Jun Shi,
Yuhua Shan*,
Yuan Tian,
Yu Wan,
Yitian Zheng and
Yangyang Feng
Advanced Catalysis and Green Manufacturing Collaborative Innovation Center, Changzhou University, Changzhou 213164, China. E-mail: yhshan@cczu.edu.cn
First published on 19th January 2016
Different (3-mercaptopropyl)trimethoxysilane (MPTS) loadings of sulfonic acid-functionalized micro-bead silica (SA-SiO2) were prepared by silylation and oxidation, and characterized by elemental analysis, SEM, FT-IR, TGA, NH3-TPD, BET N2 adsorption–desorption and 13C NMR CP/MAS. The as-prepared SA-SiO2 showed strong hydrophilic nature and excellent catalytic performance for dehydration of sorbitol to isosorbide. The selectivity to isosorbide is obviously affected by the MPTS loading. Using SA-SiO2 as a solid catalyst with 60.5% MPTS loading, 100% sorbitol conversion and 84% yield of isosorbide are achieved at 120 °C for 10 h under vacuum. The catalyst was reused 10 times without noticeable loss of activity and selectivity.
The dehydration of sorbitol to isosorbide (Scheme 1) is conventionally achieved using a liquid acid like sulfuric acid, hydrochloric acid or p-toluenesulfonic acid as a homogeneous catalyst in industrial production, which provides a yield of isosorbide (70–77%) at 130 °C under high vacuum conditions within a few hours.21 But use of homogeneous catalysts introduces many troubles such as difficulties in product separation, a corrosive hazard for the reactors and catalyst regeneration from the reaction system.20 Without adding any acid catalyst, in the dehydration of sorbitol in high temperature water (250–300 °C) for 1 h, 57% yield of isosorbide was obtained, reported by Yamaguchi et al.20,22 In addition, dehydration of pure sorbitol under microwaves at 160 °C showed sorbitol conversion and isosorbide selectivity of 100% and 60%,23 respectively.
Therefore, the research of sorbitol dehydration has been focused on developing solid acid catalysts.17,24–27 Many solid catalysts have been reported with different yields of isosorbide at suitable temperatures, such as sulfated copper oxides (67.3%, 200 °C),17 sulfated zirconia (61%, 210 °C),24 sulfated tin oxide (65%, 180 °C),25 metal phosphate (70.3%, 250 °C),28 molten salt hydrate medium (85%, 200 °C),27,29,30 and Amberlyst-15 resin (71.8%, 250 °C).31
However, most of these catalysts require comparatively high reaction temperature (200–300 °C) and the yield of isosorbide is not high. Therefore, the development of an efficient solid acid catalyst which catalyzes this reaction at moderate reaction conditions and leads to a high yield of isosorbide is highly desirable.
Polymer-supported Brønsted acid catalysts SO3H–PS–SO3H32 have been used for dehydration of sorbitol at 150 °C, with a maximum yield of 1,4-anhydro-D-sorbitol of 90% within 4 h. However, the yield of isosorbide is not high. Alternatively, a superhydrophobic mesoporous acid catalyst named as P–SO3H has been synthesized and used for dehydration of sorbitol to isosorbide,33 with a maximum yield (87%) of isosorbide at 140 °C within 10 h. Although these hydrophobic polymer-supported sulfonic acid catalysts show high yield, they are deactivated quickly because of their surface affinity to humins and cokings.
Silica sulfuric acid catalysts have been widely studied for a great number of acid-catalyzed organic reactions due to their heterogeneous nature, cheapness and availability.34 A simple in situ method to prepare water-stable acidic mesoporous sulfonated silica without any surfactant or template was reported by Hasan,35 and it has good catalytic performance for hydrolysis and alkylation reactions. Micro-bead silica has the advantages of hydrophilicity, high BET surface area and special porous structure.36,37 And the chemical bond linking method has good prospects for industrial application since it can fix organic groups on a carrier surface firmly.38,39 In this work, we report a simple and low-cost synthesis of a hydrophilic sulfonic acid-functionalized micro-bead silica catalyst named SA-SiO2 and evaluated its catalytic performance in dehydration of sorbitol to isosorbide in solvent-free condition. As we expected, SA-SiO2 gives good sorbitol conversion and isosorbide selectivity, and has excellent recyclability, compared with other reported catalysts.
By adjusting the ratio of MPTS to micro-bead silica, different SA-SiO2-X samples were obtained. For example, a catalyst loading of 60.5% MPTS is represented as SA-SiO2-60.5. The elemental analysis of the SA-SiO2 catalysts is shown in Table 1S.†
The acidity of the SA-SiO2 catalysts was determined by acid–base titration in an ultrasonic pool, according to the following definition:
Total acidity = (V1 − V0) × CNaOH/W |
Products were analyzed by HPLC (Agilent LC1260) equipped with a Zorbax NH2 column (4.6 × 200 mm, 5 μm particle size) and an evaporative light-scattering detector. The sugar products were separated by reversed-phase mode. The mobile phase was acetonitrile–water (80:
20) with a flow rate of 0.5 mL min−1. The column was thermostatically controlled at room temperature. The selectivity or yield of sorbitan or isosorbide was based on molar composition. The isolated yield is calculated by the following definition:
yieldisolated = (mol of isolated product)/(mol of sorbitol) × 100% |
The SEM images of SiO2 and SA-SiO2-60.5 are shown in Fig. 2. From the SEM images, it can be seen that the surface of SA-SiO2-60.5 catalyst is rough while that of the micro-bead silica is relatively smooth. This is caused by the sulfonic acid groups chemically bonded on the micro-bead silica.
The (CP/MAS) 13C-NMR spectra of SA-SiO2-60.5 and its precursor (unoxidized product) are shown in Fig. 3. The chemical shifts of saturated carbon are generally at 0–70, while those of the unsaturated carbon are usually at 100–165. As shown in Fig. 3A, the chemical shift peaks at 16.49 and 23.42 are associated with the saturated carbon atom attached to a silicon atom and saturated carbon atom attached to a carbon atom of both ends, respectively. The chemical shift peak at 58.95 corresponds to the saturated carbon atom attached to an oxidized sulfur atom.45 Compared with Fig. 3A, B shows the chemical shift peak at 47.61, which is associated with a saturated carbon atom attached to an unoxidized sulfur atom. Interestingly the peaks of other carbon atoms in Fig. 3B show red shifts because of the oxidation of sulfur atom. In a word, we can confirm that the product of MPTS loading is oxidized and propylsulfonic acid group is successfully connected to SiO2 by chemical bonding.
TGA of SA-SiO2-60.5 catalyst is shown in Fig. 4. Firstly, a slow weight loss in the range of 0–250 °C was noticed. Then a sharp weight loss was observed between 250 and 400 °C, which indicates desorption of the sulfonic groups on the surface of the micro-bead silica. The weight loss above 400 °C is due to the dehydration of hydroxyl groups in the pore walls of micro-bead silica.40 Thus it can be concluded that the SA-SiO2-60.5 catalyst is thermally stable and suitable for use at less than 250 °C.
The acidity, surface area, pore volume and pore size of various acid catalysts are shown in Table 1. The sulfonic resin has the maximum acidity (4530 μmol g−1) among these solid catalysts while its surface area is very poor (49.1 m2 g−1). Compared with the Amberlyst-15 resin, the SA-SiO2 has larger surface area and exhibits an appropriate high acidity and pore size. Furthermore, as shown in Table 1 (entries 1, 2, 3, 4, 5, 6, 7, 8 and 9), the acid concentration can be adjusted, and the acidity varied from 230 μmol g−1 to 1290 μmol g−1. The pore volume and pore size decreased with an increase of MPTS content. Pore size distributions of used SA-SiO2-60.5 and SBA-15-SO3H are shown in Fig. S4.†
Entry | Catalyst | SBET (m2 g−1) | Vp (cm3 g−1) | Mean Dp (nm) | Total aciditya (μmol g−1) | Acidity densityb (μmol m−2) |
---|---|---|---|---|---|---|
a Determined by NH3-TPD (Fig. 3S) except SA-SiO2-60.5.b Acidity density = total acid amount/BET surface area.c Undetectable.d After ten recycles. | ||||||
1 | Micro-bead silica | 350.9 | 0.84 | 11.3 | —c | —c |
2 | SA-SiO2-10.6 | 345.0 | 0.68 | 11.1 | 230 | 0.66 |
3 | SA-SiO2-22.5 | 310.3 | 0.51 | 10.5 | 340 | 1.10 |
4 | SA-SiO2-33.1 | 273.7 | 0.43 | 9.9 | 490 | 1.79 |
5 | SA-SiO2-40.9 | 221.1 | 0.39 | 9.6 | 610 | 2.76 |
6 | SA-SiO2-52.0 | 160.5 | 0.33 | 8.1 | 710 | 4.42 |
7 | SA-SiO2-60.5 | 140.9 | 0.29 | 7.3 | 840 | 5.96 |
8 | SA-SiO2-71.9 | 109.7 | 0.17 | 6.0 | 970 | 8.84 |
9 | SA-SiO2-83.3 | 48.3 | 0.09 | 4.9 | 1290 | 26.71 |
10 | H2SO4 | —c | —c | —c | 20![]() |
—c |
11 | Amberlyst-15 | 49.1 | 0.29 | 40.0 | 4530 | 92.26 |
12 | SBA-15-SO3H | 379.6 | 0.92 | 3.8 | 810 | 2.13 |
13 | Hβ (25) | 650.9 | 0.25 | 0.65 | 490 | 0.75 |
14 | HZSM-5 (40) | 451.7 | 0.14 | 0.54 | 320 | 0.71 |
15 | MCM-49 (30) | 457 | 0.46 | 0.71 | 435 | 0.95 |
16 | SA-SiO2-60.5d | 140.1 | 0.26 | 7.1 | 790 | 5.64 |
17 | SBA-15-SO3Hd | 303.6 | 0.69 | 3.1 | 150 | 0.49 |
Entry | Catalyst | Sorbitol conv. (%) | Yield (%) | Othersb (%) | |
---|---|---|---|---|---|
Isosorbide | Sorbitan | ||||
a Reaction conditions: 120 °C, 10 h, 1000 Pa, 50 g of sorbitol, 1.0 g of catalyst.b The by-products are 2,5-anhydro-D-sorbitol, 1,5-anhydro-D-sorbitol, and some others.c Undetectable.d After ten recycles. | |||||
1 | SA-SiO2-60.5 | 100 | 84 | 11 | 5 |
2 | H2SO4 | 100 | 90 | —c | 7 |
3 | Amberlyst-15 | 93 | 71 | 5 | 24 |
4 | SBA-15-SO3H | 95 | 70 | —c | 11 |
5 | Hβ (25) | 69 | 25 | 22 | 22 |
6 | HZSM-5 (40) | 30 | 9 | 17 | 4 |
7 | MCM-49 (30) | 73 | —c | 29 | 14 |
8 | SA-SiO2-60.5d | 100 | 75 | 20 | 5 |
9 | SBA-15-SO3Hd | 77 | 31 | 37 | 9 |
It is worth noting that although SA-SiO2-60.5 does not offer the maximum acidity, it displays good reaction performance (see Table 2). Compared with other reported hydrophilic catalysts in Table S2,† SA-SiO2-60.5 provides a higher yield of isosorbide and the reaction condition is moderate. This should be attributed to its suitable large pore size and high surface area (see Table 1). Thus it is beneficial to the mass transfer process, thereby inhibiting coke formation.
Fig. 5 shows the activity for the dehydration of sorbitol over various SA-SiO2-X at 120 °C for 10 h. With increasing MPTS loading from 10.6% to 40.9%, the yield of isosorbide increases obviously. When the MPTS loading increases from 40.9% to 60.5%, the growth rate of isosorbide yield slows down. Thereafter, with a continued increase of MPTS loading, the yield of isosorbide decreases. Yield of isosorbide and conversion of sorbitol are maximized at 60.5% MPTS loading. This phenomenon suggests that the further increase of MPTS loading (above 60.5%) will physically block the pores of micro-bead silica.40 Thus this seriously obstructs the internal diffusion and transfer of reactants and decreases the yield of isosorbide.
![]() | ||
Fig. 5 Effect of different MPTS loadings on the performance in sorbitol dehydration. Reaction conditions: 50 g sorbitol, 1.0 g catalyst, 120 °C, 10 h. |
The influence of reaction temperature on the dehydration reaction was investigated, and the results are shown in Fig. 6. As shown in Fig. 6a, the conversion of sorbitol is 100% at 100–160 °C. At 100 °C, the yield of isosorbide is low, ca. 60%. On increasing the dehydration reaction temperature to 120 °C, the yield of isosorbide reaches the highest value (84%), and then decreases with further increasing the temperature to 160 °C (see Fig. 6b), since more coke during the test is formed. But the sorbitan yield, a main intermediate product,29,46 decreases as the temperature increases (see Fig. 6c). These results demonstrate that an appropriate reaction temperature was crucial for achieving a high yield of isosorbide.47 The highest catalytic performance is observed at 120 °C for 10 h under vacuum with 2% SA-SiO2-60.5 (100% conversion of sorbitol and 84% yield of isosorbide).
Recyclability is of great importance for applying a solid catalyst in industrial production. After reaction, the SA-SiO2-60.5 was washed in ultrasonic pool with water and glycol dimethyl ether to evaluate its recyclability, and the results are shown in Fig. 7. After being reused 10 times, sorbitol conversion is still 100% and the yield of isosorbide decreases only slightly. Compared with other zeolites,48 the regeneration process does not require calcination and the synthesis does not need surfactant or template so that the cost can be reduced. All the results indicate that SA-SiO2-60.5 exhibits excellent stability and is of low cost.
As shown in Table S2,† most of the hydrophilic catalysts such as oxides, phosphates and sulfated materials showed a relatively high reactivity compared with other hydrophobic catalysts, which makes us believe that the high reactivity and stability of SA-SiO2-60.5 are derived from its high surface hydrophilicity. As observed in Fig. 8, the sorbitol and isosorbide contact angles on SA-SiO2-60.5 are about 12.1° and 34.5°, respectively. This indicates that SA-SiO2-60.5 has a highly hydrophilic surface, and its surface affinity to sorbitol is stronger than that to isosorbide. Scheme 3 shows the process of sorbitol dehydration to isosorbide. The high surface hydrophilicity and different surface affinity of SA-SiO2-60.5 catalyst are in favor of the feed (sorbitol) reaching and the product (isosorbide) leaving active sites rapidly, thus suppressing the sequential reactions which cause coke formation. Under the reaction conditions (100–160 °C, 650–1000 Pa), water, formed in the reaction, can be removed from the reactor instantly, although it is a more polar molecule than sorbitol and isosorbide. This avoids the competitive adsorption of water on the surface of the catalyst with high hydrophilicity. As a regeneration treatment in each cycle, the catalyst was washed with water and glycol dimethyl ether. Moreover, the high surface hydrophilicity of SA-SiO2-60.5 catalyst enables the deposited coke, which is hydrophobic, to be washed away easily. So the catalyst activity is restored. Therefore, the physical and chemical properties of SA-SiO2-60.5 are also stable. For example, after recycling ten times, SA-SiO2-60.5 still gave acid site at 790 μmol g−1, which is slightly lower than that of the fresh catalyst (840 μmol g−1). On the contrary, the acid site of SBA-15-SO3H almost decreased to zero, and its recyclability was very poor. From the TGA (Fig. S1†) and FT-IR curves (Fig. 1b), the curves of SA-SiO2-60.5 remain almost unchanged after ten recycles. It can be concluded that SA-SiO2-60.5 is easily regenerated and exhibits good stability.
![]() | ||
Fig. 8 Sorbitol droplet contact angle (CA) on SA-SiO2-60.5 (a) and isosorbide droplet CA on SA-SiO2-60.5 (b). |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra27510e |
This journal is © The Royal Society of Chemistry 2016 |