Switchable synthesis of 2,5-dimethylfuran and 2,5-dihydroxymethyltetrahydrofuran from 5-hydroxymethylfurfural over Raney Ni catalyst

Abstract

Raney-type metals (Cu, Co and Ni) were employed to catalyze hydrogenation of 5-hydroxymethylfurfural. Switchable synthesis of 2,5-dimethylfuran and 2,5-dihydroxymethyltetrahydrofuran was achieved with 96% and 88.5% yield respectively over Raney Ni, demonstrating high feasibility for industrialization. The excellent yields can be explained by the fact that Raney Ni facilitates the hydrogenation reaction but has limited deoxygenation ability at low temperature, while high temperature promotes the deoxygenation step. The reaction pathway was analyzed by time course experiments and HMF hydrogenation over model catalysts was performed. The reaction mechanism related to the respective catalytic sites was discussed and proposed, which has great implications in the design of efficient and non-noble metal catalysts.

---

1. Introduction

5-Hydroxymethylfurfural (HMF) is a promising platform chemical for future chemical and fuel supply, as it can be obtained from abundant bio-resources.^1,2 The C=O, C–O and furan ring of HMF make it flexible for transformations by various methodologies (e.g., hydrogenation,¹ oxidation,³ decarbonylation,⁴ etherification⁵). Many useful products including 2,5-dimethylfuran (DMF), 2,5-dimethyltetrahydrofuran (DMTHF), 2,5-dihydroxymethylfuran (DHMF), 1,2,6-hexanetriol and 2,5-dihydroxymethyltetrahydrofuran (DHMTHF) could be produced by hydrogenation of HMF. Among them, DMF has been identified as a promising liquid fuel candidate with high energy density (30 kJ cm⁻³) and octane number (RON = 119),⁶ and DHMTHF could be a bulk chemical in the manufacture of polyesters and a precursor for 1,6-hexanediol synthesis.^7–10

For the potential application of both products, it would be of great convenience for switchable synthesis of both chemicals over the same reactor, because the “real-time” switching could respond more rapidly to the market demand and enhance profitability, as well as reduce the reactor downtime.¹¹ It thus would be of great importance for tunable synthesis over the same catalyst, especially over the same non-noble catalyst. To produce DMF, the hydrogenation of aldehyde group and hydrogenolysis of C–O bonds need to be proceeded efficiently. The catalysts with both C=O hydrogenation and deoxygenation ability (e.g., bifunctional/bimetallic catalysts^6,12) could be good candidates. To produce DHMTHF, the furan ring and aldehyde group of HMF need to be saturated without further hydrogenolysis. The catalysts with strong hydrogenation and limited deoxygenation ability could satisfy the needs. Considering the different requirements, a careful development of the catalytic system is required in order to obtain the desired products selectively.

Relatively few reports were conducted on the synthesis of DHMTHF, for which monometallic catalysts are usually applied (e.g., Ru, Ni, Pd) with high efficiency.^8,10,13–16 Nevertheless, few works were further played to investigate DMF synthesis in these systems. Moreover, dehydroxylation reaction is hard to occur in Pd/C with [BMIm]Cl/H₂O system even at 200 °C,¹⁰ making the switchable synthesis impossible. They considered that the protic H₂O leads to the formation DHMF and DHMTHF, while the aprotic solvent tends to form DMF. For HMF hydrogenation to DMF, CuRu/C catalyst was firstly employed and exhibited a yield of ∼71%.⁶ Most following works were conducted over noble metal catalysts (e.g. Ru,^17,18 Pd,^19–21 Au,²² Pt²³). Recently, a metal–acid bifunctional and non-noble Ni–W₂C/AC catalyst was reported with 96% DMF yield, where W₂C played a role of acid site with high deoxygenation ability.¹² Strong acidity of catalyst would catalyze cleavage of C–O bonds of hydroxymethyl groups^24,25 and may not fit for DHMTHF synthesis. Moreover, the development of efficient and non-noble catalytic system still remains to be explored, from the aspect of large-scale industrialization.

We herein employed the non-noble Raney Ni to catalyze HMF hydrogenation and achieved tunable selectivity towards either DHMTHF (96.0%) or DMF (88.5%) over the same catalyst (Scheme 1). The catalyst saturated the C=O bonds and furan ring without broke the C–O bonds at low temperature. Instead, the high temperature compensated the limited hydrogenolysis ability of catalyst and accelerated the cleavage of C–O bonds. To our knowledge, this report possesses at least the following desirable features: (1) highly selective synthesis of DMF and DHMTHF over the same commercial non-noble catalyst, exhibiting high feasibility and profitability for industry, (2) the selective synthesis only needs different reaction temperatures, of great convenience for “real-time” switching, (3) the discussion and understanding of reaction mechanism related to the catalytic sites were proposed, which were important for the rational design of efficient non-noble catalyst for HMF hydrogenation.

Scheme 1 Switchable and efficient synthesis of DHMTHF and DMF from HMF over Raney Ni.

2. Experimental section

2.1 Chemicals

5-Hydroxymethylfurfural (Shanghai De-Mo Pharmaceutical Science and Technology Limited Company, 98%), Raney Ni, Co and Cu catalysts (Dalian Tongyong Chemical Industry Co., Ltd.), ZSM-5 (The Catalyst Plant of Nankai University) and 1,4-dioxane (Sinopharm, AR) were purchased and used as received. 5-Methylfurfural, 2,5-dihydroxymethylfuran and 2,5-dihydroxymethyltetrahydrofuran were purchased from T.C.I. Corporation. The weight percentage of metal and aluminum in the Raney metals was more than 93% and less than 7%, respectively.

2.2 Analytical method

XRD patterns were recorded by a X-ray diffractometer (MiniFlex II, Rigaku) with Cu kα radiation operating at 40 kV, with a rate of 4 °C min⁻¹.

The samples after reaction were analyzed on GC-920 equipped with a DB-WAXETR capillary column and a FID detector. The products were identified by GC-MS and comparison of retention times of pure chemicals.

2.3 Catalytic tests

The tests were performed over a 100 ml tank reactor. For a typical procedure, the reactor was fed with HMF (1.5 g), Raney metals (0.5 g) and 1,4-dioxane (35 ml), then sealed and purged by H₂ (5 times). After that, the reactor was filled with 1.5 MPa H₂ and heated to object temperature within 30 minutes. After the test, the reactor was quenched in ice-water, and then the liquid and gas products were analyzed by a GC instrument with a FID detector.

3. Results and discussion

3.1 Catalyst screening

All the Raney metals have the metal (Cu, Co and Ni) loadings higher than 93%. The fresh Raney Cu, Co and Ni catalysts were also characterized by XRD measurement (Fig. 1). The XRD patterns of the Raney Ni and Cu catalysts are consistent with the typical crystalline Ni (PDF#04-0850) and Cu (PDF#04-0836) respectively, indicating that both metals are in a complete crystalline form. Unlike Raney Ni and Cu, Raney Co exhibited the weak peaks which could be attributed to the characteristics of crystalline Co (PDF#15-0806). The result illustrated that Co has amorphous or/and microcrystalline construction and is highly dispersed. No diffraction peaks of crystalline Al were observed for all the cases.

Fig. 1 XRD patterns of fresh Raney Cu, Co and Ni catalysts.

The performance of HMF hydrogenation over these Raney metals was compared at 180 °C (Table 1, entries 1, 2 and 8). Raney Cu exhibited a low HMF conversion of 25.1% and a DMF selectivity of 42.0%. The main byproduct for Raney Cu was 5-methylfurfural (5-MF), indicating the weak C=O hydrogenation ability of the catalyst. 2,2-Methylenebis(5-methylfuran) (OMBM) and 5,5′-(oxybis(methylene))bis(2-methylfuran) (A) (2-furan, formed by the condensation of 5-MF and MFA^26,27) were also observed, as a result of the insufficient hydrogenation reactivity of Cu catalyst. Raney Co catalyst exhibited a HMF conversion of 94.3% and a DMF selectivity of 78.5%, which were greatly higher than that of Raney Cu.

Table 1 Hydrogenation of HMF over Raney metal catalysts^a

Entry	Catalyst	T/°C	Conv./%	Selectivity/%
				DMF	5-MF	MFA	DHMF	DHMTHF	2-Furan	Others
a Reaction conditions: HMF 1.5 g, catalyst 0.5 g, 1,4-dioxane 35 ml, 1.5 MPa H2, and reaction time 15 h at the defined temperatures. 2-Furan means OMBM and A. Others mainly include 2,5-DMTHF, 2-MF, 2-hexanol and C–C cracking products.b 30 h was used to optimize the DHMTHF yield.
1	Raney Cu	180	25.1	42.0	34.5	4.5	0	3.0	11.4	4.6
2	Raney Co	180	94.3	78.5	1.8	2.7	0	2.7	5.9	8.4
3	Raney Ni	100	100	1.4	0	2.0	27	67.4	0	2.2
4^b	Raney Ni	100	100	1.9	0	0	0	96.0	0	2.1
5	Raney Ni	120	100	6.0	0	5.1	18.1	65.4	0	5.4
6	Raney Ni	140	100	10.2	0	8.3	13.8	60.7	0	7.0
7	Raney Ni	160	100	46.3	0	8.6	6.1	32.6	0	6.4
8	Raney Ni	180	100	88.5	0	0	0	3.0	0	8.5
9	Raney Ni	200	100	79.8	0	0	0	3.0	0	17.2
10	Raney Ni	220	100	69.3	0	0	0	0	0	30.7

---

Raney Ni exhibited the HMF conversion of 100% and a higher DMF selectivity of 88.5% than other catalysts, indicating the superior reactivity of Ni catalyst. The Raney Ni catalyst was then detailed studied at different reaction temperatures. It was found that DHMTHF could be synthesized at 100 °C with a yield up to 96.0% and a DMF yield of 88.5% could be realized upon increasing temperature to 180 °C (Table 1, entries 4 and 8). When further increasing temperature, C–C cracking products and DMTHF were generated and DMF selectivity was decreased. Herein, this is the first report of highly selective synthesis of DHMTHF and DMF over the same non-noble catalyst, tuned simply by reaction temperatures.

3.2 Reaction mechanism

The reaction pathways were investigated by time courses at 100 °C and 180 °C to understand the tunable synthesis of DMF and DHMTHF (Fig. 2). Typical GC charts of the products at both 100 °C and 180 °C were shown in Fig. S1 and S2† respectively. In both cases, HMF was totally converted within 10 h, indicating the strong hydrogenation ability of Raney Ni. The absence of 5-hydroxymethyltetrahydrofurfural and 5-methyltetrahydrofurfural for both temperatures could also indicate that hydrogenation of aldehyde group is a fast reaction among the three functional groups (C=O, C=C, C–O), which is accordance with previous reports.^14,15 At 100 °C, DHMF yield reached a plateau at 5 h and then decreased, whereas DHMTHF yield increased monotonically with increase of reaction time. The results revealed a step-wise reaction sequence (Scheme 2, HMF → DHMF → DHMTHF). No 5-methyltetrahydrofurfuryl alcohol (MTHFA) or DMTHF was detected, indicating that hydrogenolysis of hydroxymethyl group was hard to occur. It should be noted that DHMF yield increased until HMF was fully converted, revealing that the saturation of furan ring was slow. The unconverted HMF may inhibit the DHMF conversion to DHMTHF due to the fact that η²(C,O)-type configuration endows HMF with stronger adsorption than DHMF and the competitive adsorption on the active site hinders conversion of DHMF.²⁸ On the other hand, the C=C bonds on the ring are conjugated and not easily to be hydrogenated.²⁹ Thus, HMF converted to DHMF rapidly and high yield of DHMTHF could be obtained at long reaction time. At 180 °C, trace amount of DHMTHF was detected, revealing that hydrogenation of furan ring is not preferential. A series of C–O cleavage products (MFA, 5-MF, DMF, DMTHF) were obtained, indicating that hydrogenolysis reaction could be promoted at high temperature for Raney Ni catalyst, which is in accordance with previous report.³⁰ DHMF yield exhibited a similar profile to that of 100 °C, passing through the maximum at 2 h, indicating that it was a common intermediate for both cases. 5-MF was also observed at the initial stage, revealing that hydrogenolysis of alcohol group of HMF occurs in parallel to the hydrogenation of aldehyde group at high temperature. Based on the above results and previous reports,²⁹ pathways of HMF hydrogenation at high temperature are summarized (Scheme 2). In all, DHMF is the main intermediate for both cases due to the fast hydrogenation of aldehyde group. At 100 °C, hydrogenolysis of –CH₂OH was slow and little hydrogenolysis products were detected. With increase of temperature to 180 °C, the hydrogenolysis of –CH₂OH is greatly promoted and 5 MF is observed. However, due to the weak deoxygenating ability of Raney Ni catalyst, DHMF is still the main intermediate. The behavior is dramatically different with Ni–W₂C/AC catalyst which was designed with Lewis acid sites and strong deoxygenating ability.¹²

Fig. 2 HMF conversion and product yield versus reaction time at 100 °C (a) and 180 °C (b). Reaction conditions: HMF 1.5 g, Raney Ni 0.5 g, 1,4-dioxane 35 ml, 1.5 MPa H₂.

Scheme 2 Proposed reaction pathway for HMF hydrogenation/hydrogenolysis (MTHFA was presented only when ZSM-5 was added).

At 100 °C, DHMTHF was achieved with high yield and hydrogenolysis products were not detected. We ascribed the high yield to the weak acidity of Raney Ni catalyst and proved it in Table 2 (see below). Different from reactivity at 100 °C, hydrogenolysis of hydroxymethyl group is greatly promoted at 180 °C. The slow ring hydrogenation and rapid hydrogenolysis of hydroxymethyl group contributed to the high DMF yield and trace amount of DHMTHF. It is also interesting that it was hard to stop the reaction in DHMF at 100 °C, while the ring reserved at 180 °C. According to the previous report, the polar CH₂OH groups in the DHMF molecule have more affinity to the catalyst surface and can assist in the adsorption of DHMF onto the Ni metal surface.²⁸ However, the methyl group of DMF forces the aromatic ring further away from the catalytic surface, thereby reducing the probability of a reaction.³¹ They also reported that the protective methyl group contributes to a less selectivity toward cracking in DMF hydrogenation than in 2-MF hydrogenation. Thus, DHMF is relatively more reactive than DMF and susceptible to hydrogenation. When DMF was formed, hydrogenation of furan ring became slower. Briefly, (1) the C=O bond of HMF could occupy active sites and hinder DHMF adsorption, (2) the C=C bonds are conjugated and not easily to be hydrogenated, (3) the C–O could assist in DHMF adsorption and be cleaved at high temperature. In all, the different adsorption geometry and strength of functional groups (C=O, C=C and C–O) offer a degree of flexibility to tunable DMF and DHMTHF synthesis.

Table 2 Hydrogenation of HMF over model catalysts^a

Entry	Catalyst	Yield/%
		DMF	DMTHF	DHMTHF	MTHFA	DHMF	Others
a Reaction conditions: Raney Ni: Raney Ni (0.5 g); Raney Ni + ZSM-5: Raney Ni (0.5 g) and ZSM-5 (0.1 g), HMF 1.5 g, 1,4-dioxane 35 ml, 100 °C 1.5 MPa H2, reaction time = 15 h.b Reaction time = 30 h.
1	Raney Ni	1.4	0	67.4	0	27	4.2
2	Raney Ni + ZSM-5	1.6	0.8	86.5	5.1	2.4	3.6
3	Raney Ni^b	1.9	0	96.0	0	0	2.1
4	Raney Ni + ZSM-5^b	0.8	2.9	85.9	4.5	0	5.9

---

To support the above mechanism, hydrogenation/hydrogenolysis of intermediates (DHMF and 5-MF) was performed under the working conditions (Scheme S1–S3†). DHMF is a common intermediate for both DMF and DHMTHF synthesis (Scheme S1 and S2†). The reaction temperature has great effects on the reaction pathway. At low temperature, hydrogenation of furan ring occurred and DHMTHF was formed. At high temperature, hydrogenolysis of the side –CH₂OH was promoted and DMF was achieved. The hydrogenation of 5-MF (Scheme S3†) revealed that MFA can be obtained from 5-MF hydrogenation (the hydrogenation of C=O bonds) and DMF can be obtained from MFA hydrogenolysis (the hydrogenolysis of CH₂OH group).

Pathways of HMF hydrogenation are summarized (Scheme 2). For Raney Cu, Co catalysts, some condensation intermediates (OMBM, A) were detected due to the slow conversion of 5-MF and MFA. For Ni catalyst with superior ability, reaction pathways were highly depended on reaction temperatures. The temperature influenced the occurrence of CH₂–OH hydrogenolysis and determined the product distributions. In our case, Raney Ni catalyst plays a role in the hydrogenation step but has limited deoxygenation ability at low temperature, while high temperature promotes the deoxygenation step. Therefore, a tunable synthesis of products was achieved. The key to our success could be attributed to the nature of Raney Ni and avoidance of additional acid sites, considering that a metal catalyst with high acidity could catalyze hydrogenolysis reaction more effectively.¹²

To obtain DMF, hydrogenolysis of hydroxymethyl group is needed. It is generally concluded that metal–acid bifunctional catalysts are good candidates in hydrogenolysis reaction, where acid promotes dehydration of CH₂–OH and metal acts as active phase to saturate/hydrogenate intermediates.^32,33 Indeed, W₂C acts as acid to catalyze deoxygenation reaction for DMF synthesis has been reported with high efficiency.¹² However, the acid sites may not fit for DHMTHF synthesis, in which C=O and C=C bonds are hydrogenated without further hydrogenolysis. To examine the roles of acid sites for DHMTHF synthesis, a solid acid (ZSM-5) was physically mixed with Raney Ni and employed in HMF hydrogenation (Table 2). HMF was fully converted in all cases. ZSM-5 could promote DHMTHF yield from 67.4 to 86.5% when the intermediate DHMF was not totally converted (15 h, Table 2, entry 1 and 2). However, MTHFA was formed, indicating that hydrogenolysis of hydroxymethyl group was also promoted. Besides, no obvious change was observed for the mechanical-mixed catalyst after extending the reaction time from 15 to 30 h. It might indicate the deactivation of the catalyst. This result might be also caused by the fact that acid could promote DHMF conversion to DHMTHF, but have little effect on DHMTHF conversion, considering that DHMTHF was more stable than DHMF.³⁴ The results clearly suggested that critical roles of acid sites were to promote both hydrogenation of furan ring and hydrogenolysis of CH₂–OH group. The results also suggested that DMF could be formed via dehydration of DHMF, which is similar with 2-methylfuran formation from furfuryl alcohol hydrogenolysis.³⁵ Thus, when the target compound involves C–O hydrogenolysis, control of the acidity is essential. The Raney Ni system facilitated the hydrogenation reaction at low temperature and the hydrogenolysis reaction at high temperature, achieving a switchable synthesis of DHMTHF and DMF.

3.3 Catalyst stability

Reusability of Raney Ni was also investigated under relatively severe condition (180 °C) (Fig. 3). After each test, the catalyst was washed with solvent (3 times) and recycled. After 5 runs, the conversion of HMF only decreased to 92.1% compared to 100% conversion of the fresh catalyst, demonstrating the good hydrogenation ability of metallic Ni. Nevertheless, DMF selectivity decreased from 88.5% to 74%. The deactivated Raney Ni catalyst exhibited a similar catalytic performance with Raney Cu and Co catalysts (presence of 2-furan and 5-MF), indicating a loss of hydrogenation ability. No evident growth of Ni particles was observed after 5 runs (Fig. 4). The loss of DMF yield was thus probably caused by the mass loss of catalyst during washing process and the formation of high molecular weight byproducts which adsorbed on the metal sites, as confirmed in previous reports.^12,18 Although the switchable synthesis of DMF and DHMTHF was achieved with high yields and a systematical reaction mechanism was proposed, the low reactivity and stability would hinder the applications. Thus, future work will be needed for exploring efficient and especially stable catalysts for selective hydrogenation of HMF.

Fig. 3 Recycle experiments of the Raney Ni catalyst. Conditions: HMF 1.5 g, Raney Ni 0.5 g, 1,4-dioxane 35 ml, 1.5 MPa H₂, 180 °C, 15 h.

Fig. 4 XRD patterns of fresh and used Raney Ni catalyst.

4. Conclusion

We have demonstrated a highly efficient and selective non-noble catalyst that can be used for hydrogenation of HMF. Notably, the reaction temperature offers a degree of flexibility to the process such that either DMF or DHMTHF could be obtained as the major product, due to the sufficient hydrogenation ability and weak hydrogenolysis ability of the Raney Ni catalyst. The detailed reaction mechanism and catalyst reusability were also studied. The results here demonstrated the potential of non-noble metal catalysts for HMF hydrogenation.

Acknowledgements

This work was supported by the Major State Basic Research Development Program of China (973 Program, no. 2012CB215305).

References

1. R.-J. van Putten, J. C. van der Waal, E. de Jong, C. B. Rasrendra, H. J. Heeres and J. G. de Vries, Chem. Rev., 2013, 113, 1499–1597.
2. N. Shi, Q. Liu, Q. Zhang, T. Wang and L. Ma, Green Chem., 2013, 15, 1967–1974.
3. C. Carlini, P. Patrono, A. M. R. Galletti, G. Sbrana and V. Zima, Appl. Catal., A, 2005, 289, 197–204.
4. F. M. A. Geilen, T. vom Stein, B. Engendahl, S. Winterle, M. A. Liauw, J. Klankermayer and W. Leitner, Angew. Chem., Int. Ed., 2011, 50, 6831–6834.
5. P. Che, F. Lu, J. Zhang, Y. Huang, X. Nie, J. Gao and J. Xu, Bioresour. Technol., 2012, 119, 433–436.
6. Y. Román-Leshkov, C. J. Barrett, Z. Y. Liu and J. A. Dumesic, Nature, 2007, 447, 982–985.
7. F. C. A. Figueiredo, E. Jordão and W. A. Carvalho, Appl. Catal., A, 2008, 351, 259–266.
8. T. Buntara, S. Noel, P. H. Phua, I. Melián-Cabrera, J. G. de Vries and H. J. Heeres, Angew. Chem., Int. Ed., 2011, 50, 7083–7087.
9. T. Buntara, S. Noel, P. H. Phua, I. Melián-Cabrera, J. G. de Vries and H. J. Heeres, Top. Catal., 2012, 55, 612–619.
10. H. Cai, C. Li, A. Wang and T. Zhang, Catal. Today, 2014, 234, 59–65.
11. J. G. Stevens, R. A. Bourne, M. V. Twigg and M. Poliakoff, Angew. Chem., Int. Ed., 2010, 49, 8856–8859.
12. Y. B. Huang, M. Y. Chen, L. Yan, Q. X. Guo and Y. Fu, ChemSusChem, 2014, 7, 1068–1072.
13. T. J. Connolly, J. L. Considine, Z. Ding, B. Forsatz, M. N. Jennings, M. F. MacEwan, K. M. McCoy, D. W. Place, A. Sharma and K. Sutherland, Org. Process Res. Dev., 2010, 14, 459–465.
14. Y. Nakagawa and K. Tomishige, Catal. Commun., 2010, 12, 154–156.
15. R. Alamillo, M. Tucker, M. Chia, Y. Pagán-Torres and J. Dumesic, Green Chem., 2012, 14, 1413–1419.
16. Y. Nakagawa, K. Takada, M. Tamura and K. Tomishige, ACS Catal., 2014, 4, 2718–2726.
17. Y. Zu, P. Yang, J. Wang, X. Liu, J. Ren, G. Lu and Y. Wang, Appl. Catal., B, 2014, 146, 244–248.
18. J. Jae, W. Q. Zheng, R. F. Lobo and D. G. Vlachos, ChemSusChem, 2013, 6, 1158–1162.
19. D. Scholz, C. Aellig and I. Hermans, ChemSusChem, 2014, 7, 268–275.
20. T. Thananatthanachon and T. B. Rauchfuss, Angew. Chem., Int. Ed., 2010, 122, 6766–6768.
21. M. Chatterjee, T. Ishizaka and H. Kawanami, Green Chem., 2013, 16, 1543–1551.
22. S. Nishimura, N. Ikeda and K. Ebitani, Catal. Today, 2013, 232, 89–98.
23. G.-H. Wang, J. Hilgert, F. H. Richter, F. Wang, H.-J. Bongard, B. Spliethoff, C. Weidenthaler and F. Schüth, Nat. Mater., 2014, 13, 293–300.
24. I. Gandarias, P. Arias, J. Requies, M. Güemez and J. Fierro, Appl. Catal., B, 2010, 97, 248–256.
25. M. Massa, A. Andersson, E. Finocchio and G. Busca, J. Catal., 2013, 307, 170–184.
26. M. Chidambaram and A. T. Bell, Green Chem., 2010, 12, 1253–1262.
27. M. R. Grochowski, W. Yang and A. Sen, Chem.–Eur. J., 2012, 18, 12363–12371.
28. Y. Nakagawa, H. Nakazawa, H. Watanabe and K. Tomishige, ChemCatChem, 2012, 4, 1791–1797.
29. L. Hu, L. Lin and S. Liu, Ind. Eng. Chem. Res., 2014, 53, 9969–9978.
30. A. Perosa and P. Tundo, Ind. Eng. Chem. Res., 2005, 44, 8535–8537.
31. C. Aliaga, C.-K. Tsung, S. Alayoglu, K. Komvopoulos, P. Yang and G. A. Somorjai, J. Phys. Chem. C, 2011, 115, 8104–8109.
32. Y. Kusunoki, T. Miyazawa, K. Kunimori and K. Tomishige, Catal. Commun., 2005, 6, 645–649.
33. Y. Zhu, X. Kong, X. Li, G. Ding, Y. Zhu and Y.-W. Li, ACS Catal., 2014, 4, 3612–3620.
34. S. Yao, X. Wang, Y. Jiang, F. Wu, X. Chen and X. Mu, ACS Sustainable Chem. Eng., 2013, 2, 173–180.
35. V. Vorotnikov, G. Mpourmpakis and D. G. Vlachos, ACS Catal., 2012, 2, 2496–2504.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra09550b

---