Sulfonic acid-functionalized mesoporous carbon/silica as efficient catalyst for dehydration of fructose into 5-hydroxymethylfurfural

Abstract

Mesoporous silica with desired pore structure (the appropriate channel length and pore size) was synthesized through a modified soft template method. Amorphous carbon layers were then introduced in mesoporous silica channels through infiltration of carbon precursor and carbonization processes. Finally, sulfonic acid groups (–SO₃H) attached onto the amorphous carbons by sulfonation reaction to form sulfonic acid-functionalized mesoporous carbon/silica which was used as a sulfonic acid catalyst. The morphology, pore structure, surface functional groups, and compositions of the obtained sulfonic acid-functionalized mesoporous carbon/silica were investigated by field-emission scanning electron microscopy (SEM), transmission electron microscopy (TEM), N₂ adsorption–desorption isotherm, X-ray photoelectron spectroscopy, thermogravimetric analysis and X-ray diffraction techniques. These materials are active for the catalytic dehydration of fructose into 5-hydroxymethylfurfural. The pore structure of hard template-mesoporous silica, and homogeneous coating of amorphous carbon layers are found to be critical to obtain sulfonic acid-functionalized mesoporous carbon/silica with enhanced catalytic activity. The sulfonic acid-functionalized mesoporous carbon/silica synthesized in the desired way could exhibit a high yield of 5-HMF and good recyclability, making them highly applicable in practical applications.

---

1. Introduction

Global economy development largely depends on the consumption of crude oil, which is not a renewable energy source and also suffers from the problem of greenhouse-gas effect. Utilizing abundant biomass resources for valuable chemicals production gives us a promising alternative for a sustainable energy supply. Among various biomass derived chemicals reported to date, 5-hydroxymethylfurfural (5-HMF) is a valuable intermediate, which is a critical platform compound for fine chemicals, pharmaceuticals and furan-based polymers, etc.^1,2 It has been reported that 5-HMF could be fabricated from fructose obtained directly from biomass or by the isomerization of glucose. The study of catalytic dehydration of fructose into 5-HMF has received considerable attention.^3–5 Generally, this process can be catalyzed by liquid acids such as H₂SO₄, HCl, H₃PO₄. However, the utilization of these mineral acids suffers from several serious drawbacks, such as high toxicity, corrosion, and difficulty of separation from the reaction mixture.^6–8 Therefore, developing more environmentally friendly and convenient solid acid catalysts to replace the liquid acid catalysts is highly desired. Previous studies have demonstrated the potential uses of ion exchange resins (such as Amberlyst15, DOWEX) as a green solid acid catalyst for the 5-HMF formation.^9–11 Unfortunately the reaction temperatures for these catalysts are relatively low, which greatly impedes the enhancement of 5-HMF selectivity.^9–11 Although H-form zeolites could be employed as solid acid catalysts for fructose dehydration reaction and the reaction selectivity can reach to a high level (60–90%), their practical uses in the conversion of fructose into 5-HMF has been largely limited by the lower conversion efficiency.^12,13 Recent work has suggested that low cost sulfonic acid catalysts prepared by sulfonated incomplete carbonaceous materials, which exhibit high acidity and good thermal stability because of their carbon frameworks, can function as an efficient solid acid catalyst for many reactions such as esterification, reformation, etc.^14–18 However, nonporous structure of this type of solid acid catalysts has greatly limited their catalytic activity, since the number of attached acidic groups cannot reach to a expected value due to their small active surface areas. In this sense, introducing sulfonic groups (–SO₃H groups) onto porous carbon materials with a large surface area could be an attractive way to increase the number of acidic groups.

Recently, the template method has been widely used to prepare porous carbon materials.^19–22 Among various templates reported to date, mesoporous silica have received a great deal of attention for their potential applications in catalysis due to their large surface area and pore volume, narrowly distributed and tunable pore diameters since their discovery by the Mobil oil company in 1992.²⁰ Among mesoporous silica, MCM-41 has been widely investigated in catalysis area. However, MCM-41 possesses isolated parallel channels, which greatly limits its in-pore diffusion of reactants and products and thereby decrease its catalytic turnover. This, together with its low hydrothermal stability, largely hinders further utilization of MCM-41 in the catalytic fields.²¹ Compared with MCM-41, SBA-15 with 2D-hexagonal p6mm pore structure is particularly desired since the isolated parallel channels inside the mesoporous SBA-15 is connected with small channels. Its relatively large pore diameter (which facilitates the diffusion of reactants and products) and higher hydrothermal stability make it highly applicable in catalysis. However, in order to facilitate the adsorption and desorption of reactants and products, the channel length and pore size of mesoporous silica should be well controlled. Several papers have unveiled the preparation of SBA-15 materials with short mesochannels in the submicrometer level by adding either co-surfactant, co-solvent, electrolytes, or organosilanes into the synthesis solutions^20,23,24 and pore-expanded SBA-15 by adding porogens including trimethylbenzene, triethylbenzene or triisopropylbenzene into the synthesis solutions to swell the Pluronic P123 micelles.^19,25,26 Interestingly, the potential applications of the resulting platelet SBA-15 with short mesochannels and large pore diameter in the catalysis area have not yet been well exploited. While sulfonic acid-functionalized mesoporous materials have been previously tested in biodiesel synthesis,^27,28 the pore sizes of such catalysts and the relationship between pore diameter and mass transport properties remains unexplored.

Here we present a systematic report on the impact of pore diameter and channel length on the catalytic activity of sulfonic acid-functionalized mesoporous carbon/silica as the solid acid catalysts for the catalytic synthesis of 5-HMF from fructose. Mesoporous silica with expanded pores and short channel length were used as the starting material, whose pores were firstly impregnated with carbon precursors through infiltration. The sulfonic acid-functionalized mesoporous carbon/silica was then obtained by the calcination of carbon precursors impregnated mesoporous silica and the subsequent sulfonation reaction. Through control of the pore diameter and channel length of the mesoporous silica, the thickness of carbon layer in the pores, the contact of reactant with active centers and the diffusion of product can be well regulated, thus the corresponding catalytic activity of sulfonic acid-functionalized mesoporous carbon/silica can be well optimized.

2. Experimental

2.1 Chemicals

Tetraethyl orthosilicate (TEOS), Pluronic P123 triblock copolymer (P123) and fructose were obtained from Aladdin Ltd. Concentrated sulfuric acid (H₂SO₄, 95.0–98.0%), hydrochloric acid (HCl, 37%), glucose (98%), zirconyl chloride octahydrate (ZrOCl₂·8H₂O) and trimethylbenzene (TMB) were bought from Shanghai Chemical Reagent Co. Ltd. All the chemicals were used as received without further purification. Deionized (DI) water (H₂O) through Millipore system (Milli-Q®) was used in all the experiments.

2.2 Materials synthesis

2.2.1 Preparation of mesoporous silica. Mesoporous silica was prepared by dissolving 2.00 g of P123 in 80 mL of 2 M HCl solution at 35 °C. Pre-designed amounts of ZrOCl₂ and TMB were then added and kept at 35 °C for 1 h under stirring prior to addition of 4.2 g of TEOS. The resulting mixture was stirred at 35 °C for 24 h and left static for 24 h at 100 °C in a closed Teflon bottle. The mixture was finally filtered, washed repeatedly with water, and dried at 100 °C overnight. The P123 template was removed by calcining the material at 550 °C in air for 6 h with a ramping rate of 1 °C min⁻¹. The reactant mass ratio of ZrOCl₂ : TMB : TEOS is (0–0.15) : (0–0.7) : 1.0. The prepared mesoporous silica are labeled as S(X,Y), with X representing the mass of the added TMB and Y representing the mass of ZrOCl₂ introduced.

2.2.2 Preparation of mesoporous carbon/silica. The mesoporous silica fabricated above was added to a solution containing glucose, H₂SO₄, and H₂O with a mass ratio of 2.1 : 1.0 : 4.0. The obtained mixture was stirred at room temperature for 1 h before it was heated at 100 °C for 24 h and then at 160 °C for 6 h. The resultant carbon/mesoporous silica was carbonized in a quartz tube under high vacuum at 550 °C for 3 h with a heating rate of 5 °C min⁻¹. By changing the volume ratio of glucose over the total pore volume of mesoporous silica, different mesoporous carbon/silica were prepared, and the obtained mesoporous carbon/silica are denoted as M% C-S(X,Y), where M% stands for the ratio of the volume of glucose used in the infiltration step over the total pore volume (measured from physical adsorption data).

2.2.3 Sulfonation of mesoporous carbon/silica. 2.0 g of carbon/mesoporous silica was treated with 100 mL of concentrated H₂SO₄ (>98%) at 150 °C for 10 h under the Ar atmosphere. The solids were then washed with hot water until no sulfate ions were detectable in the filtrate. The resultant sulfonated mesoporous carbon/silica are donated as M% SC-S(X,Y).

2.3 Characterization

The microscopic features of the samples were characterized with a scanning electron microscopic (SEM) on a field-emission scanning electron microscope (S-4800, Hitachi) at an operation voltage of 20.0 kV, and transmission electron microscopy (TEM) (JEM 2010, JEOL, Japan) operated at 200 kV. The porosity of the samples was investigated using physical adsorption of nitrogen at the liquid-nitrogen temperature (−196 °C) on an automatic volumetric sorption analyzer (Quantachrome, Autosorb-IQ-MP). The specific surface areas were determined according to the Brunauer–Emmett–Teller (BET) method in the relative pressure range of 0.05–0.2. The pore size distribution (PSD) curves were computed using the Barrett–Joyner–Halenda (BJH) method from the adsorption branches. The pore sizes were estimated from the maximum positions of the BJH PSD curves. The XRD patterns of samples were recorded using a X-ray diffraction technique (Bruker D8-Advance, Germany) with Cu Kα radiation. The surface functional groups were determined by X-ray photoelectron spectroscopy (XPS) on a VG ESCALAB 250 spectrometer (Thermo Electron, U.K.), using an Al Kα X-ray source (1486 eV).

2.4 Dehydration of fructose into 5-HMF

The catalytic performance of the sulfonic acid-functionalized mesoporous carbon/silica was evaluated by the dehydration reaction of fructose into 5-HMF. The reaction was carried out in a three-necked round bottom flask. Specifically, 0.5 g fructose and 6 mL DMSO were successively added into the flask. The mixture was then heated to 130 °C and 0.4 g of sulfonic acid-functionalized mesoporous carbon/silica was then added. The reaction was held at the temperature of 130 °C for 1.5 h under Ar-atmosphere for the dehydration of fructose into 5-hydroxymethylfurfural (HMF) and finally quickly quenched in a cool water bath. The solid acid catalyst was then separated from the reaction system by centrifugation. The supernatant was used for quantifying 5-HMF, and the sulfonic acid-functionalized mesoporous carbon/silica was collected after filtration and dried at 60 °C for 24 h for the next reaction cycles. Quantitative analysis of 5-HMF was carried out using a UV-VIS recording spectrophotometer (SHIMADZU, UV-2401PC). Fig. S1† shows the UV-Visible spectra of the pure 5-HMF solution, which show a strong absorption peak at 285 nm. The absorption coefficient calculated based on Fig. S1† is 2.3 × 10⁴ L mol⁻¹ cm⁻¹, indicating the high sensitivity of 5-HMF to the UV-visible spectroscopy. In addition, no obvious peaks for the aqueous solutions of formic acid or levulinic acid solution (the possible catalytic by-products of fructose), were found between the wavelength of 230 and 350 nm. This indicates that the UV-visible peak located at 284 nm is suitable for quantifying 5-HMF.

3. Results and discussion

3.1 SEM and TEM images

The mesoporous silica materials were prepared by adding small amounts of Zr(IV) ions and porogen TMB in the conventional SBA-15 synthesis solutions. The optimal mass ratio of ZrOCl₂ : TMB : TEOS in the synthesis solution was (0–0.15) : (0–0.7) : 1.0. Fig. 1a, d and g show the SEM images of the prepared mesoporous silica materials by adding small amounts of Zr(IV) ions and TMB. Compared with the conventional SBA-15, S(0,0.32) (shown in Fig. 1a) exhibits relatively small particles sizes, indicating the shortening of its channel length. Moreover, compared with S(2,0.32) (Fig. 1d), S(2,0.64) (Fig. 1g) has smaller particle size and very short mesochannels due to more Zr(IV) ions involved in its preparation process. This phenomenon is consistent with the results reported in the literature,²⁰ which demonstrated that the introduction of Zr(IV) ions can accelerate the silicate condensation around the P123 micelles, facilitating the termination of the interconnection between particles to from longer rods along the channeling direction.

Fig. 1 SEM images of (a) S(0,0.32), (b) 22% SC-S(0,0.32), (c) 77% SC-S(0,0.32), (d) S(2,0.32), (e) 22% SC-S(2,0.32), (f) 77% SC-S(2,0.32), (g) S(2,0.64), (h) 22% SC-S(2,0.64), (i) 77% SC-S(2,0.64).

In addition, compared with S(0,0.32), S(2,0.32) exhibits relatively large pore diameter, indicating the expansion of its pore structure. The results show that porogen TMB has the effect of swelling the Pluronic P123 micelles used to produce mesoporous silica, enabling the formation of expanded pore diameter. According to the above results, the prepared mesoporous silicas with desired pore structure (the appropriate channel length and pore size of mesoporous silica) can be obtained by changing the reaction parameters. The optimized pore structure of mesoporous silica will facilitate the adsorption and desorption of reactants and products, and improve the catalytic activity of the corresponding synthesized sulfonic acid-functionalized mesoporous carbon/silica shown below.

The infiltration would lead to the adsorption of glucose onto the pores of mesoporous silica, resulting in the formation of amorphous carbon layer on the channel walls of mesoporous silica after the high temperature calcination. Fig. 1b, e, and h show that the formation of amorphous carbon layer does not change the overall morphology of the mesoporous silica, indicating a high morphology retention property of the mesoporous silica during the infiltration and calcination processes. With increase of infiltrated glucose, thicker amorphous carbon layer could be grown on the walls of mesoporous silica. Fig. 1c, f and i show the aggregation of particles and the ambiguity of their edges and corners when more glucose is infiltrated for the preparation of the mesoporous carbon/silica. Especially, the morphology of 77% SC-S(0,0.32) is greatly different from S(0,0.32). Many small ball particles can be found for sample 77% SC-S(0,0.32), which might be caused by overdose glucose spilling over the pores of mesoporous silica. Fig. 2 exhibits the TEM images of 22% SC-S(2,0.32) and 77% SC-S(2,0.32). Obvious and clear pore structure can be seen in these two sulfonic acid-functionalized mesoporous carbon/silica. This indicates that the carbon layer is uniformly grown without obvious aggregation of carbon particles in the restricted mesopores of mesoporous silica. The pore structure is therefore well maintained in the final sulfonic acid catalysts, giving a high surface area and desirable structure.

Fig. 2 TEM images of (a) 22% SC-S(2,0.32), (b) 77% SC-S(2,0.32).

3.2 Pore structure

The N₂ adsorption–desorption isotherms in Fig. 3 shows that S(0,0.32) exhibits a type IV isotherm with an H₁ hysteresis loop appearing at P/P₀ = 0.6–0.8 (Fig. 3a), which clearly demonstrates its porous feature possessing cylindrical pore geometry with a high degree of pore size uniformity.²⁹ The addition of TMB would increase the pore dimension of mesoporous silica. As shown in Fig. 3a, the hysteresis loops of S(2,0.32), S(3,0.32), S(2,0.192), S(2,0.64) appear at the relatively higher pressures. This might be due to the fact that porogen TMB swells the micelles of soft template P123, facilitating the formation of mesopores with larger diameter.¹⁹

Fig. 3 N₂ adsorption/desorption isotherms of mesoporous silica and sulfonic acid-functionalized mesoporous carbon/silica. (a) S(X,Y); (b) 22% SC-S(X,Y); (c) 77% SC-S(X,Y).

The growth of the carbon layer in the pore of the mesoporous silica and the functionalization with sulfonic acid produces acid-functionalized carbon/silica composite with a mesoporous structure do not change the porous morphology of the mesoporous silica. As shown in Fig. 3b and c, all the sulfonic acid functionalized catalysts exhibit classical type IV isotherms, clearly demonstrating their mesoporous structure. This indicates that after the growth of carbon layer in the pores of mesoporous silica and the introduction of –SO₃H groups the mesoporous feature of the silica is maintained, which is helpful to enhance the specific surface area, making the acid groups well accessible to the catalytic reactions. Worth noting is that the hysteresis loops of the sulfonic acid-functionalized mesoporous carbon/silica appear at the relatively low P/P₀ compared to the corresponding mesoporous silica, indicating the decrease in the pore dimension due to the growth of the carbon layer in the pores of the mesoporous silica and the functionalization with sulfonic acid.

The specific surface areas of the mesoporous silica and the mesoporous carbon/silica before and after sulfonation reaction are summarized in Table 1. It is noticeable that the specific surface areas of S(0,0.32), S(2,0.32), and S(3,0.32) are comparable, whereas the pore sizes for S(2,0.32) and S(3,0.32) are greatly enhanced compared with S(0,0.32) due to the introduction of porogen TMB for their fabrication as shown in Fig. 4a. Table 1 and Fig. 4b and c show that the growth of carbon layer and the functionalization with sulfonic acid would reduce the specific surface area and pore size of the samples. The degree of the reduction in the specific surface area and pore size would be greater when more glucose is infiltrated for the mesoporous carbon/silica preparation. Interestingly, the sample S(0,0.32) shows the opposite trend. As shown in Table 1, the specific surface area of 77% SC-S(0,0.32) is higher than that of 22% SC-S(0,0.32), however, the pore size and pore volume of 77% SC-S(0,0.32) are smaller than those of 22% SC-S(0,0.32). This is presumably due to the reason that the pore size of S(0,0.32) is relatively smaller (around 6.5 nm) due to the absence of TMB. Therefore, excess glucose cannot be impregnated into the pore channels of mesoporous silica, forming small amorphous carbon particles out of the mesoporous carbon/silica after the calcination as observed in the SEM images shown in Fig. 1c. These extra small carbon particles could also enhance the specific surface area of final sulfonic acid functionalized catalyst.

Table 1 Pore structure and specific surface area of mesoporous silica and the corresponding catalysts

Samples	Specific surface area (m^2 g^−1)	Pore volume (cm^3 g^−1)	Samples	Specific surface area (m^2 g^−1)	Pore volume (cm^3 g^−1)
S(0,0.32)	713	1.21	S(2,0.192)	704	2.387
22% SC-S(0,0.32)	432	0.628	22% SC-S(2,0.192)	503	0.622
77% SC-S(0,0.32)	509	0.321	77% SC-S(2,0.192)	315	0.171
S(2,0.32)	681	2.34	S(2,0.32)	681	2.337
22% SC-S(2,0.32)	528	0.666	22% SC-S(2,0.32)	528	0.666
77% SC-S(2,0.32)	444	0.263	77% SC-S(2,0.32)	444	0.263
S(3,0.32)	728	2.400	S(2,0.64)	785	2.682
22% SC-S(3,0.32)	472	0.482	22% SC-S(2,0.64)	486	0.434
77% SC-S(3,0.32)	444	0.264	77% SC-S(2,0.64)	322	0.163

---

Fig. 4 BJH-PSD curves of (a and d) mesoporous silica and (b, c, e and f) sulfonic acid-functionalized mesoporous carbon/silica.

The samples of S(2,0.192), S(2,0.32), S(2,0.64) are fabricated from the solutions containing the same amount of TMB, while different concentration of Zr(IV) ions. As a result, the pore sizes of these three samples are expanded to about 10.5 nm due to the presence of TMB (Fig. 4d). In addition, the specific surface area of S(2,0.64) is higher than those of the other two. That is because more Zr(IV) ions are involved in the preparation process of S(2,0.64), which shorten the channel length. After the growth of the carbon layer in the pore of the mesoporous silica and the functionalization with sulfonic acid, the specific surface of these three samples all decline with the increase of added glucose, and the same trend can be found for the pore volume and pore size distribution (Fig. 4d–f), since the increase of glucose increases the thickness of carbon layer on the pore walls of the mesoporous silica.

3.3 TGA and XRD analysis

Fig. 5 shows the TGA curves of 22% SC-S(3,0.32) and 77% SC-S(3,0.32). According to the TGA curves the total weight loss for 22% SC-S(3,0.32) and 77% SC-S(3,0.32) between 100 to 800 °C is 67% and 37%, respectively, which could be attributed to the burning of carbon layer and attached surface groups. The higher weight loss for 77% SC-S(3,0.32) indicates a higher thickness of the carbon layer due to the impregnation of more glucose. This conclusion can also be evidenced in Fig. 4b and c, which show that the pore size distribution decreases with increase of infiltrated glucose.

Fig. 5 TGA curves of 22% SC-S(3,0.32) and 77% SC-S(3,0.32).

Fig. 6 shows the XRD patterns of the prepared mesoporous silica and sulfonic acid functionalized mesoporous carbon/silica. It can be seen that both 77% SC-S(3,0.32) and 77% SC-S(2,0.64) possess a very broad diffraction peak at around 25° (Fig. 6b), ascribable to the (002) reflection of graphic carbon and the amorphous silica framework (Fig. 6a), both of which exhibit diffraction peak at the similar position.²² In addition to the silica framework, 77% SC-S(3,0.32) and 77% SC-S(2,0.64) also showed another small peak at around 43° attributed to the (101) planes of graphitic carbon. Its weak intensity and broadening indicate amorphous feature of coated carbon layer with small graphitic domains.

Fig. 6 XRD patterns of (a) S(3,0.32), S(2,0.64) and (b) 77% SC-S(3,0.32), 77% SC-S(2,0.64).

3.4 Investigation of active groups

As reported in our previous work, glucose could be carbonized to form amorphous carbon layer through the calcination, and sulfonation reaction would facilitated the functionalization of the carbon layer with active –SO₃H groups.²² The existence of active groups was demonstrated by the XPS technique, which is a surface sensitive with a sampling depth extending from the surface to 1–5 nm. Due to the porous feature, the active –SO₃H groups in the sulfonic acid functionalized mesoporous carbon/silica mainly reside inside the pore channels. Therefore in order to intensify the detectability, a sample with relatively higher carbon loading, i.e. 200% SC-C(2,0.32), was chosen as an example to demonstrate the existence of –SO₃H groups. Fig. 7 shows that 200% SC-C(2,0.32) exhibits a peak at ∼168.8 eV, assignable to aromatic carbon atoms in connection with –SO₃H groups, which strongly demonstrate the existence of active –SO₃H groups.^22,30

Fig. 7 S2p XPS spectrum for 200% SC-C(2,0.32).

Taking all together, the preparation procedure for sulfonic acid-functionalized mesoporous carbon/silica can be illustrated in Scheme 1. Specifically, the carbon precursor (glucose) was infiltrated into the mesoporous structure of template silica. Amorphous carbon layers could then be grown on the pore channels of mesoporous silica after the high temperature calcination. The diameters of the pore channels decrease with increase of infiltrated glucose. Then sulfonic acid groups (–SO₃H) can be introduced onto the amorphous carbon layers through the sulfonation reaction. Due to the porous features making the active –SO₃H groups well accessible, acid-functionalized mesoporous carbon/silica is therefore expected to exhibit high catalytic performance.

Scheme 1 Schematic diagram showing the preparation process of sulfonic acid-functionalized mesoporous carbon/silica.

3.5 Catalytic properties

To evaluate the catalytic performance, the sulfonic acid-functionalized mesoporous carbon/silica was used as the catalysts for the dehydration of fructose to 5-HMF. Fig. 8a shows the first cycle yield of 5-HMF catalytically produced using the various sulfonic acid-functionalized mesoporous carbon/silica (control experiment shows that 5-HMF is not produced in the absence of the sulfonic acid-functionalized mesoporous carbon/silica, Fig. S2†). As shown in Fig. 8a, the catalysts fabricated with 22% and 77% glucose infiltration exhibit relatively higher catalytic performance for the 5-HMF production. As indicated in Fig. 4a and b, the catalysts fabricated with 22% glucose infiltration, such as 22% SC-S(2,0.192), 22% SC-S(0,0.32), 22% SC-S(2,0.32), 22% SC-S(3,0.32), and 22% SC-S(2,0.64) have more uniform pore size distribution, suggesting the presence of more uniform amorphous carbon layer in the mesopores of these catalysts. The N₂ adsorption/desorption isotherms in Fig. 3 have demonstrated that these catalyst have relatively high surface areas with bigger pore sizes, providing them with higher areas accessible to fructose and suitable paths for the diffusion of the reactant and product, which could be considered as the main reasons leading to their higher performance. Although the increase of the infiltration glucose to 77% would decrease the average pore sizes in the sulfonic acid-functionalized mesoporous carbon/silica as demonstrated in Fig. 4, the catalysts fabricated with 77% glucose infiltration still exhibit homogeneous pore size distributions with relative higher specific surface area (Table 1). The average pore sizes of 4 nm in these catalysts are still suitable to act as the diffusion paths of the reactants and products. Therefore, the catalysts fabricated with 77% glucose infiltration also show higher catalytic performance for the 5-HMF production. For the catalysts fabricated with a higher glucose infiltration of 200%, however, the specific surface areas and the average pore sizes are significantly lower in compared with those fabricated with 22% and 77% glucose infiltration, as demonstrated in Fig. S3,† Tables 1 and S1.† The lower specific surface areas and average pore sizes indicate smaller contact areas for the reactant accessibility and narrower diffusion paths for the reactant and product diffusion. These catalysts therefore exhibit significantly lower catalytic performance for the 5-HMF production. Indeed, for these catalysts fabricated with 200% glucose infiltration, the overdose of glucose occurs, which would result in the formation of small separate carbon particles due to the spillover of glucose out of the pores of mesoporous silica. This could further decrease the performance of these catalysts for the catalytic production of 5-HMF. Based on these results, we could attribute the higher catalytic performance of the catalysts fabricated with 22% and 77% glucose infiltration to the higher specific surface area and homogeneous coating of amorphous carbon layer in the pores, which provide the catalysts with the higher area for the reactant contact and the suitable paths for the diffusion of the reactants and the products. Compared to the previously reported non-porous ion-exchange resin³¹ and the sulfonic acid-functionalized SBA-15 type silica,^32,33 our sulfonic acid-functionalized mesoporous carbon/silica synthesized with 22% and 77% glucose infiltration exhibit much better catalytic performance. This strongly suggest the higher catalytic performance of our sulfonic acid-functionalized mesoporous carbon/silica synthesized with 22% and 77% glucose infiltration.

Fig. 8 (a) Catalytic yields of 5-HMF by the sulfonic acid-functionalized mesoporous carbon/silica catalysts in the first reaction cycle and (b) catalytic yields of 5-HMF by the sulfonic acid-functionalized mesoporous carbon/silica catalysts with higher activities after four times reaction cycles.

Fig. 1 has demonstrated that the introduction of Zr(IV) ions in the synthesis would promote the formation of the smaller catalysts with the relatively short channels and TMB would expand the pore channel. The short channel and large pore could facilitate the formation of thicker amorphous carbon layer when the same amount of glucose is infiltrated, resulting in the formation of the sulfonic acid-functionalized mesoporous carbon/silica with low specific surface areas. Due to the low specific surface areas, the sulfonic acid-functionalized mesoporous carbon/silica synthesized from mixtures containing more Zr(IV) ions exhibit low catalytic performance for the 5-HMF production. This explains why 77% SC-S(2,0.64) exhibits a slightly lower catalytic performance for the 5-HMF production than 77% SC-S(2,0.192), as shown in Fig. 8a.

Additionally, to demonstrate the potential use of the sulfonic acid-functionalized mesoporous carbon/silica with higher catalytic performance in the practical applications, their stability and reusability were measured. Fig. 8b shows that these samples could maintain their catalytic performance without noticeable declines in their performance after four cycles of the reaction. This clearly indicates the high stability and reusability of these catalysts, strongly suggesting that they are potentially usable in the practical applications.

4. Conclusions

In conclusion, by adding small amounts of Zr(IV) ions and porogen TMB, the mesoporous silica SBA-15 with short channels and large pore diameters could be successfully synthesized. These mesoporous silica were then used as the template for the synthesis of the sulfonic acid-functionalized mesoporous carbon/silica through the infiltration with carbon precursors, the high temperature calcination, and the subsequent functionalization with sulfonic acid. The obtained sulfonic acid-functionalized mesoporous carbon/silica could be used as the catalysts for the dehydration of fructose to produce 5-HMF. It is found that homogeneous growth of amorphous carbon layers in the pores of the mesoporous silica would facilitate the active groups inside the pore of the mesoporous carbon/silica well accessible to reactants, and could also promote the diffusion of the reactants and products. The sulfonic acid-functionalized mesoporous carbon/silica with amorphous carbon layers inside the pores therefore exhibit higher activity for catalytic dehydration of fructose to produce 5-HMF. This, together with their good recyclability, makes the sulfonic acid-functionalized mesoporous carbon/silica with amorphous carbon layers inside the pores highly applicable in the applicable applications. The result shown here is therefore of great significance since it introduces a new method to prepare a catalyst to produce 5-HMF with high efficiency and good recyclability and might also be helpful to promote the applications of 5-HMF in the specific areas.

Acknowledgements

This work was financially supported by the Chinese National Natural Science Foundation (No. 11474101, 21306091 and U1532139), the National Undergraduate Training Programs for Innovation and Entrepreneurship (No. 201511058006), the Zhejiang Provincial Natural Science Foundation (No. LY14B030001), the Zhejiang Provincial Public Welfare Technology Application Research Project (2015C31151), Wan weiming foundation (2016012), and the Guangdong Innovative and Entrepreneurial Research Team Program (No. 2014ZT05N200).

Notes and references

1. G. A. Kraus and T. Guney, Green Chem., 2012, 14, 1593–1596.
2. A. A. Rosatella, S. P. Simeonov, R. F. M. Frade and C. A. M. Afonso, Green Chem., 2011, 13, 754–793.
3. A. Takagaki, M. Ohara, S. Nishimura and K. Ebitani, Chem. Commun., 2009, 6276–6278.
4. S. Dutta, S. De, B. Saha and M. I. Alam, Catal. Sci. Technol., 2012, 2, 2025–2036.
5. R. Luque, J. C. Lovett, B. Datta, J. Clancy, J. M. Campelo and A. A. Romero, Energy Environ. Sci., 2010, 3, 1706–1721.
6. T. S. Hansen, J. Mielby and A. Riisager, Green Chem., 2011, 13, 109–114.
7. J. N. Chheda, Y. Roman-Leshkov and J. A. Dumesic, Green Chem., 2007, 9, 342–350.
8. D. M. Alonso, J. Q. Bond and J. A. Dumesic, Green Chem., 2010, 12, 1493–1513.
9. K.-i. Shimizu, R. Uozumi and A. Satsuma, Catal. Commun., 2009, 10, 1849–1853.
10. X. Qi, M. Watanabe, T. M. Aida and R. L. Smith, Ind. Eng. Chem. Res., 2008, 47, 9234–9239.
11. F. H. Richter, K. Pupovac, R. Palkovits and F. Schüth, ACS Catal., 2013, 3, 123–127.
12. C. Moreau, R. Durand, S. Razigade, J. Duhamet, P. Faugeras, P. Rivalier, P. Ros and G. Avignon, Appl. Catal., A, 1996, 145, 211–224.
13. E. Taarning, C. M. Osmundsen, X. Yang, B. Voss, S. I. Andersen and C. H. Christensen, Energy Environ. Sci., 2011, 4, 793–804.
14. F. Tao, H. Song and L. Chou, RSC Adv., 2011, 1, 672–676.
15. Y.-N. Li, J.-Q. Wang, L.-N. He, Z.-Z. Yang, A.-H. Liu, B. Yu and C.-R. Luan, Green Chem., 2012, 14, 2752–2758.
16. X. Qi, M. Watanabe, T. M. Aida and J. R. L. Smith, Green Chem., 2009, 11, 1327–1331.
17. C. Shi, Y. Zhao, J. Xin, J. Wang, X. Lu, X. Zhang and S. Zhang, Chem. Commun., 2012, 48, 4103–4105.
18. X. Qi, M. Watanabe, T. M. Aida and R. L. Smith Jr, Green Chem., 2010, 12, 1855–1860.
19. J. P. Dacquin, A. F. Lee, C. Pirez and K. Wilson, Chem. Commun., 2012, 48, 212–214.
20. S.-Y. Chen, C.-Y. Tang, W.-T. Chuang, J.-J. Lee, Y.-L. Tsai, J. C. C. Chan, C.-Y. Lin, Y.-C. Liu and S. Cheng, Chem. Mater., 2008, 20, 3906–3916.
21. S.-Y. Chen, T. Yokoi, C.-Y. Tang, L.-Y. Jang, T. Tatsumi, J. C. C. Chan and S. Cheng, Green Chem., 2011, 13, 2920–2930.
22. X. N. Tian, L. Luo, Z. Jiang and X. S. Zhao, Surf. Rev. Lett., 2011, 18, 229–239.
23. P. Kipkemboi, A. Fogden, V. Alfredsson and K. Flodström, Langmuir, 2001, 17, 5398–5402.
24. H. Zhang, J. Sun, D. Ma, X. Bao, A. Klein-Hoffmann, G. Weinberg, D. Su and R. Schlögl, J. Am. Chem. Soc., 2004, 126, 7440–7441.
25. L. Cao, T. Man and M. Kruk, Chem. Mater., 2009, 21, 1144–1153.
26. D. Chen, Z. Li, Y. Wan, X. Tu, Y. Shi, Z. Chen, W. Shen, C. Yu, B. Tu and D. Zhao, J. Mater. Chem., 2006, 16, 1511–1519.
27. R. Liu, X. Wang, X. Zhao and P. Feng, Carbon, 2008, 46, 1664–1669.
28. M. A. Jackson, I. K. Mbaraka and B. H. Shanks, Appl. Catal., A, 2006, 310, 48–53.
29. M. Kruk and M. Jaroniec, Chem. Mater., 2001, 13, 3169–3183.
30. S. Suganuma, K. Nakajima, M. Kitano, D. Yamaguchi, H. Kato, S. Hayashi and M. Hara, J. Am. Chem. Soc., 2008, 130, 12787–12793.
31. X. Qi, M. Watanabe, T. M. Aida and J. R. L. Smith, Green Chem., 2008, 10, 799–805.
32. Y.-Y. Lee and K. C. W. Wu, Phys. Chem. Chem. Phys., 2012, 14, 13914–13917.
33. A. J. Crisci, M. H. Tucker, M.-Y. Lee, S. G. Jang, J. A. Dumesic and S. L. Scott, ACS Catal., 2011, 1, 719–728.

---

Footnote

† Electronic supplementary information (ESI) available: UV-Vis spectra of the 5-HMF solutions at the different concentrations and UV-Vis spectra of formic acid or levulinic acid; plot of absorbance at 284 nm vs. the 5-HMF concentration; UV-Vis spectra of the solutions in the present and absence of the sulfonic acid-functionalized mesoporous carbon/silica (77% SC-S(3,0.32)); pore structure and specific surface area of the mesoporous silica and the corresponding catalysts; BJH pore size distribution of 200% SC-S(2,0.32). See DOI: 10.1039/c6ra20304c

---