Catalytic oxidative C–C bond cleavage route of levulinic acid and methyl levulinate

Abstract

Recently, obtaining value-added chemicals from biomass resources has attracted considerable attention. Levulinic acid is one of the most important biomass platform compounds, which could be obtained from carbohydrate biomass. In this work, levulinic acid was selectively converted into C4 product, including succinic anhydride, via catalytic oxidation with a manganese catalyst in acetic anhydride. Moreover, an unexpected product of maleic anhydride was obtained, which greatly differs from that of levulinate ester. The pathway for formation of maleic anhydride was studied by monitoring and confirming intermediates α-angelica lactone and its derivative 2-methyl-5-oxotetrahydro-2-furanyl acetate. Based on the obtained mechanistic information, the different behaviour between the oxidative cleavage of levulinic acid and levulinate ester was further discussed.

---

Introduction

Succinic acid is a versatile platform chemical that can be utilized for preparation of a variety of value-added downstream products such as succinate ester, succinamide, γ-butyrolactone, 1,4-butanediol, tetrahydrofuran, 2-pyrrolidone and N-methyl-2-pyrrolidone.^1,2 Succinic anhydride is the precursor of succinic acid, both of which are important monomers for biodegradable material poly(butylene succinate).^3–5 Succinic acid and succinic anhydride were traditionally produced from the hydrogenation of maleic anhydride,^6–12 which was obtained from catalytic oxidation of fossil derived feedstocks such as benzene^13–15 and butane.^16–18 However, the raw materials applied in these procedures are short of oxygen atom; therefore, extra oxygen resource was needed.

As a renewable oxygen-containing carbon resource, biomass has attracted extensive attention in recent years. Several studies have focused on the conversion of saccharides, such as cellulose, hemicellulose, glucose, and fructose, to value-added chemicals and platform chemicals,^19–27 among which levulinic acid is an important C5 biomass platform compound that can be obtained by hydrolysis of cellulose, glucose and 5-hydroxymethylfurfural.^28–32 Levulinic acid could be converted to various useful chemicals, including levulinate esters,³³ γ-valerolactone^34,35 and 1,4-pentadiol.^36,37

Levulinic acid contains carbonyl and carboxyl group and is structurally similar to succinic acid.³⁸ Therefore, production of succinic acid from oxygenation of levulinic acid is preferable considering atom economy.^39,40

To date, there are a few reports on catalytic conversion of levulinic acid to succinic acid or succinic anhydride using molecular oxygen. The first example was a patent in 1954 that described V₂O₅ catalysed succinic anhydride and succinic acid production from levulinic acid at high temperatures of 200–400 °C.⁴¹ Recently, Podolean et al. reported the use of noble metal ruthenium for the conversion of levulinic acid to succinic acid at 150 °C.⁴²

However, it is still a challenge to develop a catalytic procedure characterized by using dioxygen under mild conditions. In 2013, our group reported the first example of manganese-catalysed selective oxidation of levulinate to succinate in 58.6% yield with O₂ under mild conditions. Mn(III) acetate was disclosed as an efficient catalyst for the oxidative C–C cleavage of methyl levulinate at the terminal methyl-carbonyl position.³⁸ The oxidation of the substrate methyl levulinate in our previous research was conducted in acetic anhydride. However, the use of methyl levulinate is not arbitrary. When levulinic acid was oxidized under the similar conditions, remarkably different results were observed. Succinic anhydride (1), as a succinate series C4 product, decreased dramatically compared to methyl levulinate, whereas a certain amount of maleic anhydride (2) was detected as a maleate series C4 product. In addition, the route for levulinic acid oxidation is vague as well.

Hitherto, no report has been published on oxidation of levulinic acid to succinic anhydride with manganese compounds as a catalyst; moreover, no relevant study on the significant difference of the oxidation behaviour between the two levulinates has been investigated thoroughly. On the basis of the previous study, herein, we investigated the reaction process and confirmed the possible intermediates. Moreover, the difference between the oxidation of levulinic acid and levulinate ester was also discussed.

Experimental

Materials and methods

Levulinic acid, dimethyl malonate and dimethyl oxalate were obtained from Aladdin chemistry Co. Ltd. Methyl levulinate and α-angelica lactone were obtained from TCI (Shanghai) Development Co. Ltd. Succinic acid was purchased from Sigma-Aldrich. Dimethyl succinate and dimethyl maleate were supplied by Alfa Aesar. Maleic anhydride was purchased from Shenyang chemical reagent Co. Ltd. Succinic anhydride was obtained from Sinopharm chemical reagent Co. Ltd. 2-Methyl-5-oxotetrahydro-2-furanyl acetate was separated from the experiments by column chromatography and characterized on a Bruker 400 MHz spectrometer. Methanol, acetic acid, ethyl acetate, DMF, DMSO, dichloromethane, acetonitrile and cyclohexane were purchased from Tianjin Kermel chemical reagent Co. Ltd. Methanol, acetic acid, ethyl acetate, DMF, DMSO, dichloromethane, acetonitrile and cyclohexane were of analytical grade and dried using activated 3 A molecular sieve before use. Mn(OAc)₃·2H₂O and Mn(acac)₃ were purchased from Alfa Aesar. Mn(OAc)₂·4H₂O was obtained from Aladdin chemistry Co. Ltd. Deionized water was purified by Milli-Q system (Millipore). All other reagents were commercially available and used as received.

Typical procedure for catalytic oxidation

In a typical experiment, levulinate (2.5 mmol), Mn(OAc)₃·2H₂O (5 mol%) and 2 mL solvent were added into a 25 mL Teflon-lined stainless steel autoclave equipped with a magnetic stirrer, a pressure gauge and an automatic temperature control apparatus. After purging for 6 times to exclude air, O₂ was charged to 0.5 MPa. The autoclave was heated to 90 °C and kept for the desired reaction time.

Product analysis

The autoclave was cooled to room temperature after the reaction. The liquid products were transferred into a volumetric flask followed by addition of an internal standard for analysis. The liquid reaction mixture was analysed by GC calibration curve.

As for the substrate of methyl levulinate, all the liquid reaction mixture was transferred into a 50 mL round-bottomed flask, and then BF₃·Et₂O (150 mg) and absolute methanol (20 mL) were added and reflux for 6 h. This method is verified experimentally and similar methods are also employed for analysis of carboxylic acid in the previous reports.^38,43 GC measurements were conducted on an Agilent 7890A GC equipped with an auto sampler and a flame ionization detector. DB-225 (30 m × 250 μm × 0.25 μm) and DB-17 (30 m × 320 μm × 0.25 μm) capillary columns were employed for analysis of reaction mixtures of levulinic acid and methyl levulinate, respectively. Identification of main products was based on a GC-MS (Agilent 7890A/5975C) system equipped with an Agilent HP-5ms (30 m × 250 μm × 0.25 μm) as well as by comparison with authentic samples. 1,2,4,5-Tetramethylbenzene (TMB) was used as the internal standard. The product distribution was shown on the molar basis.

The conversion (mol%) of levulinates and yield (mol%) of main products were calculated as follows:

Pi: succinic anhydride, maleic anhydride, angelica lactones, 2-methyl-5-oxotetrahydro-2-furanyl acetate, dimethyl succinate, dimethyl malonate, dimethyl oxalate and dimethyl maleate.

Results and discussion

Comparison of oxidation of levulinic acid and methyl levulinate

Our group's previous study focused on the manganese(III) catalysed aerobic oxidation of methyl levulinate to dimethyl succinate in acetic anhydride, and 95.3% conversion with 58.6% yield of succinate was obtained.³⁸ In this study, when levulinic acid was oxidised under the same reaction conditions, 97.6% levulinic acid was converted, whereas the yield of 1 decreased dramatically to 6.8% (Table 1, entry 1). In addition, reaction catalysed by Mn(acac)₃ showed similar results: 97.7% conversion with negligible yield of 1 (6.6%) was detected compared with 92.4% conversion and 49.8% yield of succinate from that of methyl levulinate. 2 was observed in both systems. Based on these results, it was suggested that there might exist a different conversion pathway between the two substrates, while methyl levulinate may be more preferable for the production of succinate through catalytic oxidation compared with levulinic acid.

Table 1 Catalytic oxidation of levulinic acid^a

Entry	Catalyst	Conv. [mol%]	Yield of main products [mol%]					Others
			1	2	3 + 4	5	C4–C5 products
a Reaction conditions: levulinic acid (2.5 mmol), catalyst (5 mol% Mn), acetic anhydride (2 mL), 90 °C, 0.5 MPa O2, 10 h. Data were analysed by GC.b 10 mol% Mn.c 20 h.d Substrate: 5 (2-methyl-5-oxotetrahydro-2-furanyl acetate). Others refer to those that cannot be quantified through calibration curve but conformed by GC-MS, including 1,2,3,5-tetra-O-acetylpentofuranose, acetoxyacetic acid, 1,3-cyclopentanedione, 1,1-ethanediol diacetate, 4-hydroxy-2-pentenoic acid and 1,1,2-ethanetriol triacetate.
1	Mn(OAc)3·2H2O	97.6	6.8	14.8	5.8	38.3	65.7	31.9
2	Mn(acac)3	97.7	6.6	15.3	6.5	31.5	59.9	37.8
3	None	96.1	0.9	2.0	3.5	83.7	90.1	6.0
4	Mn(OAc)2·4H2O	96.8	4.9	12.7	1.4	44.2	63.2	33.6
5^b	Mn(OAc)3·2H2O	97.8	5.3	8.8	3.9	61.7	79.7	18.1
6^c	Mn(OAc)3·2H2O	97.9	9.4	26.0	1.2	1.0	37.6	60.3
7^d	Mn(OAc)3·2H2O	61.8	7.1	17.0	4.7	—	28.8	33.0

---

In order to investigate the differences, controlled experiments were conducted. In the oxidation of levulinic acid catalysed by Mn(OAc)₃·2H₂O and Mn(acac)₃, 2 was obtained unexpectedly as one of the main C4 products with the yield of 14.8% and 15.3%, respectively (Table 1, entries 1 and 2). In contrast, succinate was given as the only C4 product in the oxidation of methyl levulinate,³⁸ which further confirmed that different reaction routes existed for the oxidation of levulinic acid and methyl levulinate. According to the GC-MS results, α-angelica lactone (α-AL, 3), β-angelica lactone (β-AL, 4) and its derivative 2-methyl-5-oxotetrahydro-2-furanyl acetate (5) were detected during the oxidation (Fig. S1, S3 and S13†). About 6% of ALs (α,β-AL) with more than 30% of 5 was obtained after 10 h of reaction under the catalysis of Mn(III) (Table 1, entries 1 and 2). To further explore the reaction mechanism, comparative experiments of these two substrates were conducted under the same conditions while samples were taken periodically at different time along the reaction course, whereas possible intermediates and products were monitored and confirmed by GC (Fig. 1) and GC-MS (Fig. S1–S3†). The products of oxidation of levulinic acid and methyl levulinate were traced and are shown in Fig. 1(a) and (b); suspected intermediates 3 and its isomer 4 as well as its derivative 5 were generated during the reaction of levulinic acid. The dramatic difference between the oxidation behaviour of levulinic acid and methyl levulinate illustrated that there exists a “non-oxidative” consumption of levulinic acid (Scheme 1). During that process, it generates 3 through dehydration cyclization with the aid of acetic anhydride. This was followed by an addition reaction with acetic acid that derived from the dehydration process of acetic anhydride to form 5. Moreover, the yield of 5 was in relatively large quantity particularly without any catalysts (83.7%, Table 1, entry 3). This is in agreement with previous studies. Mascal et al. recently reported the dehydration of levulinic acid under the catalysis of solid acid to give 3.⁴⁴ The treatment of levulinic acid and methyl levulinate with acetic anhydride under a nitrogen atmosphere (Fig. 2) showed that products 3, 4 and 5 were generated due to the dehydration process of levulinic acid, whereas no such products were obtained for that of methyl levulinate. These results indicated that acetic anhydride promoted the dehydration of levulinate substrates, whereas the ester group of methyl levulinate prohibited the process from generating intermediate 3, 4 and 5. On the basis of former reports and our results, we can come to the conclusion that the dehydration of levulinic acid generates 3 that undergoes isomerization to 4 ⁴⁵ and further addition reaction with acetic acid to form 5. However, the reaction of methyl levulinate does not undergo this route, nor does it give any ALs or its derivatives as intermediates under the same conditions (Fig. 1(b) and 2).

Fig. 1 (a) GC trace of the oxidation of levulinic acid at different reaction time. The retention time corresponds to the following compounds: 7.37 min (α-angelica lactone), 10.42 min (maleic anhydride), 11.68 min (β-angelica lactone), 15.83 min (2-methyl-5-oxotetrahydro-2-furanyl acetate), and 16.06 min (succinic anhydride). (b) GC trace of the first period of the oxidation of methyl levulinate (methyl levulinate retention time: 10.6 min).

Scheme 1 Dehydration of levulinic acid to angelica lactones and its derivative.

Fig. 2 Controlled experiments of dehydration of levulinic acid and methyl levulinate. The retention time corresponds to the following compounds: 7.37 min (α-angelica lactone), 7.92 min (inner standard: 1,2,4,5-tetramethylbenzene (TMB)), 10.41 min (methyl levulinate), 11.68 min (β-angelica lactone), and 15.83 min (2-methyl-5-oxotetrahydro-2-furanyl acetate).

As shown in Fig. 1(a), ALs formed instantly by the dehydration of levulinic acid in acetic anhydride during the reaction, thus resulting in the increase of ALs in the first 20 min followed by the decrease to a relatively stable level as the reaction proceeded to 2 h. Moreover, the quantity of the intermediate 5 rapidly increased to a high level within the first 1 or 2 h. When the reaction time was prolonged to 3 h and longer, an increasing amount of the dehydration product 3 was detected after 10 or 15 h, whereas the amount of 5 almost remained in a relatively stable level. As for this abnormal phenomenon, it is supposed that this was partially due to the decreasing consumption of 5 due to the deactivation of the catalyst Mn(III) by trace water formed during the reaction process. This can be demonstrated by the decrease in quantity of the intermediates 3 and 5 as well as the increase of products 1 and 2 by adding fresh catalyst to the system and prolonging the reaction for another 5 h. Doubling the catalyst amount at the beginning of the reaction did not increase any of the main C4 products except the by-product 5 (Table 1, entry 5). Moreover, prolonging the reaction time showed some positive effects particularly on 2 (26.0%) with nearly trace amount of 5 detected (Table 1, entry 6), which indicated that intermediate 5 could be further converted to 1, 2 and 3. This was also evidenced by the oxidation of 5 (as the substrate) under our reaction system (Table 1, entry 7) in which a similar product distribution was obtained; other side products can be confirmed by GC-MS (Fig. S3–S12†).

The studies of the oxidation behaviour of levulinic acid and methyl levulinate proved that 3 and its derivatives were key intermediates during the reaction in the presence of acetic anhydride. However, the formation route of 2 is still vague. In order to figure out the relationship between the formation of 2 and the generation of the intermediate 3, extra controlled experiments are required. When 3 was subjected to similar conditions, about 13.1% of 2 and less than 1% of 1 and 5 were detected within 1 h, which was in quite resemblance with the oxidation of levulinic acid. Neither of which could be associated with the oxidation of methyl levulinate wherein succinate (58.6%) consisted exclusively the majority of the final products without any maleate or other detectable C4 products.³⁸ As for the oxidation of 3, the conversion reached more than 99%. We assumed that these results may be due to its high activity and instability, particularly under specific conditions with a high concentration. As shown in Fig. 1(a), during the oxidation of levulinic acid, quite a small amount of ALs were formed through the dehydration process and the concentration of which remained relatively low level during the entire reaction course. The concentration difference of ALs between the two systems and the high activity and instability of 3 might account for the low selectivity of the target C4 product. Some of the products were confirmed and shown in Fig. S14–S18.†

Consequently, a plausible reaction route for the Mn(III)-catalysed oxidation of levulinic acid and methyl levulinate in acetic anhydride was proposed, as summarized in Scheme 2. The dehydration of levulinic acid with the aid of solvent acetic anhydride played an important role, in which the substrate was dehydrated to ALs followed by the subsequent oxidation ring opening to afford 2. The controlled experiments of 3 further supported the hypothesis that 3 and its isomer 4, as well as derivative 5 were the specific intermediates in the catalytic oxidation of levulinic acid, which differed from that of methyl levulinate.

Scheme 2 Plausible protocol of the oxidation of levulinic acid and methyl levulinate.

The solvent effect on the oxidation of levulinate

Since the generation of intermediates and 2 was closely related to the dehydration role of acetic anhydride solvent, the solvent effect on the manganese catalytic oxidation of levulinic acid and methyl levulinate should be considered. The results in different solvents are summarized in Table 2, in which products were analysed after esterification with excess of methanol in order to facilitate detection and comparison. When the oxidation of levulinic acid was conducted in acetic acid and DMF separately, only 2.5% and 7.2% of dimethyl maleate with trace amount of dimethyl succinate were obtained though the conversion did not vary significantly (82.1% and 74.8%, respectively) (Table 2, entries 1 and 2). In other solvents such as ethyl acetate, methanol, acetonitrile, cyclohexane, DMSO and dichloromethane, the conversion of levulinic acid ranges from 18.0% to 46.3%, whereas the yield of succinate was less than 3% with trace or even no amount of maleate. As for methyl levulinate, when the oxidation was conducted in solvents such as acetic acid, DMF and methanol,³⁸ the conversion dramatically reduced to 26.9%, 29.8% and 17.9%, respectively, from 95.3% that was obtained in acetic anhydride, and the yield of succinate was less than 6.0% (Table 2, entries 1, 2 and 4). Results in other solvents such as ethyl acetate, acetonitrile, cyclohexane, DMSO and dichloromethane (Table 2, entries 3–8) indicated that the majority of methyl levulinate remained unconverted and negligible succinate and even no maleate were detected as the C4 products. Both levulinic acid and methyl levulinate yielded some portion of oxalate or malonate as by-products in these solvents.

Table 2 Solvent effect on the oxidation of levulinic acid and methyl levulinate^a

Entry	Solvent	Substrate	Conv. [mol%]	Yield of important products^b [mol%]
				Succinate	Maleate	Malonate	Oxalate
a Reaction conditions: levulinate (2.5 mmol), Mn(OAc)3·2H2O (5 mol% Mn), solvent (2 mL), 90 °C, 0.5 Pa O2, 10 h.b Data detected by GC after esterification.c Data in the second row of entry 4 was from ref. 38.
1	Acetic acid	Levulinic acid	82.1	Trace	2.5	—	—
		Methyl levulinate	26.9	5.2	—	3.9	8.5
2	DMF	Levulinic acid	75.9	—	7.2	0.5	—
		Methyl levulinate	29.8	0.4	—	—	1.4
3	Ethyl acetate	Levulinic acid	24.1	2.5	1.4	—	1.2
		Methyl levulinate	11.4	0.1	1.9	—	—
4^c	Methanol	Levulinic acid	46.3	2.1	—	—	7.9
		Methyl levulinate	17.9	3.5	—	—	14.0
5	Acetonitrile	Levulinic acid	19.6	2.0	—	—	1.8
		Methyl levulinate	Trace	—	—	—	—
6	Cyclohexane	Levulinic acid	31.4	2.2	—	—	6.4
		Methyl levulinate	5.7	1.7	—	0.6	2.0
7	DMSO	Levulinic acid	18.0	0.2	—	—	—
		Methyl levulinate	7.9	1.0	—	—	—
8	Dichloromethane	Levulinic acid	19.5	2.5	—	—	3.5
		Methyl levulinate	3.8	1.3	—	—	—

---

Compared with other solvents, therefore, acetic anhydride played a key role in the oxidation of levulinic acid for the generation of dehydration products 3, which further oxidized to 2. In addition, acetic anhydride was the appropriate solvent for the Mn(III)-catalytic oxidation of methyl levulinate to succinate, since small amount of water produced during the oxidation process could be absorbed by acetic anhydride, thus keeping the Mn(III) catalyst effective by avoiding it from disproportionation.⁴⁶ This was confirmed by the results of replacing Mn(OAc)₃·2H₂O with Mn(OAc)₂·4H₂O with which the conversion of methyl levulinate dropped to 12.9% from 95.3% and yield of succinate decreased from 58.6% to 7.7%.³⁸ To the best of our knowledge, this is the first report on insights of the comparison of the mechanism of Mn(III)-catalysed aerobic oxidation of levulinic acid and methyl levulinate under mild conditions.

Conclusions

In summary, catalytic oxidative cleavage of the biomass derived platform compound levulinic acid was conducted using manganese(III) as a catalyst with molecular oxygen to obtain succinic anhydride. Moreover, maleic anhydride was obtained. This process differs greatly from the oxidation of methyl levulinate under the same reaction conditions, which only afforded the C4 product succinate. The pathway for levulinic acid oxidation was disclosed by confirming the intermediates α and β-angelica lactone and 2-methyl-5-oxotetrahydro-2-furanyl acetate for the generation of maleic anhydride and low selectivity of succinic anhydride. Acetic anhydride was proved to be a key solvent in the oxidation conversion reaction for dehydration of levulinic acid and maintaining the activity of the Mn(III) catalyst. This study provides an insight into the comparison of the catalytic oxidation of levulinic acid and methyl levulinate, thus indicating the protection of ester group of methyl levulinate in its oxidative conversion to succinate. Further study on the transformation of levulinic acid is currently underway.

Acknowledgements

This study was supported by the National Natural Science Foundation of China (21233008 and 21473184), the Liaoning Natural Science Foundation of China (2015020587) and the “Strategic Priority Research Program-Climate Change: Carbon Budget and Related Issues” of the Chinese Academy of Sciences (XDA05010203).

Notes and references

1. C. Delhomme, D. Weuster-Botz and F. E. Kühn, Green Chem., 2009, 11, 13–26.
2. B. Cok, I. Tsiropoulos, A. L. Roes and M. K. Patel, Biofuels, Bioprod. Biorefin., 2014, 8, 16–29.
3. I. Bechthold, K. Bretz, S. Kabasci, R. Kopitzky and A. Springer, Chem. Eng. Technol., 2008, 31, 647–654.
4. Y. Tachibana, T. Masuda, M. Funabashi and M. Kunioka, Biomacromolecules, 2010, 11, 2760–2765.
5. L. W. Ren, Y. S. Wang, J. Ge, D. N. Lu and Z. Liu, Macromol. Chem. Phys., 2015, 216, 636–640.
6. C. I. Meyer, S. A. Regenhardt, A. J. Marchi and T. F. Garetto, Appl. Catal., A, 2012, 417, 59–65.
7. J. Li, W. P. Tian and L. Shi, Catal. Lett., 2011, 141, 565–571.
8. Y. H. Feng, H. B. Yin, A. L. Wang, T. Xie and T. S. Jiang, Appl. Catal., A, 2012, 425–426, 205–212.
9. D. Wang, Y. Zhang, H. T. Li, L. L. Zhao, H. X. Zhang and Y. X. Zhao, Chin. J. Catal., 2012, 33, 1229–1235.
10. S. F. Guo and L. Shi, Catal. Today, 2013, 212, 137–141.
11. W. T. Huo, C. L. Zhang, H. J. Yuan, M. J. Jia, C. L. Ning, Y. Tang, Y. Zhang, J. H. Luo, Z. L. Wang and W. X. Zhang, J. Ind. Eng. Chem., 2014, 20, 4140–4145.
12. Y. Ma, Y. Q. Huang, Y. W. Cheng, L. J. Wang and X. Li, Catal. Commun., 2014, 57, 40–44.
13. B. Dmuchovs, M. C. Freerks, E. D. Pierron, R. H. Munch and F. B. Zienty, J. Catal., 1965, 4, 291–300.
14. A. Bielanski and M. Najbar, Appl. Catal., A, 1997, 157, 223–261.
15. C. Uraz and S. Atalay, Chem. Eng. Technol., 2007, 30, 1708–1715.
16. L. M. Madeira and M. F. Portela, Catal. Rev., 2002, 44, 247–286.
17. H. G. Lintz and A. Reitzmann, Catal. Rev., 2007, 49, 1–32.
18. M. J. Cheng and W. A. Goddard, J. Am. Chem. Soc., 2013, 135, 4600–4603.
19. A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107, 2411–2502.
20. D. M. Alonso, S. G. Wettstein and J. A. Dumesic, Chem. Soc. Rev., 2012, 41, 8075–8098.
21. P. Gallezot, Chem. Soc. Rev., 2012, 41, 1538–1558.
22. Y. B. Huang and Y. Fu, Green Chem., 2013, 15, 1095–1111.
23. M. Besson, P. Gallezot and C. Pinel, Chem. Rev., 2014, 114, 1827–1870.
24. C. Chatterjee, F. Pong and A. Sen, Green Chem., 2015, 17, 40–71.
25. J. J. Wang, J. X. Xi and Y. Q. Wang, Green Chem., 2015, 17, 737–751.
26. J. P. Ma, W. Q. Yu, M. Wang, X. Q. Jia, F. Lu and J. Xu, Chin. J. Catal., 2013, 34, 492–507.
27. J. H. Lan, Z. Q. Chen, J. C. Lin and G. C. Yin, Green Chem., 2014, 16, 4351–4358.
28. S. Van De Vyver, J. Geboers, P. A. Jacobs and B. F. Sels, ChemCatChem, 2011, 3, 82–94.
29. R. Weingarten, W. C. Conner and G. W. Huber, Energy Environ. Sci., 2012, 5, 7559–7574.
30. Z. Sun, M. X. Cheng, H. C. Li, T. Shi, M. J. Yuan, X. H. Wang and Z. J. Jiang, RSC Adv., 2012, 2, 9058–9065.
31. R. Weingarten, Y. T. Kim, G. A. Tompsett, A. Fernandez, K. S. Han, E. W. Hagaman, W. C. Conner, J. A. Dumesic and G. W. Huber, J. Catal., 2013, 304, 123–134.
32. W. P. Deng, Q. H. Zhang and Y. Wang, Sci. China: Chem., 2015, 58, 29–46.
33. Y. B. Huang, T. Yang, B. Cai, X. Chang and H. Pan, RSC Adv., 2016, 6, 2106–2111.
34. F. Liguori, C. Moreno-Marrodan and P. Barbaro, ACS Catal., 2015, 5, 1882–1894.
35. J. M. Tukacs, B. Fridrich, G. Dibo, E. Szekely and L. T. Mika, Green Chem., 2015, 17, 5189–5195.
36. L. Corbel-Demailly, B. K. Ly, D. P. Minh, B. Tapin, C. Especel, F. Epron, A. Cabiac, E. Guillon, M. Besson and C. Pinel, ChemSusChem, 2013, 6, 2388–2395.
37. M. X. Li, G. Y. Li, N. Li, A. Q. Wang, W. J. Dong, X. D. Wang and Y. Cong, Chem. Commun., 2014, 50, 1414–1416.
38. J. X. Liu, Z. T. Du, T. L. Lu and J. Xu, ChemSusChem, 2013, 6, 2255–2258.
39. S. Dutta, L. Wu and M. Mascal, Green Chem., 2015, 17, 2335–2338.
40. Y. R. Wang, F. Vogelgsang and Y. Román-Leshkov, ChemCatChem, 2015, 7, 916–920.
41. A. P. Dunlop and S. Smith, US 2676186, 1954.
42. L. Podolean, V. Kuncser, N. Gheorghe, D. Macovei, V. I. Parvulescu and S. M. Coman, Green Chem., 2013, 15, 3077–3082.
43. A. Atlamsani, J. M. Bregeault and M. Ziyad, J. Org. Chem., 1993, 58, 5663–5665.
44. M. Mascal, S. Dutta and I. Gandarias, Angew. Chem., Int. Ed., 2014, 53, 1854–1857.
45. J. Xin, S. Zhang, D. Yan, O. Ayodele, X. Lu and J. Wang, Green Chem., 2014, 16, 3589–3595.
46. J. M. Anderson and J. K. Kochi, J. Am. Chem. Soc., 1970, 92, 2450–2460.

---

Footnote

† Electronic supplementary information (ESI) available: Details of characterization of products and the calculation methods for the experiments. See DOI: 10.1039/c6ra16149a

---