Experimental and Kinetic Studies on the Conversion of Glucose to Levulinic Acid Catalyzed by Synergistic Cr/HZSM-5 in GVL/H₂O Biphasic System

Abstract

In this paper, modified HZSM-5 catalysts with different ratios of chromium (Cr/HZSM-5) were synthesized and the solvent effect of gamma valerolactone (GVL) on the enhancement of levulinic acid (LA) yield was investigated. Characterization of the Cr/HZSM-5 catalyst revealed that the introduction of Cr did not change the structure of HZSM-5. The LA yield was increased from 42.5% (H₂O solvent system) to 51.4% (GVL/H₂O solvent system) under optimal conditions. The influence of GVL on the reaction mechanism was investigated through kinetic analysis, revealing that the incorporation of GVL reduces the activation energy barrier for the conversion of glucose to LA, thereby enhancing the glucose dehydration process. The effect of GVL on the product (LA) was studied, based on molecular dynamics. It was found that the addition of GVL squeezes the water in the solvent system into the second solvation shell layer, which causes GVL to distribute around the carbonyl, hydroxyl, and carboxyl groups of LA, and reduces the likelihood of LA side reactions, thus increasing the yield of LA.

1. Introduction

In the background of global fossil energy scarcity and environmental degradation, the development of renewable energy is given priority in the current trajectory of social progress [1,2,3]. As the exclusive renewable carbon source, it has attracted significant scrutiny from researchers owing to its environmentally friendly and sustainable characteristics. Cellulose, a natural polymer composed of glucose units linked by β-1,4-glycosidic bonds, represents one of the most abundant biomass resources on Earth. Glucose is broadly acknowledged as a pivotal precursor for synthesizing a wide range of high-value chemicals, including 5-hydroxymethylfurfural (HMF) [4], levulinic acid (LA), and 5-ethoxymethylfurfural (EMF) [5]. Among these distinguished chemicals, LA has been categorized as one of the 12 biomass-based platform compounds acknowledged by the Department of Energy, United States, and it is extensively employed in the pharmaceutical and agricultural fields [6,7]. It can undergo further synthesis, producing valuable compounds like γ-valerolactone, resin, 2-methyltetrahydrofuran, etc. [8,9]. Consequently, the conversion of glucose to LA holds profound significance in addressing the energy crisis and mitigating environmental pollution [10].

The conversion of glucose into LA involved a series of complex, multi-step cascade reactions, where the catalyst played a pivotal role in orchestrating key reactions and optimizing target products [11,12]. In the non-homogeneous catalytic systems extensively employed in the glucose synthesis of LA, the catalyst carriers were usually carbon, metal oxides, and zeolite molecular sieves [13,14]. These carriers were subsequently loaded with active centers tailored to meet specific requirements. In the realm of catalyst carriers, zeolite molecular sieve was a prevalent choice, characterized by its impressive specific surface area, ample pore structure, and effortless loading of active sites [15]. These characteristics eased the diffusion of the reaction substrate into the pore structure and increased the contact possibility with the active sites [16]. Thus, molecular sieve-based catalysts with strong catalytic activity are frequently utilized in materials science for the high-value transformation of biomass.

Research has indicated that metal chromium exhibits a favorable catalytic effect in the reaction of glucose to LA [17]. Consequently, solid acid catalysts constructed by loading metal chromium onto zeolite molecular sieves were of interest. Nevertheless, when chromium was loaded onto molecular sieves as active sites, the interaction between metal and molecular sieve skeleton led to increasing load and clogging pore structure, hindering the contact between reactants and active sites, and thereby diminishing the product yield [18,19,20,21].

Confronted with these challenges, the solvent can be tailored to the reaction process, which can provide tunability and control the stable and ordered generation of products [22]. Mellmer et al. discovered that polar nonprotonic solvents containing chloride ions can envelop the reaction substrate and this configuration facilitated the protonation reactions [23]. Polar nonprotonic solvents could establish hydrophilic domains surrounding the reaction substrate. Notably, γ-valerolactone (GVL) as a quintessential polar nonprotonic solvent was extensively used in biomass conversion and diverse organic reactions, distinguished by its green and renewable properties [24,25]. Song et al. found that when the reaction system changed from pure water into solution of 75% GVL with 25% H₂O, the reaction rate constant decreased from 0.0034 min⁻¹ to 0.0020 min⁻¹ and the energy barrier decreased from 117 kJ/mol to 96 kJ/mol [26]. Therefore, the addition of a suitable solvent, such as GVL, to the reaction system could enhance the reaction rate, reduce the energy barrier, and thereby facilitate the overall reaction process. Despite extensive discourse on solvent effect at the macroscopic level, there has been scant research addressing the interaction mechanism between solvent system and reactants, and the inhibition of by-products at the microscopic level [27].

In this study, Cr/HZSM-5 catalysts were synthesized by incorporating chromium metal into HZSM-5 molecular sieves as precursor materials. The catalytic performance of Cr/HZSM-5 catalysts was assessed using LA derived from glucose conversion as a model reaction. Additionally, the impact of various solvent systems on the catalytic efficiency of Cr/HZSM-5 catalysts was examined. In addition, reaction kinetics and molecular dynamics were employed to delve deeper into the mechanism by which solvent enhancement impacts the catalytic performance of Cr/HZSM-5 catalysts. This study aims to prepare catalysts for efficient conversion of glucose to LA, and elucidate the fundamental interactions between the solvent and product at the microscopic level to further explain the mechanism of solvent-improving catalytic effect.

2. Results and Discussion

2.1. Catalyst Characterization

Figure 1A shows the XRD patterns of Cr/HZSM-5 catalysts with different loading ratios. As depicted in Figure 1A, it was apparent that Cr/HZSM-5 retained several noticeable diffraction peaks at 7.9°, 8.8°, 23.0°, and 23.9°, which were the distinctive peaks associated with MFI-type molecular sieves [28]. The characteristic peaks of MFI-type molecular sieves weakened upon loading Cr metal into the HZSM-5 catalyst, likely due to the structural changes induced by chromium incorporation. To confirm the chemical valence state of Cr in HZSM-5, XPS analysis was performed (Figure 1B), revealing that Cr species on the Cr/HZSM-5 catalyst surface existed as Cr₂O₃ with a binding energy of 576.62 eV. These findings confirm that elemental Cr was effectively loaded onto the HZSM-5 surface via a straightforward wet impregnation method.

Figure 2 shows the infrared spectra of HZSM-5 and Cr/HZSM-5. Figure 1 reveals characteristic peaks at 456 cm⁻¹, 546 cm⁻¹, 794 cm⁻¹, 1066 cm⁻¹, and 1644 cm⁻¹ for both Cr/HZSM-5 and HZSM-5 across various loadings. These results show that the HZSM-5 framework remains constant with increasing chromium loading, which is consistent with the XRD findings. Vibrational peaks at 456 cm⁻¹ and 1066 cm⁻¹ were observed on both HZSM-5 and Cr/HZSM-5 catalysts, corresponding to the symmetric vibrations of SiO₄ and AlO₄ tetrahedra within the five-membered ring of HZSM-5 [29]. The characteristic peak at 546 cm⁻¹ is attributed to the exo-bicyclic vibrations of the zeolite framework [30]. The vibrational peak at 794 cm⁻¹ corresponds to the structure-sensitive band of the molecular sieve framework [31].

Figure 3 presents the TEM, SEM, and EDS analyses of HZSM-5 and 15% Cr/HZSM-5 catalysts. As observed in Figure 3, the HZSM-5 structure retained its prismatic crystalline form after Cr loading. To confirm the successful Cr loading, EDS-mapping images of Cr/HZSM-5 were obtained, demonstrating that Cr was effectively loaded and well-dispersed on the surface of HZSM-5.

The metal content and structural changes of HZSM-5 following the introduction of Cr were analyzed using ICP-OES and N₂ adsorption–desorption techniques, with the results presented in

Table 1. Cr content, pore parameters, and acidic properties of HZSM-5 and Cr/HZSM-5 catalysts.

Catalysts	Cr (wt%)	Surface Area (m2/g)	Pore Size (nm)	Weak Acidity Amount (mmol/g)	Moderate Acidity Amount (mmol/g)	Total Acidity (mmol/g)
HZSM-5	-	266	5.11	0.89	0.76	1.650
5% Cr/HZSM-5	0.97	262	5.16	2.813	1.154	3.967
10% Cr/HZSM-5	1.65	246	4.97	2.946	1.398	4.344
15% Cr/HZSM-5	2.48	231	3.50	2.994	1.581	4.575
20% Cr/HZSM-5	3.09	190	1.87	3.223	1.553	4.776

. It was observed that the Cr content in the Cr/HZSM-5 catalysts increased nearly proportionally with the increase in Cr loading. The increase in Cr loading led to varying reductions in the surface area and pore size of HZSM-5. As the Cr content in the Cr/HZSM-5 catalyst increased from 0.97 wt% to 3.09 wt%, the total specific surface area of the catalyst decreased from 262 m²/g to 190 m²/g, and the pore size decreased from 5.16 nm to 1.87 nm. This can be attributed to the Cr loading occupying the pore structure, reducing specific surface area and pore size.

The acidity of the Cr/HZSM-5 catalysts was systematically evaluated via NH₃-TPD, with the resulting data presented in Figure S3. As depicted in Table 1, all catalysts exhibited a distinct peak at approximately 110 °C, corresponding to the weak acid sites, such as hydroxyl (OH) or carboxyl (COOH) groups. The strong acid sites observed in HZSM-5, characterized by a peak around 400 °C, are attributed to the Brønsted acid centers within the Si-OH-Al structure. The incorporation of Cr metal induces ion exchange between metal cations and the Brønsted acid sites, thereby enhancing the concentration of strong acid sites within the catalyst [32]. The aforementioned results demonstrate that, with increasing Cr metal loading, the total acidity of the Cr/HZSM-5 catalyst also rises proportionally. The 20% Cr/HZSM-5 catalyst exhibits the highest acidity, quantified at 4.776 mmol/g. This suggests that a greater amount of metal is incorporated into the pore structure of the HZSM-5 relative to other Cr/HZSM-5 catalysts, thereby providing a higher density of active acid sites.

2.2. Activity Comparison of HZSM-5 Catalyst with Different Cr Loading Ratios

The glucose conversion and downstream product distributions as a function of HZSM-5 catalyst with different Cr loading ratios are shown in Figure 4. The primary downstream products of glucose were FA and LA. The isomerization of glucose to fructose played a crucial role in the sequential conversion of glucose to LA. The interactions between the metal and oxygen atoms (Cr⁺³-OH) actively promoted the isomerization of glucose [33,34]. In the blank control experiment, the glucose conversion and LA yield only reached 56.8% and 20.1%, respectively. Yet, with the progressive increase in metal loading (Cr) on the Cr/HZSM-5 catalyst, both the glucose conversion and LA yield first increased and then declined. The highest values of LA yield of 36.5% and glucose conversion of 95.6% were reached when the catalyst was 15% Cr/HZSM-5. It indicated that an elevated number of acidic sites did not consistently promote the reaction process from glucose to LA. Excessive Cr, functioning as Lewis acid sites, could potentially expedite concurrent side reactions, leading to the degradation of glucose and reaction intermediates into by-products, such as humins [35].

2.3. Influence of Reaction Conditions on Catalytic Performance

The effects of 15% Cr/HZSM-5 catalyst on the conversion of glucose to LA under different reaction conditions were investigated. Research has demonstrated that temperature plays a crucial role in influencing the downstream products of glucose [36]. As shown in Figure 5A, both the desired product LA and the accompanying product FA exhibited a trend of initial increase followed by a subsequent decrease with the elevated temperature. Upon elevating the temperature from 140 °C to 180 °C, the yields of LA and FA exhibited an ascent from 4.8% and 19.2% to 38.7% and 58.3%, respectively. Simultaneously, glucose conversion rose from 24.7% to 97%, As the temperature increased further, the yields of LA and FA decreased. The maximal yields of LA and FA were obtained at 180 °C. This might because HMF generation was more favorable within the temperature range of 140–160 °C, and lower temperature was not conducive to the spontaneous advancement of this cascade reaction [37].

The influence of reaction time on the conversion of glucose to LA is illustrated in Figure 5B. It is evident that both fructose and HMF exhibited a gradual decrease when the reaction time was extended from 2 to 5 h, suggesting that these unstable intermediates undergo further rehydration to yield LA and FA (Figure 6). The yield of LA demonstrated a trend of initially increasing and subsequently decreasing with the prolongation of reaction time. This observation suggests that LA is more prone to form humic acid through aldol condensation, acetal cyclization, and dehydration processes under prolonged high-temperature conditions [38].

As shown in Figure 5C,D, the yields of both LA and FA exhibited a trend of initial increase followed by subsequent decrease as the glucose concentration was elevated along with the substrate concentration. This phenomenon could be attributed to the escalating concentration of glucose in the reaction system. As glucose concentration rose, the sole catalytic site offered by the catalyst became increasingly encapsulated by hydroxyl groups within the glucose. Consequently, this singular catalytic site became incapable of facilitating the subsequent dehydration of generated HMF, leading to the formation of LA and FA. The yield of HMF exhibited a gradual decline with the augmentation of catalyst dosage. This phenomenon could be attributed to the dynamics of glucose hydrolysis, wherein an increased quantity of catalysts supplied additional acidic sites. These sites played a pivotal role in sustaining the ongoing dehydration of HMF, leading to the production of LA. Additionally, as depicted in Figure 5C, there was nearly a 12% reduction in the LA yield when the substrate concentration escalated from 30 g/L to 50 g/L. This could be attributed to the heightened concentrations of glucose inducing more dynamic intermolecular movements and the generation of additional intermediates (1,2-enediols, etc.) These intermediates possessed a furan ring structure and were highly prone to generating undesirable by-products, such as humins. This occurred through intermolecular etherification reactions and electrophilic substitution of furfuryl alcohols [39].

Consequently, the optimal LA yield of 42.5% was achieved under conditions of 180 °C temperature, 5 h reaction time, 30 g/L substrate concentration, and 1 g catalyst dosage. The accompanying FA yield peaked at 52.1%. Calculations based on the equation revealed that the production of LA was concomitant with the production of an equivalent number of moles of FA. Throughout this reaction, the molar ratio of LA to FA was 13:16, where the surplus FA likely originated from the gradual degradation of glucose. In this reaction system, equilibrium was essentially attained in the reaction process, resulting in few by-products. Hence, the dosage of 15% Cr/HZSM-5 catalyst proved advantageous for the cascade conversion of glucose to LA.

2.4. Effect of Different Co-Solvents on the Conversion of Glucose

Figure 7 shows the effect of different co-solvents on glucose conversion and LA yield. The addition of GVL and THF as co-solvents had a boosting effect on the LA yield. In a pure aqueous system, the LA yield reached 42.5%. However, the incorporation of two polar nonprotonic solvents led to an enhancement in the LA yield, ca., 51.4% and 46.5%, respectively. The primary explanation for this lies in the ability of polar nonprotonic solvents to augment interaction with the reaction substrate, which can substantially increase the yield of LA. The introduction of acetonitrile as a co-solvent resulted in a reduction in LA yield, which is possibly attributed to the inherent differences in individual solvents. Furthermore, the inclusion of methanol and ethanol, serving as typical polar proton solvents, adversely impacted the yield of LA. A substantial distinction existed in the characteristics of these two solvent types. Polar nonprotonic solvents possessed the capability to accept electrons and exhibited a reduced propensity to form hydrogen bonds with the solute. Consequently, this accelerated the mass transfer rate and diminished the reaction energy barriers, thereby facilitating the reaction process. By contrast, polar protonated solvents were prone to accepting electrons, leading to the formation of hydrogen bonds with the solute and resulting in stronger interactions [40]. These interactions may potentially influence the catalytic efficiency of the overall reaction system. A comprehensive analysis conclusively identified GVL as the most effective solvent.

2.5. Effect of Different GVL Contents in Solvent Systems on the Conversion of Glucose

Figure 8 shows the pattern of the effect of different GVL contents in solvent systems on the conversion of glucose to LA. As the GVL content increased, the glucose conversion consistently remained at approximately 99% (±1%). The yields of LA and FA exhibited an initial increase followed by a decline, whereas the yield of HMF steadily increased. At a GVL/H₂O ratio of 20/80, the LA yield reached 51.4%, approximately 10% higher than that in the pure water system. This finding suggests that the addition of GVL effectively mitigates side reactions caused by the condensation of the target product with the reaction substrate. However, further increasing the GVL proportion failed to enhance the LA yield, likely due to GVL’s coating effect on HMF, which impedes the rehydration reaction. Moreover, due to glucose’s higher solubility in water, the use of water as a co-solvent enhances substrate solubility and reduces the viscosity of the reaction medium (GVL/H₂O), thereby accelerating the mass transfer rate [41]. Based on a comprehensive analysis, the optimal GVL/H₂O ratio was identified as 20/80.

2.6. Kinetic Study of the Response of GVL to Glucose to LA

Table 2. Kinetic parameters of glucose dehydration and HMF rehydration in H₂O and GVL/H₂O (1:4) solvent systems.

Solvent Systems	Tem. (°C)	Glucose Dehydration KG	R2	Ea (kJ/mol)	HMF Rehydration KH	R2	Ea (kJ/mol)
H2O	160	0.09578	0.96322	59.19	0.07813	0.95625	52.62
	180	0.20125	0.94285		0.15054	0.96437
	200	0.38468	0.97468		0.26869	0.99518
20%GVL/H2O	160	0.11245	0.9788	27.02	0.08351	0.99087	26.27
	180	0.13284	0.94533		0.13683	0.91705
	200	0.21312	0.97582		0.15417	0.98887

summarizes the kinetic parameters for glucose dehydration and HMF rehydration in various solvent systems over the 15% Cr/HZSM-5 catalyst. The reaction energy barrier (Ea) was subsequently determined from the Arrhenius equation. As indicated in Table 1, the reaction energy barrier for glucose dehydration in pure aqueous solvent was 59.19 kJ/mol, whereas that for HMF rehydration was 52.62 kJ/mol.

The specific role of GVL as a co-solvent in the conversion of glucose to LA has not been adequately elucidated. Researchers have classified the impacts of incorporating co-solvents into two categories: physical type and chemical type [42]. The former aimed to safeguard biomass molecules from undesirable side effects, while the latter enabled the co-solvents as catalysts to modify the reaction energy barriers or directly engage in the reaction. To deeply understand the influence of GVL co-solvent on the reactions of glucose dehydration and HMF rehydration, kinetic analyses of the solvent system with a GVL/H₂O ratio of 20/80 were conducted. It is evident from Table 1 that the energy barriers for glucose dehydration and HMF rehydration in GVL/H₂O solvent exhibited a substantial reduction, in comparison to those in pure water (27.02 kJ/mol < 59.19 kJ/mol and 26.27 kJ/mol < 52.62 kJ/mol). This implied that the incorporation of GVL had a positive impact on both glucose dehydration and HMF rehydration, thereby augmenting the yield of the desired product LA. Under the same temperatures, the glucose dehydration rate constant (K_G) in the GVL/H₂O system exhibited a significant increase compared to that in the pure water system. This further emphasized the promotional impact of GVL on glucose dehydration. In a word, the inclusion of GVL reduced the energy barrier for this secondary reaction, signifying that the action of GVL as a co-solvent in the production of LA from glucose primarily fell into the chemical category.

2.7. MD Simulation Analysis of Glucose to LA Under Solvent Action

Molecular dynamics simulations provided additional insights into the interaction between the target product LA and the solvents (H₂O, GVL/H₂O). The examination of the distribution, position, and arrangement of solvent molecules around the product involved calculating spatial density distribution maps. These maps, represented in red for GVL and blue for water, were generated around glucose (Figure 9). Additionally, the radial distribution functions (RDF, Figure 10) of solvent molecules’ centers of mass were calculated for further analysis.

Figure 9 illustrates the spatial density distribution of LA in both water and the mixed solvent system. As depicted in the figure, within the GVL/H₂O system, water molecules surrounding LA are relocated to the second solvation shell due to the competitive interactions between GVL and water. Within the first solvation shell, GVL primarily surrounds the carbonyl, hydroxyl, and carboxyl groups of LA, and this coordination plays a pivotal role in minimizing side reactions, such as esterification and condensation, which can occur with LA in the presence of excess water. Consequently, in the GVL/H₂O biphasic solvent system, the introduction of GVL provides a protective effect on the functional groups of LA while simultaneously reducing the mobility of glucose molecules within the reaction system. These synergistic effects facilitate the efficient catalytic conversion of glucose to LA. To elucidate the solvation shell layer more specifically, we simulated the solvation peaks across different solvent systems, as shown in Figure 10.

In a pure aqueous solvent, the solvation peak for water was 0.187 nm, whereas the solvation peak in the biphasic solvent reached 0.618 nm with the addition of 20% GVL to the system. The radial distribution function (RDF) curves between LA and the biphasic solvent exhibit a pronounced first solvation peak, indicating robust interactions between the biphasic solvent molecules and LA. This suggests that within the biphasic solvent system, GVL molecules actively compete with water to enter LA’s first solvation shell.

2.8. Catalyst Reusability

To evaluate the stability of the 15%Cr/HZSM-5 catalyst, recovery tests were conducted. As depicted in Figure 11A, the 15%Cr/HZSM-5 catalyst exhibited excellent reusability. After four reaction cycles under identical conditions, the catalytic activity remained largely unchanged. In the fifth cycle, the target product LA achieved a yield of 45.7%, representing a 4.7% decline compared to the fresh 15%Cr/HZSM-5 catalyst. Figure 11B shows that the FT-IR characterization of the catalyst after five cycles revealed characteristic peaks at 456 cm⁻¹, 546 cm⁻¹, and 794 cm⁻¹. However, the weakened intensity of these peaks after five cycles suggests partial structural degradation of the catalyst, thereby reducing its stability.

3. Experimental

3.1. Materials

Zinc Chloride (ZnCl₂, AR), Ferric Chloride (FeCl₃, AR), Chromium Chloride (CrCl₃, AR), Glucose (C₆H₁₂O₆, AR), Fructose (C₆H₁₂O₆, GR), 5-hydroxymethylfurfural (C₆H₆O₃, GR), Levulinic acid (C₅H₇O₃, GR), and HZSM-5 were obtained from Biotest Technology, Tianjin, China.

3.2. Catalyst Preparation

In this paper, a wet impregnation method was used to synthesize solid acid catalysts. For the preparation of the 15% Cr/HZSM-5 catalyst, 1.5 g of CrCl₃·6H₂O was accurately weighed and dissolved in 100 mL of deionized water to ensure complete dissolution. Subsequently, 10 g of HZSM-5 was precisely weighed and uniformly mixed with the solution. The resulting mixture was stirred at room temperature for 3 hours, followed by drying in an oven at 105 °C overnight. Finally, the dried mixture was calcined in air to yield a solid acid catalyst, which was heated to 550 °C at a rate of 2 °C/min and calcined for 6 h. The obtained solid acid catalyst was ground and passed through a 200-mesh sieve.

3.3. Catalytic Reactions

The conversion of glucose to LA was carried out in a 250 mL intermittent reactor equipped with a magnetic stirrer. At the beginning of the experiment, 1 g of glucose, 1 g of Cr/HZSM-5 catalyst, and 50 mL of water were added to the reactor. The optimal concentrations of glucose solution and catalyst dosage were subsequently determined under varying reaction conditions. The reaction was carried out at 140–220 °C for 1–7 h. After the reaction, the reactor was placed in ice water to stop the reaction. Each set of responses was repeated three times.

After the reaction, the catalyst was recovered by centrifugation and thoroughly washed with water and ethanol. The catalyst was subsequently dried overnight in an oven and reused in the next cycle.

3.4. Analytical Characterization

The product distribution in the reaction solution was detected by HPLC (LC-20A), with a refractive detector using a column (BioRad Aminex HPX-87H, 300 × 7.8 mm) at 65 °C and 0.05 M H₂SO₄ as the mobile phase at a flow rate of 0.6 mL/min.

The conversion of glucose, the yield of fructose, HMF, LA, and FA were calculated as following:

$$\mathrm{Glucose}\mathrm{conversion}=\left({\displaystyle \frac{\mathrm{mole}\mathrm{of}\mathrm{glucose}\mathrm{reacted}}{\mathrm{mole}\mathrm{of}\mathrm{initial}\mathrm{glucose}}}\right)\times 100\%$$

$$\mathrm{Products}\mathrm{yield}=\left({\displaystyle \frac{\mathrm{mole}\mathrm{of}\mathrm{fructose},\mathrm{HMF},\mathrm{LA},\mathrm{and}\mathrm{FA}}{\mathrm{mole}\mathrm{of}\mathrm{initial}\mathrm{glucose}}}\right)\times 100\%$$

3.5. Characterization Test

The morphology and elemental composition of Cr/HZSM-5 catalysts were examined using a high-resolution field-emission scanning electron microscope (SEM, TESCAN MIRA LMS) coupled with an energy-dispersive spectrometer (EDS, TESCAN MIRA LMS). The equipment was acquired by TESCAN, Brno, Czech Republic. The crystalline structure was analyzed using an X-ray diffractometer (XRD, Rigaku Smartlab 9KW) with a scanning range of 5° to 90° and a scanning speed of 10°/min. The facility was acquired by Nippon Rigaku Corporation, Akishima, Tokyo, Japan. Functional groups in the samples were identified using a Fourier-transform infrared spectrometer (FT-IR, Perkin Elmer Spectrum One) over a wavenumber range of 400–4000 cm⁻¹. The facility was acquired by PerkinElmer AG, Waltham, Massachusetts, USA. The crystalline structure was characterized using a Rigaku Smartlab 9KW X-ray diffractometer (XRD) with a scanning range of 5° to 90° at a speed of 10°/min. The X-ray source employed a rotating copper anode operating at 40 kV and 40 mA. The facility was acquired by Nippon Rigaku Corporation, Akishima, Tokyo, Japan. Metal concentrations in the catalysts were detected using an inductively coupled plasma emission spectrometer (ICP-OES, Agilent 5110). The device was acquired by Agilent Technologies, Inc. of California, USA. The specific surface area and pore volume of the prepared catalysts were detected using a specific surface area and porosity analyzer (ASAP 2020 PLUS). The device was acquired by Micromeritics of Norcross, GA, USA. Metal valence states in the catalysts were detected using x-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha). The facility was acquired by Thermo Fisher Scientific, Waltham, MA, USA.

3.6. Molecular Dynamics Simulation

Molecular dynamics (MD) simulations were executed employing the GROMACS 2021 package. Files pertaining to formic acid (FA) and LA molecules were acquired by Sobtop and employed in subsequent MD simulations. Periodicboundarycondition (PBC) was used for all three dimensions of the system. Modeling of FA and LA molecules involved the Generalized Amber Force Field (GAFF) force field, while the water molecule was optimized by the TIP4P model. Upon the completion of system construction, an energy minimization process was executed by using the steepest descent method for a duration of 5000 steps and the conjugate gradient method for an additional 5000 steps. This meticulous approach ensured the conventional structure and rational geometric configuration for molecules. After the system was at the energy minimization point, the real dynamics simulation could be started. A 1 ns NVT ramp-up kinetics phase was performed initially, followed by a 1 ns NPT balancing phase, and culminating in a 50 ns NPT synthesis simulation. The Leapfrog method with a time step of 2 fs was employed as the integral solution algorithm. Long-range electrostatic interactions were calculated using the Cut-Off method, employing a truncation distance of 1.0 nm for non-bonded interactions. Pressure coupling was achieved through the isotropic Parrinello–Rahman method, maintaining a controlled pressure of 1 bar. Temperature coupling was analyzed by the Velocity-rescale heat bath method. All bonds were constructed by the Linear Constraint Solver (LINCS) algorithm. The simulation trajectory data with an interval of 2 ps were analyzed by GROMACS 2021.6 software package, and visualized by VMD 1.9.3 software [43].

4. Conclusions

In this study, we synthesized Cr/HZSM-5 catalysts with varying Cr loadings to evaluate their catalytic performance in the conversion of glucose to LA. It was found that at a reaction temperature of 180 °C, a reaction time of 5 h, a catalyst dosage of 1 g, and a glucose concentration of 30 g/L, the glucose conversion reached 99% and the LA yield was 42.5%. Building on this foundation, the impact of solvent enhancement on the catalytic performance of Cr/HZSM-5 catalysts was investigated, identifying GVL as the most effective solvent. The addition of 20% GVL into the solvent system led to an LA yield of 51.4%, representing an 8.9% increase compared to using water as the solvent. Reaction kinetics and molecular dynamics simulations were conducted to further elucidate the mechanism of solvent potentiation. Based on reaction kinetics calculations, it was demonstrated that the solvent system composed of GVL and water reduced the activation energy barrier for the conversion of glucose to LA, compared to the pure water system (78.89 kJ/mol < 86.21 kJ/mol). Molecular dynamics simulation results indicate that the incorporation of GVL forces water molecules in the solvent system into the secondary solvation shell, thereby diminishing the presence of water around the LA. Concurrently, GVL predominantly localizes around the carbonyl, hydroxyl, and carboxyl groups of LA, thereby minimizing the likelihood of side reactions and consequently enhancing the LA yield.