Efﬁcient Conversion of Monosaccharides into 5-Hydroxymethylfurfural Using Acidic Deep Eutectic Solvents

: Inthisstudy,aquick,simple,greenmethodofconvertingcarbohydratesinto5-hydroxymethylfurfural (HMF) with the use of deep eutectic solvents (DESs) was reported on. We synthesized 12 DESs for HMF conversion from carbohydrates which were studied under different conditions. Under optimal conditions, oxalic acid and citric acid with a choline chloride-based DES produced a maximum yield of HMF at 59 ± 2% and 62 ± 3% in 5 min at 120 ◦ C, respectively. The efficiency of converting glucose to HMF in a short time (5 min) at 140 ◦ C using CrCl 3 with a choline chloride-based DES was around 37 ± 1%, which was higher than in previous work. This study demonstrates the signiﬁcant potential of DESs as a combination for the continuous catalytic transformation of biomass in the synthesis of platform chemicals.


Introduction
Due to the rapid depletion of fossil fuel sources and increasing global warming concerns, the study of alternative biofuels and value-added chemicals from sustainable resources has attracted more attention [1][2][3][4]. As a result of these problems, scientists from all over the world are working to find and develop new, clean, and safe energy resources in order to reduce reliance on fossil fuels. Carbohydrates are a promising replacement for fossil resources in the production of valuable chemicals and fuel components because they are renewable resources. Among many types of high-value chemicals, 5-hydroxymethylfurfural (HMF) is one of the most promising alternative chemicals because it is an important building block for conversion into valuable derivatives, such as 2,5-dimethylfuran, 2,5-diformylfuran, 1,6-hexanediol, and levulinic acid [5][6][7][8][9]. The most straightforward route to produce HMF is via direct conversion from fructose due to its high yield and selectivity. Carbohydrate sources such as glucose or cellulose are preferred for the synthesis of HMF. Recently, the synthesis of HMF from carbohydrates has received significant attention. For example, Ly et al. demonstrated the continuous conversion of fructose into HMF in a home-built continuous flow system employing conventional homogeneous acid catalysts such as p-TsOH and Brønsted acidic ionic liquid. An HMF yield of 46% was obtained for p-TsOH/DMSO at the temperature of 130 • C within 5 min [10]. Souzanchi and co-workers reported niobium phosphate, synthetic sulfated niobium, and Amberlyst 36 were active and selective, resulting in HMF yields of 54-60% under ideal working conditions [11]. However, mineral acids involve equipment damage, energy costs, and environmental difficulties. To reduce negative environmental impacts, the study of 36 were active and selective, resulting in HMF yields of 54-60% under ideal working conditions [11]. However, mineral acids involve equipment damage, energy costs, and environmental difficulties. To reduce negative environmental impacts, the study of deep eutectic solvents (DESs) as a catalyst has attracted scientists. DESs are composed of massive, nonsymmetric ions with low lattice energy and, as a result, low melting points. They are typically synthesized by complexing a quaternary ammonium salt with a metal salt or hydrogen bond donor (HBD). The charge delocalization caused by hydrogen bonding between, for example, a halide ion and a hydrogen-donor moiety is responsible for the mixture's lower melting point in comparison to the melting points of the separate components [12]. DESs usually have lower thermal stability and leach more easily, but are less costly and easier to prepare than ionic liquids. At present, DESs have received much attention in the conversion of carbohydrates into HMF [13][14][15][16]. Chen et al. reported using sulfuric acid and DES catalyst systems to convert glucose into HMF under optimal conditions with an efficiency of 34.86%; however, the use of liquid catalyst makes it challenging to recover and reuse the catalyst system [17]. Vigier et al. (2012) examined the influence of natural deep eutectic solvents (NADESs) based on betaine hydrochloride on fructose dehydration. A mixture of betaine hydrochloride, choline chloride, and water (10/0.5/2) generated the highest HMF yields from fructose (84%) and inulin (52%) at 110 °C for 1 h [18]. In 2017, Delbecq et al. investigated HMF production using a water/MIBK biphasic system with a combination of betaine and formic acid as a catalyst. From fructose, the yield of HMF obtained was 82% at 160 °C for 60 min; from starch, 45% at 190 °C for 60 min; from glucose, 55% at 190 °C for 60 min; and from microcrystalline cellulose, 26% at 200 °C for 80 min [19]. Low yield and difficulty in product separation are two issues that arise throughout the process of converting glucose into HMF [20][21][22][23]. Glucose undergoes isomerization to form fructose, then fructose dehydration to form HMF (Scheme 1). Therefore, converting glucose into HMF has proven to be challenging due to the fact that the product of the reaction creates a by-product, which causes the yield to drop and results in a lack of selectivity. Scheme 1. The procedure of synthesizing HMF from fructose/glucose. The synthesis of HMF from carbohydrates (fructose and glucose) employing deep eutectic solvents (DESs) as a catalyst/solvent system under typical heating conditions was presented. The use of DESs in combination with conventional heating methods made it possible to convert carbohydrates into HMF in a short time while maintaining a high yield. Our work provides a simple method for the synthesis of DESs for HMF conversion. The synthesis of HMF from carbohydrates (fructose and glucose) employing deep eutectic solvents (DESs) as a catalyst/solvent system under typical heating conditions was presented. The use of DESs in combination with conventional heating methods made it possible to convert carbohydrates into HMF in a short time while maintaining a high yield. Our work provides a simple method for the synthesis of DESs for HMF conversion.  Figure 1.The choline chloride spectrum shows a characteristic signal for the C-N junction vibration at 1137 cm −1 of the quaternary ammonium salt. The signals at 2911 and 2895 cm −1 are assigned to stretching vibrations of H-C sp3 bonds, and signals at 1543 cm −1 are ascribed to alkyl groups. The presence of oxygen is confirmed by the absorption fringes of the C-O junction oscillation at 1031 cm −1 . In addition, the -OH group has a stretching vibration in the region from 3000 to 3500 cm −1 [24][25][26]. The FT-IR spectrum of oxalic acid shows the appearance of peaks at 1186 cm −1 , which is typical of the C-O bond oscillations. The peak at 1720 cm −1 is typical of the stretching vibrations of C=O. In addition, a wide bulb absorption band characteristic of the O-H group is observed at 3354 cm −1 [27]. Therefore, [CholineCl][Oxalic acid] was synthesized from choline chloride and oxalic acid in a 1:1 ratio; the FT-IR results showed the presence of characteristic oscillations for both components. The region ranges from 3100 to 3650 cm −1 bulb and is wider than that of both individual components. Similar to [CholineCl][Oxalic acid], [CholineCl][Citric acid] is synthesized from choline chloride and citric acid in 1:1 ratio, including characteristic signals of both components when compared on FT-IR spectra through the figure below. The carbonyl group at position 1722 cm −1 and the C-O group at position 1211 cm −1 are special signals from citric acid. In particular, the signal specific to the C-N junction vibration at 1142 cm −1 of quaternary ammonium salt is confirmed. The signal at 2920 cm −1 is assigned to the stretching oscillation of the H-Csp 3 . The presence of acid and alcohol O-H groups was observed at 3100-3650 cm −1 with a broad signal. In particular, the appearance of hydrogen bonding is noted by the shift of this signal towards a lower wave number [28]. When the FT-IR spectrum of [CholineCl][CrCl 3 .6H 2 O] and choline chloride was studied, it was found that the OH group shifted between 3500 and 3000 cm −1 . This was noticed while comparing the two spectra; it is hypothesized that these alterations take place in [CholineCl][CrCl 3 .6H 2 O] because of the hydrogen interactions that take place between the functional groups [29]. Deep eutectic solvents were identified via FT-IR as shown in Figure 1. The choline chloride spectrum shows a characteristic signal for the C-N junction vibration at 1137 cm −1 of the quaternary ammonium salt. The signals at 2911 and 2895 cm −1 are assigned to stretching vibrations of H-Csp3 bonds, and signals at 1543 cm −1 are ascribed to alkyl groups. The presence of oxygen is confirmed by the absorption fringes of the C-O junction oscillation at 1031 cm −1 . In addition, the -OH group has a stretching vibration in the region from 3000 to 3500 cm −1 [24][25][26]. The FT-IR spectrum of oxalic acid shows the appearance of peaks at 1186 cm −1 , which is typical of the C-O bond oscillations. The peak at 1720 cm −1 is typical of the stretching vibrations of C=O. In addition, a wide bulb absorption band characteristic of the O-H group is observed at 3354 cm −1 [27]. Therefore, [CholineCl][Oxalic acid] was synthesized from choline chloride and oxalic acid in a 1:1 ratio; the FT-IR results showed the presence of characteristic oscillations for both components. The region ranges from 3100 to 3650 cm −1 bulb and is wider than that of both individual components. Similar to [CholineCl][Oxalic acid], [CholineCl][Citric acid] is synthesized from choline chloride and citric acid in 1:1 ratio, including characteristic signals of both components when compared on FT-IR spectra through the figure below. The carbonyl group at position 1722 cm −1 and the C-O group at position 1211 cm −1 are special signals from citric acid. In particular, the signal specific to the C-N junction vibration at 1142 cm −1 of quaternary ammonium salt is confirmed. The signal at 2920 cm −1 is assigned to the stretching oscillation of the H-Csp 3 . The presence of acid and alcohol O-H groups was observed at 3100-3650 cm −1 with a broad signal. In particular, the appearance of hydrogen bonding is noted by the shift of this signal towards a lower wave number [28]. When the FT-IR spectrum of [Cho-lineCl][CrCl3.6H2O] and choline chloride was studied, it was found that the OH group shifted between 3500 and 3000 cm −1 . This was noticed while comparing the two spectra; it is hypothesized that these alterations take place in [CholineCl][CrCl3.6H2O] because of the hydrogen interactions that take place between the functional groups [29].    Figure 2. According to the data provided by the TGA, the mass does not change much (approximately 10%) even at the first thermal decomposition step when the temperature is raised to 130 • C; nevertheless, previous work [27,28,30] has shown that when the temperature is increased from 200 • C to 300 • C, this fully disintegrates. At the final thermal degradation step, [CholineCl][Citric acid] completely decomposed in the temperature range from 300 • C to 450 • C, while the structure of [CholineCl][Oxalic acid] completely decomposed above 300 • C. Therefore, with the temperature that our study determined, the DES was still stable and had not decomposed. Moreover, 1 H NMR and 13 C NMR were used to identify DESs, which provided detailed Supporting information.  Figure 2. According to the data provided by the TGA, the mass does not change much (approximately 10%) even at the first thermal decomposition step when the temperature is raised to 130 °C; nevertheless, previous work [27,28,30] has shown that when the temperature is increased from 200 °C to 300 °C, this fully disintegrates. At the final thermal degradation step, [CholineCl][Citric acid] completely decomposed in the temperature range from 300 °C to 450 °C, while the structure of [CholineCl][Oxalic acid] completely decomposed above 300 °C. Therefore, with the temperature that our study determined, the DES was still stable and had not decomposed. Moreover, 1 H NMR and 13 C NMR were used to identify DESs, which provided detailed Supporting information.

Effect of Brønsted Acidic Additives on the Conversion of Fructose into HMF
To evaluate the effect of Brønsted acidic additives, we conducted the conversion of HMF from fructose, using DESs as a catalyst system ( Figure 3) with fructose (1 mmol) and DESs (5 mmol), at 100 °C, and various reaction times (5, 10, 20, 30 and 60 min). The reaction was carried out with fructose (1 mmol, 180 mg) and DESs (5 mmol . This could be caused by the resonance effect in the fumaric acid structure, which made the Brønsted acid sites less active. As a result, their ability to dehydrate fructose was greatly reduced, and the product formation efficiency was significantly lower than [CholineCl][Malonic acid].

Effect of Brønsted Acidic Additives on the Conversion of Fructose into HMF
To evaluate the effect of Brønsted acidic additives, we conducted the conversion of HMF from fructose, using DESs as a catalyst system ( Figure 3) with fructose (1 mmol) and DESs (5 mmol), at 100 • C, and various reaction times (5, 10, 20, 30 and 60 min). The reaction was carried out with fructose (1 mmol, 180 mg) and DESs (5 mmol (20,30, and 60 min). Interestingly, with the increased amount of DES loading, the higher yield of HMF was observed up to 5 mmol with no further significant increase beyond 10 mmol. The main reason why increasing the amount of DES reduced the yield of the product could be due to the rise in the by-product from the polymerization of HMF. As the amount of catalyst increased, dehydration to produce the desired product was given priority. It seems evident that the higher yields of HMF were mainly observed in the presence of [ , was investigated across a range including 1, 5, and 10 mmol in the synthesis of HMF from fructose over five different lengths of time (5,10,20,30, and 60 min) for 100 °C. As illustrated in Figure 5, the use of 5 mmol catalysts [CholineCl][Oxalic acid] and [CholineCl][Citric acid] provided higher yield than those of 1 mmol and 10 mmol over different lengths of time (20,30, and 60 min). Interestingly, with the increased amount of DES loading, the higher yield of HMF was observed up to 5 mmol with no further significant increase beyond 10 mmol. The main reason why increasing the amount of DES reduced the yield of the product could be due to the rise in the by-product from the polymerization of HMF. As the amount of catalyst increased, dehydration to produce the desired product was given priority. It seems evident that the higher yields of HMF were mainly observed in the presence of [   In this case, the yield of HMF tended to decline with increased reaction time because HMF was polymerizing or becoming humin [31]. As expected, the rate of conversion accelerated with rising temperature. Lower temperatures, such as 80 and 100 °C, caused the reaction time to be extended to approximately 1 h, and the HMF conversion efficiency to begin to decline. Simultaneously, after only 10 min at 120 °C, the HMF conversion efficiency began to decline. The findings showed that the higher the temperatures that were used, the higher the yields that were obtained over In this case, the yield of HMF tended to decline with increased reaction time because HMF was polymerizing or becoming humin [31]. As expected, the rate of conversion accelerated with rising temperature. Lower temperatures, such as 80 and 100 • C, caused the reaction time to be extended to approximately 1 h, and the HMF conversion efficiency to begin to decline. Simultaneously, after only 10 min at 120 • C, the HMF conversion efficiency began to decline. The findings showed that the higher the temperatures that were used, the higher the yields that were obtained over short reaction times, and the activity of DESs decreased when the reaction time was prolonged. According to the aforementioned evaluation, a temperature of 120 • C was chosen as the most optimal temperature for future investigation of HMF conversion. short reaction times, and the activity of DESs decreased when the reaction time was prolonged. According to the aforementioned evaluation, a temperature of 120 °C was chosen as the most optimal temperature for future investigation of HMF conversion. The production of HMF was limited by low selectivity and yield, particularly from glucose or cellulose substrates. Thus, various pathways have been studied as processes for preparing DFF from glucose or cellulose using different catalysts under various conditions. Therefore, the use of several DESs has recently been investigated for their potential as catalysts for biomass-to-HMF conversion. As shown in Figure 7, we evaluated the conversion of glucose to HMF using two different DESs (quaternary ammonium salt/metal chloride hydrate and quaternary ammonium salt/hydrogen bond donor). The HMF yield of the [CholineCl][Oxalic acid] solvent was very low at both temperatures, with the highest at 100 °C (5%). However, as the reaction time and temperature increased, the yields of HMF were reduced. This can be explained by the weak and easy decomposition of hydrogen bonds in these DESs at high temperatures. On the other hand, the obtained HMF yield in the [CholineCl][CrCl3] was significantly higher than from the previous DESs. At 100 °C, the HMF yield obtained was 19% after just 5 min. Furthermore, increasing the temperature to 140 °C (36%) for 5 min improved performance significantly. This type of DES is made up of coordination bonds to form a stable Cr 3+ complex. The production of HMF was limited by low selectivity and yield, particularly from glucose or cellulose substrates. Thus, various pathways have been studied as processes for preparing DFF from glucose or cellulose using different catalysts under various conditions. Therefore, the use of several DESs has recently been investigated for their potential as catalysts for biomass-to-HMF conversion. As shown in Figure 7, we evaluated the conversion of glucose to HMF using two different DESs (quaternary ammonium salt/metal chloride hydrate and quaternary ammonium salt/hydrogen bond donor). The HMF yield of the [CholineCl][Oxalic acid] solvent was very low at both temperatures, with the highest at 100 • C (5%). However, as the reaction time and temperature increased, the yields of HMF were reduced. This can be explained by the weak and easy decomposition of hydrogen bonds in these DESs at high temperatures. On the other hand, the obtained HMF yield in the [CholineCl][CrCl 3 ] was significantly higher than from the previous DESs. At 100 • C, the HMF yield obtained was 19% after just 5 min. Furthermore, increasing the temperature to 140 • C (36%) for 5 min improved performance significantly. This type of DES is made up of coordination bonds to form a stable Cr 3+ complex.
A recycling test of DESs was necessary for industrial processes. Thus, the recycling tests were carried out under optimal conditions with fructose (1 mmol, 180 mg) and The results are shown in Figure 8. After completion of the reaction, the desired HMF was separated using a 1:1 mixture of diethyl ether and ethyl acetate (6 × 5 mL). The HMF yield of DESs produced from organic acids declined dramatically over the reusability periods (approximately 5-10%). This is due to the fact that when the reaction creates water, the water content of the DESs increases, and these DESs deteriorate with time, lowering the product conversion activity. The HMF yield was also somewhat lowered while utilizing DESs  Figure 8. After completion of the reaction, the desired HMF was separated using a 1:1 mixture of diethyl ether and ethyl acetate (6 × 5 mL). The HMF yield of DESs produced from organic acids declined dramatically over the reusability periods (approximately 5-10%). This is due to the fact that when the reaction creates water, the water content of the DESs increases, and these DESs deteriorate with time, lowering the product conversion activity. The HMF yield was also somewhat lowered while utilizing DESs

Mechanism Proposal
The conversion of glucose into HMF is composed of two processes, including the isomerization of glucose to fructose and the dehydration of fructose to HMF. The mechanism for fructose conversion into HMF through dehydration has been reported in previous literature [14,21,32]. In this work, we strongly proposed the mechanism for glucose conversion into HMF using [CholineCl][CrCl3.6H2O] as a catalyst (Scheme 2). Direct coordination of glucose with the Lewis acidic sites of CrCl3 supports the open-ring form of

Mechanism Proposal
The conversion of glucose into HMF is composed of two processes, including the isomerization of glucose to fructose and the dehydration of fructose to HMF. The mechanism for fructose conversion into HMF through dehydration has been reported in previous literature [14,21,32]. In this work, we strongly proposed the mechanism for glucose conversion into HMF using [CholineCl][CrCl 3 .6H 2 O] as a catalyst (Scheme 2). Direct coordination of glucose with the Lewis acidic sites of CrCl 3 supports the open-ring form of glucose. Moreover, Cr 3+ complexes temporarily self-organize into dimers during the catalytic process, which aids in the glucose-to-fructose isomerization phase that ultimately determines HMF selectivity. In order to produce high HMF yields [33,34], it is necessary to have a reaction environment that allows for the free movement of the Cr complexes and the Cl − sites [35] Table 1 illustrates the comparison of this work on the conversion of glucose/fructose into HMF with previous work. Generally, the previous work provided a method for synthesizing HMF with different catalysts. For instance, Hou et al. reported a catalyst system that was synthesized from biomass and applied to converting fructose into HMF with an HMF yield of 81% after 60 min at 120 °C (entry 1). For glucose conversion (entries 2-5), the study investigated the catalytic system for synthesizing HMF with a yield of 27.5 to 51% HMF. It can be seen that glucose has to isomerize into fructose with a Lewis acid as a catalyst. In the next step, fructose was dehydrated to HMF using Brønsted acidic. Interestingly, our work provided a simple method for the conversion of glucose/fructose into HMF with a high yield in a short time using a green catalytic system (entry 7  Table 1 illustrates the comparison of this work on the conversion of glucose/fructose into HMF with previous work. Generally, the previous work provided a method for synthesizing HMF with different catalysts. For instance, Hou et al. reported a catalyst system that was synthesized from biomass and applied to converting fructose into HMF with an HMF yield of 81% after 60 min at 120 • C (entry 1). For glucose conversion (entries 2-5), the study investigated the catalytic system for synthesizing HMF with a yield of 27.5 to 51% HMF. It can be seen that glucose has to isomerize into fructose with a Lewis acid as a catalyst. In the next step, fructose was dehydrated to HMF using Brønsted acidic. Interestingly, our work provided a simple method for the conversion of glucose/fructose into HMF with a high yield in a short time using a green catalytic system (entry 7). The yields of HMF obtained from fructose conversion were 61% and 59% using [CholineCl][Citric acid] or [CholineCl][Oxalic acid], respectively. Moreover, the glucose conversion remarkably yielded 37% HMF at 140 • C after 5 min.
Thermogravimetric analysis (TGA) with model TGA/DSC (Mettler, Toledo, Switzerland) was used to analyze and evaluate the thermal stability of DESs. Samples were calcined from room temperature to 800 • C with a scanning speed of 10 Kpm in N 2 . Fouriertransform infrared spectroscopy (FT-IR) was recorded on a Bruker's VERTEX 70 series FT-IR spectrometers with a measuring range of 600 to 4000 cm −1 .
The HMF was quantified via an Agilent Technologies 1260 Infinity HPLC with a DAD detector at wavelength 285 nm. An Inert Sustain C18 (5 µm, 4.6 mm × 150 mm) column was used to separate the components of the reaction mixture. The column temperature was maintained at 30 • C. A mixture of methanol (A) and 2.5 mM sulfuric acid (B) was used as eluent with a flow rate of 0.7 mL min −1 , gradient as follows: 0-2.50 min, 100% B; 2.50-2.51 min, 85% B; 2.51-10.00 min, 50% B; 10.00-17 min, 100% B.

Procedure of Synthesizing HMF
All reactions were carried out in a 10 mL glass vial equipped with a stirring bar. In the typical experiment, a mixture of fructose (1 mmol, 180 mg) and DES (5 mmol) was charged into the vial and heated via a magnetic hot plate stirrer with a closedcap reaction tube.  (5,10,20,30, and 60 min), catalyst dosage (1, 5, and 10 mmol), and substrate (fructose and glucose), were investigated to find optimal reaction conditions. After a set reaction time, the sample was weighed and diluted with ultrapure water, then filtered through 0.45 µm. Hence, the yield of HMF was determined via HPLC with a DAD detector. The calculation of HMF yield using an external standard calibration curve method is as follows (1): % HMF = (mole of HMF/mole of initial fructose or initial glucose) × 100 (1)

Conclusions
In summary, we have successfully created a reliable process for the synthesis of 5-HMF from glucose/fructose by the combined use of DESs. An essential role for DESs in this biomass catalytic process was discovered, and the hypothesized mechanism for the synergistic action of quaternary ammonium ionic liquids and inorganic/organic salts was confirmed. For both carbohydrates evaluated, the best conditions for HMF synthesis were identified as (for fructose with [CholineCl][Citric acid] or [CholineCl][Oxalic acid] 120 • C for 5 min and glucose with [CholineCl][CrCl 3 ] at 140 • C, 5 min). The catalytic system could be reused three times without changing the yield of HMF. Therefore, our work provided a simple procedure to synthesize HMF from glucose/fructose in a short time with a good yield.  Data Availability Statement: In this manuscript, our characterizations were FT-IR, TGA, 1 H NMR, 13 C NMR, and HPLC. All data have been reported as images.