Green catalytic conversion of bio-based sugars to 5-chloromethyl furfural in deep eutectic solvent, catalyzed by metal chlorides

Abstract

5-Chloromethylfurfural (5-CMF), a biomass-derived platform chemical with great potential applications, was synthesized by a novel method from sugars, using metal chlorides as catalysts in a deep eutectic solvent (DES). AlCl₃·6H₂O was verified as the most effective catalyst among various metal chlorides, and provided a 5-CMF yield of 50.3% along with 8.1% 5-HMF yield at 120 °C in 5 h. By this green, mild and cost-effective approach, the dependence of 5-CMF production on the large amount and high concentration of hydrochloric acid in previous studies was eliminated.

---

Renewable lignocellulosic biomass has been utilized for the production of fuels and chemicals in consideration of the depletion of fossil resources worldwide.^1–3 Much effort has been devoted to the conversion of lignocellulose into versatile platform chemicals, such as 5-hydroxymethylfurfural (5-HMF), which can be produced by acid-promoted dehydration of cellulose, and has been regarded as one of the most important cellulose-derived products. However, 5-HMF is highly active and instable, especially in the presence of acid catalysts which could further promote the formation of humins and degradation of 5-HMF to levulinic acid (LA).^4,5

5-Chloromethylfurfural (5-CMF), which can be directly generated with a high yield up to 80% from cellulosic biomass, is also acknowledged as a novel and promising platform molecule owing to its higher stability compared with 5-HMF.^4,6–9 Moreover, highlighted by Mascal's research, the chlorine atom in 5-CMF is rather active for substitution reactions of 5-CMF,⁴ which facilitates the conversion of 5-CMF into various chemicals such as 5-aminolevulinic acid (5-ALA),¹⁰ 2,5-dimethylfuran (DMF)¹¹ and pharmaceutical chemicals.¹² According to previous reports, 5-CMF was typically produced in a biphasic reaction system, in which concentrated HCl solution was adopted for the generation of 5-CMF via hydrolysis, dehydration and halogenation process followed by simultaneous extraction with dichloromethane.¹³

Herein we proposed the “one-pot” conversion of carbohydrates to 5-CMF in a biphasic reaction system (Fig. S1†), which was composed of methyl isobutyl ketone (MIBK) and deep eutectic solvent (DES).^14,15 DES is widely acknowledged as a new class of ionic liquid (IL), under mild conditions without the utilization of concentrated acid. As illustrated in Scheme 1, fructose which generated DES with choline chloride (ChCl) is firstly dehydrated to 5-HMF in DES phase with the catalysis of AlCl₃·6H₂O. The resulting 5-HMF is rapidly extracted by MIBK from the DES phase. The following halogenation of 5-HMF to 5-CMF in organic phase is catalyzed by HCl, which might be in situ generated by the hydrolysis of AlCl₃·6H₂O. In this work, a highest 5-CMF yield over 50% was obtained with a high feed concentration (over 15 wt%) in a DES-MIBK system in the absence of concentrated HCl. It is noted that carbohydrates used here were both reactants and part of DES.

Scheme 1 Plausible mechanism for the conversion of fructose into 5-CMF in DES-MIBK system catalyzed by AlCl₃·6H₂O.

As moderate acids, metal chlorides were adopted in this study to eliminate the dependence of 5-CMF production on concentrated HCl. Various of metal chlorides were investigated for the catalytic production of 5-CMF in DES-MIBK system (Table S1†). It is noted that the strength of Lewis acid greatly affected the conversion of fructose and the selectivity to 5-CMF and 5-HMF. Metal chlorides with stronger Lewis acidity provided much higher 5-CMF yields than 5-HMF. A highest 5-CMF yield of 37% was obtained in the presence of AlCl₃·6H₂O at 110 °C in 2 h. Thus, AlCl₃·6H₂O was proven to be the most effective among various metal chlorides and was applied for further experiments.

When the reaction was performed in an isolated DES phase, a severe coking process was observed and the yields of 5-CMF and 5-HMF were quite low. Thus, several aprotic solvents including acetonitrile (MeCN), methyl isobutyl ketone (MIBK), γ-valerolactone (GVL), dimethyl formamide (DMF) and dimethyl sulfoxide (DMSO) were introduced for the in situ extraction of products from DES reaction mixture to prevent polymerization and coking. As shown in Table 1, the addition of MeCN, DMF and DMSO effectively inhibited the coking process and 5-HMF yields ranged 44.9 to 46.8% were achieved. However, almost no 5-CMF was detected in the above mentioned cases. As for GVL, a 5-CMF yield of 18.9% was obtained. Interestingly, MIBK strongly facilitated the generation of 5-CMF and provided a yield up to 37.1% at 110 °C in 2 h.

Table 1 Effects of extractants on the yields of 5-CMF and 5-HMF^a

Extractant	Y5-CMF^a (%)	Y5-HMF^b (%)
a Reaction conditions: 5 mmol fructose, 5 mmol AlCl3·6H2O, 25 mmol ChCl, 30 ml extractant, 110 °C, 2 h. ^aThe yield of 5-CMF; ^bthe yield of 5-HMF; ^cextracted by 30 ml MIBK after the reaction mixture was cooled down to room temperature.
Blank^c	0	0
MeCN	3.7	44.9
MIBK	37.1	16.7
GVL	18.9	32.0
DMF	0	44.8
DMSO	0	46.8

---

The effects of temperature and reaction time on 5-CMF and 5-HMF production in the biphasic system were also studied. According to Fig. 1(a), higher temperature promoted the generation of 5-CMF and a highest yield up to 50.3% was achieved at 120 °C in 5 h. The yield was then decreased to 44.8% when the reaction time was prolonged to 8 h, which was probably attributed to further conversion of 5-CMF into humins and other by-products. Oppositely, the yield of 5-HMF decreased as the reaction proceeded (Fig. 1(b)), from which we can deduce that fructose was dehydrated into 5-HMF and then halogenated into 5-CMF in the biphasic system. The cascade generation of 5-HMF and 5-CMF was proved by GC-MS detected product distribution, as shown in Fig. S2.†

Fig. 1 Effects of reaction time and temperature on the conversion of fructose into 5-CMF (a) and 5-HMF (b); reaction conditions: 5 mmol fructose, 5 mmol AlCl₃·6H₂O, 25 mmol ChCl, 30 ml MIBK.

The effect of the catalyst dosage was studied, as shown in Fig. 2. In a control experiment, 20.9% of 5-HMF yield was obtained and no 5-CMF was detected. 5-CMF yield of 23.6% and 5-HMF yield of 13.4% were obtained when the mole ratio of AlCl₃·6H₂O to fructose was 0.5. As the mole ratio of AlCl₃·6H₂O to fructose increased to 1.0, a sharp increase of 5-CMF yield to 50.3% was observed while 5-HMF yield decreased from 20.9% to 8.1%. However, 5-CMF yield was reduced to 40% as the mole ratio of AlCl₃·6H₂O to fructose was further increased to 4.0, probably due to the overmuch acid sites which would promoted the formation of undesired by-products such as humins.

Fig. 2 Effect of catalyst loading on the conversion of fructose into 5-CMF & 5-HMF; reaction conditions: 5 mmol fructose, 25 mmol ChCl, 30 ml MIBK, 120 °C, 5 h. Catalyst loading is relative to fructose.

As Table 2 showed, low concentration of ChCl helps to reduce the costs of 5-CMF production which is desirable for the practical production. Hence, the effect of the initial ChCl concentration on 5-CMF and 5-HMF yields was studied at a reaction temperature of 120 °C in 5 h with 5 mmol fructose and 5 mmol catalyst. As demonstrated in Table 2, a fructose to ChCl weight ratio of 1 : 1 resulted in the predomination of 5-HMF in the products. However, 5-CMF yield gradually increased from 17.4% to 50.3% when the weight ratio of ChCl to fructose increased from 1 to 4, and the 5-HMF yield sharply descended from 30.8% to 7.4%. This was ascribed to that suitable amount of ChCl promoted the conversion process from fructose to 5-CMF. It was also noted that high content of ChCl have a promoting function on the total yield of 5-HMF and 5-CMF.

Table 2 Effect of fructose concentration on the conversion of fructose into 5-CMF & 5-HMF^a

Fructose : ChCl weight ratio	Y5-CMF (%)	Y5-HMF (%)	Total (%)	Cfructose (%)
a Reaction conditions: 5 mmol (0.9 g) fructose, 5 mmol AlCl3·6H2O, 30 ml MIBK, 120 °C, 5 h.
1 : 1	17.4	30.8	48.2	87.3
1 : 2	25.8	21.9	47.7	95.5
1 : 3	32.6	18.6	51.2	100
1 : 4	50.3	8.1	58.4	100
1 : 5	50.2	7.4	57.9	100

---

Other kinds of ionic liquids, such as [BMIM]Cl and [BMIM]BF₄, were applied to compare with ChCl-based DES for 5-CMF production, as shown in Table S2.† 5-CMF was obtained with a highest yield of 50% in the presence of ChCl while the yield was only 6% and 0% in the cases of [BMIM]Cl and [BMIM]BF₄ respectively. This indicated that the composition of DES reaction system was crucial for the production of 5-CMF from fructose.

As our previous works shown, fructose was firstly converted into 5-HMF rapidly, and then transformed into 5-CMF. In order to investigate the forming process from 5-HMF to 5-CMF, a series of experiments were conducted with 5-HMF as substrate. As demonstrated in Table S3,† low yield of 5-CMF and 5-HMF were obtained with the addition of AlCl₃·6H₂O or ChCl respectively (entry 1, 4). However, 5-HMF was totally converted and 86% 5-CMF yield was obtained in the presence of both AlCl₃·6H₂O and ChCl. Moreover, HCl was detected from the MIBK phase, which was formed in the DES with the existence of both AlCl₃·6H₂O and ChCl. It was supposed that 5-CMF could be produced from 5-HMF in the presence of HCl. Unfortunately, the source of chloride anions in 5-CMF has not been exactly determined yet.

The continuous solvent extraction method highlighted by Mascal has greatly facilitated the conversion of carbohydrates into 5-CMF.⁶ When MIBK was used as the extractant and formed a biphasic system with DES, as shown in Fig. S3 and S4,† 5-CMF yield decreased significantly with an enhanced 5-HMF yield, whereas the total yield of furfuran products was maintained at about 60%, corresponding with our aforementioned work. This could be explained by that the transfer of extraction solvent would lead to the loss of HCl in the organic phase, which might affect the further halogenation reaction of 5-HMF in MIBK. Moreover, those experiment results indicated that the organic extraction solvent also participated in the process of 5-CMF production.

To study the recycle performance of the catalyst, the catalyst and ChCl were reused with water-washing and distillation after being separated from the reaction mixture (the HPLC analysis result was showed in Fig. S4†). A slight descending of 5-CMF yield was observed in the successive 2 cycles. A 5-CMF yield of 41.3% and a 5-HMF yield of 11.7% were still achieved after 5 h in the 3rd run (Fig. 3).

Fig. 3 Performance of recycled catalyst mixture. Reaction conditions: 5 mmol (0.9 g) fructose, 4.8 g catalyst mixture (ChCl and AlCl₃·6H₂O), 30 ml MIBK, 120 °C, 5 h. The used catalyst mixture was washed by deionized water and dried in a rotary evaporator before reutilization.

In addition to fructose, other biomass derived carbohydrates were also tested to produce 5-CMF in this MIBK/DES biphasic system. As shown in Fig. 4, 5-CMF yields of 16.93%, 22.66% and 17.86% were obtained with glucose, inulin and sucrose as substrates, respectively. Compared with the high yield of 5-CMF achieved from fructose, these carbohydrates showed low reaction activity in the production of 5-CMF. It was probably due to that those carbohydrates could not be converted into 5-HMF easily in the DES-biphasic system, in which the isomerization into fructose was not favored and the formation of 5-CMF was further impeded. Thus, our further study would focus on the conversion of various carbohydrates into 5-CMF with high yields in a green DES-biphasic system.

Fig. 4 Conversion of various carbohydrates into 5-CMF & 5-HMF in the presence of AlCl₃·6H₂O and ChCl. Reaction conditions: 0.9 g different carbohydrate, 5 mmol AlCl₃·6H₂O, 25 mmol ChCl, 30 ml MIBK, 120 °C, 5 h.

Conclusions

In this work, 5-CMF was effectively produced in a DES/MIBK biphasic reaction system. Although the 5-CMF yield obtained in this work was lower than the yield reported by other literatures, the major advantages of the present work was that a combination system of metal chlorides and ChCl was developed for the conversion of fructose into 5-CMF, which eliminated the reliance on concentrated hydrochloric acid. This method provided a moderate and green condition for the production of 5-CMF with a high yield. In conclusion, this study presents a new green method to efficiently produce important platform chemical 5-CMF from biomass derived carbohydrates in a one pot/two-step mild biphasic reaction system.

Acknowledgements

This work was supported financially by the Natural Science Foundation of China (No. 21506177), the Natural Science Foundation of Fujian Province of China (2015J05034), the Fujian province development and reform commission and the Key Program for Cooperation between Universities and Enterprises in Fujian Province (2013N5011) and the Fundamental Research Funds for the Xiamen University (20720160077).

Notes and references

1. A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek, J. Cairney, C. A. Eckert, W. J. Frederick, J. P. Hallett, D. J. Leak and C. L. Liotta, Science, 2006, 311, 484–489.
2. Y. Román-Leshkov, J. N. Chheda and J. A. Dumesic, Science, 2006, 312, 1933–1937.
3. J. B. Binder and R. T. Raines, J. Am. Chem. Soc., 2009, 131, 1979–1985.
4. M. Mascal, ChemSusChem, 2015, 8, 3391–3395.
5. F. Liu, J. Barrault, K. De Oliveira Vigier and F. Jérôme, ChemSusChem, 2012, 5, 1223–1226.
6. M. Mascal and E. B. Nikitin, Angew. Chem., 2008, 120, 8042–8044.
7. M. Mascal and E. B. Nikitin, ChemSusChem, 2009, 2, 423–426.
8. M. Brasholz, K. von Kanel, C. H. Hornung, S. Saubern and J. Tsanaktsidis, Green Chem., 2011, 13, 1114–1117.
9. J. Howard, D. W. Rackemann, Z. Zhang, L. Moghaddam, J. P. Bartley and W. O. Doherty, RSC Adv., 2016, 6, 5240–5248.
10. M. Mascal and S. Dutta, Green Chem., 2011, 13, 40–41.
11. M. Mascal and S. Dutta, in Selective Catalysis for Renewable Feedstocks and Chemicals, Springer, 2014, pp. 41–83.
12. S. Dutta, L. Wu and M. Mascal, Green Chem., 2015, 17, 3737–3739.
13. S. Breeden, J. Clark, T. Farmer, D. Macquarrie, J. Meimoun, Y. Nonne and J. Reid, Green Chem., 2013, 15, 72–75.
14. A. P. Abbott, D. Boothby, G. Capper, D. L. Davies and R. K. Rasheed, J. Am. Chem. Soc., 2004, 126, 9142–9147.
15. E. L. Smith, A. P. Abbott and K. S. Ryder, Chem. Rev., 2014, 114, 11060–11082.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra00267f

---