Construction of Modified Silica Gel Catalysts and Their Enhancement of Fructose Dehydration for 5-HMF Production

Abstract

To address the challenges of difficult recovery, significant environmental hazards associated with homogeneous catalysts, and insufficient catalytic activity of heterogeneous supports in the catalytic dehydration of fructose to produce 5-hydroxymethylfurfural (5-HMF), this study employs a straightforward nitric acid modification method to prepare an acid-activated silica gel catalyst for application in this reaction system. Through systematic investigation of the influence of modification conditions on catalyst performance and economic benefits, optimal reaction conditions were determined: DMSO as the solvent, nitric acid-modified silica gel as the catalyst, a reaction temperature of 120 °C, a solid–liquid ratio of 1:30 (g∙mL⁻¹), and a fructose-to-catalyst mass ratio of 1:1. Under these conditions, the maximum 5-HMF yield reached 91.6%. Characterization via specific surface area, pore size analysis, and acid/base site characterization (NH₃-TPD) revealed that nitric acid modification preserved the silica gel’s pore structure. Through oxidative cleaning, etching to expose silanol groups, and inducing surface defects, this process significantly increased the number of acid sites on the silica gel surface, thereby enhancing catalytic activity. This study presents a low-cost, easily recoverable, and environmentally friendly heterogeneous catalytic strategy for the efficient conversion of fructose into 5-HMF. It also provides experimental guidance for the targeted functionalization of silica-based catalytic materials, holding significant implications for advancing the high-value utilization of biomass resources.

1. Introduction

The advancement of global sustainable development strategies and the widespread adoption of green chemistry have made the substitution of fossil fuels with renewable biomass resources a core focus in chemical engineering and biomass conversion research [1,2,3]. 5-Hydroxymethylfurfural (5-HMF), with its functionalized structure of an aldehyde, hydroxymethyl and furan ring, has become a pivotal ‘bridge molecule’ linking biomass conversion and petrochemical industries. Through targeted synthesis via oxidation, hydrogenation, and condensation reactions, it yields key fine chemicals and polymer monomers, such as 2,5-furanedicarboxylic acid (FDCA) and 2,5-dimethylfuran (DMF), showing industrial potential in bio-based polyesters and green fuels [4,5,6].

Among existing synthetic pathways for 5-HMF, the catalytic dehydration reaction using carbohydrates such as fructose is widely recognized as one of the most direct and efficient methods. This is due to its streamlined reaction steps and excellent atom economy [7]. However, this reaction system still faces two core bottlenecks in practical industrial applications. Firstly, the polyhydroxy structure in fructose molecules is prone to degradation, cross-linking polymerization, and other side reactions under the high-temperature conditions required for the reaction. This generates soluble polymers or insoluble humin-like substances, significantly reducing the selectivity of 5-HMF [8]. Secondly, the homogeneous acid catalysts currently in widespread use (including inorganic acids such as H₂SO₄ and HCl, as well as organic acids like p-toluenesulfonic acid) exhibit high catalytic activity. However, they present significant environmental and engineering challenges, including severe equipment corrosion, difficulty in catalyst recovery and recycling, complex product separation and purification, and the generation of substantial volumes of acidic wastewater post-reaction. These issues severely constrain their large-scale industrial deployment [9,10,11].

To address the aforementioned challenges, the development of heterogeneous catalysts that combine high catalytic efficiency, excellent chemical stability, and recyclability has become a research focus in the field of 5-HMF synthesis. Silica gel, as a typical porous adsorbent material, is characterized by its high specific surface area and open pore structure. Its chemical formula is typically represented as mSiO₂·nH₂O, where ‘n’ denotes the variable amount of water in the structure. It is prepared through the partial hydrolysis and condensation of silica compounds, followed by washing, drying, and heating to remove excess water. At the microstructural level, the framework of silica gel consists of Si-O tetrahedra as fundamental building blocks. These tetrahedra are distorted and arranged in an amorphous, disordered manner, with the interstitial spaces between the tetrahedral vertices forming the material’s pore network. This structural arrangement provides essential support for its adsorption capabilities and carrier functions. It is important to note that the water content in the chemical formula primarily exists as bound water [12]. Specifically, -OH form chemical bonds with silicon atoms and uniformly coat the silica gel surface. The presence of surface hydroxyl groups critically influences the surface reactivity of silica gel and subsequent functionalization modification processes. Furthermore, silica gel exhibits physical and chemical properties such as insolubility in water and common organic solvents, chemical stability, and being non-toxic and odourless. These advantages enable its widespread use as a catalyst support in the field of catalysis, providing reliable assurance for the efficient loading of active components and the smooth progression of catalytic reactions [13,14,15,16,17,18]. However, it should be noted that the surface of unmodified silica gel possesses few acidic sites with relatively weak acidity (predominantly weak Brønsted acids), failing to meet the stringent requirements for a strongly acidic catalytic environment in fructose dehydration reactions. Consequently, the rational chemical modification to construct highly efficient acidic sites on its surface constitutes the core challenge in preparing high-performance silica-based heterogeneous catalysts. Nitric acid, an inorganic acid possessing both strong oxidizing and strong acidifying properties, has been validated in carbon material surface modification studies. It can effectively increase oxygen-containing functional groups (such as -COOH and -OH) on carbon surfaces through oxidative etching. Drawing upon this approach, this study proposes utilizing nitric acid for liquid-phase modification of silica gel. This process is anticipated to achieve dual objectives: Firstly, nitric acid can cleanse the silica gel surface through oxidation, removing adsorbed organic impurity molecules and residual salts within the pores (such as sodium or potassium salts left over from the preparation process), while simultaneously mildly etching the silica gel framework to expose more surface -SiOH [19]. Secondly, the strong oxidizing power of nitric acid induces structural defect sites on the silica gel surface. These defects not only enhance the Brønsted acid strength of surrounding silanol groups but may also generate coordination-unsaturated Si⁴⁺ Lewis acid sites, ultimately constructing abundant and highly efficient acidic catalytic sites on the silica gel surface [20].

Based on the aforementioned research approach, this study aims to employ a straightforward nitric acid modification method to prepare a novel acidic activated silica gel catalyst and apply it in the reaction system for the catalytic dehydration of fructose to produce 5-HMF. This study will systematically investigate the effects of conditions such as nitric acid concentration, modification temperature, and modification time on the catalyst’s surface acidity (characterized by NH₃-TPD to determine acid site quantity and strength) and catalytic performance (5-HMF yield, selectivity, and catalyst cycling stability). Concurrently, key reaction system parameters (e.g., solvent type, reaction temperature, and reaction time) will be optimized. This research not only provides a low-cost, easily recoverable, and environmentally friendly heterogeneous catalytic strategy for the efficient conversion of fructose into 5-HMF but also offers theoretical foundations and experimental references for the targeted design and functional modification of silica-based catalytic materials. It holds significant importance for advancing the high-value utilization of biomass resources.

2. Results

2.1. BET and TPD Analysis

This study employed macroporous silica gel with a mesh size of 30–60 mesh as the primary experimental material. Its pore structure parameters were characterized using the nitrogen adsorption–desorption method (BET method), with results presented in

Table 1. The characterization of the catalysts based on silica gel.

Types of Catalysts	Macroporous Silica Gel	Activated Silica Gel	Aminated Silica Gel
Specific surface area (m2∙g−1)	372	376	338
Average pore size (nm)	12.3	12.1	12.5
Total acidity (mmol∙g−1)	0.009	0.037	0

: The raw silica gel exhibited a high specific surface area of 372 m²∙g⁻¹ and an average pore diameter of 12.3 nm, significantly outperforming the pore structure characteristics of typical silica-alumina molecular sieve catalysts (conventional silica-alumina molecular sieves typically possess specific surface areas below 300 m²∙g⁻¹ and average pore diameters less than 5 nm). However, analysis of its acidity via ammonia temperature-programmed desorption (NH₃-TPD) revealed an extremely low total acidity of merely 0.006 mmol∙g⁻¹, rendering it inadequate to satisfy the active site requirements for acid-catalyzed reactions.

Following nitric acid activation of pristine silica gel, BET analysis showed that its specific surface area and average pore diameter remained essentially unchanged, indicating that the treatment did not damage the pore structure. However, NH₃-TPD measurements revealed a pronounced increase in the total acidity of the activated silica gel. This enhancement can be attributed to two main factors. First, nitric acid effectively removes adsorbed organic impurities and residual salts originally present on the silica surface and within the pore channels. These impurities occupy potential acidic sites and hinder the interaction between NH₃ and surface silanols, thereby contributing to the low measurable acidity of the untreated silica gel. Second, under strongly acidic conditions, partial activation and hydrolysis of surface Si-O bonds can occur, generating additional -SiOH groups. Since silanols are typical weak Brønsted acid sites, an increase in their population directly leads to higher overall acidity.

According to the literature, silica surfaces contain several types of silanols—isolated or vicinal silanols, hydrogen-bonded silanols, and the less common bridging silanols—each exhibiting distinct acid strengths. Hydrogen-bonded silanols generally possess stronger Brønsted acidity than isolated silanols, whereas true bridging silanols are predominantly associated with crystalline aluminosilicates (e.g., zeolites) and are rarely present on amorphous silica gel [21].

In this work, the nitric acid-activated silica gel exhibited two NH₃ desorption regions at 50–150 °C and 300–500 °C, corresponding to weak and moderately strong acid sites, respectively (Figure 1). The appearance of these moderately strong acid sites suggests that HNO₃ activation not only increases the number of surface silanols but also introduces defect-associated or hydrogen-bonded silanols with higher acidity. These newly formed or exposed acidic sites are likely responsible for the significantly improved catalytic performance of the activated silica gel in fructose dehydration.

Following surface modification of silica gel with the silane coupling agent KH-550, the alkaline -NH₂ within the KH-550 molecule successfully grafted onto the silica gel surface. This transformation altered the surface properties from acidic to alkaline, consequently yielding no detectable acidic signals in subsequent NH₃-TPD testing. To confirm its alkaline character, analysis was conducted via carbon dioxide programmed temperature desorption (CO₂-TPD) testing. The results revealed that the modified silica gel exhibited a total base content of 0.048 mmol∙g⁻¹, with a single desorption peak observed only near 160 °C. Considering the correlation between CO₂-TPD desorption peak temperature and base strength (low-temperature desorption peaks typically correspond to weakly basic sites), this indicates that only weakly basic sites exist on the surface of this amino-modified silica gel. Furthermore, BET testing results confirmed that the specific surface area and average pore size of the modified silica gel remained largely unchanged compared to the original silica gel. This indicates that the KH-550 surface modification process did not disrupt the pore structure of the silica gel.

2.2. Effect of Catalyst Type on 5-HMF Yield

Different catalysts exhibit varying promotional or inhibitory effects on the yield of 5-HMF. This section compares the catalytic performance of the prepared 65 wt% HNO₃, native silica gel, aminated silica gel and activated silica gel, as shown in Figure 2a. Under HNO₃ catalysis, 5-HMF reached a peak yield of 76.1% after 2 h of fructose reaction, followed by a rapid decline in 5-HMF concentration. This indicates reduced 5-HMF stability within the nitric acid catalytic system. Faranak’s research also indicates that acidic environments can promote the formation of 5-HMF, but excessively high acidity reduces its stability [22]. Furthermore, the liquid nature of nitric acid hinders its separation for recycling. The initial reaction rate under native silica gel is relatively slow, with a maximum 5-HMF yield of 43.9% achieved after a prolonged reaction time of 9 h. The initial reaction rate under native silica gel was low due to its limited surface acidity. After reacting for a period in DMSO solvent at 120 °C, some soluble adsorbates desorbed, increasing the acidity of the native silica gel. Consequently, Figure 2 shows a marked increase in the 5-HMF formation rate after 10 min of reaction. Subsequently, as the fructose concentration decreased, the 5-HMF formation rate began to decline. Under the action of amino silica gel, the 5-HMF yield is only 5%. This may be attributed to the relatively high pH of the system, which hinders the progression of the dehydration process. Under activated silica gel catalysis, the 5-HMF yield from fructose reached a maximum of 89.7% after 2 h of reaction. This yield surpassed those achieved with the other three catalysts, and 5-HMF remained stable within this reaction system. This stems from the weak acid sites on the silica gel surface effectively catalyzing the conversion of fructose to 5-HMF. The absence of high H⁺ concentrations within the system further permits stable 5-HMF retention, constituting a primary factor for achieving elevated 5-HMF yields [23]. The effects of aminated silica gel, nitric acid solution and activated silica gel also can preliminarily explain the successful preparation of activated silica gel and the environment required for dehydration reaction is weak acid.

2.3. Effect of Solvent on 5-HMF Yield

In addition to the catalyst, the type of solvent is also a significant factor influencing the yield and separation of 5-HMF. The furanose form, such as fructose, is more stable in non-protonic solvents [24]. Therefore, the primary focus of this section of the experiment was to investigate the effect of the interaction between activated silica gel and organic solvents on the yield of 5-HMF. As shown in Figure 2b, activated silica exhibits the highest catalytic activity in DMSO solvent, whereas 5-HMF yields detected in DMAc and methyl ethylene glycol were both below 5%. In the NMP solvent system, the 5-HMF yield progressively increased with reaction time. However, after 300 min of reaction, the 5-HMF yield from fructose remained only around 30%, significantly lower than that achieved in DMSO. Consequently, DMSO is identified as the optimal solvent for activated silica gel-catalyzed 5-HMF production from fructose. This result can be attributed to DMSO’s high polarity, which effectively dissolves fructose and its intermediates, thereby promoting reaction progress. It may also aid in dissolving the generated 5-HMF, thereby reducing side reactions. In contrast, the lower polarity of ethylene glycol monomethyl ether and DMAc fails to provide sufficient solvency, limiting reaction efficiency. NMP exhibits polarity intermediate between DMSO and DMAc, thereby moderately enhancing reaction efficiency and yielding a 5-HMF yield second only to DMSO [25,26].

2.4. Synergistic Effect of Temperature and Catalyst Dosage on 5-HMF Yield

To further enhance reaction efficiency and target product yield, key reaction parameters such as the mass ratio of substrate to catalyst and the solid–liquid ratio require further optimization. This section of optimization experiments examined the influence of substrate/catalyst mass ratio at different temperatures to identify the optimal temperature and ratio combination that balances economic viability with high efficiency. As shown in Figure 3a, at 140 °C, increasing the catalyst loading significantly accelerated the initial formation rate of 5-HMF, manifested as a gradual increase in initial yield. Simultaneously, the maximum yield of 5-HMF also exhibited an upward trend with increasing catalyst dosage. Specific data indicates that when the mass ratio of fructose to catalyst was progressively adjusted from 4:1 to 2:1, the maximum 5-HMF yield increased from 82.03% to 89.7%, representing an improvement of 7.67 percentage points; However, when this ratio was further adjusted to 1:2, the maximum 5-HMF yield increased only marginally to 90.3%, with the improvement rate significantly narrowing (by merely 0.6 percentage points). It is noteworthy that while higher catalyst loading promotes 5-HMF formation, it also accelerates side reactions (such as polymerization and degradation) during the later stages of the reaction at elevated temperatures. This reduces the stability of 5-HMF in the system, and the resulting byproducts may clog the surface of the activated silica gel, thereby diminishing its activity. Based on these findings, and considering that appropriately increasing catalyst dosage can enhance the maximum 5-HMF yield, experiments further explored optimizing 5-HMF yield by raising catalyst dosage at lower temperatures. Figure 3b,c show the effects of catalyst dosage on 5-HMF yield at 120 °C and 100 °C, respectively. Comparing the data from both figures reveals that at 100 °C, the slower fructose conversion rate necessitates a longer reaction time to reach equilibrium. Consequently, increasing the catalyst dosage yields only marginal improvements in 5-HMF yield at this temperature, demonstrating limited optimization potential. At 120 °C, however, when the fructose-to-catalyst mass ratio was 1:1, the 5-HMF yields at 20 min, 40 min, and 60 min reached 91.3%, 96.3%, and 98.9%, respectively. This represented a significant improvement compared to the yields obtained at this temperature with fructose-to-catalyst ratios of 4:1 and 2:1. Further increasing the catalyst dosage (adjusting the mass ratio to 1:2) only marginally elevated the maximum 5-HMF yield from 98.9% to 99.2%, offering negligible improvement potential while incurring higher catalyst costs and increased risk of side reactions. Comprehensively evaluating the impact of reaction temperature and catalyst loading on 5-HMF yield, alongside reaction economics and product stability, this study concludes that within the activated silica gel catalytic system, a reaction temperature of 120 °C and a fructose-to-catalyst mass ratio of 1:1 represent optimal conditions balancing 5-HMF yield, reaction efficiency, and cost. The solid-to-liquid ratio at this point is 1:60.

2.5. Effect of Substrate Concentration on 5-HMF Yield

During the preparation of 5-HMF, substrate concentration primarily influences mass transfer rates by altering the viscosity of the reaction system, with variations in mass transfer efficiency subsequently affecting 5-HMF yield [27]. From an industrial production perspective, maximizing substrate concentration while maintaining target product yield represents a key strategy for reducing production costs and enhancing efficiency. Consequently, investigating the impact of substrate concentration on 5-HMF yield holds significant practical importance [28]. To elucidate the relationship between fructose concentration (represented by the solid–liquid ratio of fructose to organic solvent) and 5-HMF yield, this study varied the substrate concentration within the reaction system by adjusting the fructose dosage while maintaining a constant catalyst concentration. The resulting effects on 5-HMF yield are presented in Figure 3d. As depicted in Figure 3d, the 5-HMF yield exhibited a progressive decline with increasing fructose concentration (i.e., rising solid–liquid ratio). This phenomenon may arise from increased system viscosity at high fructose concentrations, elevating mass transfer resistance. This not only reduces the contact efficiency between reactive sites and substrates but may also promote side reactions such as fructose self-polymerization and 5-HMF repolymerization, ultimately diminishing the 5-HMF yield [29]. At a solid–liquid ratio of 1:6 (g∙mL⁻¹) between fructose and organic solvent, the 5-HMF yield barely exceeded 70%. As the solid–liquid ratio decreased (i.e., fructose concentration decreased), the 5-HMF yield significantly improved. At a solid–liquid ratio of 1:60 (g∙mL⁻¹), the 5-HMF yield reached 98.9%; even at 1:30 (g∙mL⁻¹), the yield remained at a high level of 91.6%. Comprehensively evaluating factors including 5-HMF yield, reaction time (higher substrate concentrations reduce processing time), and substrate concentration (higher concentrations better align with industrial-scale production requirements), the optimal solid–liquid ratio of fructose to organic solvent under these experimental conditions was determined to be 1:30 (g∙mL⁻¹): This condition ensures an ideal 5-HMF yield of 91.6% while maintaining a relatively high fructose concentration, balancing product yield with industrial applicability. Based on the above analysis, the optimal reaction conditions are established as follows: solvent system DMSO, catalyst activated silica gel, temperature 120 °C, solid–liquid ratio 1:30 (g∙mL⁻¹), fructose/catalyst ratio 1:1. Under these conditions, a maximum yield of 91.6% can be achieved. This yield exceeds the highest yields obtained by Hou [30] and Huang [31] (81% and 73.71%, respectively).

2.6. Recyclability of Catalysts

For the industrial application of catalytic reaction systems, alongside reaction condition optimization and target product yield, the catalyst’s service life and recyclability are equally pivotal evaluation metrics, directly impacting production costs and process feasibility [32]. Consequently, this study investigated the recycling performance of the activated silica gel catalyst under the optimal reaction conditions identified above (120 °C, fructose-to-catalyst mass ratio of 1:1, solid-to-liquid ratio of 1:30). The procedure involved: after the reaction concludes, the spent catalyst is filtered and vacuum-dried at 80 °C for 5 h before being directly reused in the subsequent catalytic cycle. This process is repeated to evaluate its activity stability. This process was repeated to assess its activity stability. As shown in Figure 4, when the activated silica gel catalyst was directly recycled after simple separation and drying, its catalytic activity exhibited a marked decline with increasing number of cycles. This decrease accelerated significantly after five recycling cycles. Accompanying this decline in catalytic activity, the catalyst’s surface colour progressively darkened. Considering the characteristics of the catalytic reaction system, the observed decrease in activity from 91.6% to 71.6% is likely attributable to the formation of humins on the catalyst surface. During the reaction, polymeric by-products and carbon precursor compounds can adsorb onto the silica gel, gradually accumulating into a humins layer that blocks active sites and thus diminishes the overall catalytic capability. However, after simple ethanol washing followed by drying, the catalyst exhibits a noticeable recovery in activity, indicating that the deactivation is primarily due to reversible humin deposition rather than permanent structural degradation of the catalyst.

2.7. Regeneration of Catalyst Activity

To validate the aforementioned hypothesis and investigate the possibility of catalyst regeneration, this study subjected deactivated silica gel catalysts after five cycles of reuse to a regeneration process: the deactivated catalysts were placed in ethanol and boiled, filtered, and this procedure was repeated three times. Following drying, the catalysts were reused for the preparation of 5-HMF from fructose under optimal conditions. Experimental results demonstrated that the regenerated catalyst restored the 5-HMF yield to 90.8%, approaching the catalytic efficiency of the fresh catalyst. This outcome confirms that the deactivation of the activated silica gel catalyst does not stem from structural damage to the active components. Rather, it primarily arises from the adsorption of by-products generated during the reaction (as opposed to simple carbon fouling) onto its surface. These adsorbed substances block the active silanol sites on the catalyst surface, hindering effective contact between fructose and the active sites, ultimately leading to a decline in catalytic activity. Boiling ethanol treatment effectively desorbs these surface-adsorbed substances, thereby restoring the catalyst’s activity.

2.8. Reaction Kinetics for the Preparation of 5-HMF from Fructose on Silica Gel Catalyst

Having established the influence of solvent systems on conversion efficiency, further attention must be directed towards the reaction pathway characteristics governing fructose conversion to 5-HMF. This transformation fundamentally constitutes a series of reactions: under catalytic action, fructose undergoes a directed conversion to the target product 5-HMF while concurrently undergoing multiple side reactions. Notably, 5-HMF exhibits high reactivity, readily undergoing hydrolysis to yield acetoacetic acid (LA) and formic acid (FA). Concurrently, both fructose and 5-HMF may polymerize to form humins primary cause of significant colour darkening observed in most reaction systems during later stages [33].

Regarding the kinetic characteristics of fructose and 5-HMF conversion, a relatively strong consensus has emerged within the academic community: most researchers believe that both the conversion of fructose to 5-HMF and the subsequent side reactions of 5-HMF (hydrolysis, polymerization) conform to first-order reaction kinetics models [34]. This conclusion has been supported by experimental verification from some researchers. Consequently, to focus on core research objectives, this paper refrains from conducting experiments to determine the reaction order in the kinetic investigation section. Instead, drawing upon existing research consensus, the processes of ‘fructose conversion to 5-HMF’ and ‘5-HMF undergoing side reactions’ are simplified into a single-step reaction for kinetic modelling and data analysis. This approach streamlines computational processes while highlighting the research value of key kinetic parameters. The reaction model for fructose dehydration can be expressed as:

The degradation reaction of 5-HMF may also be treated as a first-order reaction, in which case its rate equation can be expressed as:

$${\displaystyle \frac{d{C}_{HMF}}{dt}}=k_3·C_{HMF}$$

(1)

$${\displaystyle \frac{d{C}_{fru}}{dt}}=k_0·C_{fru}$$

(2)

$${\displaystyle \frac{d{C}_{HMF}}{dt}}=k_1·C_{fru}-k_3·C_{HMF}$$

(3)

$$k_0=k_1+k_2$$

(4)

At a given temperature, k₁, k₂, and k₃ are all constant values.

From Equation (1), we obtain:

$$-ln\left(1-x_3\right)=k_3·t$$

(5)

In the equation, x₃ denotes the conversion rate of 5-HMF. Performing linear regression of $-ln\left(1-x_3\right)$ against t yields a slope k₃.

From Equations (2) and (3), we obtain:

$$-ln\left(1-x_0\right)=k_0·t$$

(6)

$$Y_{HMF}={\displaystyle \frac{k_1}{k_3-k_0}}\left(e^{-k_0·t}-e^{-k_3·t}\right)$$

(7)

In the equation, x₀ and Y_HMF denote the conversion rate of fructose and the yield of 5-HMF, respectively. Linear regression of $-ln\left(1-x_0\right)$ against t yields k₀, while linear regression of Y_HMF against ${\displaystyle \frac{1}{k_3-k_0}}\left(e^{-k_0·t}-e^{-k_3·t}\right)$ yields k₁. k₂ is obtained according to Equation (4).

Taking the logarithm of both sides of the Arrhenius equation $k=A·e^{-{\displaystyle \frac{Ea}{RT}}}$ yields:

$$\mathit{ln}k=-{\displaystyle \frac{Ea}{RT}}+lnA$$

(8)

The activation energy Ea and pre-exponential factor A were determined by linear regression of ln k versus ${\displaystyle \frac{1}{RT}}$, based on the k values calculated from

Table 2. The reaction rate constants of each reaction pathway of fructose degradation at different temperatures.

Temperature (°C)	k0 (min−1)	k1 (min−1)	k2 (min−1)	k3 (min−1)
100	0.03265	0.02953	0.00312	8.079 × 10−7
120	0.08326	0.07339324	0.00987	5.680 × 10−6
140	0.1453	0.126216	0.01911	1.652 × 10−5
160	0.2796	0.238565529	0.04102	4.924 × 10−5

, as shown in Table 2.

As demonstrated by the calculated reaction rate constants presented in Table 2, the rate constant for the conversion of fructose to 5-HMF increases significantly with rising temperature, corresponding to a concurrent acceleration in reaction rate. Simultaneously, the rate constants for both the degradation of 5-HMF and the conversion of fructose to humic acid also exhibit an upward trend, leading to a synchronous increase in the rates of these two side reactions. This phenomenon can be quantitatively explained by the Arrhenius equation ($k=A·e^{(-E_a/RT)}$): the greater the activation energy, the more sensitive the reaction rate becomes to temperature. As the reaction temperature increases, the absolute value of the exponential term ($-E_a/RT$) decreases, causing the reaction rate constant k to increase exponentially. As shown in

Table 3. Activation energy and pre-exponential factor.

Activation Energy, Ea (kJ∙mol−1)			Pre-Exponential Factor, A (min−1)
Ea1	Ea2	Ea3	A1	A2	A3
45.88	56.55	90.47	8.24 × 104	2.79 × 105	4.48 × 106

, the activation energy for the 5-HMF degradation reaction is the highest. Compared to the other two reactions with lower activation energies, k₃ increases by two orders of magnitude, representing a significant change. From a reaction mechanism perspective, elevated temperatures intensify molecular thermal motion, substantially increasing the proportion of molecules reaching the activation energy threshold within the system [35]. This enhances the frequency of effective collisions, thereby concurrently accelerating multiple thermally activated reactions: fructose dehydration to form 5-HMF, the subsequent degradation of 5-HMF, and the incorporation of fructose into humic acid formation.

Figure 5a,b present the conversion rate curves of 5-HMF and fructose at different temperatures, fitted using the data from Table 3. It can be observed that the experimental values show good correlation with the fitted curves. Based on the data from Table 3, the activation energies and pre-exponential factors for the primary and secondary reactions converting fructose to 5-HMF under conditions employing 0.5 g of activated silica gel catalyst were determined. Consequently, the rate equations for each reaction involved in the conversion of fructose and 5-HMF can be expressed as follows:

$$k_1 = 8.24 \times {10}^4{·C}_{\mathit{fru},0}·{\mathit{e}}^{\left({\displaystyle \frac{-45.88 \times 1000}{\mathit{RT}}}\right)}$$

(9)

$$k_2=2.79 \times {10}^5{·C}_{\mathit{fru},0}·{\mathit{e}}^{\left({\displaystyle \frac{-56.55 \times 1000}{\mathit{RT}}}\right)}$$

(10)

$$k_3=4.48 \times {10}^6{·C}_{\mathit{HMF},0}·{\mathit{e}}^{\left({\displaystyle \frac{-90.47 \times 1000}{\mathit{RT}}}\right)}$$

(11)

2.9. Catalytic Mechanism of Activated Silicas

The mechanism of the reaction for preparing 5-HMF from fructose, catalyzed by silica gel, can be summarized as shown in Figure 6. First, a hydrogen proton from the Si-OH on the silica gel surface migrates to the carbonyl oxygen of fructose. This protonates the carbonyl oxygen to form a hydroxyl group, while the silica gel oxygen loses a proton and becomes negatively charged. Subsequently, this negatively charged silica gel oxygen abstracts a proton from the terminal carbon atom of the fructose molecule, completing the proton transfer cycle. During the subsequent multi-step dehydration process, silica gel continues to drive progressive dehydration of intermediates through analogous proton transfer mechanisms-namely, the release and reacquisition of protons from the silanol group-ultimately yielding 5-HMF with high efficiency. This mechanism aligns with the solid acid-catalyzed fructose dehydration pathways extensively reported in the literature, particularly emphasizing the pivotal role of Brønsted acid sites in facilitating proton transfer and dehydration cyclization. Notably, silica gel not only provides abundant surface acidic sites but also exhibits a pore structure that exerts a confining effect on reactants and intermediates. This enhances the selectivity of the reaction pathway and minimizes side reactions. Consequently, silica gel fulfills dual functions in this transformation: supplying protons and providing structural guidance. This provides a theoretical basis for the further design of highly efficient and stable catalysts for fructose dehydration.

3. Materials and Methods

3.1. Experimental Materials

Silica microspheres (30–60 mesh, purchased from Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), nitric acid (analytical grade, Sinopharm), distilled water (laboratory-prepared), 3-Aminopropyltriethoxysilane (KH-550, analytical grade, purchased from Shanghai Aladdin Biochemical Technology Co., Ltd.), Molecular Sieves (Sinopharm Chemical Reagent Co., Ltd., Shanghai, China), Toluene (analytical grade, purchased from Shanghai Aladdin Biochemical Technology Co., Ltd.), Acetone (analytical grade, purchased from Beijing Chemical Factory Co., Ltd., Beijing, China), Nitrogen gas (≥99.99%, purchased from Beijing Haipu Gas Co., Ltd., Beijing, China), D-fructose (analytical grade, purchased from Shanghai Aladdin Bio-Chem Technology Co., Ltd.), dimethyl sulfoxide (DMSO, analytical grade, purchased from Shanghai Aladdin Bio-Chem Technology Co., Ltd.), ethylene glycol monomethyl ether (analytical grade, purchased from Shanghai Aladdin Bio-Chem Technology Co., Ltd.), dimethylacetamide (DMAc, analytical grade, purchased from Shanghai Aladdin Bio-Chem Technology Co., Ltd.), N-methylpyrrolidone (NMP, analytical grade, purchased from Shanghai Aladdin Bio-Chem Technology Co., Ltd.), concentrated sulphuric acid (analytical grade, China National Pharmaceutical Group Chemical Reagent Co., Ltd., Beijing, China), methanol (chromatography grade, purchased from Shanghai McLean Bio-Chem Technology Co., Ltd., Shanghai, China).

3.2. Preparation of Activated Silica Gel

Place 30–60 mesh microsphere silica gel in a 4 mol∙L⁻¹ nitric acid solution, degas by ultrasonication, then immerse. Subsequently, filter the silica gel and wash repeatedly with deionised water until the filtrate is neutral to remove residual acid. Thereafter, place the silica gel in deionised water once more, treat with ultrasonication, then immerse overnight. Following filtration and washing, the material was vacuum-dried for 5 h. The resulting sample was stored for subsequent use.

3.3. Preparation of Aminated Silica Gel

Add 10 g of KH-550 and 50 mL of toluene (dehydrated by molecular sieve and degassed by ultrasonication) to a 100 mL three-neck flask. Mix thoroughly under magnetic stirring Subsequently, add 10 g of activated silica gel to the system and react at 90 °C under a nitrogen atmosphere for 24 h. After cooling the reaction mixture, filter to separate the solvent. Thoroughly wash the solid product with acetone to remove unreacted KH-550 and impurities, then vacuum dry at 100 °C for 4 h to obtain the final product.

3.4. Catalyst Characterization

The specific surface area and pore size distribution of the samples were determined using a Thermo Electron Sorptomatic 1990 surface analyser (Thermo Fisher Scientific, Waltham, MA, USA) via the N₂ adsorption method. Measurements of NH₃-TPD and CO₂-TPD were conducted using a Thermo Scientific TPD1100 (Thermo Fisher Scientific, Waltham, MA, USA) fully automated physical adsorption instrument: First, 50 mg of catalyst was pretreated under a helium atmosphere (30 mL·min⁻¹) by heating from room temperature to 150 °C at a rate of 10 °C·min⁻¹, followed by holding at this temperature for 120 min to desorb physically adsorbed species. After cooling to 50 °C, the sample was exposed to 10% NH₃/He (NH₃-TPD) or CO₂ (CO₂-TPD) at a flow rate of 30 mL·min⁻¹ for 30 min to achieve adsorption. The system was then purged with helium until a stable baseline was achieved. Subsequently, the temperature was raised to 500 °C at a heating rate of 10 °C·min⁻¹, and desorption curves were continuously recorded via the thermal conductivity detector (TCD) (Bei Shi De. Beijing, China). Acid sites and base sites were quantified based on the integrated values of the NH₃ and CO₂ desorption signals, respectively.

3.5. Preparation of Dehydrated Fructose from Fruit and Quantitative Analysis of 5-HMF

Add 30 mL DMSO to a three-neck flask. After heating to 120 °C, accurately weigh 1 g of fructose and 1 g of catalyst and add them to the flask. Following a reaction period, filter the mixture through a filter for subsequent chromatographic analysis. Quantitative analysis of fructose and 5-HMF content was conducted using a 1260 high-performance liquid chromatograph (Agilent Technologies, Santa Clara, CA, USA). Chromatographic conditions: RID detector (Agilent Technologies, Santa Clara, CA, USA), HPX H column (300 × 7.7 mm), mobile phase 5 mM H₂SO₄ for fructose quantification; UV detector (Agilent Technologies, Santa Clara, CA, USA) (280 nm), C18 column (250 mm × 4.6 mm), mobile phase methanol/water (6/4, v/v) for 5-HMF determination. Fructose conversion rate, 5-HMF yield, and selectivity may be calculated using the following equations:

$$\mathit{Fructose} \mathit{conversion} (\%)={\displaystyle \frac{\left(\mathit{\text{Final fructose moles−Initial fructose moles}}\right)}{\mathit{Initial} \mathit{fructose} \mathit{moles}}}\times 100\%$$

(12)

$$\mathit{\text{5-HMF}} \mathit{yield} (\%)={\displaystyle \frac{\mathit{\text{5-HMF}} \mathit{moles} \mathit{of} \mathit{product} }{\mathit{Initial} \mathit{fructose} \mathit{moles}}}\times 100\%$$

(13)

$$\mathit{\text{5-HMF}} \mathit{selectivity} (\%)={\displaystyle \frac{\mathit{\text{5-HMF}} \mathit{yeild}}{\mathit{Fructose} \mathit{conversion}}}\times 100\%$$

(14)

4. Conclusions

In this study, a nitric acid-modified silica gel catalyst was successfully prepared and applied to the dehydration of fructose to produce 5-HMF. Compared with native and amino-functionalized silica gels, the activated silica gel exhibited markedly improved catalytic performance, mainly due to the increased Brønsted acidity and the exposure of more accessible surface silanol groups. Structural characterization confirmed that nitric acid treatment preserved the pore structure while enhancing the density of acidic sites. The influence of reaction parameters and solvent effects was systematically analyzed. Among the solvents examined, DMSO afforded the highest 5-HMF yield owing to its strong solvation ability, although its non-green nature indicates that further exploration of more sustainable solvents is needed. Under the optimized conditions (DMSO, 120 °C, fructose-to-catalyst ratio of 1:1), a maximum 5-HMF yield of 91.6% was achieved. Recycling experiments showed that catalyst deactivation was mainly caused by humin deposition on the surface, which could be largely removed by simple ethanol washing, restoring catalytic activity. Overall, while limitations remain-particularly in solvent greenness-the nitric acid-modified silica gel provides a practical, low-cost, and recyclable heterogeneous catalyst, offering useful insights for the design of silica-based solid acids and the valorization of biomass-derived carbohydrates.