Alkyl Levulinates from Furfuryl Alcohol Using CT151 Purolite as Heterogenous Catalyst: Optimization, Puriﬁcation, and Recycling

: Commercially available Purolite CT151 demonstrated to be an efﬁcient acid catalyst for the synthesis of alkyl levulinates via alcoholysis of furfuryl alcohol (FA) at mild temperatures (80–120 ◦ C) and short reaction time (5 h). Reaction conditions were ﬁrst optimized for the synthesis of ethyl levulinate and then tested for the preparation of methyl-, propyl-, isopropyl-, butyl, sec -butyl- and allyl levulinate. Preliminary scale-up tests were carried out for most of the alkyl levulinates (starting from 5.0 g of FA) and the resulting products were isolated as pure by distillation in good yields (up to 63%). Furthermore, recycling experiments, conducted for the preparation of ethyl levulinate, showed that both the Purolite CT151 and the exceeding ethanol can be recovered and reused for four consecutive runs without any noticeable loss in the catalyst activity. Synthesis of methyl levulinate in autoclave : Furfuryl alcohol g, 50.97 was dissolved in 100 mL of methanol in the presence of CT151 (2.50 g, 50% wt.) as heterogeneous acid catalyst. The reaction mixture was heated at 80 ◦ C for 5 h with an autogenous pressure of 2 bar. The puriﬁcation proceeds by distillation as previously reported.


Introduction
Alkyl levulinates (AlkLevs) are biobased chemicals that occupy a strategic role in biomass valorization as potential substitutes of several oil-based compounds [1,2]. Their applications in the fields of biofuels (additives), plasticizers, lubricants, flavoring agents, fragrances, and green solvents have been extensively investigated [3,4].
As depicted in Figure 1, several synthetic routes to AlkLevs are reported in the literature; raw biomass or monomers derived from biomass can be both used as starting materials. The related synthetic procedures are based on an alcoholysis reaction promoted by (Brønsted) acid catalyst where the product selectivity depends on the substrate used [4]. In fact, direct conversion of biomass such as cellulose leads to AlkLevs with low or moderate yields due to a combination of solubility issues and drastic conditions required for the depolymerization step [10]. In this view, starting from simple C6 carbohydrates such as D-Glucose or D-Fructose is more convenient, although once again this acid-catalyzed methodology suffers from some solubility issues and the need for relative high temperature (120-200 • C) [11][12][13][14][15][16][17][18][19][20][21]. It should also be mentioned that, when D-Glucose is employed as substrate, an isomerization step is required, and this negatively affects the AlkLev yield. Furthermore, using simple carbohydrates as starting materials often resulted in the formation of a certain amount of humins due to the presence of 5-(hydroxymethyl) furfural (HMF) as reaction intermediate, a biobased platform chemical well known for its unstable nature.  In contrast, conversion of levulinic acid (LA) into alkyl levulinates via esterification is a high yielding process; however, since LA is derived from biomass, (poly)saccharides would be preferable starting materials [22][23][24].
Taking into account these considerations, a pentose sugar would be a more suitable starting material for AlkLev preparation ( Figure 1). Pentose sugars such as D-xylose, obtained from hemicellulose, are easily converted into furfural via acid-catalyzed dehydration; furfural can then be promptly reduced to furfuryl alcohol (FA) via hydrogenation. As more than 60% of the furfural produced worldwide is transformed into FA, this precursor is low cost and available in large amounts. Together with LA, FA provides the most promising substrate for the preparation of AlkLevs [25].
Numerous acid catalysts were investigated for alcoholysis of FA to alkyl levulinates. Initial studies focused on the use of homogeneous catalysts such as strong mineral acids or metal salts. [26] However, the difficult product separation, catalyst toxicity, and reaction container corrosion led to the exploitation of heterogeneous catalysts, including acidic ion exchange resins [27][28][29], zeolites [30][31][32], (sulfated) zirconia-based catalysts [33][34][35], (supported) heteropolyacids [36][37][38][39], activated carbon [40][41][42], functionalized organosilica materials [43][44][45][46][47][48], ionic liquids [49,50], and several sulfonic acid functionalized materials [51][52][53][54]. Some studies were also conducted employing continuous-flow apparatus [32,44] or microwave-assisted procedures [26,42]. Despite this extensive research on the FA alcoholysis reaction, there are still some issues to be addressed in order to set up a cost-effective and green procedure for its conversion into AlkLevs. In fact, most of the abovementioned published works focused on the catalytic system highlighting its preparation, effi- Taking into account these considerations, a pentose sugar would be a more suitable starting material for AlkLev preparation (Figure 1). Pentose sugars such as D-xylose, obtained from hemicellulose, are easily converted into furfural via acid-catalyzed dehydration; furfural can then be promptly reduced to furfuryl alcohol (FA) via hydrogenation. As more than 60% of the furfural produced worldwide is transformed into FA, this precursor is low cost and available in large amounts. Together with LA, FA provides the most promising substrate for the preparation of AlkLevs [25].
In this prospect, the work presented here encompasses a systematic study on the alcoholysis of FA to alkyl levulinates focusing on:

•
Efficiency of commercially available acid catalysts for the preparation of a library of AlkLevs that optimize the reaction conditions; • Preliminary investigations on a larger scale reaction (up to 5.0 g of FA) to address the purification of AlkLevs; • Recycling of the catalyst and of the reagent used in stoichiometric excess.
Results collected showed that, among the catalysts tested, Purolite CT151 was the most efficient for the synthesis of several alkyl levulinates in mild conditions. Methyl-, ethyl-, propyl-, isopropyl-, butyl-, sec-butyl-, and allyl levulinate were all prepared in good yields. The pure compounds were isolated as colorless or light yellowish liquids by distillation and fully characterized. Moreover, recycling experiments demonstrated that the Purolite CT151 can be reused for four runs without any noticeable loss in activity.

Materials and Methods
All reagents and solvents were purchased from Sigma-Aldrich; zeolite employed was also from Sigma-Merk (code: 96096). Purolites were kindly provided by Purolite ® . The reactions were monitored through GC-MS (GC System 6890N, Agilent Technologies mass selective detector 5973) with chromatography column HP-5MS. Compounds were injected with microsyringe Hamilton (10 µL). NMR spectra were recorded with a Bruker 300 MHz and 400 MHz spectrometer, in CDCl 3.
HR-MS spectra were acquired by means of a Bruker compact QTOF with a mass resolution of 30,000 in positive polarity mode. The mass calibration was conducted using a sodium formate cluster's solution and the data were processed in HPC mode. The acquisition was conducted in full scan mode in the range 50 to 500 m/z, with a 4.0 L min −1 at 180 • C of source dry gas. The ion formula of each compound was calculated with the Smart formula tool within the Bruker software platform by using a 4 mDa of mass confidence and considering the isotopic pattern ratio.

General Synthetic Procedure for the Synthesis of Alkyl Levulinates in One Gram Scale
In a 100 mL round bottom flask equipped with a water condenser and a magnetic stir bar, furfuryl alcohol (1.0 g, 10.2 mmol) was dissolved in 20 mL of the corresponding alcohol in the presence of CT151 (0.5 g, 50% wt.) as heterogeneous acid catalyst. The reaction mixture was heated at reflux temperature for 5 h under magnetic stirring. Afterward, the reaction was cooled to room temperature and the mixture was filtered on a layer of basic alumina and washed with a minimum amount of acetone. The solvent was evaporated by reduced pressure (rotavapor). The reaction crude was weighed (g crude ) and then analyzed via proton NMR. A comparison among NMR spectra of the pure samples (FA, EtLev, and EtMF) with the reaction mixture allowed the identification of residual traces of the reagent (FA), along with EtLev and EtMF. Integration of the distinctive NMR signals of these compounds gave the percent composition of the reaction mixture, i.e., %FA, %EtLev and %EtMF. Conversion was the %FA observed. The yields of EtLev and EtMF were calculated considering the recovered amount of the reaction mixture (g crude ) and the %EtLev and %EtMF evaluated via NMR. As an example, the yield of EtLev was assessed according to the equations g EtLev = %EtLev × g crude and thus Yield EtLev = (g EtLev/ g theorical ) × 100. The same calculation allowed for achieving the EtMF yield (Tables 1-3).

General Synthetic Procedure for the Synthesis of Alkyl Levulinates in Five Gram Scale
In a 250 mL round bottom flask equipped with a water condenser and a magnetic stir bar, furfuryl alcohol (5.0 g, 50.9 mmol) was dissolved in 100 mL of the corresponding alcohol in presence of CT151 (2.5 g, 50% wt.) as heterogeneous acid catalyst. The reaction mixture was heated at solvent reflux temperature for 5 h under magnetic stirring. Afterwards, the reaction was cooled to room temperature and the mixture was filtered on a layer of basic alumina and washed with a minimum amount of acetone. The organic phase recovered was transferred (in two tranches) to a 100 mL double-necked round bottom flask equipped with distillation apparatus; the solvent was distilled at atmospheric pressure. In order to completely remove any solvent residue, the mixture was transferred into a 25 mL double-necked round bottom flask equipped with distillation apparatus connected to a vacuum pump (40-60 mbar). The alkyl levulinates were then purified through distillation at reduced pressure as transparent or light-yellow liquids (Table 4).

Ethyl Levulinate as Model Study: Optimization of the Reaction Conditions
Ethyl levulinate (EtLev) was selected as a model compound to test the efficiency of the heterogenous and homogeneous catalysts for the alcoholysis of furfuryl alcohol (Scheme 1, Table 1).

Ethyl Levulinate as Model Study: Optimization of the Reaction Conditions
Ethyl levulinate (EtLev) was selected as a model compound to test the efficiency of the heterogenous and homogeneous catalysts for the alcoholysis of furfuryl alcohol (Scheme 1, Table 1).  In a typical experiment, FA was reacted with an excess of ethanol, used as solvent and reagent, in the presence of an acid catalyst. The reaction was conducted under magnetic stirring at reflux temperature (80 • C) for 3 h. Thus, prior to solvent evaporation, the mixture was filtered on a thin layer of basic alumina to remove acid residues that in some cases led to the formation of insoluble black plastic material (Supplementary Figure S1). The product yield was calculated via proton NMR (see Materials and Methods), as this technique allowed for certain identification of the signals related to furfuryl alcohol, ethyl levulinate, and 2-(ethoxymethyl)furan (EtMF) that forms as reaction intermediate.
The first test was conducted in the absence of a catalyst (#1; Table 1); FA was recovered unaltered from the reaction mixture. Amberlyst-15 was used as referring catalyst (#2; Table 1) as it has been extensively investigated for the alcoholysis of furfuryl alcohol [29,30]. In our reaction conditions, Amberlyst-15 led to a quantitative conversion of FA; however, EtLev formed only in 22% yield.
Purolite exchange resins, available in our laboratory, were then tested (#3-8; The Purolites selected were employed both wet and after being oven dried at 100 • C overnight (#3-8; Table 1). The presence of interstitial water in these sulfonated resins resulted in a lower efficiency with discrete values of FA conversion and product yield; in contrast, dried Purolites led to more promising results as EtLev yields ranged from 25 to 34%.
Some homogeneous acid catalysts were investigated (#11-16; Table 1). Experiments conducted in the presence of AlCl 3 and H 2 SO 4 allowed quantitative conversion of FA and discrete EtLev yield. However, these catalysts were discarded in consideration of their well-known toxicity/corrosivity, and foremost due to the formation of an insoluble black plastic material at the end of the reactions as consequence of the abovementioned concurrent polymerization reactions taking place in strong acidic conditions.
Despite the higher acidity of trifluoroacetic acid compared to acetic acid, both these homogeneous catalysts showed a negligible efficiency in promoting EtLev formation (#13-14; Table 1). Similar discouraging results were achieved for the experiments conducted in the presence of oxalic acid and BF 3 -O(Et) 2 (#15-16, Table 1); the latter showing quantitative conversion but only limited yield.
In conclusion, although all Purolites displayed similar catalytic performances, the oven dried Purolite CT151 showed the best result (#8; Table 1) and was selected for further optimization studies. Its observed enhanced efficiency might be ascribed to the smaller surface area and pore volume of Purolite CT151 compared to the other heterogenous catalysts employed (Supplementary Table S1); these characteristics might allow an optimal interaction with FA.
The effect of the amount of catalyst was next investigated; experiments were conducted by varying the amount of Purolite CT151 (Table 2). Amberlyst-15 was also tested in the same conditions as referring catalyst (#1-8; Table 2). In all experiments, Purolite CT151 confirmed its superior catalytic activity compared to Amberlyst-15 and the latter was no further investigated in the present study.
Although the best results for the Purolite catalyst were obtained when used in 100% wt., the difference in EtLev yield between the reaction conducted with 50% wt. catalyst was too small to justify doubling the amount of catalyst. Thus, further optimization reactions were conducted in the presence of 50% wt. Purolite CT151.
Several experiments were carried out varying the reaction time; FA ethanolysis was carried out for 1, 2, 3, 5, and 7 h after the reaction mixture reached the operating temperature of 80 • C; the results achieved are depicted in Figure 2.
FA conversion reached quantitative value already after 2 h at reflux temperature; however, EtLev yield continued to increase, reaching a maximum at 5 h (71%). This discrepancy might be attributed to the formation of at least one reaction intermediate different from EtMF that was clearly visible in the proton NMR spectrum, although we were not able to identify or isolate it. When the reaction was prolonged to 7 h, a lower yield of EtLev was observed (62%), possibly due to concurrent side reactions and/or product degradation.
To validate the data so far achieved, a larger scale reaction was conducted on 5.0 g of FA; the experiment led to a slightly higher EtLev yield (77% yield) compared to the one achieved on a 1.0 g scale. This result is a good indication that higher scale reactions might perform even better than smaller scale reactions.   Additional optimization attempts were conducted by varying the operating te ature and the FA concentration, which was previously set arbitrarily at 0.50 M (1.0 20 mL of EtOH). The reaction carried out at lower temperatures, i.e., from room tem   Figure 3 reports a possible reaction mechanism for the formation of EtLev from FA. A comprehensive investigation on the alcoholysis of furfuryl alcohol is still lacking, most likely due to the difficulties in identifying all the concurrent synthetic pathways and the related reaction intermediates; thus, further investigations are needed to clarify these aspects. Additional optimization attempts were conducted by varying the operating temperature and the FA concentration, which was previously set arbitrarily at 0.50 M (1.0 gr in 20 mL of EtOH). The reaction carried out at lower temperatures, i.e., from room temperature up to 60 °C, led to reduced substrate conversion and lower EtLev yield (Supplementary Table S2) Table S3). It can be assumed that at high dilution there is a reduced interaction between FA and ethanol. Meanwhile, working at high concentration, although faster interactions between the reagents took place, unwanted secondary reactions such as polymerization and degradation might also occur more rapidly, affecting product yield. Additional optimization attempts were conducted by varying the operating temperature and the FA concentration, which was previously set arbitrarily at 0.50 M (1.0 gr in 20 mL of EtOH). The reaction carried out at lower temperatures, i.e., from room temperature up to 60 • C, led to reduced substrate conversion and lower EtLev yield (Supplementary Table S2 Table S3). It can be assumed that at high dilution there is a reduced interaction between FA and ethanol. Meanwhile, working at high concentration, although faster interactions between the reagents took place, unwanted secondary reactions such as polymerization and degradation might also occur more rapidly, affecting product yield.
In conclusion, optimization investigations indicated that the best reaction condition encompassed a 0.5 M solution of FA in ethanol heated at reflux temperature (80 • C) for 5 h; analysis of the reaction mixture indicated an EtLev yield of 71% (77% in larger scale).

Alkyl Levulinates Syntheses, Products Isolation, and Catalyst Recycling
With the best-found reaction conditions known, the scope of this procedure was next investigated to develop a library of alkyl levulinates. Thus, FA alcoholysis was conducted in the presence of primary (methanol, 1-propanol, 1-butanol), secondary (2-propanol, 2-butanol), and unsaturated (allyl alcohol) alcohol (Scheme 2, Table 3).

Sustain. Chem. 2021, 2, FOR PEER REVIEW
In conclusion, optimization investigations indicated that the best reactio encompassed a 0.5 M solution of FA in ethanol heated at reflux temperature ( h; analysis of the reaction mixture indicated an EtLev yield of 71% (77% in lar

Alkyl Levulinates Syntheses, Products Isolation, and Catalyst Recycling
With the best-found reaction conditions known, the scope of this procedu investigated to develop a library of alkyl levulinates. Thus, FA alcoholysis was in the presence of primary (methanol, 1-propanol, 1-butanol), secondary (2-p butanol), and unsaturated (allyl alcohol) alcohol (Scheme 2, Table 3).   Table 3, AlkLevs were achieved in yields ranging from 30% according to 1 H NMR data. Among them, methyl levulinate (MeLev) was the uct formed in a rather low yield (#1 , Table 3). Thus, an additional experiment out in an autoclave apparatus at 80 °C, doubling the amount of the reagents a (#2, Table 3). In these conditions an autogenous pressure of 2 bar was obser eventually favored the FA methanolysis reaction, leading to 50% yield of Me Scheme 2. Synthesis of alkyl levulinates in presence of acid catalyst Purolite CT151. As reported in Table 3, AlkLevs were achieved in yields ranging from 30% up to 71% according to 1 H NMR data. Among them, methyl levulinate (MeLev) was the only product formed in a rather low yield (#1 , Table 3). Thus, an additional experiment was carried out in an autoclave apparatus at 80 • C, doubling the amount of the reagents and catalyst (#2, Table 3). In these conditions an autogenous pressure of 2 bar was observed, which eventually favored the FA methanolysis reaction, leading to 50% yield of MeLev. A further autoclave test was conducted at 160 • C (#3, Table 3); in this case the autogenous pressure increased constantly during the reaction up to 44 bar due to the unwanted formation of dimethyl ether. MeLev yield was calculated to be 32%; most likely the reduced yield observed is imputable to degradation of the substrate at this high temperature.

As reported in
FA alcoholysis in the presence of the other selected alcohols was tested both at 80 • C, and at their reflux temperature (#5-13, Table 3). The latter experiments resulted in the formation of the related AlkLevs in higher yields (40-63%).
Note also that to be best of our knowledge, this is the first time that allyl levulinate (AllLev, #12-13, Table 3) was prepared via FA alcoholysis. Previous published reports focused mainly on levulinic acid esterification [58].
AllLev is an interesting derivative as the presence of the allyl group within the levulinate backbone can allow further chemical modification of this compound.
Having established the robustness of our procedure, the best experiments-for most of the AlkLevs reported in Table 3-were repeated in larger scale starting from 5.0 g of FA ( Table 4) to address isolation of the products from the reaction mixture. After several attempts, we found that the best purification procedure encompassed the filtration of reaction mixture through a thin layer of basic alumina followed by a two-step distillation: (i) a first at atmospheric pressure to recover the residual alcohol, followed by (ii) a second distillation at reduced pressure to recover the pure AlkLev.
As shown in Table 4, the isolated yields of MeLev, EtLev, iPrLev, BuLev, sBuLev, and AllLev are fully consistent with the related values calculated by proton NMR analysis, confirming the validity of the analytical method used. It is interesting to point out that when the methanolysis of FA was conducted in larger scale, MeLev was achieved in higher yield (65%) compared to the 2.0 g scale procedure (50%), similar to that previously observed for the preparation of EtLev ( Figure 2).
All AlkLevs were isolated as transparent or yellowish liquid and completely characterized to confirm the proposed structures (see Supplemental Materials).
Furthermore, as part of this investigation, catalyst recycling was also studied, focusing on the case of EtLev preparation.
Purolite CT151, a heterogeneous catalyst, can be recovered from the reaction mixture by simple filtration; however, due to magnetic stirring employed to ensure the reaction mixture homogeneity, the tiny catalyst spheres were reduced to powder. To solve this issue, the ethanolysis of FA was performed without stirring, although the EtLev yield decreased to 23%. Further attempts were carried out inserting nitrogen flow via a glass pipette directly into the reaction mixture to ensure the mixing of the solution. Although in this case EtLev yield improved to 57%, the continuous nitrogen inlet caused a rapid ethanol evaporation that had to be constantly added to the reaction mixture. In contrast, when a 100% wt. of Purolite CT151 (compared to FA) was employed without stirring and/or nitrogen flow, EtLev was achieved with a yield comparable to our previous results, i.e., 66% (Supplementary Table S4).
Thus, the experiments on catalyst recycling were conducted employing FA:Purolite CT1515 in 1:1 weight ratio. At the end of each run, the catalyst was recovered, washed with the minimal amount of acetone, and dried overnight at 100 • C. According to the data collected ( Figure 4a) the catalyst can be reused for a maximum of four consecutive reactions; in fact, in the fifth run an evident decrease of EtLev yield was observed (44%).
Restoration of the catalyst acid sites was attempted by immersion of the spheres in an acid solution of water/methanol; the Purolite was then washed with acetone and dried overnight before being used in a new reaction. However, analysis of the reaction mixture showed that catalyst activity was not reestablished as EtLev yield was even lower than the fifth run (32%). This result was ascribed to the surface poisoning of the catalyst due to deposition of carbonaceous material and polymers that formed during the degradation of furfuryl alcohol. In fact, as the recycling runs proceeded, the Purolite spheres assumed a dark black color consistent with the one of the degradation products (polymers and humins) previously observed.
Sustain. Chem. 2021, 2, FOR PEER REVIEW 10 overnight before being used in a new reaction. However, analysis of the reaction mixture showed that catalyst activity was not reestablished as EtLev yield was even lower than the fifth run (32%). This result was ascribed to the surface poisoning of the catalyst due to deposition of carbonaceous material and polymers that formed during the degradation of furfuryl alcohol. In fact, as the recycling runs proceeded, the Purolite spheres assumed a dark black color consistent with the one of the degradation products (polymers and humins) previously observed.
(a) (b) At an industrial level, catalyst recycling is as relevant as the recovery and reuse of the reagents/solvents used in stoichiometric excess. Thus, some recycling tests were conducted on the five-gram-scale FA ethanolysis; both the catalyst and the exceeding ethanol were recycled for four consecutive runs (Figure 4b). For each recycling test the catalyst was recovered by filtration; the ethanol was recovered by distillation at atmospheric pressure and the product and EtLev purified via distillation at reduced pressure, as previously described. Ethanol distillation was operated in the presence of a mild flow of nitrogen to speed the procedure, although this caused a loss of about 30% (30 mL) of the reagent that, in the following run, had to be replenished with new reagent.
As a result, the large-scale reaction for the ethanolysis of FA allowed isolation of pure EtLev with a constant yield of ca. 50% in four consecutive runs where both ethanol and Purolite were recovered and reused (Figure 4b).

Conclusions
This work focuses on different aspects inherent to biomass valorization. Our intent was to fully investigate the process for the alcoholysis of furfuryl alcohol, starting from the selection of an efficient commercially available and low-cost homogenous/heterogenous acid catalyst, optimizing the reaction conditions, and exploring the scope of the reaction on different alcohols to achieve a library of alkyl levulinates isolated as pure compounds.
In this view, Purolite CT151-used in 50% wt. compared to FA-demonstrated to be an effective catalyst for the preparation of methyl, ethyl-, propyl-, isopropyl-, butyl-, secbutyl, and allyl levulinate from moderate to good yield. Performing, in the best-found reaction conditions, five grams scale alcoholysis reactions allowed setting up a purification procedure for all the investigated alkyl levulinates that were recovered in good yield via distillation and fully characterized. Furthermore, recycling tests conducted on the ethanolysis of FA showed that the catalyst and the exceeding ethanol can be recycled up to At an industrial level, catalyst recycling is as relevant as the recovery and reuse of the reagents/solvents used in stoichiometric excess. Thus, some recycling tests were conducted on the five-gram-scale FA ethanolysis; both the catalyst and the exceeding ethanol were recycled for four consecutive runs (Figure 4b). For each recycling test the catalyst was recovered by filtration; the ethanol was recovered by distillation at atmospheric pressure and the product and EtLev purified via distillation at reduced pressure, as previously described. Ethanol distillation was operated in the presence of a mild flow of nitrogen to speed the procedure, although this caused a loss of about 30% (30 mL) of the reagent that, in the following run, had to be replenished with new reagent.
As a result, the large-scale reaction for the ethanolysis of FA allowed isolation of pure EtLev with a constant yield of ca. 50% in four consecutive runs where both ethanol and Purolite were recovered and reused (Figure 4b).

Conclusions
This work focuses on different aspects inherent to biomass valorization. Our intent was to fully investigate the process for the alcoholysis of furfuryl alcohol, starting from the selection of an efficient commercially available and low-cost homogenous/heterogenous acid catalyst, optimizing the reaction conditions, and exploring the scope of the reaction on different alcohols to achieve a library of alkyl levulinates isolated as pure compounds.
In this view, Purolite CT151-used in 50% wt. compared to FA-demonstrated to be an effective catalyst for the preparation of methyl, ethyl-, propyl-, isopropyl-, butyl-, sec-butyl, and allyl levulinate from moderate to good yield. Performing, in the best-found reaction conditions, five grams scale alcoholysis reactions allowed setting up a purification procedure for all the investigated alkyl levulinates that were recovered in good yield via distillation and fully characterized. Furthermore, recycling tests conducted on the ethanolysis of FA showed that the catalyst and the exceeding ethanol can be recycled up to four times before Purolite CT151 started losing its efficiency.
These encouraging results may lead the way to future investigation of this catalyst on more complicated substrates, such as monosaccharide (D-Glucose and D-Fructose) and disaccharides (sucrose and cellobiose), eventually also employing a continuous-flow apparatus.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/suschem2030027/s1, Figure S1: Examples of the substrate/product degradation observed without previous filtration on basic alumina; Table S1: Purolite physical and chemical characteristics; Table S2: Synthesis of EtLev starting from FA at different temperatures; Table S3: Synthesis of EtLev starting from FA employing different solvent amount; Table S4: Synthesis of EtLev starting from FA: Preliminary investigation for catalyst recycling; Scheme S1: Reaction scheme of EtLev synthesis.