction time and temperature, the catalyst loading and the solvent composition on the dehydration reaction. PA–PBI exhibited good activity, high HMF selectivity, and reusability for the dehydration of high concentration fructose.

Scheme 1 (a) Synthesis of PA–PBI; (b) dehydration of D-fructose to HMF.

2. Experimental

2.1. Reagents and materials

3,3′-Diaminobenzidine (DAB), isophthalic acid, and diphenyl-m-phthalate (DPIP), D₂O, and CDCl₃ were purchased from Sigma-Aldrich. Poly(phosphoric acid) (PPA), D-fructose, glucose, sucrose, starch, and all solvents used in this work were purchased from Aladdin.

2.2. Characterization

The ¹H NMR spectra were collected on a Bruker DDMX 300 spectrometer (300 MHz) at 25 °C in D₂O and CDCl₃. ³¹P NMR spectra were collected on a Bruker DDMX 300 spectrometer (300 MHz) at 25 °C with D₂O as the solvent. ³¹P-solid state NMR spectra were collected on a Bruker AVANCEIII 300 WB spectrometer (400 MHz). FT-IR spectra in the range of 4000–400 cm⁻¹ were recorded on a Bruker VERTEX 70 Fourier transform infrared spectrometer using KBr disc. CHN elemental analysis was carried out using a vario EL CUBE. SEM was performed using a Hitachi S-4800 electron microscope.

2.3. Preparation of PA–PBI

17.13 g (80 mmol) of DAB and 13.29 g (80 mmol) of isophthalic acid were mixed with 856 g of PPA (PPA/DAB = 10.7 g mmol⁻¹) in a three-neck 1000 mL flask equipped with a mechanical stirrer, a nitrogen inlet and an allochroic silicagel drying tube. The mixture was heated at 240 °C for 9 h under a nitrogen atmosphere. The resulting dark brown solution was poured into deionized water. The precipitate was filtered and cooked with deionized water (180 °C) in a pressure flask twice. After filtration, PA–PBI was dried at 120 °C in vacuum for 2 h (Scheme 1).

2.4. Synthesis of PBI via melt polymerization

DAB (0.3214 g, 15 mmol) and DPIP (0.2492 g, 15 mmol) were added to a three-neck flask, which was equipped with a mechanical overhead stirrer and a nitrogen inlet and outlet. The mixture was heated to 180 °C in an oil bath under a nitrogen atmosphere and stirred for 1 h. The viscosity slowly built up during the polymerization and stirring was stopped. The temperature was then increased to 270 °C in 1 h and the mixture was held at 270 °C for 2 h. The foamed polymer was pulverized into powder and then the powder was heated at 350 °C for 1 h to give PBI (Scheme 1).²⁷

2.5. General procedure for the conversion of fructose to HMF

All dehydration experiments were performed in a 100 mL pressure tube under magnetic stirring. In a typical procedure, D-fructose (2.5 g), deionized water (2.5 mL), MIBK (5 mL) and PA–PBI (0.25 g) were charged into the flask. The reaction mixture was stirred and heated in the oil bath at 180 °C for 1 h. After reaction, the concentrations of HMF and fructose were determined by ¹H NMR spectroscopy using p-dihydroxybenzene as the internal standard.²⁸ The amount of HMF was obtained by combining HMF in both organic and water phases. The conversion of fructose and the yield and selectivity of HMF are defined as follows:

2.6. Recyclability tests

The recyclability of the catalyst was investigated under the same reaction conditions as described above for the fresh catalyst. After each reaction, the catalyst was recovered by filtration, washed with water for four times and then MIBK for four times, dried at 120 °C for 2 h under vacuum, and reused in the next cycle.

3. Results and discussion

3.1. Synthesis and characterization of PA–PBI

PA–PBI was synthesized by polycondensation of 3,3′-diaminobenzidine (DAB) and isophthalic acid in poly(phosphoric acid) (PPA) at 240 °C.²⁹ The polymerization mixture was directly poured into deionized water to give dark brown fibers with a diameter of ∼1 mm. The fibers were washed in boiling water for 10 times and then cooked with deionized water (180 °C) in a pressure flask twice. PA was not detected by ³¹P NMR spectroscopy in the final washed water (Fig. S1†), indicating the absence of free PA in PA–PBI. The cross section was obtained after breaking the PA–PBI fiber under liquid nitrogen. The surface and cross section morphology of the PA–PBI was characterized with scanning electron microscopy (SEM), and the SEM images in Fig. S2† showed the non-porous nature in the PA–PBI. Meanwhile, N₂ adsorption experiment showed that pulverized PA–PBI powders didn't have pores. Solid state ³¹P NMR analysis of PA–PBI confirmed the presence of P in the polymer and revealed that P was present in the form of PA (Fig. 1(a)). Elemental analysis of PA–PBI showed that PA–PBI contained 2.71 wt% of P (or 0.87 mmol g⁻¹) and 14.79 wt% of N. The mole ratio of P to N was 0.08/1, corresponding to 0.16 PA per imidazole unit.

Fig. 1 (a) Solid state ³¹P NMR spectrum of PA–PBI; (b) FT-IR spectra of PBI and PA–PBI.

PA interacted with imidazole through strong acid–base interaction. FT-IR measurements were carried out to clarify the nature of the interaction between PA and PBI in PA–PBI, as shown in Fig. 1(b). For comparison, PBI was synthesized via melt polycondensation of DAB and diphenyl m-phthalate (DPIP).²⁷ Compared to the spectrum of PBI, absorption bands at 497, 875, 960, 1070 and 1165 cm⁻¹ related to PA were observed in the spectrum of PA–PBI. In addition, a very broad absorption band ranging from 2000 to 3600 cm⁻¹ was also observed and attributed to the acid–base interactions between PA and imidazole groups.^30,31

3.2. Influence of initial fructose concentrations on dehydration of fructose

Dehydration of fructose to HMF was catalyzed by PA–PBI fibers in a biphasic water–methyl isobutyl ketone (MIBK) system. We performed experiments to determine the optimal initial fructose weight concentration. Different initial concentrations including 30, 50, and 70 wt% fructose in water phase were studied and their corresponding amounts of fructose were 2.16, 5.00, 11.67 g, respectively. As shown in Table 1, the amount of consumed fructose increased from 1.36 g to 5.60 g with the increase of the fructose concentration from 30 wt% to 70 wt%, but the percentage conversion decreased from 63% to 48%. Since the amount of catalyst was same, it was obvious that consumed fructose was proportional to the initial concentration of fructose. The yield of HMF also showed the same trend as the consumption of fructose: the amount of produced HMF increased from 0.82 g to 2.29 g, while the percentage yield dropped from 54% to 28%. The selectivity of HMF (defined as the moles of HMF produced divided by the moles of fructose reacted) was used to compare the efficiency of the dehydration system. The highest HMF selectivity (94%) was obtained when the fructose concentration was 50 wt%. There were no detectable LA or formic acid in ¹H NMR (Fig. S3†) and MS (126.9000: [M + H]⁺) analysis for all the experiments, and we observed insoluble humins as the only byproduct. Tao et al. also found the similar trend: the highest selectivity was obtained when the fructose concentration was 50% in the water phase.³²

Table 1 Effect of the initial fructose concentration on fructose dehydration^a

Initial fructose concentration (wt%)	Fructose consumption		YieldHMF		Selectivity (%)
	(g)	(%)	(g)	(%)
a Reaction conditions: H2O (5 mL), MIBK (10 mL), PA–PBI (0.50 g), 180 °C, 1 h.
30	1.36	63	0.83	54	86
50	2.65	53	1.75	50	94
70	5.60	48	2.29	28	58

---

There are two passways to form humins: condensation of fructose and dehydration of HMF.³³ In our system, PA–PBI not only catalyzes the dehydration of fructose to HMF but also promotes the dehydration of HMF to humins. At the same time, direct condensation of fructose at high temperature also consumed fructose to yield humins. The competition of the dehydration of HMF and the condensation of fructose resulted in the highest selectivity at 50 wt% fructose concentration.^6,34 Roman-Leshkov et al. reported a process for the selective dehydration of fructose catalyzed by 0.25 M HCl at high fructose concentration (50 wt%). They found that phase modifiers promoted efficient production of HMF from fructose and a highest selectivity (77%) was obtained.⁶ Fan et al. studied the dehydration of fructose (50 wt%) catalyzed by a solid heteropolyacid salt, Ag₃PW₁₂O₄₀, and ∼87% HMF selectivity was achieved at ∼70% fructose conversion.³⁵ In comparison, dehydration of fructose catalyzed by PA–PBI had a higher selectivity (94%), exhibiting better balance between activity, yield, and selectivity at high initial fructose concentration. Accordingly, an initial fructose concentration of 50 wt% was employed as the reaction conditions in the following studies.

3.3. Influence of reaction time and temperature on dehydration of fructose

We also performed experiments to investigate the effect of the reaction temperature and time on the dehydration reaction (Table 2). The conversion of fructose increased with the increase of the reaction temperature from 160 °C to 200 °C, but the highest yield of HMF was obtained at 180 °C when the reaction was performed for 1 h. The severe side reaction might occur at 200 °C and the HMF selectivity dropped significantly from 94% (180 °C) to 67% (200 °C). When the reaction time was increased from 0.5 h to 2 h at 180 °C, both the conversion of fructose and the HMF yield increased while the HMF selectivity decreased from 96% to 83%. As the reaction proceeded, the concentration of fructose decreased and the rate of HMF formation also decreased while the rate of the side reaction was proportional to the concentration of HMF, which resulted in the decrease of HMF selectivity. Although the highest selectivity (96%) was obtained for 0.5 h, the conversion of fructose (14%) and HMF yield (13.4%) was too low. A higher conversion of fructose (53%) and HMF yield (50%) as well as selectivity (94%) were obtained for 1 h. Accordingly, the following reactions were all performed at 180 °C for 1 h to balance both the selectivity and HMF yield (Table 2, run 2).

Table 2 Effect of reaction time and temperature on fructose dehydration^a

Run	Temperature (°C)	Time (h)	Conv. (%)	YieldHMF (%)	Selectivity (%)
a Reaction conditions: fructose (5.00 g, 50 wt%), H2O (5 mL), MIBK (10 mL), PA–PBI (0.50 g).
1	160	1	7.5	7.0	93
2	180	0.5	14	13.4	96
3	180	1	53	50	94
4	180	2	86	71	83
5	200	1	63	42	67

---

3.4. Influence of solvent ratio and catalyst loading on dehydration of fructose

The effect of the ratio of MIBK to water on the performance of PA–PBI was investigated (Table 3). Compared to the pure water system, addition of the same volume of MIBK into the reaction system improved the HMF selectivity from 68% to 82% because HMF was continuously extracted into the organic phase from the reactive aqueous solution to reduce the side reaction (run 2).⁶ Further increase of the ratio of MIBK to water from 1/1 to 2/1 led to the improvement of the HMF yield and selectivity from 32% and 82% to 50% and 94%, respectively. When the ratio of MIBK to water reached to 3/1, HMF yield and selectivity (52% and 93%) did not show further improvement.

Table 3 Effect of the solvent ratio and catalyst loading on the dehydration of fructose^a

Run	PA–PBI (wt%)	MIBK/H2O (v/v)	Conv. (%)	YieldHMF (%)	Selectivity (%)
a Reaction conditions: fructose (5.00 g, 50 wt%), H2O (5 mL), 180 °C, 1 h.
1	10	0/1	34	23	68
2	10	1/1	39	32	82
3	10	2/1	53	50	94
4	10	3/1	56	52	93
5	5	2/1	47	44	94
6	20	2/1	69	54	78

---

Moreover, the loading of catalyst also affected the dehydration of fructose. Increase of the catalyst loading from 5 wt% to 10 wt% only led to a slight improvement from 47% to 53% in the fructose conversion, and the selectivity was almost same. Further increase of the catalyst loading to 20 wt% resulted in an even higher fructose conversion (69%), but the HMF selectivity dropped from 94% to 78% due to the obvious side reaction. In terms of the balance of yield and selectivity, a better performance was achieved when 10 wt% of PA–PBI was used. Therefore, the following studies were performed with a 2/1 ratio of MIBK to water and 10 wt% catalyst loading (Table 3, run 3).

Table 4 showed the features of different catalyst systems for the fructose dehydration into the HMF with a 2/1 ratio of MIBK to water at 180 °C for 1 h. When the dehydration reaction of fructose was performed without any catalyst, 8.4% fructose conversion was obtained and the HMF selectivity was as high as 95% (Table 4, run 3). Addition of PBI did not show any improvement (run 2), suggesting that PBI was essentially inactive in the fructose dehydration and the PA moiety in PA–PBI was the active species for dehydration of fructose. In comparison, fructose dehydration was also carried out in 0.087 mol L⁻¹ H₃PO₄ aqueous phase solution which had an equivalent PA concentration of PA–PBI in the aqueous phase solution under the same reaction conditions. A higher fructose conversion (73%) but lower HMF yield (48%) was obtained when the reaction was catalyzed by the H₃PO₄ solution. The available catalytic PA sites in PA–PBI are significantly lower than the number calculated from elemental analysis since PA–PBI doesn't contain pores. Roman-Leshkov et al. also studied the dehydration of fructose catalyzed by H₃PO₄ at a lower temperature (90 °C). The HMF selectivity (76%) was significantly lower than that in the PA–PBI catalyzed system (Table 4, run 1).⁶ Compared to other homogenous and heterogenous catalysts, PA–PBI afforded as high as 94% HMF selectivity with high fructose conversion.

Table 4 Performance of different catalysts for the dehydration of fructose^a

Run	Catalysts	Conv. (%)	YieldHMF (%)	Selectivity (%)
a Reaction conditions: fructose (5.00 g, 50 wt%), H2O (5 mL), MIBK (10 mL), 180 °C, 1 h.
1	PA–PBI	53	50	94
2	PBI	8.5	8.1	95
3	None	8.4	8.0	95
4	PA	73	48	66

---

In our experiments, insoluble humins were the only observed byproduct. In other words, the formation of humins was the only side reaction. Kuster et al. demonstrated that in the dehydration of fructose, insoluble humins mainly formed via the self-polymerization of HMF or the cross-polymerization between HMF and fructose.³⁶ The dehydration and polymerization reactions both occur in the water phase under acid catalysis. If HMF can be continuously removed from the aqueous phase, the formation of humans will be reduced. Compared to the single solvent system, the use of a biphasic solvent system can efficiently extract HMF from the water phase to the organic phase, which significantly minimizes the byproduct formation. In addition, the initial fructose concentration has a pronounced effect on the humin formation. When a high initial fructose concentration is used, the collision probability of reactive compounds is also large, leading to fast formation of humins. Meanwhile, the undesirable formation of humins is also inevitable at high temperature with prolonged reaction time. Optimization of reaction conditions, e.g. fructose concentration, temperature, time and solvent ratio results in the formation of HMF with high selectivity.

3.5. Recycling of PA–PBI

PA–PBI showed excellent stability in the reactive systems and could be easily recycled and reused for multiple times. The catalyst was recovered after each use by successively washing with deionized water and MIBK for five times each and drying at 120 °C for 2 h. ³¹P NMR analysis of the aqueous solution of the 4th run showed that there was no detectable amount of phosphoric acid in the solution, indicating that phosphoric acid did not dissociate from PBI during reaction (Fig. 2). Meanwhile, elemental analysis also proved that PA–PBI was stable under the reaction conditions. PA–PBI before use and after being used four times (original/4th run) contained 60.26/63.94 wt% C, 4.60/4.75 wt% H, 14.74/13.34 wt% N, and 2.71/2.64 wt% P. The ration of N to P did not change. The slight increase of C and H contents might be attributed to the deposition of small amount of humins on PA–PBI fibers, which also resulted in the slight decrease of P and N contents.

Fig. 2 ³¹P NMR spectrum of the water phase after reaction.

We also compared the catalytic performance of the PA–PBI catalyst for fructose dehydration for each use. The catalyst was recycled for three times and the results of the four runs were summarized in Fig. 3. The PA–PBI catalyst did not show activity loss after recycling. The conversion of fructose, the yield of HMF and the selectivity were almost same for the original and recycled catalysts. It is reasonable to claim the PA–PBI catalyst can be recycled and reused for even more cycles before it loses the activity.

Fig. 3 Conversion, yield and selectivity of fructose dehydration catalyzed by PA–PBI (1st run) and recycled PA–PBIs (2nd–4th runs). Reaction conditions: fructose (5.00 g, 50 wt%), H₂O (5 mL), MIBK (10 mL), 180 °C, 1 h.

3.6. Dehydration of various saccharides to HMF

Moreover, dehydration of various saccharides, e.g. glucose, sucrose, starch and cellulose, was also tested using the PA–PBI catalyst. The formation of HMF was observed for all saccharides (Fig. 4). Compared to fructose, the HMF yield (∼10%) in glucose dehydration was significantly lower (Fig. S4†). It was reported that glucose first isomerized to fructose and then fructose dehydrated to form HMF and isomerization of glucose to fructose was the slow rate-determining step.¹⁸ The HMF yield for glucose and starch were very close because the hydrolysis of starch to glucose was much faster than the dehydration of glucose (Fig. S5†). The high HMF yield (∼22%) in sucrose dehydration was mainly attributed to the formation of one fructose and one glucose in sucrose hydrolysis (Fig. S6†). The poor solubility of cellulose in the reaction system limited its hydrolysis to form low molecular saccharides or glucose, which resulted in very low HMF yield (∼3%) (Fig. S7†). These results indicated that PA–PBI was useful for hydrolysis of di- or polysaccharide to monosaccharide (glucose or fructose) and their dehydration to HMF.

Fig. 4 Dehydration of various saccharides to HMF catalyzed by PA–PBI. Reaction conditions: saccharide (5.0 g), H₂O (5 mL), PA–PBI (0.5 g), MIBK (10 mL), 180 °C, 1 h.

4. Conclusions

We reported an excellent solid catalyst, PA–PBI, for the dehydration of fructose to HMF. The dehydration was operated at a very high initial fructose concentration (50 wt%) with high HMF selectivity (94%). PA–PBI could be easily recycled and reused for four times without the loss of catalytic activity. PA–PBI is also applicable for the hydrolysis and dehydration of a variety of saccharides to form HMF, which makes it a superior catalyst for environment-friendly and cost-effective conversion of biomass to HMF.

Acknowledgements

We gratefully thank the funding from the National Natural Science Foundation of China (No. 51303175, 51203154, 51173181, and 51373166), “The Hundred Talents Program” from the Chinese Academy of Sciences, the Program of Scientific Development of Jilin Province (No. 20150204027GX) and the International S&T Cooperation Program from Department of Science and Technology of Jilin Province (No. 20160414032GH).

References

1. A. Corma, O. de la Torre, M. Renz and N. Villandier, Angew. Chem., Int. Ed., 2011, 50, 2375–2378.
2. C. O. Tuck, E. Pérez, I. T. Horváth, R. A. Sheldon and M. Poliakoff, Science, 2012, 337, 695–699.
3. J. R. Regalbuto, Science, 2009, 325, 822–824.
4. Y. Roman-Leshkov, C. J. Barrett, Z. Y. Liu and J. A. Dumesic, Nature, 2007, 447, 982–U985.
5. Y. Z. Gen Yong and J. Y. Ying, Angew. Chem., Int. Ed. Engl., 2008, 47, 9345–9348.
6. Y. Roman-Leshkov, J. N. Chheda and J. A. Dumesic, Science, 2006, 312, 1933–1937.
7. J. N. Chheda, Y. Roman-Leshkov and J. A. Dumesic, Green Chem., 2007, 9, 342–350.
8. F. S. Asghari and H. Yoshida, Ind. Eng. Chem. Res., 2006, 45, 2163–2173.
9. H. Yan, Y. Yang, D. Tong, X. Xiang and C. Hu, Catal. Commun., 2009, 10, 1558–1563.
10. Y. Shen, J. Sun, Y. Yi, M. Li, B. Wang, F. Xu and R. Sun, Bioresour. Technol., 2014, 172, 457–460.
11. V. V. Ordomsky, J. van der Schaaf, J. C. Schouten and T. A. Nijhuis, J. Catal., 2012, 287, 68–75.
12. Z. Zhang, Q. Wang, H. Xie, W. Liu and Z. Zhao, ChemSusChem, 2011, 4, 131–138.
13. Z. Zhang, B. Liu and Z. Zhao, Carbohydr. Polym., 2012, 88, 891–895.
14. A. Dutta, D. Gupta, A. K. Patra, B. Saha and A. Bhaumik, ChemSusChem, 2014, 7, 925–933.
15. G. K. Parshetti, M. S. Suryadharma, T. P. T. Pham, R. Mahmood and R. Balasubramanian, Bioresour. Technol., 2015, 178, 19–27.
16. X. Qi, M. Watanabe, T. M. Aida and J. R. L. Smith, Green Chem., 2008, 10, 799–805.
17. X. Qi, M. Watanabe, T. M. Aida and R. L. Smith, Ind. Eng. Chem. Res., 2008, 47, 9234–9239.
18. J. E. H. Haibo Zhao, H. Brown and Z. Conrad Zhang, Science, 2007, 316, 1597–1600.
19. S. Xiao, B. Liu, Y. Wang, Z. Fang and Z. Zhang, Bioresour. Technol., 2014, 151, 361–366.
20. Y. Qu, C. Huang, J. Zhang and B. Chen, Bioresour. Technol., 2012, 106, 170–172.
21. Z. Huang, W. Pan, H. Zhou, F. Qin, H. Xu and W. Shen, ChemSusChem, 2013, 6, 1063–1069.
22. B. Liu and Z. Zhang, RSC Adv., 2013, 3, 12313–12319.
23. Z. Zhang, B. Du, L.-J. Zhang, Y.-X. Da, Z.-J. Quan, L.-J. Yang and X.-C. Wang, RSC Adv., 2013, 3, 9201–9205.
24. X. Qi, H. Guo, L. Li and R. L. Smith, ChemSusChem, 2012, 5, 2215–2220.
25. P. M. Hergenrother, Angew. Chem., Int. Ed. Engl., 1990, 29, 1262–1268.
26. T.-K. Ahn, M. Kim and S. Choe, Macromolecules, 1997, 30, 3369–3374.
27. M. M. Han, G. Zhang, Z. G. Liu, S. A. Wang, M. Y. Li, J. Zhu, H. T. Li, Y. Zhang, C. M. Lew and H. Na, J. Mater. Chem., 2011, 21, 2187–2193.
28. S. De, S. Dutta and B. Saha, Green Chem., 2011, 13, 2859–2868.
29. L. X. Xiao, H. F. Zhang, E. Scanlon, L. S. Ramanathan, E. W. Choe, D. Rogers, T. Apple and B. C. Benicewicz, Chem. Mater., 2005, 17, 5328–5333.
30. J. A. Asensio and P. Gómez-Romero, Fuel Cells, 2005, 5, 336–343.
31. R. Bouchet and E. Siebert, Solid State Ionics, 1999, 118, 287–299.
32. F. Tao, H. Song and L. Chou, RSC Adv., 2011, 1, 672–676.
33. H. Y. Mu Zhang, Y. Liu, X. Sun, D. Zhang and D. Xue, Nanoscale Res. Lett., 2012, 7, 38.
34. Q. Zhao, L. Wang, S. Zhao, X. Wang and S. Wang, Fuel, 2011, 90, 2289–2293.
35. C. Fan, H. Guan, H. Zhang, J. Wang, S. Wang and X. Wang, Biomass Bioenergy, 2011, 35, 2659–2665.
36. B. F. M. Kuster, Starch/Staerke, 1990, 42, 314–321.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra00473c

---