Direct Conversion of Glucose to 5-Hydroxymethylfurfural over H3PW12O40/TEOS Heterogeneous Catalyst

5-Hydroxymethylfurfural (5-HMF) is a valuable bio-based intermediate designed from carbohydrate resources such as glucose (hexose) or fructose. In this work, direct conversion of glucose into 5-HMF was studied by analysing the activity of solid acid catalyst namely silica-supported phosphotungstic acid (H3PW12O40/TEOS) as a heterogeneous catalyst. The reactions were conducted in a three-neck conical flask using dimethyl sulfoxide (DMSO) as reaction solvent under different reaction time (1, 3 and 5 hours) and temperature (100, 115 and 130°C). The effect of phosphotungstic acid loading was also studied in this literature (5, 12.5 and 20% H3PW12O40). Thus, this paper aims to study the optimum reaction time, temperature and H3PW12O40 loading to give the maximum yield of 5-HMF via direct catalytic dehydration process. The prepared catalyst 20% H3PW12O40/TEOS shows promising results by displaying a yield of 5-HMF as high as 62% after 3 hours at 130°C reaction temperature in the presence of DMSO solvent. Since heteropoly acid is highly soluble in DMSO, thus H3PW12O40 supported in TEOS (H3PW12O40/TEOS) is a promising solid catalyst for the conversion of glucose into 5-HMF. The prepared catalyst can also be recovered and recycled easily without significant loss of performance.


INTRODUCTION
The process of acid-catalyzed dehydration of dissimilar carbohydrates such as fructose, glucose, sucrose, cellulose or inulin are usually done in order to produce 5-Hydroxymethylfurfural (5-HMF). Although 5-HMF is popularly known as an intermediate element, it is yet produced in large industrial scale due to their high production cost. Luckily, glucose carbohydrate which is the most abundant and the cheapest monosaccharide are easily available although they give a low yield of 5-HMF, which is credited to their stable pyranoside ring glucose structure (Gomes et al. 2015).
Recently, biomass is treated as a way to substitute fossil fuel in the production of chemical substances. However people are currently focusing on the production of furan derivatives from biomass, a valuable platform chemical for the production of other series of chemicals. Thus, less development is directed towards the production of 5-HMF from biomass. If biomass can be utilized into 5-HMF, it can be the key intermediate in the development of biomass-based products such as organic acids, polymer precursors and also biofuels.
A catalyst is generally defined as a substance that can accelerate the rate of a chemical reaction to achieve balance without having to undergo any change after the reaction. (de Souza et al. 2012, Gomes et al. 2017. Since the synthesis of 5-HMF is very complex and less stable depending on the reaction conditions, 5-HMF may be converted to levulinic and formic acid during the dehydration process resulting in loss of yield. Thus, solid acid catalyst is used to lead a huge influence to the yield of 5-HMF during the dehydration of glucose carbohydrate. FIGURE 1. Mechanism of dehydration of glucose to 5-HMF and its parallel reactions (Gomes et al. 2015) In this study, the dehydration reaction was carried out in the presence of solid acid catalyst phosphotungstic acid (H 3 PW 12 O 40 ¬ ) supported on silica, which is derived from tetraethoxysilane (TEOS) and dimethylsulfoxide (DMSO) as a reaction solvent to further enhance the reaction. Heteropoly acid catalyst synthesis was first done before glucose dehydration reaction by producing heteropoly acid catalyst in different catalyst-to-support percentage (H 3 PW 12 O 40 /TEOS). DMSO was used as the reaction solvent since biphasic and organic media have higher advantages for 5-HMF production over aqueous medium. Among a lot of solvents used in the literature, compound with four carbon interaction have been identified to produce better yield of 5-HMF (Huang et al. 2014, Jimenez-Morales et al. 2014, Safri et al. 2017. In this study, DMSO showed similar results to solvents with four carbon by reducing the parallel reaction and oxidation possibility of 5-HMF in the same medium. However, the acid strength of the catalyst can also give a huge influence to the selectivity of 5-HMF. The use of heterogeneous catalysts require a longer residence time for glucose conversion, where it can increase the polymerization reaction of both substrates and products.

SYNTHESIS OF THE CATALYST
According to Izumi et al. (2009), the catalyst can be produced through various methods. In this study, 0.72 g of water, 2.96 g of 1-butanol and 0.48 g of phosphotungstic acid was added into 8.3 g of tetraethyl orthosilicate (TEOS), before they were agitated at 80°C for 3 hours. The solution was stirred for another 1.5 hours before the dehydrated hydrogel form can be seen. The resulting sample was dried at 100°C for 3 hours in an oven and labeled as 20% H 3 PW 12 O 40 /TEOS. The experiment was then repeated with different reaction time and catalyst to support loading to produce 5% H 3 PW 12 O 40 /TEOS and 12.5% H 3 PW 12 O 40 /TEOS as in Table 2.  Figure 1. As seen in 20% H 3 PW 12 O 40 /TEOS catalyst spectrum, the Si-O-Si bond spotted at 1100 cm-1 has bigger and wider peak than others catalyst. It obviously showed the bend of Si-O-Si with higher composition in H 3 PW 12 O 40 affected Si and O from TEOS was chemically bonded with each other (Isahak et al. 2011).

CATALYTIC TESTS OF GLUCOSE DEHYDRATION
Performance of dehydration reaction of glucose into 5-HMF by H 3 PW 12 O 40 /TEOS catalysts was evaluated in dimethyl sulfoxide (DMSO) as an aprotic solvent. According to the literature by Shimizu et al. (2009) andvan Putten et al. (2013), DMSO is a good solvent that would defeat unwanted side reactions in hexose dehydration. However, for the formation of 5-HMF, the main byproducts reported in hexose dehydration are humins, soluble polymers, furfural, levulinic acid and formic acids, where these compounds are products from a few of reaction which is cross-polymerization, condensation and rehydration reactions.
The use of 20% H 3 PW 12 O 40 /TEOS catalyst has resulted in maximum yield of 62% 5-HMF with reaction time and temperature of 3 hours and 130 o C. This result is comparable to Fan et al. (2011) who used heteropolyacid catalyst Ag 3 PW 12 O 40 which results in 5-HMF yield of 70-78% with water/MIBK biphasic solvent system at 120 o C in 60-120 min. The yield is also identified to be effected by initial feedstock concentration and temperature. This is proven though Figure  5(a) and 5(b), where dehydration yield is directly proportional to catalyst (H 3 PW 12 O 40 ) loading and dehydration temperature. The sample with 20 % catalyst loading perform better than other sample. For reacting temperature, it was observed that the yield was the highest at 130°C reacting temperature, regardless of catalyst loading. From the relationship of temperature and 5-HMF yield, we can also see the significant effect of time where 3 h is the optimum dehydration time for all sample.
Based on overall experiment data, it is difficult to confirms the effect of 5-HMF yield changes unless it is at According to the XRD analysis, the presence of silica compound gave the amorphous state of the catalyst that is produced by sol-gel technique. From XRD pattern in Figure  3, a broader peak displayed at diffraction peak around 28° has represented the Si-O-Si bond as the major component in H 3 PW 12 O 40 /TEOS catalyst while the element of tungsten can be obtained at diffraction angle of 41° and 54° (Firoozi et al. 2014). Since the intensity of silica is higher than tungsten in the spectrum, in can be concluded that the sample is mainly in amorphous state. Figure 3 below clearly demonstrated the almost similar shape of FTIR spectrum for all samples of 5-HMF regardless of the catalyst to support loading. The major analysis excluded on the bonding of conjugate ketone (C = O) bond which has the physical characteristic of the molecular structure of 5-HMF.

FTIR ANALYSIS OF 5-HMF
Among the 5-HMF spectrum, sample 1, 11 and 13 have been selected as the best sample for this analysis due to their highest peak of C = O bond detected at 1662.5 cm-1, whish is bigger and wider than other samples. It was indicated that the higher composition of C = O bond was due to reaction of C and O molecules in glucose (Nybacka 2016). the optimum reaction time and temperature. Furthermore, the increase in temperature gives less obvious impression on the 5-HMF results (van Der Vis et al. 1993, Tong et al. 2011, Tao et al. 2014, Toftgaard et al. 2015. Therefore we prepare Figure 5(c) to prove that the effect of temperature and time was only significant at their optimum state which is at 103°C and 3 h. Given the optimum temperature as seen in figure, the yield of 5-HMF was perpendicular to the time regardless of catalyst loading, however until at 3h. Beyond that, the yield decreases. According to Kuster et al., formation increases with increasing enolization levels as well as increased levels of acyclic and fructose furanose forms at higher temperatures, which it may be disclosed that rising temperatures play a positive role to 5-HMF results. However, this attention is only used in case of less than 10 min of follow-up. After 20 min, there is almost no change to apply to 5-HMF results with rising temperatures. Therefore, it can be concluded that 5-HMF yield is directly proportional to the function of optimum time and temperature.
Previous studies also concluded the maximum yield of 5-HMF obtained at the optimum condition was 62% at 130°C for 3 hours (Tahvildari et al. 2011, Wang et al. 2014, Su et al. 2016. Meanwhile, Zhao et al. (2009) figured out about the effect of reaction time, aqueous to organic solvents ratio and catalyst loading. They concluded that higher yields can be obtained at longer reaction time, greater amounts of DMSO and lower catalyst loadings. 5% PA content only gives 5% of 5-HMF, but when PA content increases to 20%, it gives higher product yield of 62%. For the effect of reaction time, the highest yield of 5-HMF was obtained at 3 hours of reaction, however the yield decrease into 3.5% after 5 h reaction time. It was noted that excessive reaction time promoted the decomposition of 5-HMF into smaller molecules such as formic acid and levulinic acid which subsequently decrease the yield of 5-HMF (Pagan-Torres et al. 2012;Rosatella et al. 2011). As shown in Figure 5, the reaction temperature plays an important role on the outcome of 5-HMF. At higher reaction temperature, the 5-HMF also have increased rate of reaction as well as 5-HMF yield. 5-HMF yield increase from 6.8% at 100 o C to 62% at 130 o C after 3 hours of reaction. Shimizu et al. was reported the removal of water is a mild evacuation suppresses into two side reactions which are the hydrolysis of 5-HMF to levulinic acid and the partially reaction of dehydrated intermediates to condensation products. The effect of phosphotungstic acid (PA) content also showed a significant improvement of the catalytic performance where Normally, the mechanism reaction was more a cyclic cycle of dehydration of fructose. Figure 6 explains the molecular mass of 145 gmol -1 was the result of dehydration of fructose, which is a molecule (4R, 5R) -4-hydroxy-5-hydroxymethyl-4 49 or 5-dihydrofuran-2-carbaldehyde known as intermediates for the dehydration of fructose to 5-HMF in a solvent DMSO (Yue et al. 2016). The molecular mass of 163 gmol -1 showed that it was the product dehydrogenated of compound 1, namely 2,5-anydro-hexose involved in the dehydration of fructose (Yan et al. 2009;Yue et al. 2014). This dehydrogenated product is a byproduct related to the stoichiometric oxidation of the alcohol group of tungsten oxide catalyst into the by-product dehydrogenated stable and sturdy. 5-HMF can be obtained where the presence of two compounds with a molecular masses of 145 gmol -1 (compound 2) and 163 gmol -1 (compound 3) shows the unconverted sugar as monomer and oligomer. Figure 7 shows the reaction scheme of intermediates in the dehydration of glucose to 5-HMF. Conversion of carbohydrates-rich biomass into chemicals that may be used as direct substitutes for non-renewable source compounds was done. The study evaluated the synthesis of 5-HMF (5-Hydroxymethylfurfural) by dehydration of glucose in the presence of aprotic solvent, DMSO. The prepared sample 20% H 3 PW 12 O 40 /TEOS gives the highest yield of 5-HMF production (62% of 5-HMF yield at 130 o C in 3 h). The synthesized solid acid catalyst, set by sol-gel procedures also displayed good physical and catalytic properties, since the heteropoly acid has turned into insoluble compound by substituting a fraction of their proton apart from being supported on TEOS.