The functional groups of these samples were identified by FT-IR analysis, and the corresponding FT-IR spectra were depicted in Fig. 2. In the case of the MMT, the peak around 3600–3235 cm–1 is associated with O—H stretching vibration and the peak at 1480 cm–1 is related to the symmetric stretching vibration of Si—OH (Shirini et al., 2011). For LMT-C, a new vibration bands C＝O (1610 cm–1) emerged and the intensity of —OH peak around 3600–3235 cm–1 was amplified, due to the covalent attachment of lignin to the MMT. After sulfonation, the existence of sulfonic —SO3H group in the samples was confirmed by the presence of stretching modes at 1150 cm–1, which were considered as the evidence for S＝O asymmetric and symmetric stretching modes (Kolvari et al., 2015).
Figure 2. The FT-IR spectra of montmorillonite (MMT), carbonized lignin-montmorillonite complex (LMT-C) and sulfonic acid-functionalized lignin-montmorillonite complex (LMT-SO3H)
The TG/DTG was performed for the characterization of LMT-SO3H in comparison with LMT-C, and the results are shown in Fig. 3. The initial weight loss of 6.2% at about 110 ℃ corresponding to the loss of the physically adsorbed water was observed for LMT-C (Fig. 3a). However, a great additional weight loss appeared at 200 ℃–600 ℃ with an endothermic DTG signal at 250 ℃, which can be attributed to the dehydroxylation of montmorillonite (Shirini et al., 2011; Varadwaj et al., 2016). Compared with the observation of the LMT-C, the DTG curve pattern of LMT-SO3H (Fig. 3b) exhibits a strong endothermic DTG signal occurred at 450 ℃, which could be attributed to the decomposition of SO3H group, and agrees with the reported literature (Zhou et al., 2019). The TG/DTG results show that the LMT-SO3H catalyst has great thermal stability and applicability in organic reaction with temperatures below 220 ℃.
Firstly, the catalytic performance of different acid catalysts for the synthesis of the BL was investigated (Table 1). The results reveal that oxalic acid as the active acid type shows the lowest catalytic activity for the conversion of LA, while the LMT-SO3H affords the highest catalytic activity as compared with other acid catalysts. Interestingly, by using A- and R- TiO2NTs-SO3H catalysts, the LA conversion values of 73.4% and 74.0% were obtained, respectively, whereas the yield of BL production was below 20%, probably due to the production of various by-products.
Entry Catalyst Conversion (%) BL selectivity (%) BL yield (%) 1 LMT-SO3H 99.8 98.7 98.5 2 Pine-SO3H 98.1 97.5 95.6 3 A-TiO2NTs-SO3H 73.4 17.6 12.9 4 R-TiO2NTs-SO3H 74.0 13.1 9.7 5 HOOC-COOH 48.5 85.5 41.5 6 H3PO4 70.5 96.7 68.2 7 ZnCl2 51.3 78.5 40.3 8 CuSO4 89.3 83.0 74.1 Notes: reaction conditions are 15 mmol LA, 21 mmol BuOH, 10 wt% catalyst amount, 10 mL cyclohexane, and reflux reaction at 120 ℃ for four hours. BL, n-butyl levulinate.
Table 1. Catalytic performance of different catalysts in esterification of levulinic acid (LA)
Figure 4 shows the effect of reaction time on the esterification of LA in the presence of LMT-SO3H. The conversion of LA rose from 40.7% to 99.8% with increasing reaction time from 0.5 h to 4.0 h, the selectivity of BL was increased from 63.9% to 98.7%. In addition, an intermediate p-BL during the reaction process was identified by GC-MS techniques. When the reaction time increased from 0.5 h to 4.0 h, the selectivity of p-BL significantly decreased from 36.1% to 1.3%. It indicated that the intermediates p-BL that produced during the reaction gradually transformed into BL with increasing reaction time and the path of LA to BL was: LA→p-BL→BL under the catalysis of LMT-SO3H. When increasing reaction time from 2 h to 4 h, there was an insignificant change in the selectivity of BL and p-BL, therefore the conversion process of p-BL to BL was the reaction rate-determining step. Based on the results above, we proposed a possible reaction pathway for LA to be converted to BL over the LMT-SO3H.
Since the esterification reaction is reversible, the choice of water-carrying agent is more important. In this paper, the yield of esterification of LA was studied in the presence of different water-carrying agents. As shown in Table 2, when using benzene as the water-carrying agent, the yield of the product is significantly lower than that of toluene and cyclohexane. Furthermore, considering that benzene is more toxic and the reflux temperature of cyclohexane is lower, cyclohexane is a suitable water-carrying agent for the reaction.
Entry Water-carrying agent T (℃) t (h) Yield (%) 1 Benzene 80 4 74.0 2 Toluene 110 4 96.3 3 Cyclohexane 80 4 98.6 Notes: reaction conditions are 15 mmol LA, 21 mmol BuOH, 10 wt% catalyst amount and 10 mL water-carrying agent.
Table 2. Effect of water-carrying agent on yield of BL at reflux temperature
The effect of cyclohexane amount was investigated ranging from 2 mL to 10 mL, and the results are shown in Fig. 5. Under the same amount of water-carrying agent reaction conditions, the yield of BL gradually increases with the reaction time. Further increasing reaction time from 3 h to 4 h, the yield of BL has unchanged, which should be attributable that the conversion of LA achieves chemical balance. The BL yield rises with the increase of water-carrying agent in the same reaction time. When the amount of cyclohexane is 8 mL, the yield of BL is up to 99.3%. However, further increasing the amount of cyclohexane, the BL yield remains basically unchanged. Taking the cost and the efficiency into consideration, the optimal condition is 8 mL cyclohexane.
Then, the catalyst amount was optimized and the results are presented in Fig. 6. As can be observed, the yield of BL increases from 61.1% to 99.3% as the amount of catalyst increases from 2 wt% to 10 wt%. At higher catalyst concentration, due to the availability of more active sites, more reactants were transformed to give higher conversion. However, improving the catalyst amount to 12 wt%, the yield of BL decreased to 98.0%. The inverse effect of the further increase in catalyst amount on the yield of BL is probably due to the increase of the available active sites of the catalyst which facilitates the reverse reaction (Chermahini and Nazeri, 2017). Therefore, the optimum catalyst amount of 10 wt% was selected for further experiments.
The results obtained from the study of the effect of molar ratio (LA: n-butyl alcohol) on the esterification of LA are depicted in Fig. 7. The yield of BL was found to be increased from 90.2% to 99.3% with an increase in the molar ratio from 1꞉1.0 to 1꞉1.4, respectively. However, the noteworthy decrease in the yield of BL is observed when the molar ratio beyond 1꞉1.4. The decrease in the yield for the molar ratio 1꞉1.6 to 1꞉1.8, may be attributed to the fact that the excessive amount of alcohol is accumulated on the surface of the catalyst and blocks the catalytic active sites (Das and Parida, 2007). Hence, the optimum LA to n-butanol molar ratio of 1꞉1.4 was selected for further experiments.
To develop the application range of the catalyst, the esterification reaction of LA was investigated when different alcohols were used as substrates. As shown in Table 3, most alcohols trans formed to afford the corresponding esters in excellent yields. The yield of BL decreased with the increase of steric hindr ance, and the following order of substrate reactivity was obse rved: primary alcohol > secondary alcohol > tertiary alcohol (Table 3, entries 1, 2, 4, 5 and 8), which is in agreement with the general rules of esterification reaction. Interestingly, when using primary alcohol as a substrate, the yield of BL is generally high (Table 3, entries 1, 3, 5, 6 and 7). In connection with the possible reaction mechanism proposed below, this phenomenon may be due to that primary alcohols are more likely to attack carbocations than other alcohols.
Entry Alcohol Yield (%) 1 n-propanol 98.2 2 isopropanol 58.5 3 n-butanol 99.3 4 isobutanol 61.5 5 n-pentanol 98.6 6 n-hexanol 98.2 7 n-heptyl alcohol 98.5 8 tert-butanol 9.5 Notes: reaction conditions are 15 mmol LA, 21 mmol alcohol, 10 wt% catalyst amount, 8 mL cyclohexane, and reflux reaction at 120 ℃ for 4 h.
Table 3. Effect of different alcohols on LA esterification
The recycle experiment was carried out to investigate the reu sability of the catalyst under the optimum reaction conditions. After the reaction, the reaction solution was removed by centrifugation to recover the catalyst, washed by cyclohexane three times and dried at 60 ℃ before the usage for the next time. As shown in Fig. 8, the recycled catalyst could be reused for at least five times without observing a decrease in yield, which shows that the catalyst maintains high activity and good stability. After the sixth cycle, the conversion of LA decreased from 93.7% to 84.7%, which may be due to the partial —SO3H groups on the catalyst surface falling off.
According to above-obtained results, the mechanism proposed for the conversion of LA to BL is shown in Fig. 9. It is similar that recently reported by Ciptonugroho et al. (2016) and Enumula et al. (2017).
Firstly, the enol form of LA can undergo intramolecular lactonization to afford α-angelica lactone, which can undergo nucleophilic addition of an alcohol to give as a product. Sub sequently, the protonation of p-BL facilitates another nucleo philic addition and the intermediate I was formed. Then, the intermediate I underwent ring opening under the action of the LMT-SO3H catalyst to produce hemiacetal material (inter mediate II). After that, BL was generated by the elimination reaction of alcohols.
To highlight the superiority of synthesized LMT-SO3H, some previous catalytic results are presented in Table 4. Comparing with these reported catalysts employed for the synthesis of BL, the present LMT-SO3H catalyst exhibits high activity without using an excessive amount of alcohols and the reaction time is the shortest. More importantly, LMT-SO3H has a good sta bility which can be repeatedly used for more than nine times. Besides, the BL yield of 99.3% indicates that LMT-SO3H can be superior to other materials. In summary, the present catalyst shows the obvious advantages of low cost, easy separation and satisfactory catalytic activity.
Entry Catalyst T (℃) t (h) Mole ratio (LA꞉BuOH) Yield (%) Catalyst reuse References 1 H-BEA 120 4 1꞉7 82.2 5 times (Maheria et al., 2013) 2 H3PW12O40/ZrO2-Si(PhSi)-1.0 120 3 1꞉7 82.8 3 times (Su et al., 2013) 3 H-ZSM-5 120 5 1꞉6 98.0 6 times (Nandiwale and Bokade, 2015) 4 HPA/K10 120 4 1꞉6 97.0 3 times (Dharne and Bokade, 2011) 5 WO3-SBA-16 250 10 1꞉7 94.0 10 times (Enumula et al., 2017) 6 CALB@nanoflowers 50 24 N/A 85.5 10 times (Gao et al., 2017) 7 Al-MCM41 120 8 1꞉5 90.0 Several runs (Najafi Chermahini et al., 2017) 8 ZSM-5 120 5 1꞉10 86.0 4 times (Morawala et al., 2019) 9 Sulfonated carbon 100 4 1꞉5 90.5 5 times (Yang et al., 2018) 10 NER@3DOM/m-OS 40 12 1꞉10 74.6 9 times (Zhou et al., 2018) 11 Ten sulfonic resins 115–155 4–6 1꞉3 72.3–82.4 N/A (Iborra et al., 2019) 12 WO x/ZrO2 120 2 1꞉2 66.0 N/A (Ciptonugroho et al., 2016) 13 Zr-MOF 120 5 1꞉6 99.0 3 times (Cirujano et al., 2015) 14 TNTs-SO3H 120 8 1꞉11.2 82.7 5 times (Zhou et al., 2019) 15 HPA/C-Sil-1 100 5 1꞉5 92.0 4 times (Manikandan and Cheralathan, 2017) 16 LMT-SO3H 120 4 1꞉1.4 99.3 9 times This work
Table 4. Summary of catalytic results obtained for synthesis of BL from LA over different catalysts