Preparation of Cs–La–Sb/SiO₂ catalyst and its performance for the synthesis of methyl acrylate by aldol condensation

Abstract

A variety of cesium supported catalysts with amorphous SiO₂ as a carrier were prepared by vacuum impregnation and ultrasonic impregnation methods, which were used for the synthesis of methyl acrylate by aldol condensation of methyl acetate and formaldehyde. The as-prepared catalysts were characterized by XRD, N₂ adsorption–desorption, NH₃-TPD, CO₂-TPD, XPS, ICP and TG. The results indicated that the bifunctional Cs–La–Sb/SiO₂ showed a high conversion of methyl acetate as well as a high yield of methyl acrylate, which was attributed to the weak acid–base sites and the formation of basic Cs–O–Si species with the addition of antimony and lanthanum, respectively. Furthermore, Cs–La–Sb/SiO₂-(V) prepared by vacuum impregnation showed a high initial catalytic activity, but its activity sharply decreased due to the loss of active species. Comparatively, Cs–La–Sb/SiO₂-(U) prepared by ultrasonic impregnation showed a good stability owing to the cavitation effect of ultrasonic waves. The conversion of methyl acetate and yield of methyl acrylate remained above 20% and 9.0% (based on methyl acetate) for 100 h without obvious deactivation. Moreover, carbon deposition was the main factor for the deactivation of Cs–La–Sb/SiO₂-(U), and the coked catalyst could be regenerated by calcination in air.

---

Introduction

As a fundamental industrial monomer, acrylic ester is widely used in the production of plastics, adhesives and coatings^1–5 owing to the conjugated unsaturated system C=C–C=O.⁶ Among the acrylates, methyl acrylate (MA) is the essential monomer for the synthesis of PAN fibres.^7–9 Also, adipic acid, which is an intermediate to manufacture nylon-66, is produced by tail-to-tail dimerization of methyl acrylate.^10–12 Commercially, methyl acrylate is produced by esterification of acrylic acid and methanol, prior to the synthesis of acrylic acid by the gas phase catalytic oxidation of propylene (Scheme 1).^13–17 However, with the declining of propylene resources based on petrochemical industry, it is crucial to develop a new synthetic process of methyl acrylate. The development of coal chemistry industry makes the synthesis of methyl acrylate by vapor phase aldol condensation of methyl acetate and formaldehyde which are the downstream products as a potential way. The mechanism for the reaction as shown in Scheme 2 was reported by Vitcha and Sims in 1966.¹⁸

Scheme 1 Commercial production of methyl acrylate.

Scheme 2 Mechanism of vapor phase aldol condensation.

Two types of heterogeneous catalyst are applied for the formation of acrylates by vapor phase aldol condensation. One is considered as solid acid catalysts (e.g. V₂O₅, P₂O₅);^19–26 the other is solid base catalyst, such as oxides or hydroxides of alkali metals and alkaline earth metals supported on porous materials (e.g. SiO₂, ZrO₂, Al₂O₃, TiO₂, MgO and ZSM-5).^26–35 Nevertheless, more attentions are paid to solid base catalysts for vapor phase aldol condensation because of the easy preparation and high catalytic activity. SiO₂ supported cesium catalyst was reported to be a good solid base catalyst for vapor phase aldol condensation. Mamoru Ai²⁷ studied the vapor-phase aldol condensation of methyl propionate with formaldehyde over a series of silica-supported alkali and alkaline earth metal catalysts and found that silica-supported cesium catalyst showed a higher selectivity than other catalysts. SiO₂ modified by mixed metals was developed to further improve the catalytic performance for vapor phase aldol condensation. Zr–Mg–Cs/SiO₂ catalyst,³⁰ Zr–Fe–Cs/SiO₂ catalyst³¹ and Bi–Cs/SiO₂ catalyst³⁵ exhibited good reusability and long cycle life for the aldol condensation of methyl propionate and formaldehyde.

Ordered mesoporous silica material SBA-15 now received much attention as support for solid base catalyst used in vapor phase aldol condensation thanks to its special two-dimensional hexagonal structure,^31,36,37 and well catalytic performance^31,38 was obtained over this kind of catalyst. Compared to SBA-15, common amorphous SiO₂ was easily synthesized with lower cost. If the catalyst prepared using common amorphous SiO₂ shows the similar catalytic performance, the catalyst would have higher industrial application value.

A large number of papers have reported the preparation of methacrylic acid^26,29,34 and methyl methacrylate^{19,27,28,30,31,33,35} by vapor phase aldol condensation, while little has studied the formation of methyl acrylate by aldol condensation of formaldehyde and methyl acetate. However, vanadium phosphorus oxides as solid acid catalyst was studied for the aldol condensation of formaldehyde and methyl acetate by Ai M.^20,25 and Feng X. et al.²⁴ For the source of formaldehyde, formalin is much more attractive than trioxymethylene and methylal because of the price and available. This paper investigated the catalytic performance of vapor phase aldol condensation of formaldehyde and methyl acetate using the aqueous solution of 36% HCHO as the source of formaldehyde over amorphous SiO₂ supported cesium catalyst prepared by vacuum impregnation and ultrasonic impregnation. Then, lanthanum and antimony species were added to improve the catalytic activity which was the first time for the catalyst used in vapor phase aldol condensation. Also, the deactivation mechanisms of the catalysts prepared by different impregnation methods were discussed.

Experimental

Catalyst preparation

The spherical amorphous SiO₂ support was purchased from Aote Catalysts Company of Jiangyan. Cesium carbonate (Cs₂CO₃, 99.9%, Aladdin), lanthanum nitrate hydrate (La(NO)₃·6H₂O, 99%, Aladdin) and antimony chloride (SbCl₃, 99.0%, Sinopharm Chemical Reagent CO., Ltd) were used as the precursor of cesium, lanthanum and antimony, respectively.

The concentration of the metal salt was determined by the desired metal-loading (4.7 wt% based on the quality of the support for Cs, 2.5 wt% for La, and 2.5 wt% for Sb) and the amount of water (ethanol for SbCl₃) that the support can absorb during the impregnation. The catalysts prepared by vacuum impregnation under 6 × 10⁻² Pa were donated as Cs/SiO₂-(V) and Cs–La–Sb/SiO₂-(V). The catalysts prepared by ultrasonic impregnation for 2 h were donated as Cs/SiO₂-(U), Cs–Sb/SiO₂-(U) and Cs–La–Sb/SiO₂-(U). For Cs–Sb/SiO₂ and Cs–La–Sb/SiO₂ catalysts, the impregnation of antimony, lanthanum and cesium was finished step by step. All the impregnated catalysts were dried overnight in air at 100 °C and then calcined in flowing air at 600 °C for 1 h.

Catalyst characterization

X-ray diffraction (XRD) patterns of the catalysts were characterized by a Bruker Foucs D8 diffractometer with Cu Kα radiation at 40 kV and 2θ region from 10° to 70°.

The N₂ adsorption–desorption isotherm of catalyst was measured by a Surface Area Analyzer (NOVA 2200e, Quantachrome) at 77 K in order to investigate the textural properties of catalysts. The samples were firstly outgassed at 50 °C for 30 min and then at 300 °C (150 °C for deactivated catalysts) for 3 h before measurement. The total surface area was calculated by multi-points BET method using thee adsorption data in the relative pressure range from 0.05 to 0.30. The total pore volume was determined from the amount absorbed at the relative pressure of about 0.99. The pore size distribution curve and mean pore diameter were calculated by the BJH model. Micropore surface area and micropore volume were calculated by t-plot method.

The amount of cesium, lanthanum and antimony in the catalysts was identified by inductively coupled plasma (ICP, OPTIMA2100DV, Perkin Elmer).

The acidity and basicity of the catalysts were measured by temperature-programmed desorption of ammonia (NH₃-TPD) and carbon dioxide (CO₂-TPD) on a PX200 TPD/TPR instrument equipped with a TCD (CHEMBET 3000).

X-ray photoelectron spectroscopy (XPS) measurements were conducted on AXIS Ultra DLD spectrophotometer in order to analyze the surface properties of the catalysts.

Thermogravimetric (TG) analysis of the deactivated catalysts was carried out on a STA409 PC analyzer. The samples were heated from room temperature to 800 °C with a rate of 10 °C min⁻¹ under air in a flow of 30 mL min⁻¹.

GC-MS (Agilent 5973N) was used to determine the composition of soluble coke in deactivated catalysts. The framework of deactivated catalyst was eliminated using 15% HF, and then the soluble coke was extracted with CH₂Cl₂.

Catalyst reaction

Methyl acetate (≥99.0%), formaldehyde (aqueous solution of 36% HCHO) and methanol were analytical grade and all purchased from Sinopharm Chemical Reagent CO., Ltd.

The aldol condensation reaction was carried out in a single pass, fixed bed flow reactor made of a stainless steel tube mounted vertically in the furnace at atmospheric pressure. Formalin, an aqueous solution of HCHO (36%), was used as a source of formaldehyde (HCHO). HCHO was dissolved in methyl acetate (MeOAc) with a molar ratio of 1/10, and methanol (HCHO/CH₃OH = 1, wt) was added into the reaction feed in order to avoid the polymerization of formaldehyde and prevent the formation of side product acetic acid (AA) from hydrolysis of MeOAc. The mixed solution was injected in from the top of reactor using a syringe pump, and nitrogen was fed in from the top of the reactor. The reaction was carried out at 380 °C and 1.0 h⁻¹ WHSV. The product was collected and analyzed by an HP5890 gas chromatograph equipped with a 30 m HP Innowax capillary column and an FID detector.

The activity of catalyst was measured by testing the conversion of MeOAc; the selectivity and yield of MA and acetone which were calculated by equations below.

C_MeOAc (conversion of MeOAc) = (MeOAc_in,mol − MeOAc_out,mol)/MeOAc_in,mol × 100%

S_MA (selectivity of MA) = MA_out,mol/(MeOAc_in,mol − MeOAc_out,mol) × 100%

Y_MA (yield of MA) = C_MeOAc × S_MA

Results and discussions

Physicochemical properties of the prepared catalysts

Fig. 1 shows the XRD patterns of the SiO₂ supported cesium catalysts and SiO₂ support. All the SiO₂ supported cesium catalysts showed a broad peak between 17° and 38° similar to parent SiO₂ support which can be assigned to amorphous silica.³⁹ No new diffraction peak appeared after modified by metal species (cesium, lanthanum and antimony oxides). This indicated that the structure of SiO₂ was preserved well. Also, there were no new phases produced during impregnation process. However, the intensity was slightly decreased for the modified catalysts which indicated that the metallic species were dispersed on the support.

Fig. 1 XRD patterns of SiO₂ and SiO₂ supported cesium catalysts.

The N₂ adsorption–desorption isotherm and pore size distribution curve of SiO₂ support are shown in Fig. 2. The N₂ adsorption–desorption isotherm of SiO₂ support belonged to classic type-IV according to the classification of the International Union of Pure and Applied Chemistry (IUPAC).⁴⁰ There was a clear H3-type hysteresis loop in the isotherm curve at the relative pressure of 0.7–1.0, which indicated the mesoporosity of the SiO₂ support. The pore size distribution curve of SiO₂ support indicated a sharp peak centred at 17.95 nm. The textural properties (surface area, pore volume and mean pore diameter) of SiO₂ supported cesium catalysts and parent SiO₂ are listed in Table 1. After modification of SiO₂ support with metal species, the surface area and pore volume both decreased, which indicated that the metal oxide components were loaded into the channels.³⁰ However, the mean pore diameter of SiO₂ supported cesium catalysts, Cs/SiO₂-(V) and Cs/SiO₂-(U), increased slightly from 17.95 nm to 17.97 nm and 17.99 nm. It was because that the cesium species blocked some micropores of SiO₂ support,³¹ which also caused the decrease of micropore surface and micropore volume. There was a significant reduction in the surface area and pore volume for Cs–La–Sb/SiO₂ owing to the more modification by cesium, lanthanum and antimony. For different catalysts prepared by vacuum and ultrasonic impregnation, catalysts prepared by ultrasonic impregnation possessed slightly lower surface area (114.3 m² g⁻¹, 105.8 m² g⁻¹) and pore volume (0.708 cm³ g⁻¹, 0.671 cm³ g⁻¹) than catalysts prepared by vacuum impregnation (114.9 m² g⁻¹ and 106.4 m² g⁻¹, 0.706 cm³ g⁻¹ and 0.674 cm³ g⁻¹). However, the pore volume of Cs/SiO₂-(U) was higher than that of Cs/SiO₂-(V) due to the instrumental and analysis error. The amount of active components supported on SiO₂ is summarized in Table 2. The ICP results showed that the actual amount of active components supported on SiO₂ was closed to the theoretical loading amount. It can be seen that the amount of the active components in the catalysts prepared by ultrasonic impregnation was higher than the catalysts prepared by vacuum impregnation. Wang et al.⁴¹ studied influences of different impregnation methods for Cs/HX catalyst, and reported that ultrasonic impregnation benefited the sufficient diffusion of impregnation solution into the channels of support. More metal active components supported on SiO₂ also caused lower pore structural parameters of the catalysts prepared by ultrasonic impregnation.

Fig. 2 N₂ adsorption–desorption isotherm and pore size distribution of SiO₂ support.

Table 1 Textural properties of the parent SiO₂ and SiO₂ supported cesium catalysts

Catalysts	Surface area (m^2 g^−1)		Pore volume (nm)		Mean pore diameter (nm)
	Total	Micropore	Total	Micropore
SiO2	156.2	35.28	0.840	0.017	17.95
Cs/SiO2-(V)	114.9	30.75	0.706	0.013	17.97
Cs/SiO2-(U)	114.3	30.01	0.708	0.014	17.99
Cs–Sb/SiO2-(U)	111.2	28.76	0.683	0.012	17.72
Cs–La–Sb/SiO2-(V)	106.4	25.70	0.674	0.012	17.52
Cs–La–Sb/SiO2-(U)	105.8	24.96	0.671	0.012	17.48

---

Table 2 Elemental analysis of the SiO₂ supported cesium catalysts

Catalysts	Cs (wt%)	La (wt%)	Sb (wt%)
Cs/SiO2-(V)	4.46	—	—
Cs/SiO2-(U)	4.67	—	—
Cs–La–Sb/SiO2-(V)	4.48	2.12	2.18
Cs–Sb/SiO2-(U)	4.60	—	2.19
Cs–La–Sb/SiO2-(U)	4.62	2.26	2.19

---

The acidity and basicity of SiO₂ supported cesium catalysts were determined respectively by temperature-programmed desorption of NH₃ and CO₂. Fig. 3 depicts the desorption profiles of NH₃ on different SiO₂ supported cesium catalysts. The amount of acid sites can be obtained from the area of desorption peak. All catalysts exhibited a peak below 300 °C which was assigned to desorption of NH₃ adsorbed on weak acid sites. The NH₃-TPD profiles of Cs/SiO₂ catalysts also showed a high temperature peak at around 350 °C corresponding to strong acid sites generated by the impurities on the surface of SiO₂. Moreover, Cs–Sb/SiO₂-(U) catalyst exhibited more weak and medium acid sites than Cs/SiO₂-(U) attributed to the medium Lewis and Bronsted acidity possessed by antimony species.⁴² While, lanthanum species of Cs–La–Sb/SiO₂ catalysts covered the strong acid sites and resulted in no high temperature desorption peak of NH₃ over Cs–La–Sb/SiO₂ catalysts. Fig. 4 shows the CO₂-TPD results of the SiO₂ supported cesium catalysts. The peak in the range of 150 °C and 300 °C, corresponding to weak basic sites, was exhibited on cesium-supported SiO₂ catalysts. It is worth noting that desorption peak of CO₂ on Cs–La–Sb/SiO₂ catalyst shifted to higher temperature. Meanwhile, the catalyst prepared by ultrasonic impregnation showed the desorption peak at higher temperature than the catalyst prepared vacuum impregnation due to more active components supported on SiO₂ (as shown in Table 2).

Fig. 3 NH₃-TPD profiles of SiO₂ supported cesium catalysts.

Fig. 4 CO₂-TPD profiles of SiO₂ supported cesium catalysts.

The XPS measurement of SiO₂ supported cesium catalyst was investigated to determine the chemical state of metallic species on SiO₂. The wide XPS spectra (Fig. 5A) confirmed that Cs/SiO₂ catalysts contain cesium (724 eV), silicon (103.4 eV) and oxygen (531.6 eV). Meanwhile, lanthanum (834.9 eV) was detected on Cs–La–Sb/SiO₂ catalysts. An overlapping for O 1s and Sb 3d_5/2 photoelectron lines would exist at around 532 eV,⁴³ so the XPS spectra of antimony was not clearly showed in the wide XPS spectra of Cs–La–Sb/SiO₂. The high-resolution Cs 3d_5/2 spectra of cesium-supported catalysts (Fig. 5B) showed three peaks at 724.5 eV, 725.2 eV and 726 eV which can be assigned to Cs₂O₂,⁴⁴ Cs₂O⁴⁵ and Cs₁₁O₃,⁴⁶ respectively. Moreover, all Cs 3d_5/2 spectra of SiO₂ supported cesium catalysts showed a peak at 724 eV corresponded to Cs in SiO₂,⁴⁵ Yan et al. also reported Si–O–Cs species in Cs/SBA-15 catalyst.³⁸ The percentage contents of these cesium species determined by normalization method were listed in Table 3. From the NH₃-TPD profile of Cs–La–Sb/SiO₂ catalyst in Fig. 3, it can be easily seen that the strong acid site on the surface of SiO₂ were neutralized by lanthanum species. The content of base Cs–O–Si structure remarkably increased after addition of lanthanum which changed the surface properties of SiO₂ support as shown in Table 3.

Table 3 The percentage contents of different cesium species on SiO₂ supported cesium catalysts from XPS analysis

Catalyst	Cs–O–Si (wt%)	Cs2O2 (wt%)	Cs2O (wt%)	Cs11O3 (wt%)
Cs/SiO2-(V)	23.34	31.29	39.86	5.51
Cs/SiO2-(U)	28.75	22.60	41.93	6.72
Cs–Sb/SiO2-(U)	30.78	24.51	40.22	4.49
Cs–La–Sb/SiO2-(V)	65.19	16.58	16.40	1.83
Cs–La–Sb/SiO2-(U)	58.01	21.79	18.91	1.29

---

Fig. 5 The wide XPS survey spectra (A) and Cs 3d_5/2 spectra (B) of SiO₂ supported cesium catalysts.

Aldol condensation of formaldehyde and methyl acetate

The catalytic performance of the as-prepared SiO₂ supported cesium catalysts for the vapor phase aldol condensation of formaldehyde and methyl acetate is shown in Fig. 6. Formalin, an aqueous solution of HCHO was chosen as the source of formaldehyde. The hydrolysis of methyl acetate was unavoidable in the presence of water and the basic catalysts. As can be seen, for Cs/SiO₂ catalyst, Cs/SiO₂-(U) catalyst exhibited better catalytic performance than Cs/SiO₂-(V) owing to the more cesium active species supported on the SiO₂. The conversion of methyl acetate and the yield of methyl acrylate both increased over Cs–Sb/SiO₂-(U) catalyst compared with Cs/SiO₂-(U) catalyst. This result comes from that Cs–Sb/SiO₂-(U) catalyst showed more acid–basic sites according to the acidity–basicity analysis of catalysts. In addition, aldol condensation reaction was facilitated by combination of Bronsted acid sites and base sites on the catalyst surface,^19,29 while antimony species brought Bronsted acid sites over the catalyst. Moreover, basic Cs–O–Si structure was formed on Cs–La–Sb/SiO₂ catalyst with the addition of lanthanum, which was benefit for the vapor phase aldol condensation.³⁸ The conversion of methyl acetate at about 20% and the yield of methyl acrylate at above 8.5% were obtained over Cs–La–Sb/SiO₂ catalysts. For Cs–La–Sb/SiO₂ catalyst, Cs–La–Sb/SiO₂-(U) catalyst showed higher yield of methyl acetate owing to its high basicity. From textural properties of catalysts, the surface area of Cs–La–Sb/SiO₂-(U) was lower than Cs–La–Sb/SiO₂-(V), so the diffusion of reactants was more difficult in the channels of Cs–La–Sb/SiO₂-(U) catalyst, which resulted in a slight lower conversion of methyl acetate.

Fig. 6 Catalytic performance of SiO₂ supported cesium catalysts: conversion of MeOAc; yield of MA.

Deactivation and regeneration of Cs–La–Sb/SiO₂

Fig. 7 shows the catalytic stability of Cs–La–Sb/SiO₂-(V) and Cs–La–Sb/SiO₂-(U). As can be seen, an evident decline of catalytic activity over Cs–La–Sb/SiO₂-(V) was observed after time on stream of 5 h, with the conversion of methyl acetate from 18.05% to 16.16% and the yield of methyl acrylate from 7.71% to 5.50%. However, for Cs–La–Sb/SiO₂-(U) catalyst, the conversion of methyl acetate and the yield of methyl acrylate remained at about 20.0% and 90.0% until 100 h, respectively. This result suggested that Cs–La–Sb/SiO₂-(U) showed better stability for aldol condensation of methyl acetate and formaldehyde.

Fig. 7 Catalytic stability of Cs–La–Sb/SiO₂ catalysts for aldol condensation of HCHO and MeOAc: (A) Cs–La–Sb/SiO₂-(V); (B) Cs–La–Sb/SiO₂-(U).

In order to clarify the deactivation mechanism, the crystal structure and pore structure of used catalysts were characterized by XRD and N₂ adsorption–desorption analysis. It can be seen from Fig. 8 that the frameworks of used Cs–La–Sb/SiO₂ catalysts were not destroyed during the reaction and no new phases occurred. Table 4 lists the textural properties (surface area, pore volume and mean pore diameter) of used catalysts. The surface area and pore volume of both the used Cs–La–Sb/SiO₂-(U) and Cs–La–Sb/SiO₂-(V) catalysts reduced because of the carbon deposition on catalysts.

Fig. 8 XRD patterns of the used Cs–La–Sb/SiO₂ catalysts.

Table 4 Textural properties of the used Cs–La–Sb/SiO₂ catalysts

Catalyst	Surface area (m^2 g^−1)		Pore volume (cm^3 g^−1)		Mean pore diameter (nm)
	Total	Micropore	Total	Micropore
Used Cs–La–Sb/SiO2-(V)	96.2	6.27	0.489	0.002	9.572
Used Cs–La–Sb/SiO2-(U)	84.7	5.89	0.403	0.002	9.063

---

From the discussion above, coke deposited on the catalysts was probably a main factor responsible for the deactivation for the deactivation of the catalysts. The thermo stability properties of the coke on the deactivation catalysts were investigated by a TG analyzer. As shown in Fig. 9, there were two distinct weight loss steps displayed on the deactivated catalysts. The first weight loss from room temperature to 300 °C was due to the desorption of physically adsorbed water, and the second weight loss occurred from 300 °C to 550 °C, which was definitely owing to the combustion of deposited coke on deactivated catalysts. The second weight loss in the temperature from 300 °C to 550 °C could be considered as the coke amount on the deactivated catalyst. So, the amount of coke deposition for used Cs–La–Sb/SiO₂-(V) was 1.4 wt%, while 2 wt% coke deposition was obtained over used Cs–La–Sb/SiO₂-(U). The coke may be removed through carbon combustion under 500 °C in air. Also, the nature of soluble coke on catalysts was further analyzed by GC-MS. Fig. 10 showed the compositions of soluble coke. The composition of soluble coke in two deactivated catalysts was mainly aromatic hydrocarbon. Toluene, DMBs, TriMBs and hextraMB were all detected.

Fig. 9 TG profiles of the deactivated Cs–La–Sb/SiO₂ catalysts.

Fig. 10 The composition of soluble coke on the used Cs–La–Sb/SiO₂.

The deactivated catalysts were regenerated at 500 °C for 3 h under a steam of air. The catalytic performances of regenerated Cs–La–Sb/SiO₂-(U) and Cs–La–Sb/SiO₂-(V) catalysts were shown in Fig. 11. It clearly indicated that activity of Cs–La–Sb/SiO₂-(U) was restored. The conversion of methyl acetate was up to 20.36% and the yield of methyl acrylate is up to 9.24%. The conversion of methyl acetate and the yield of methyl acrylate were approximately 12.34% and 4.90%, respectively over regenerated Cs–La–Sb/SiO₂-(V) catalyst. So, the deactivated Cs–La–Sb/SiO₂-(U) catalyst was regenerated completely and the deactivation of Cs–La–Sb/SiO₂-(U) mainly resulted from carbon deposition.

Fig. 11 Catalytic performance of regenerated Cs–La–Sb/SiO₂ catalysts for aldol condensation of HCHO and MeOAc.

However, Cs–La–Sb/SiO₂-(V) catalyst still exhibited unsatisfactory catalytic performance after removing the carbon deposition. Besides carbon deposition, the loss of active component was another important factor that might cause the deactivation of catalysts. The metal contents of used Cs–La–Sb/SiO₂ catalysts were determined by ICP. The results were listed in Table 5, showing the metal content of used Cs–La–Sb/SiO₂-(U) catalyst was almost in accord with the fresh catalyst. While the cesium content of Cs–La–Sb/SiO₂-(V) catalyst sharply decreased after reaction. So, the loss of active component was the main factor of the deactivation of Cs–La–Sb/SiO₂-(V) catalyst prepared by vacuum impregnation. According to the studies of the water solubility and diffusion in alkali silicate melts,⁴⁷ there was a tendency to leach in the presence of water for alkali metal ions in the silicate. The alkali oxide crystallites tended to agglomerate on the support and the agglomeration of cesium compounds was quite hygroscopic⁴⁸ and prone to dissolving in water, which resulted in the loss of cesium. From Table 5, it could easily seen that Cs–La–Sb/SiO₂-(U) prepared by ultrasonic impregnation could prevent the agglomerate of cesium compounds. The physical and chemical effects of ultrasonic cavitation during ultrasonic impregnation had an impact on the property of catalysts. Ultrasonic cavitation could reduce the agglomeration between particles and improve the dispersion of cesium on the surface. In addition, the ultrasonic cavitation could enhance the interaction between the cesium species and the silica. This indicated that the Cs–La–Sb/SiO₂-(U) prepared by ultrasonic impregnation had a good reusability.

Table 5 Elemental analysis of the used Cs–La–Sb/SiO₂ catalysts

Catalysts	Cs (wt%)	La (wt%)	Sb (wt%)
Used Cs–La–Sb/SiO2-(V)	2.84	2.06	2.14
Used Cs–La–Sb/SiO2-(U)	4.53	2.11	2.14

---

Conclusions

Vapor phase aldol condensation of methyl acetate and formaldehyde was investigated over Cs/SiO₂ and Cs–La–Sb/SiO₂ catalyst in a fixed-bed reactor. Better catalytic performance was obtained over the bifunctional Cs–La–Sb/SiO₂ catalyst than Cs/SiO₂ catalyst. The addition of lanthanum resulted in the formation of basic Cs–O–Si structure which was beneficial for the aldol condensation by changing the surface properties of SiO₂ support. Also, lanthanum species covered strong acid sites on catalysts and antimony provided weak acid sites especially for Bronsted acid site. Vacuum impregnation and ultrasonic impregnation methods were applied to prepare Cs–La–Sb/SiO₂ catalyst. Cs–La–Sb/SiO₂-(V) prepared by vacuum impregnation rapidly deactivated because of the loss of cesium species in reaction. However, the catalytic performance was obtained with a good stability over Cs–La–Sb/SiO₂-(U) prepared by ultrasonic impregnation. The cavitation effect of ultrasonic wave could enhance the interaction between the cesium species and SiO₂ support in addition to improving the diffusion of cesium species in silica pores. The conversion of methyl acetate and the yield of methyl acrylate remained at above 20% and 9.0%, respectively. The reason for deactivation of Cs–La–Sb/SiO₂-(U) was carbon deposition formed during the reaction. Furthermore, the catalytic activity of the deactivated Cs–La–Sb/SiO₂-(U) could be regenerated completely by calcination at 500 °C for 3 h in air.

Acknowledgements

This work was financially supported by National Natural Science Foundation of China 20873091.

Notes and references

1. X. Xu, J. Lin and P. Cen, Chin. J. Chem. Eng., 2006, 14, 419–427.
2. A. Niesbach, J. Daniels, B. Schröter, P. Lutze and A. Górak, Chem. Eng. Sci., 2013, 88, 95–107.
3. A. J. Straathof, S. Sie, T. Franco and L. M. van der Wielen, Appl. Microbiol. Biotechnol., 2005, 67, 727–734.
4. J. Dziczkowski, V. Dudipala and M. D. Soucek, Prog. Polym. Sci., 2012, 73, 308–320.
5. I. Degirmenci, T. F. Ozaltın, O. Karahan, V. Van Speybroeck, M. Waroquier and V. Aviyente, J. Polym. Sci., Part A: Polym. Chem., 2013, 51, 2024–2034.
6. P. Vlček and L. Lochmann, Prog. Polym. Sci., 1999, 24, 793–873.
7. E. Hauptman, S. Sabo-Etienne, P. S. White, M. Brookhart, J. M. Garner, P. J. Fagan and J. C. Calabrese, J. Am. Chem. Soc., 1994, 116, 8038–8060.
8. M. Brookhart and S. Sabo-Etienne, J. Am. Chem. Soc., 1991, 113, 2777–2779.
9. A. Behnia, B. Tamaddoni Jahromi and A. Nemati Kharat, J. Coord. Chem., 2014, 67, 3018–3025.
10. E. A. Morris, M. C. Weisenberger, S. B. Bradley, M. G. Abdallah, S. J. Mecham, P. Pisipati and J. E. McGrath, Polymer, 2014, 55, 6471–6482.
11. S. Kim, Y.-S. Kuk, F.-L. Jin and S.-J. Park, Bull. Korean Chem. Soc., 2014, 35, 407–414.
12. Y. Wang, S. Wang and J. Liu, Key Eng. Mater., 2014, 575–576, 151–155.
13. M. M. Lin, Appl. Catal., A, 2001, 207, 1–16.
14. E. Balcells, F. Borgmeier, I. Grißtede, H. G. Lintz and F. Rosowski, Appl. Catal., A, 2004, 266, 211–221.
15. S. Ishikawa, X. Yi, T. Murayama and W. Ueda, Catal. Today, 2014, 238, 35–40.
16. M. Hävecker, S. Wrabetz, J. Kröhnert, L.-I. Csepei, R. Naumann d'Alnoncourt, Y. V. Kolen'ko, F. Girgsdies, R. Schlögl and A. Trunschke, J. Catal., 2012, 285, 48–60.
17. R. Naumann d'Alnoncourt, L.-I. Csepei, M. Hävecker, F. Girgsdies, M. E. Schuster, R. Schlögl and A. Trunschke, J. Catal., 2014, 311, 369–385.
18. J. F. Vitcha and V. A. Sims, I&EC Prod. Res. Dev., 1966, 5, 50–53.
19. J. J. Spivey, M. R. Gogate, J. R. Zoeller and R. D. Colberg, Ind. Eng. Chem. Res., 1997, 36, 4600–4608.
20. M. Ai, J. Catal., 1987, 107, 201–208.
21. M. R. Gogate, J. J. Spivey and J. R. Zoeller, Catal. Today, 1997, 36, 243–254.
22. M. Ai, Appl. Catal., 1989, 54, 29–36.
23. M. Ai, J. Catal., 1990, 124, 293–296.
24. X. Feng, B. Sun, Y. Yao, Q. Su, W. Ji and C.-T. Au, J. Catal., 2014, 314, 132–141.
25. M. Ai, Appl. Catal., 1989, 48, 51–61.
26. M. Ai, H. Fujihashi, S. Hosoi and A. Yoshida, Appl. Catal., A, 2003, 252, 185–191.
27. M. Ai, Appl. Catal., A, 2005, 288, 211–215.
28. M. Ai, Catal. Today, 2006, 111, 398–402.
29. J. Tai and R. J. Davis, Catal. Today, 2007, 123, 42–49.
30. B. Li, R. Yan, L. Wang, Y. Diao, Z. Li and S. Zhang, Catal. Lett., 2013, 143, 829–838.
31. B. Li, R. Yan, L. Wang, Y. Diao, Z. Li and S. Zhang, Ind. Eng. Chem. Res., 2014, 53, 1386–1394.
32. J. Li, J. Tai and R. J. Davis, Catal. Today, 2006, 116, 226–233.
33. P. Wierzchowski and L. Zatorski, Catal. Lett., 1991, 9, 411–414.
34. O. H. Bailey, R. A. Montag and J. S. Yoo, Appl. Catal., A, 1992, 88, 163–177.
35. J. S. Yoo, Appl. Catal., A, 1993, 102, 215–232.
36. R. Martín-Aranda and J. Čejka, Top. Catal., 2010, 53, 141–153.
37. Z. Y. Wu, Q. Jiang, Y. M. Wang, H. J. Wang, L. B. Sun, L. Y. Shi, J. H. Xu, Y. Wang, Y. Chun and J. H. Zhu, Chem. Mater., 2006, 18, 4600–4608.
38. J. Yan, C. Zhang, C. Ning, Y. Tang, Y. Zhang, L. Chen, S. Gao, Z. Wang and W. Zhang, J. Ind. Eng. Chem. DOI:10.1016 /j.jiec.2014.11.014.
39. H. Wang, M. Yu, C. Lin, X. Liu and J. Lin, J. Phys. Chem. C, 2007, 111, 11223–11230.
40. K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl. Chem., 1985, 57, 603–619.
41. W. Qunlong, Z. Guoqing, L. Junhui, W. Yanan, C. Huanhui and Z. Zhirong, Acta Pet. Sin., 2014, 30, 785–791.
42. M. Baca, A. Pigamo, J. L. Dubois and J. M. M. Millet, Catal. Commun., 2005, 6, 215–220.
43. T. Honma, R. Sato, Y. Benino, T. Komatsu and V. Dimitrov, J. Non-Cryst. Solids, 2000, 272, 1–13.
44. L. Soriano, L. Galán and F. Rueda, Surf. Sci., 1991, 251–252, 1075–1080.
45. P. A. W. van der Heide, Surf. Sci., 2000, 447, 62–72.
46. C. C. Phillips, A. E. Hughes and W. Sibbett, J. Phys. D: Appl. Phys., 1984, 17, 611.
47. M. G. Mesko and J. E. Shelby, Phys. Chem. Glasses: Eur. J. Glass Sci. Technol., Part B, 2001, 42, 173–178.
48. J. R. Monnier, J. L. Stavinoha Jr and R. L. Minga, J. Catal., 2004, 226, 401–409.

---