Optimization of the synthesis of levulinic acid and levulinic acid derivatives from sugarcane bagasse using ionic liquids

Abstract

Globally, the effects of climate change due to natural sources and human activities that releases greenhouse gases has led to a greater need for a sustainable and renewable resource called biomass. In South Africa (KwaZulu-Natal) there is an excess of sugarcane bagasse (SB) therefore this study was undertaken using SB. SB was valorised to replace chemicals obtained from fossil fuel processing. Levulinic acid (LA) was identified by the National Renewable Energy Laboratory (NREL) as a chemical that can be produced from biomass. LA is a platform chemical therefore many compounds are produced from LA hence in this work three LA derivatives namely diphenolic acid (DPA), γ-valerolactone (GVL) and ethyl levulinate (EL) production were studied. The main purpose of this work was to optimize the production of LA from depithed sugarcane bagasse (DSB) using ionic liquids (ILs), which are environment benign compared to sulfuric acid which is currently used in commercial production of LA from biomass. Firstly, the optimal reaction conditions to produce LA from DSB using 1-ethyl-3- methylimidazolium hydrogen sulfate [EMim][HSO4] ionic liquid (IL) were investigated. The effect of temperature (100-220 oC), reaction time (2-12 h), and ionic liquid loading (1- 4 g) was assessed using response surface methodology (RSM) based on the Box-Behnken design (BBD). The optimum conditions were found to be 100 oC, 7 h, and 4 g of IL, which yielded 54.6 % of LA from DSB. The analysis of variance (ANOVA) indicated that the design model was significant at the 95 % confidence level. The pareto chart revealed that IL loading had the most significant effect on the production of LA, followed by temperature and reaction time. The P-values also showed that these were the significant model terms. There is a strong correlation between temperature and IL loading. Solvent optimization revealed that the type of solvent used in the LA production has a significant effect on LA yield. Water was used as the control solvent for this study. Methyl isobutyl ketone (MIBK) yielded the highest LA (62 %) from all the solvents that were used. Secondly, an environmentally friendly catalyst: 1-butyl-2,3-dimethylimidazolium tetrafluoroborate ([BMMim][BF4]) was used to optimize the LA production from DSB. The Box-Behnken design (response surface methodology) was used to design the set of experiments with three variables, namely, time, temperature, and catalyst loading. The investigated conditions were temperature (100 - 220 oC), time (2 – 12 h), and catalyst loading (1 – 4 g). The optimum condition was found to be at 100 oC, 7 h and 4 g of a catalyst which yielded a maximum amount of 44.8 % of LA from DSB. This study also showed that the IL is capable of theoretically producing 62.1 % of LA. The reusability study showed that [BMMim][BF4] can be used for up to four times without losing it activity. Thirdly, an environmentally benign method to produce a LA derivative called ethyl levulinate (EL) was investigated. EL can be produced from LA using heterogeneous catalysts replacing the highly active, corrosive, and toxic homogeneous catalyst sulfuric acid. This work uses an environmentally friendly homogeneous catalyst: methanesulfonic acid (MsOH) as the control catalyst and ILs. The esterification of commercial LA into EL was first optimized using MsOH and response surface methodology. The optimum condition for the esterification of commercial LA into EL was 5.25 h, 90 oC, and 2.75 g of MsOH loading. The EL yield and selectivity obtained were 92.2 % and 94 %, respectively, at a LA conversion of 98 %. The optimized conditions were then utilized to produce EL from DSB derived LA with a EL yield of 75% using MsOH. Fourthly, gamma-valerolactone (GVL) production was optimized by a Box Behnken design from commercial LA with an environmentally friendly catalyst methanesulfonic acid. The optimum parameters were a temperature of 112.5 oC, reaction time of 6 h, and catalyst loading of 2.75 g yielding 78.6 % GVL with 97 % LA conversion and an 81 % selectivity. Thereafter the optimised conditions were used to produce GVL from LA derived from DSB. The hydrogen required for the reduction of LA to GVL was formed in-situ by formic acid and triethylamine in the presence of MsOH. Different solvents (including water and alcohols) were also tested to determine their effect on GVL yield, water yielded the highest GVL of 78.6 %. Different types of catalysts which included mineral acids and ionic liquids were used to determine their effect on GVL yield, and to provide a benchmark against MsOH. Sulfuric acid gave the highest GVL yield (80.9 %). Fifthly, the production of diphenolic acid (DPA) which is one of the LA derivatives was studied. DPA has a potential to replace bisphenol A, a plasticizer. To determine the optimum conditions for DPA production, commercial LA was used with a mild environmentally benign acid namely, methanesulfonic acid. The optimized reaction parameters were time (6 h), temperature (75 oC), and catalyst loading (5.5 g) yielding 65.8 % DPA at 90 % LA conversion. The response surface methodology study indicated that the temperature had the most significant effect on DPA yield followed by time and catalyst loading. The analysis of variance (ANOVA) revealed that the model was able to satisfactorily predict the DPA yield. To determine the effect of catalyst on DPA production from commercial LA, ionic liquids (ILs), MsOH, and sulfuric acid were used. IL catalysts produced 59-68 % of DPA, MsOH produced 65.6 % of DPA, and sulfuric acid produced the maximum DPA of 74 %. The study also investigated the effect of the LA: phenol ratio using the optimised reaction conditions. The LA: phenol ratio of 2:5 yielded the most DPA (86.35%). The optimized reaction conditions were then used to produce DPA from LA derived from depithed sugarcane bagasse (DSB), which yielded 64.5 % of DPA. This indicates that the DSB derived LA is a good starting material for DPA production. The LA production from depithed sugarcane bagasse was successfully optimized by using two ionic liquids namely 1-ethyl-3-methylimidazolium hydrogen sulfate and 1-butyl-2,3- dimethylimidazolium tetrafluoroborate, where for both ILs the same conditions yielded a maximum LA yield. MIBK was found to be the optimum solvent for both ILs, giving a higher LA yield compared to when water is used as a solvent. This study also revealed that at the optimum conditions there was no formation of humins, but humins were observed at maximum reaction conditions. Both ILs showed that they can be reused up to four times, which is very important for any catalysts especially for the ILs because they are known to be expensive. This study also illustrates the first-time optimization of three LA derivatives namely EL, GVL, and DPA using methanesulfonic acid.