Synthesis of Methyl Methacrylate by Aldol Condensation of Methyl Propionate with Formaldehyde Over Acid–Base Bifunctional Catalysts

Abstract

Supported cesium catalysts with various carriers (SiO₂, Al₂O₃, TiO₂, MgO) were prepared and characterized by X-ray diffraction, BET nitrogen adsorption–desorption, NH₃ and CO₂-TPD methods and thermogravimetric analysis. Experimental results showed that the Zr–Mg–Cs/SiO₂ catalyst exhibited moderate activity for aldol condensation of methyl propionate with formaldehyde (FA) to produce methyl methacrylate. Though the activity of Zr–Mg–Cs/SiO₂ catalyst decreased with the time-on-stream, the deactivated catalyst was completely regenerated by calcination. The catalyst was regenerated 16 times and total operation time was over 500 h, its activity was identical with that of the fresh catalyst.

Graphical Abstract

Gas aldol condensation reaction of methyl propionate with FA to produce MMA is an environmentally friendly technology; it doesn’t need toxic raw materials. Cesium-impregnated silica exhibited better catalytic activity and high selectivity for this reaction, and this catalyst had good reusability and long cycle life. So it is a promising catalyst for industrial application.

1 Introduction

Methyl methacrylate (MMA) and methacrylic acid (MAA) are high-value polymer intermediate, which are widely used for producing polymethyl methacrylate (so-called organic glasses) or producing polymer dispersions for paints and coating [1]. Most of MAA and MMA are produced by the traditional Acetone Cyanohydrin process (ACH). In this route, acetone and hydrogen cyanide react to product cyanohydrin, which subsequently react with sulfuric acid to form methacrylamide sulfate. The methacrylamide sulfate is either hydrolyzed to MAA and ammonium sulfate or converted to a mixture of MAA and MMA in a combined hydrolysis–esterification step with methanol. However, the ACH route needs to utilize highly toxic hydrogen cyanide and corrosive, concentrated sulfuric acid that are extremely harmful to human and environment. In recent years, the environmental pressures against shipment of HCN have forced many manufacturers to abandon utilizing HCN. Besides, disposal of spent sulphuric acid/ammonium sulphate has been another source of headaches which needs to the installation of device for thermal decomposition of the salt and recovery of sulphuric acid for recycling [2, 3]. Therefore, many researchers make a lot of efforts to develop an environmentally friendly product of MMA. Therefore, much patented work has been devoted to investigation on the aldol-condensation reaction of propionate derivatives (propionic acid and methyl propionate) with formaldehyde to produce MAA and MMA [4–6].

$${\text{CH}}_{ 3} {\text{CH}}_{ 2} {\text{COOH + HCHO}} \to {\text{ CH}}_{ 2} {\text{=C(CH}}_{ 3}^{{}} ) {\text{COOH + H}}_{ 2} {\text{O}}$$

$${\text{CH}}_{ 3} {\text{CH}}_{ 2} {\text{COOCH}}_{ 3} {\text{ + HCHO }} \to {\text{ CH}}_{ 2} {\text{=C(CH}}_{ 3} ) {\text{COOCH}}_{ 3} {\text{ + H}}_{ 2} {\text{O}}$$

It is known that the synthesis of MAA and MMA is much more difficult than that of acrylic acid and methyl acrylate obtained by aldol condensation of acetic acid and methyl acetate, respectively, with HCHO. Both acid and base catalysts are known to be active in the reaction. About the reaction between PA and HCHO, plenty of papers and patents have been published [7–16]. On the other hand, there are limited reports on the reaction between MP and HCHO. Gogate et al. [17] reported a 10.5 % yield of MMA (based on MP) in the aldol condensation of MP with HCHO over a V–Si–P ternary catalyst. Ai [18] reported the conversion reached 14.2 % (based on MP) and 85.9 % of selectivity to MMA over a silica-supported cesium hydroxide catalyst. The mechanism for the reaction is as follows [17]:

• Activation of reactant by basic sites of the catalyst

  $${\text{R}}{-}{\text{CH}}_{ 2}{-}{\text{X + B}} \to {\text{R}}{-}{}^{ - }{\text{CH}}{-}{\text{X + BH}}^{ + } ( {\text{B = O}}{}^{ - }{\text{ or OH}}^{ - } , {\text{ R = CH}}_{ 3} , {\text{ X = COOCH}}_{ 3} )$$

• Activation of HCHO by acidic sites of the catalyst

  $${\text{HCHO + H}}^{ + } \to {\text{H}}_{ 2} {\text{C}}^{ + } {\text{OH}}$$

• Reaction of two intermediate molecules to form an aldol followed by dehydration

  $${\text{R}}{-}{}^{ - }{\text{CH}}{-}{\text{X + H}}_{ 2} {\text{C}}^{{^{ + } }} {\text{OH}} \to {\text{H}}_{ 2} {\text{C(OH)}}{-}{\text{C(H)R}}{-}{\text{X}} \to {\text{H}}_{ 2} {\text{C = CR}}{-}{\text{X + H}}_{ 2} {\text{O}}$$

Although the condensation route involves fewer steps than the ACH process, a major disadvantage is that the single-pass yield of MMA in the condensation reaction is relatively low, resulting in extensive recycle of the unreacted starting materials. Here our contribution is aimed to prepare more effective catalysts for the production of MMA.

2 Experimental

2.1 Catalyst Preparation

Silica and γ-alumina were obtained from commercial companies. For the Zr–Mg–Cs/SiO₂ catalyst, silica was dried first at 500 °C for 3 h and then impregnated with an aqueous solution containing desired amount of zirconium nitrate, magnesium nitrate and cesium nitrate. This mixture was oscillated at 40 °C for 24 h and then dried in an oven. The resulting solid was calcined in flowing air at 350, 400, 450, 500, 550 and 600 °C for 3 h, respectively. Cesium-impregnated alumina (Zr–Mg–Cs/Al₂O₃) was prepared in a similar method and calcined in flowing air at 450 °C. For Zr–Mg–Cs/TiO₂ catalyst, aqueous ammonia (28 %) was added dropwise with vigorously stirring into a solution of 0.2 mol/L Ti(SO₄) to obtain hydroxides until pH reached 8. The solution containing precipitation was kept in a water bath warmed at 70 °C for 5 h, and then filtered and washed with deionized water, and dried at 110 °C for 12 h. The resulting powder was impregnated with a mixed aqueous solution of zirconium nitrate, magnesium nitrate and cesium nitrate under continuous stirring at room temperature for 5 h and then dried at 110 °C for 12 h. The prepared precursor was calcined in flowing air at 450 °C. The Zr–Mg–Cs/MgO catalyst was prepared by the similar process with Zr–Mg–Cs/TiO₂ catalyst. Adding elements magnesium and zirconium, can effectively improve the performance of the catalyst. For all these catalysts, the content of zirconium and magnesium were 0.033 and 0.025 wt%, respectively [13]. The loading amounts of the catalysts were based on the quality of the carriers.

2.2 Catalyst Characterization

X-ray diffraction (XRD) patterns of catalysts were recorded by X’Pert PRO MPD diffractometer operated at an accelerating voltage of 40 kV and an emission current of 40 mA with Cu Kα radiation. The angle (2θ) was measured in steps of 0.41778^o/s between 5^o and 90^o.

Weight loss and temperature associated with phase transformation were determined by thermogravimetry and differential thermal analysis (TG/DTA) on a Shimadzu DTG-60H analyzer. The fresh samples of catalysts before calcination were heated from room temperature to 800 °C at rate of 10 °C/min in nitrogen flow (30 mL/min).

The BET specific surface area and the pore size were obtained from nitrogen absorption and desorption isotherm method at liquid nitrogen temperature on a Quanta Chrome Instrument NOVA 2000. Prior to analysis, samples were degassed at 300 °C for 10 h.

The adsorption of CO₂ and NH₃ was measured on an Autochem II 2920 apparatus from Micromeritics. In a typical experiment, 40 mg of samples were placed in a quartz tube and was pretreated in a helium flow at 450 °C at a heating rate of 10 °C/min and the sample was kept at this temperature for 1 h. The temperature was then reduced to 50 °C prior to adsorption. 10 % NH₃–He or 10 % CO₂–He passed over the samples for 30 min at 50 °C. After purging with pure helium at 50 °C for 1 h until the baseline was stable, the desorption profile was measured by the thermal conductivity detector in helium flow at a heating rate of 10 °C/min to 500 °C.

The cesium, magnesium and zirconium in the Zr–Mg–Cs/SiO₂ were measured using inductively coupled plasma (ICP). A SHIMADZU ICPE-9000 was used for the measurement.

2.3 Reactivity Measurement

SiO₂ (20–30 mesh) was obtained from Qingdao Hailang Silica Gel Drier Factory. Methyl propionate (≥99.0 %) and paraformaldehyde (≥94.0 %), Magnesium nitrate (≥99.0 %) and zirconium nitrate (≥99.0 %), ammonia solution (25 %) and ammonium sulfate (≥99.0 %) are analytical grade. Titanium(IV) sulfate (≥96.0 %) is chemical grade. Cesium nitrate (99.9 %) is industrial grade.

The reaction was carried out in a fixed-bed reactor at atmospheric pressure. The reactor was made of a stainless steel tube, 42 cm long and 0.8 cm inside diameter, mounted vertically in the furnace. The amount of catalyst was 4 mL and was placed in the middle of reactor. Quartz sands were placed both under and above the catalyst sample. The reaction temperature was in the range of 300–400 °C. The mixed solution of MP, FA and methanol was fed in from the top of the reactor by an advection pump with the feed rate of 0.1 mL/min. The MP/FA/CH₃OH molar ratio was fixed at 1/2/2, unless otherwise indicated. Nitrogen was fed in from the top of the reactor with a fixed feed rate. The contact time was 1.9 s. The samples were collected after 100 min on stream.

In order to reduce the impact of water in the reaction and avoid the hydrolysis of MP and MMA, paraformaldehyde was dissolved in methanol to obtain mixture of formaldehyde and methanol, which was used as a source of HCHO instead of formalin.

The effluent gas from the reactor was led into a condenser and the liquid samples were collected per 45 min interval. After 100 min on run, the reactants and products of gas phase and liquid phase were analyzed by gas chromatography (Agilent 7890A GC). N-Heptane was added to the products as an internal standard. The activity of catalyst was measured by testing the conversion of MP; the selectivity and yield of MMA and MAA which were calculated in Eqs. 1–5.

$${\text{Conversion of MP}} = \frac{{{\text{MP}}_{\text{in,mol}} - {\text{MP}}_{\text{out,mol}} }}{{{\text{MP}}_{\text{in,mol}} }} \times 100 $$

(1)

$${\text{Selectivity to MMA}} = \frac{{{\text{MMA}}_{\text{out, mol}} }}{{{\text{MP}}_{\text{in,mol}} - {\text{MP}}_{\text{out,mol}} }} \times 100 $$

(2)

$${\text{Selectivity to MAA}} = \frac{{{\text{MAA}}_{\text{out, mol}} }}{{{\text{MP}}_{\text{in,mol}} - {\text{MP}}_{\text{out,mol}} }} \times 100 $$

(3)

$${\text{Yield of MMA}} = {\text{Conversion of MP}} \times {\text{Selectivity of MMA}} $$

(4)

$${\text{Yield of MAA}} = {\text{Conversion of MP}} \times {\text{Selectivity of MAA}} $$

(5)

3 Results and Discussion

3.1 Catalysts Characterization

3.1.1 XRD Results

Figure 1a presents the XRD patterns of the prepared catalysts with different supports (SiO₂, Al₂O₃, TiO₂ and MgO). The catalysts were prepared with 15 wt% cesium-loading and calcined at 450 °C for 3 h. The cesium nitrate with hexagonal were observed in all catalysts, but few peaks of cesium nitrate were exhibited on the Zr–Mg–Cs/Al₂O₃ catalysts. Figure 1b presented the XRD patterns of Zr–Mg–Cs/SiO₂ catalyst at different calcination temperatures. The cesium nitrate with hexagonal phase was observed in all catalysts, but the intensity of cesium nitrate decreased with increasing calcination temperature. This was attributed to cesium nitrate decomposed and volatilized with increase of the calcination temperature; this phenomenon was also confirmed by TG/DTA analysis.

Fig. 1

XRD patterns of the catalysts with different supports (a) and calcined at different temperatures (b): (hash) CsNO₃, hexagonal, (asterisk) Al₂O₃, rhombohedral (dollar) MgO, cubic, (ampersand) TiO₂, tetragonal

3.1.2 Thermal Analysis (TG/DTA)

The thermostability of the Zr–Mg–Cs/SiO₂ catalysts was investigated by TG/DTA method. The thermal stability of cesium nitrate was studied and Fig. 2 showed the TG/DTA curves of CsNO₃. The first endothermic DTA peak turned up at 158 °C with no weight loss, which attributed to melt of impurities in the cesium nitrate. The second big endothermic DTA peak at 408 °C with no weight loss was due to melting of CsNO₃. It was closed to the melting point of CsNO₃ which was 414 °C reported in the literature. The weight loss was 97 % between 500 and 800 °C accompanied with the two broad endothermic DTA peak, which was attributed to decomposition and volatilization of CsNO₃. For the sample of Zr–Mg–Cs/SiO₂ calcined at 450 °C (Fig. 3a), the weight loss was 1.5 % between 50 and 150 °C with a small endothermic DTA peak at 55 °C, which was attributed to the loss of physical adsorbed water molecules. The second weight loss was 5 % between 500 and 650 °C accompanied a broad endothermic DTA peak at 550 °C, which was attributed to decomposition and volatilization of CsNO₃ in the Zr–Mg–Cs/SiO₂ catalyst. For the sample of Zr–Mg–Cs/SiO₂ calcined at 600 °C (Fig. 3b), there was no visible weight loss between 500 °C and 800 °C, this indicates the Zr–Mg–Cs/SiO₂ catalyst calcined at 600 °C, the CsNO₃ had completely decomposed. Thus the thermogravimetry studies revealed that the Zr–Mg–Cs/SiO₂ catalyst should be prepared and used below 500 °C.

Fig. 2

TG/DTA curves of CsNO₃

Fig. 3

TG/DTA curves of the Zr–Mg–Cs/SiO₂ calcined at: a 450 °C; b 600 °C

3.1.3 BET Specific Surface Area Analysis

The physical properties (BET specific surface area, pore volume and mean pore diameter) of the catalysts were presented in Tables 1, 2 and 3, respectively. Table 1 showed the commercial SiO₂ and γ-Al₂O₃ had large specific surface area, while the specific surface area of prepared TiO₂ and MgO were relatively small. The specific surface area of Zr–Mg–Cs/SiO₂ decreased with increasing the calcination temperature, while the mean pore diameter gradually increased (see Table 2). The specific surface area slightly decreased below 500 °C, but the specific surface area sharply declined over 500 °C. This phenomenon may be due to the destruction of pores of SiO₂ at higher temperature. In Table 3, the specific surface area of the catalysts decreased with increasing the content of cesium nitrate. It indicated that cesium nitrate loaded in SiO₂ occupies some channels.

Table 1 Textural and acid–base property of the catalysts with different supports

Table 2 Textural and acid–base property of the Zr–Mg–Cs/SiO₂ catalysts calcined at different temperatures

Table 3 Textural and acid–base property of the Zr–Mg–Cs/SiO₂ catalysts loaded with different amount of cesium

3.1.4 Acidity and Basicity Analysis

The quantitative analysis of base and acid sites on the surfaces of catalysts were determined by stepwise temperature-programmed desorption (TPD) of CO₂ and NH₃. The peaks showed in each of the profiles correspond to desorption of CO₂ and NH₃ related to the base and acid sites of the catalysts surface. The desorption temperature indicates the base and acid strength of the catalysts. The higher temperature of desorption, the stronger base and acid strength of the catalysts [19, 20].

The characteristic CO₂ and NH₃ TPD patterns for each of the four catalysts in Table 1 had been compared in Fig. 4. Figure 4a gave the CO₂ TPD patterns (measuring basicity). The four catalysts had a distinct CO₂ desorption peak between 50 and 250 °C which corresponded to the weak strength base sites. The number density of base sites was greater on the γ-Al₂O₃ and TiO₂ samples than on the SiO₂ sample. Figure 4b showed the acidity of catalysts by NH₃ TPD patterns. The TiO₂ sample presented a desorption peak at 130 °C; the SiO₂ and Al₂O₃ samples presented desorption peaks at 150 °C; and the desorption peak of MgO sample was at 220 °C; which all corresponded to the weak strength acid sites, however, the MgO sample showed a little stronger acidity than the other ones. This may be due to the following reasons. The MgO carrier had minimum specific surface area (14.7 m²/g); which could be coated by cesium nitrate, so the base sites of MgO were completely covered by cesium nitrate. Cesium nitrate presented a weak acidic substance. Therefore, the acidity of the Zr–Mg–Cs/MgO sample corresponded to the acid sites of cesium nitrate. The γ-Al₂O₃ sample presented another NH₃ desorption peaks (300–500 °C), which corresponded the middle strength acid sites. The acid sites density on the TiO₂ sample was greater than the other samples.

Fig. 4

TPD curves of the catalysts with different supports in atmosphere of: a CO₂; b NH₃

The characteristic CO₂ and NH₃ TPD patterns for Zr–Mg–Cs/SiO₂ catalysts at different calcination temperature were compared in Fig. 5. Figure 5a showed the desorption profiles of CO₂ from Zr–Mg–Cs/SiO₂ catalysts. All catalysts exhibited a prominent CO₂ desorption peak in the range of 50–250 °C, the basic sites on these catalysts were considered to be weak. Increasing calcination temperature of catalyst decreased the CO₂ adsorption capacity, indicating some base sites were disappeared on the catalyst surface. Figure 5b showed the desorption profiles of NH₃ from Zr–Mg–Cs/SiO₂ catalysts. The NH₃ adsorption capacity increased with increasing calcination temperature.

Fig. 5

TPD curves of the Zr–Mg–Cs/SiO₂ catalyst calcined at different temperatures in atmosphere of: a CO₂; b NH₃

Figure 6 compared the desorption profiles of CO₂ and NH₃ from SiO₂ supported different loadings of cesium catalysts. Figure 5a gave the CO₂ TPD patterns. Adding more cesium to the SiO₂ increasing its CO₂ adsorption capacity significantly, indicating new base sites were formed on the surface. Figure 5b gave the NH₃ TPD patterns. Adding cesium only slightly affect the NH₃ adsorption capacity. The total amount of acid and base sites on the catalysts surface were summarized in Tables 1, 2 and 3, respectively. From the view of acid–base analysis, the presence of the low-temperature peak in the range of 50–250 °C directly related to the MMA yield, the amount of base and acid sites directly corresponded to the MP conversion.

Fig. 6

TPD curves of the Zr–Mg–Cs/SiO₂ catalysts with different amount of cesium in atmosphere of: a CO₂; b NH₃

3.2 Aldol Condensation of MP with FA

3.2.1 Effects of the Supports

The effects of the supports on the catalytic activity were summarized in Table 4. The catalysts were prepared with 15 wt% cesium-loading and reaction was performed at 360 °C. The samples were collected after 100 min on stream. In this paper, 15 wt% cesium-loading were based on the quality of the carriers. In fact, the theoretical content of cesium was 12.3 % in the entire catalyst. The actual cesium-loading in different carriers were measured by XRF and the results were presented in Table 5. The actual amounts of Cs contained in the SiO₂ and MgO samples were 11.5 and 12.1 %, respectively. These were closed to the theoretical content of cesium (12.3 %).

Table 4 Effect of the supports on the aldol condensation of MP with FA

Table 5 Property of the catalysts with different carriers

The Zr–Mg–Cs/SiO₂ catalyst exhibited the best activity for the reaction, and the Zr–Mg–Cs/Al₂O₃ catalyst showed moderate activity. Although the Zr–Mg–Cs/TiO₂ catalyst was active for the conversion of MP, its selectivity to MMA was very poor. The Zr–Mg–Cs/MgO catalyst exhibited little activity for the reaction. In order to maintain constant of residence time, the volume of catalyst should be uniform, so 4 mL of catalysts were used for all reaction. However, due to differences in density of the carriers, the qualities of 4 mL catalysts with different carriers were diverse. Their quality and cesium-loading were summarized in Table 5. For the sake of comparing catalytic activity of active component (Cs) in different carriers, the conversion of MP was also calculated by the unit mass active component (Cs) of the catalyst and the results were showed in Table 5. The conversion of MP on per gram cesium was 0.084 mol/h in the SiO₂ catalyst, which was higher than the other ones. It was indicated that the Zr–Mg–Cs/SiO₂ catalyst had the best catalytic activity. From the results of BET, the SiO₂ and Al₂O₃ samples had large specific surface area and mesoporous about 10 nm; while the specific surface area of TiO₂ and MgO samples was relatively small. It indicated the large specific surface area corresponded to the high yield. From the acid/base analysis, these four catalysts had almost the same strength of base sites, but the strength of acid sites was significantly different. Compared with the Zr–Mg–Cs/SiO₂ sample, the acid strength of Zr–Mg–Cs/TiO₂ sample was weak, while the acid strength of Zr–Mg–Cs/MgO sample was stronger. The mechanism of condensation clearly implied that acidic and basic properties were both needed for a moderate catalyst; and the condensation reaction may be facilitated by a combination of a Brønsted acid site and a base site on the catalyst surface [2]. Our results suggested there may be a possible correlation between the acid–base site distribution and the condensation yield, the acid sites at description temperature of 50–300 °C was desired for the selectivity to MMA. In the cases of TiO₂ and MgO catalysts, large amounts of MP were consumed without forming MMA. So we used silica as support to conduct the research on the aldol condensation of MP with FA.

3.2.2 Effects of Calcination Temperature

Calcination temperature was an important factor in preparing catalysts since it significantly affected the structure and catalytic properties of catalysts. The Zr–Mg–Cs/SiO₂ catalysts with 15 wt% cesium-loading were calcined at 350, 400, 450, 500, 550 and 600 °C for 3 h, respectively. The reaction was performed at 360 °C.

The effect of calcination temperature on catalytic activity was showed in Fig. 7. The conversion of MP changed slightly at the calcination temperature below 500 °C and was around 55 %, but more than 500 °C, the conversion of MP decreased rapidly. The selectivity of MMA was not affected by the calcination temperature and was around 73 %. From the textural analysis, the specific surface area decreased slightly on the catalysts at the calcination temperature below 500 °C, while it decreased sharply over 500 °C. From the view of TG/DTA, acidity and basicity analysis, cesium nitrate will decompose when the calcination temperature was over 500 °C, and the total amount of base sites on the catalysts surface decreased when increasing the calcination temperature. The decline of surface area and the total amount of base sites with increasing of the calcination temperature might be main factor for the decrease of catalytic activity on the conversion of MP.

Fig. 7

Effects of calcination temperature on the catalytic performance. Reaction was performed at 360 °C; molar ratio of MP/HCHO was 1/2; feed rate of 0.1 mL/min

3.2.3 Effects of the Amount of Cesium Supported on Silica

The reaction was performed on the Zr–Mg–Cs/SiO₂ catalysts with six different cesium loading, namely, 0, 3, 5, 10, 15 and 20 wt%. All catalysts were calcined at 450 °C. The catalytic activity of the silica catalysts with different amounts of cesium were showed in Fig. 8. The conversion of MP increased rapidly with increasing the loading of the cesium when the cesium loading was below 15 wt%. It appeared to reach a plateau at 15 wt% loading. The conversion of MP increasing slightly when the cesium loading from 15 wt% up to 20 wt%. However, Bailey et al. [21] reported that the catalytic activity for aldol condensation of PA with FA went through a maximum at a cesium loading about 5 % by weight. Tai and Davis [2] also observed that the conversion of PA over the Cs-loaded silica catalysts appeared to reach a plateau at 4 wt% loading for the aldol condensation of PA with FA. This difference in loading of cesium was attributed to the following reasons. First, the silica which we used had a larger surface area than those of Bailey and Tai. Second, it was believed that the aldol condensation reaction with MP was more difficult than that with PA. From the textural, acidity and basicity analysis, the amount of acid/base sites increased with the increasing of cesium. It provided more active sites for the reaction. These may be the main reason of high conversion for catalysts with high cesium content. The specific surface area decreased with increasing cesium loading, while the catalytic activity gradually strengthened. It indicated the specific surface area was not a key factor for the reaction in our experimental condition.

Fig. 8

Effects of the amount of cesium on the catalytic performance. Reaction was performed at 360 °C; molar ratio of MP/HCHO was 1/2; feed rate of 0.1 mL/min

3.2.4 Effects of Reaction Temperature

The reaction was performed on the Zr–Mg–Cs/SiO₂ catalyst which was calcined at 450 °C with 15 wt% cesium. The molar ratio of MP/FA/CH₃OH was 1/1/1.5; the reaction temperature range from 300 to 400 °C. The result was showed in Fig. 9. The conversion of MP increased with increasing of the reaction temperature, while the selectivity proceeded with the opposite trend. This was attributed to that the more by-products were produced at the higher reaction temperature, so the selectivity of MMA decreased. The maximum yield of MMA was 32.9 % at 380 °C and the selectivity of MMA was 70.3 %.

Fig. 9

Effects of reaction temperature on the catalytic performance. molar ratio of MP/HCHO was 1/1; feed rate of 0.1 mL/min

3.2.5 Effects of the MP/HCHO Molar Ratio

The reaction was performed at 360 °C at Zr–Mg–Cs/SiO₂ catalysts with 15 wt% cesium. The effect of the MP/HCHO molar ratio on the reaction was summarized in Table 6. With the increasing of the MP/HCHO molar ratio, the conversion of MP increased from 36.2 to 55.4 %. When the MP/HCHO molar ratio was 1/1, the selectivity to MMA was the best and achieved to 80.3 %.

Table 6 Effect of the MP/HCHO molar ratio on the aldol condensation of MP with FA

3.3 Catalyst Stability and Reusability

The stability and reusability of Zr–Mg–Cs/SiO₂ catalyst (cesium loading: 15 wt% and calcined: 450 °C) was also investigated. The reaction was performed at 320 °C; the molar ratio of MP/FA was 1/1. Figure 11a showed the yield of MMA based on the MP feed as function of the time-on stream. The activity dropped by nearly 50 % after reacting of 32 h. The deactivation mechanism may be due to the formation of carbonaceous deposits on the surface of the catalyst. In fact, the catalyst turned black after use [22, 23]. The carbon content of the fresh and used catalysts was measured by carbon/sulfur analyzer (CS-344, US LECO Corporation). The carbon content of the fresh catalyst was 0.07787 %, while the carbon content of the used catalyst was 3.8461 %. The used catalyst was characterized by TG/DTA (Fig. 10), different from the fresh catalyst, the used catalyst had a distinct weight loss from 300 to 450 °C accompanied with a big exothermic DTA peak at 405 °C, which mainly ascribed to the coke combustion with O₂.

Fig. 10

TG/DTA curves of the used Zr–Mg–Cs/SiO₂ catalyst after reaction

In order to investigate the reusability of the catalyst, the used catalyst was regenerated by calcination in a stream of mixture of oxygen and nitrogen at 450 °C for 3 h, the volume of oxygen was 1 %. The regenerated catalyst was investigated according to the same conditions with the fresh catalyst by reaction 32 h. The catalyst was repeatedly regenerated 16 times and total operation time was over 500 h. Figure 11b showed the catalytic activity of the repeatedly regenerated catalyst. The catalytic properties can be completely restored after regeneration. From the ICP analysis, the content of cesium, magnesium and zirconium in the fresh catalyst and regenerated catalyst (after the 16th regenerated) was showed in Table 7. It indicated that the components of regenerated catalyst did not lose over 500 h. So after the sixteenth regeneration, its activity was still the same as that of the fresh catalyst, which demonstrated that the Zr–Mg–Cs/SiO₂ catalyst had good reusability.

Fig. 11

Yield of MMA on the fresh (a) and regenerated catalyst (b). Reaction was performed at 320 °C; molar ratio of MP/HCHO was 1/1; feed rate of 0.1 mL/min

Table 7 Element content of the fresh and regenerated catalyst

4 Conclusions

The gas aldol condensation reaction of MP with FA to produce MMA was investigated over various catalysts. The cesium-impregnated TiO₂ and MgO carriers presented little activity for the reaction. The conversion and selectivity over the Zr–Mg–Cs/Al₂O₃ catalyst were relatively low. The cesium-impregnated SiO₂ exhibited better catalytic activity and high selectivity for the reaction. The characterization results suggested that the calcination temperature and content of cesium were main factors to affect the textural properties, surface acidity and basicity and catalytic activity of the catalysts. Although the single life of this catalyst was slight short due to carbon deposition on the surface of catalyst, its catalytic activity was completely regenerated by a simple calcination. The Zr–Mg–Cs/SiO₂ catalyst which had good reusability and long cycle life was easily to prepare. Therefore, the Zr–Mg–Cs/SiO₂ catalyst was a promising catalyst for the aldol condensation of MP with FA to produce MMA.