Highly Dispersed ZrO₂ Modified Cs/SiO₂-Based Catalyst via Ligand–Assisted Strategy for Aldol Condensation of Methyl Acetate with Formaldehyde

Abstract

A kind of highly dispersed ZrO₂ modified Cs/SiO₂-based catalyst was prepared via ligand–assisted impregnation strategy. The catalyst was then used to the aldol condensation reaction of methyl acetate (MAc) with formaldehyde (FA) to produce methyl acrylate (MA). The effect of highly dispersed ZrO₂ promotion on Cs/SiO₂ catalyst selectivity and activity was studied as well. Various techniques such as XRD, FT-IR, TEM, XPS, NH₃-TPD, CO₂-TPD, N₂ physical adsorption, and solid-state MAS NMR were used to characterize the structural features, Cs species states, and acid–base properties of these catalysts. The use of Zr(acac)₄ as a zirconium source for modifying Cs/SiO₂ could be considered optimal for ligand-assisted impregnation. By modifying the highly dispersed ZrO₂, more Cs–O–Si species and a suitable balance of acid–base sites were formed in the Zr–Cs/SiO₂ catalyst. As a result, the catalyst showed significant improvements in selectivity and yield of MA under optimized reaction conditions. This study is beneficial for the development and optimization of catalysts for aldol condensation and provides a reference for future industrial applications.

Keywords:

• Cs/SiO₂ catalyst
• ZrO₂ modification
• Aldol condensation
• Methyl acrylate
• Acid–base synergy

Introduction

Vapor phase aldol condensation of methyl acetate (MAc) with formaldehyde (FA) to prepare methyl acrylate (MA) is regarded as a high-potential MA production process, because it can offer significant economic and environmental benefits compared with other traditional process routes such as the acetone cyanohydrin (ACH) process (Ma et al. 2021; Wang et al. 2022). The feedstock of this process can be acquired easily through a range of sources such as coal chemicals, petrochemical, biomass and so on, which makes this process generally better insulated from oil-price fluctuations and depletion of oil resources (Bao et al. 2018; Zhao et al. 2016).

Obviously, the aldol condensation of MAc with FA emerges as a much more thermodynamically favorable reaction with an apparent negative Gibbs free energy change (Δ_rG_T^θ) of the reaction (Guan et al. 2020). However, it still needs to be carried out at 350–380 °C with the participation of suitable catalysts because of the more difficult α-deprotonation of MAc (Wang et al. 2022). Various attempts have been made to find suitable catalyst systems for this aldol condensation, which could be sorted into V-P-O base catalysts and Cs-base catalysts.

Due to its unique interfacial structure and properties, VPO-based catalysts can activate the C–H bonds of short-chain hydrocarbons and have been found to be one of the most effective catalysts in aldol condensation reactions (Feng et al. 2014; Hu et al. 2016). The composition of the active phase and the acidity and basicity of the V-P-O have a significant impact on the conversion and selectivity of aldol condensation, which can be controlled by the P:V atomic ratio, calcination temperature and activation atmosphere (Guo et al. 2017; Shen et al. 2019) Nevertheless, the applicability of the V-P-O base catalysts is unsatisfactory due to the restriction of the catalyst scale-up preparation and rapid deactivation (Wang et al. 2022).

Among a series of alkali metal catalysts, Cs-base catalysts possess excellent catalytic performance in the aldol condensation reaction. Due to the limitation of the low specific surface area, Cs species must be loaded on porous catalyst supports for high dispersion and stability (Ma et al. 2021). Common porous Cs species supports include SiO₂ (Deng et al. 2022)_, Al₂O₃ (Wang and Cai 2020) and zeolite (Yan et al. 2015), which could provide the surface –OH to form Si–O–Cs or Al–O–Cs groups with Cs species to activate MAc (Wu et al. 2022). With the advancement of research, scientists have conducted a series of studies on acid–base bifunctional Cs base catalysts and found that the acid–base balance of the supported Cs species catalyst could be regulated effectively by adding promoter elements, for example Cs–P/γ–Al₂O₃ (Wang and Cai 2020), Cs–Zr/SiO₂ (Li et al. 2019), Cs–Si/Al₂O₃ (Wu et al. 2022). The promoter elements not only promote the generation of Cs species on the surface of supports, but also increase the weak/strong acids and bases ratio which are crucial for MA production (Darabi Mahboub et al. 2018).

Among them, ZrO₂, as an amphoteric metal oxide, is often used for surface modification of catalysts (Dagle et al. 2020). The forms and dispersion of ZrO₂ will determine the quantity and distribution of surficial active sites on the catalyst, which are influenced by catalyst modification and preparation methods (Salinier et al. 2009). Ligand-assisted preparation method is an extremely efficient approach for controlling the size and dispersion of loaded metals or metal oxides (Costa and Rossi 2012; Zhang et al. 2017). Therefore, organic ligands are employed to hinder the growth process of ZrO₂ particles, control the size of ZrO₂, and maintain its high dispersibility through the steric hindrance effect (Salinier et al. 2009). However, the effect of highly dispersed ZrO₂ on aldol condensation over Cs/SiO₂ catalysts has rarely been studied.

Herein, we fabricated highly dispersed ZrO₂ modified Cs/SiO₂ catalysts by ligand–assisted impregnation strategy to enhance the catalytic activity of aldol condensation reaction of MAc with FA. Compared to traditional impregnation process using Zr(NO₃)₄, in ligand–assisted impregnation strategy, Zr(acac)₄ and Zr(O-n-Bu)₄ were used as zirconium sources. The physical − chemical features and acid–base properties were comprehensively analyzed through a series of characterization methods. Cs₂O and ZrO₂ loading and reaction conditions were optimized and the catalyst demonstrates excellent catalytic performance and stability. A comprehensive investigation was conducted to analyze and characterize the performance and properties of the catalyst.

1. Experimental Section

1.1. Materials and Chemicals

The spherical SiO₂ support was purchased from Qingdao Hailang Silicone Co. Ltd. The spherical SiO₂ was ground into 20–40 meshes and pretreated at 823 K for 6 h in ambient atmosphere before use. All the chemicals used were selected from the Sinopharm Chemical Reagent Company and used without further purification.

1.2. Catalyst Preparation

Normal ZrO₂ modified SiO₂ catalysts were prepared via incipient wetness impregnation with zirconium nitrate [Zr(NO₃)₄] in water for 10 h. Two kinds of highly dispersed ZrO₂ modified SiO₂ catalysts were prepared via ligand–assisted incipient wetness impregnation with solutions of zirconium acetylacetonate [Zr(acac)₄] in methanol, zirconium n-butoxide [Zr(O-n-Bu)₄] in n-butanol for 10 h, respectively. The obtained samples were treated by drying at 80 °C for 3 h and calcination 500 °C for 4 h in air with a 5 °C/min heating ramp rate.

Cs₂O supported Zr/SiO₂ catalysts were prepared by wet impregnation with aqueous cesium nitrate (CsNO₃) solution for 10 h. Then, the samples after impregnation was dried at 100 °C for 3 h and calcined at 500 °C for 4 h in air with a 5 °C/min heating ramp rate.

The obtained samples were labeled as Zr(A)-xCs/SiO₂, Zr(B)-xCs/SiO₂, and Zr(N)-xCs/SiO₂, where x represents the concentration of Cs₂O (wt.%, g/g). A, B and N represents Zr(acac)₄, Zr(O-n-Bu)₄ and Zr(NO₃)₄, respectively.

1.3. Materials Characterization

XRD diffractometer (Rigaku Smart Lab) was employed for the crystalline analysis within the scan range of 5°–80°. X-ray photoelectron spectroscopy (XPS) analyses were performed using an ESCALAB 250Xi (Thermo Fisher Scientific) spectrometer with Al Kα radiation at θ = 90°. The binding energy of Zr 3d and Cs 3d spectra were calibrated by the C 1s peak at 284.8 eV. Fourier transform infrared (FTIR) measurements were conducted with a Nicolet-380 spectrometer (Thermo Fisher Scientific) utilizing KBr as a diluting agent. By employing a transmission electron microscope (TEM) (JEOL Ltd., JEM-2010), the microstructural characteristics of the sample could be observed. N₂ adsorption–desorption isotherms were measured at 77 K using an ASAP 2020 system. Solid‐state magnetic resonance (NMR) spectra were tracked in an AVANCE III HD 500 MHz (BRUKER BIOSPIN AG, CH) instrument. To evaluate the acid–base sites of the samples, NH₃/CO₂-temperature-programmed desorption (TPD) were obtained through the utilization of a Micromeritics Auto Chem II 2920 device. During the CO₂-TPD experiments, the introduced CO₂ will not react with Cs₂O to form CsCO₃, which affects the experimental results, as demonstrated in Figure S1.

1.4. Catalytic Activity Evaluation

The above-prepared ZrO₂ modified Cs/SiO₂ catalysts were applied to the aldol condensation of MAc with FA. The stainless-steel micro-fixed bed integral reaction device was used to carry out catalyst evaluation experiments. 6.0 mL catalyst samples (3.0 g) were filled in the reactor with 10 mm inner diameter. Trioxane is used as a source of formaldehyde. A constant-flux pump was used to feed a mixture of MAc, FA, and CH₃OH (molar ratio 1:1:2) with the reactor. The reaction was performed at 340–380 °C in the atmosphere, and liquid feed was pumped into the reactor with liquid hourly space velocity (LHSV) of 0.2–2 h⁻¹. After the reaction reached steady-state (4 h), product was condensed and collected in a liquid storage tank. The reaction mixture was quantitatively analyzed using an Agilent 7890GC fitted with DB-WAX capillary column (30 m × 0.25 nm × 0.25 μm).

The conversion of MAc (X_MAc), selectivity and yield of MA(S_MA and Y_MA) can be calculated by the following equations: (1) $$X_{\text{MAc}}\left(\text{\%}\right)\text{=}\left({\mathrm{n}}_{\text{MAc,in}}‐{\mathrm{n}}_{\text{MAc,out}}\right)/{\mathrm{n}}_{\text{MAc,in}}\mathrm{\times }100$$(1) (2) $$S_{\text{MA}}\left(\text{\%}\right)\text{=}{\mathrm{n}}_{\text{MA,out}}/\left({\mathrm{n}}_{\text{MAc,in}}‐{\mathrm{n}}_{\text{MAc,out}}\right)\mathrm{\times }100$$(2) (3) $$Y_{\text{MA}}\left(\text{\%}\right)\text{=}{X}_{\text{MAc}}\mathrm{\times }{S}_{\text{MA}}$$(3)

In the above formulas, ${\mathrm{n}}_{\text{MAc},\text{in}}$ is the amount of substance of MAc in feed, ${\mathrm{n}}_{\text{MAc},\text{out}}$ is the amount of substance of leftover MAc in the liquid phase products, and ${\mathrm{n}}_{\text{MA},\text{out}}$ is the amount of substance of generated MA in the liquid phase products.

2. Results and Discussion

2.1. Catalyst characterization

2.1.1. Structure and morphology

The XRD patterns of three kinds of ZrO₂-modified Cs/SiO₂ catalysts are shown in Figure 1. All of the samples presented a broad peak at around the values of 22.0°, which were assigned to the characteristic of amorphous silica (Gao et al. 2018). 10Cs/SiO₂ demonstrated well distinguished characteristic diffraction peaks at around the values of 19.86°, 28.28°, 34.84°, 40.37°, 45.40° and 50.08°, assigned to the (1 1 1), (1 1 2), (0 0 3), (2 2 2), (3 0 3) and (1 1 4) crystallographic planes of CsNO₃ (PDF No. 09-0403), respectively (Ibrahim et al. 2019). After introducing a similar content of ZrO₂ as shown in Table S1, the characteristic diffraction peaks of CsNO₃ disappears in Zr(N)-10Cs/SiO₂, Zr(B)-10Cs/SiO₂ and Zr(A)-10Cs/SiO₂ samples. The above results verified that the modification of ZrO₂ tended to promote the decomposition of CsNO₃ in the SiO₂ during calcination and the homogeneous dispersion of Cs species. There was no new characteristic peak of ZrO₂ appeared either, indicating the well dispersed Zr compound over SiO₂ (Sohn and Lim 2005). By comparison, when the loading of Cs reached 15 wt.%, those CsNO₃ peaks appeared again, and Zr(A)-15Cs/SiO₂ displayed the weakest intensity of CsNO₃ compared to Zr(N)-15Cs/SiO₂ and Zr(B)-15Cs/SiO₂. This demonstrated that using Zr(acac)₄ as zirconium source was more conducive to the process of ZrO₂ modification than using Zr(O-n-Bu)₄ and Zr(NO₃)₄, which could promote the CsNO₃ decomposition and Cs species formation and dispersion.

Figure 1. XRD patterns of prepared SiO₂, 10Cs/SiO₂, Zr(A)-xCs/SiO₂, Zr(B)-xCs/SiO₂, and Zr(N)-xCs/SiO₂.

Similar findings can be arrived at by the spectral analysis of FT $‐$ IR (Figure 2). The strong absorption peaks of all the samples in 470, 805, 1,105, and 1,231 cm⁻¹ were attributed to the presence of amorphous silica (Yan et al. 2015), which implied the main structure of SiO₂ remain unchanged after introducing Zr and Cs primarily. The characteristic peak of the Si − OH groups band at 970 cm⁻¹ disappeared after the incorporation of Cs and Zr species, probably corresponding to the consumption of Si − OH to formation of Si − O−Zr and Si − O−Cs species (Gao et al. 2018). The N − O bonding antisymmetric stretching vibrations in the CsNO₃ could be observed at 1,384 cm⁻¹ in the 10Cs/SiO₂ sample, but disappeared in ZrO₂ modified 10Cs/SiO₂, which suggested the introduction of ZrO₂ could promote the decomposition of CsNO₃ in the SiO₂. As the content of Cs increased to 15 wt.%, this characteristic signal appeared again, and Zr(A)-15Cs/SiO₂ displayed the weakest relative intensity. The above results further implied using Zr(acac)₄ as zirconium source to modify SiO₂ was in favor of the uniformity of Cs species and interaction of Cs and SiO₂, consistent with that of XRD analysis.

Figure 2. FT $‐$ IR spectra of prepared SiO₂, 10Cs/SiO₂, Zr(A)-xCs/SiO₂, Zr(B)-xCs/SiO₂, and Zr(N)-xCs/SiO₂.

The surface morphological characteristics were observed by TEM to discover the effect of ZrO₂ modification on Cs/SiO₂ catalysts. Figure 3(a and b) revealed that 10Cs/SiO₂ did not exhibit crystal characteristics which was consistent with amorphous SiO₂ phase. After ZrO₂ modification, the crystal particles of ZrO₂ can be observed on the surface of SiO₂ in Figure 3(c–h). The different organic ligand could influence the morphology of Zr species in the Cs/SiO₂ catalyst, and using organic Zr(acac)₄ and Zr(O-n-Bu)₄ as zirconium source was more propitious to form highly dispersed ZrO₂ on the surface of SiO₂ compared with using inorganic Zr(NO₃)₄. Moreover, highly dispersed ZrO₂ was instrumental in strengthening interaction between Cs species and SiO₂, which has been proved to be contributed to the enhanced catalytic performance by numerous studies (Salinier et al. 2009).

Figure 3. TEM images of prepared (a, b) 10Cs/SiO₂, (c, d)Zr(A)-10Cs/SiO₂, (e, f)Zr(B)-10Cs/SiO₂, and(g, h) Zr(N)-10Cs/SiO₂.

2.1.2. Chemical states analysis

The surface chemical states of three kinds of ZrO₂ modified Cs/SiO₂ catalysts were revealed by XPS spectra. As shown in Figure 4(a), the high-resolution Cs 3d spectrum showed two peaks at 738.7 eV and 724.7 eV, which were referred to Cs 3d_3/2 and Cs 3d_5/2 of Cs⁺, respectively (Wu et al. 2022). For three kinds of ZrO₂ modified Cs/SiO₂ catalysts, these peaks shifted slightly to the higher binding energy. The shift can be ascribed to the decreased electron density on Cs, which indicated the ZrO₂ modification enhanced the interaction between Cs and SiO₂. The Zr(A)-10Cs/SiO₂ and Zr(B)-10Cs/SiO₂ samples displayed a more noticeable binding energy shift of Cs and Zr than Zr(N)-10Cs/SiO₂. This phenomenon indicated that when the organic Zr(acac)₄ and Zr(On-Bu)₄ were used as zirconium sources, the modified ZrO₂ will show good dispersion, thus leading to the superior interaction of Cs with ZrO₂-modified SiO₂ than that with pristine SiO₂. In the Zr 3d narrow scan spectrum (Figure 4(b)), the peaks at 184.9 eV and 182.5 eV can be in accord with to the Zr 3d_3/2 and Zr 3d_5/2 of Zr⁴⁺ (La Parola et al. 2022). The ZrO₂ content in Cs/SiO₂ catalysts modified using different Zr resource was obtained by XRF characterization, as shown in Table S1. The binding energy of Zr 3d in Zr(A)-10Cs/SiO₂, Zr(B)-10Cs/SiO₂ and Zr(N)-10Cs/SiO₂ was quite similar, which can be attributed to these three catalysts having a similar ZrO₂ loading.

Figure 4. Cs 3d spectra(a), Zr 3d spectra(b) of prepared 10Cs/SiO₂, Zr(A)-10Cs/SiO₂, Zr(B)-10Cs/SiO₂, and Zr(N)-10Cs/SiO₂.

To further investigate the chemical environment of Cs in ZrO₂ modified Cs/SiO₂ catalysts, solid-state magic angle spinning (MAS) NMR analysis was conducted. Figure 5 shows the chemical shift of ¹³³Cs for 10Cs/SiO₂ (−8 ppm), Zr(A)-10Cs/SiO₂ (−35 ppm), Zr(B)-10Cs/SiO₂ (−32 ppm), and Zr(N)-10Cs/SiO₂ (−20 ppm). Generally, a more negative chemical shift of ¹³³Cs indicates stronger chemical combination between Cs and $‐$ OH of SiO₂ (Kim et al. 2015). Accordingly, the more negative shift of ¹³³Cs for the Zr(A)-10Cs/SiO₂ sample implied that highly dispersed ZrO₂ modification was better for the strong combination between Cs species and $‐$ OH in SiO₂ carriers, which matched with the conclusion of XPS spectra of Cs 3d.

Figure 5. ¹³³Cs MAS NMR spectra of prepared 10Cs/SiO₂, Zr(A)-10Cs/SiO₂, Zr(B)-10Cs/SiO₂, and Zr(N)-10Cs/SiO₂.

2.1.3. BET and acid–base property analysis

The Brunauer–Emmett–Teller (BET) specific surface area and pore structure information of all catalysts are displayed in Table 1. The SiO₂ carriers processed a substantial BET specific surface area (269 m² g⁻¹) and pore sizes (9.2 nm). Due to the introduction of the Cs, the 10Cs/SiO₂ showed a relatively smaller BET specific surface area and pore volume, while average pore sizes displayed the opposite tendency. This trend can be explained by the aggregation of excessive Cs species, thereby filling some porous structures. Interestingly, after modification of ZrO₂, the impact of Cs on the SiO₂ carrier is significantly suppressed. Three kinds of ZrO₂ modified Cs/SiO₂ catalysts possessed a larger BET specific surface area and pore volume, and smaller average pore sizes, which is more advantageous to the aldol condensation reaction. Due to the similar ZrO₂ content, the Zr(A)-10Cs/SiO₂, Zr(B)-10Cs/SiO₂ and Zr(N)-10Cs/SiO₂ show the similar specific surface area and pore structure. The results above implied that modification of ZrO₂ was beneficial for inhibiting the aggregation of Cs on the surface of SiO₂, and slowing down the destructive effect on SiO₂ pore structure. Highly dispersed ZrO₂ modification is more conducive to improving pore structure of Cs/SiO₂.

Table 1. The BET specific surface areas and pore structure information of prepared SiO₂, 10Cs/SiO₂, Zr(A)-10Cs/SiO₂, Zr(B)-10Cs/SiO₂, and Zr(N)-10Cs/SiO₂.

Catalysis	Surface areas (m^2 g^−1)	Average pore diameter (nm)	Pore volume (cm^3 g^−1)
SiO2	269	9.2	0.84
10Cs/SiO2	200	9.9	0.69
Zr(A)-10Cs/SiO2	239	9.1	0.71
Zr(B)-10Cs/SiO2	238	9.4	0.70
Zr(N)-10Cs/SiO2	231	9.6	0.69

It has been validated that the acidity and alkalinity of catalysts play a crucial role in aldol condensation reactions (Guan et al. 2020). Herein, NH₃-TPD and CO₂-TPD tests were performed to characterize the strength of acid and base properties, as shown in Figure 6. Generally, the strength of acid and base sites of catalyst could be reflected by the temperature range during the desorption process of NH₃ and CO₂. The acid and base sites in the range of 100–250 °C, 250–400 °C, 400–600 °C are defined as follows weak sites, middle-strong sites and strong sites, respectively (Zheng et al. 2022). In Figure 6(a), it could be discovered that the base sites signal of 10Cs/SiO₂ were in the form of two broad peaks at 120–200 °C and 200–400 °C, indicating that the generation of weak and medium-strong base sites in 10Cs/SiO₂ catalyst. With the modification of ZrO₂, the total base amount of Zr(A)-10Cs/SiO₂, Zr(B)-10Cs/SiO₂, and Zr(N)-10Cs/SiO₂ catalysts showed an obvious increase, especially weak base sites, which could be assigned to the alkalinity of ZrO₂. In Figure 6(b), 10Cs/SiO₂ catalyst displayed a broad peak indicating the non-uniform of the acid sites. Moreover, medium strong acid sites occupied a big proportion of acid sites, which could speculatively be referred to the Lewis acid of Cs⁺. After introducing ZrO₂ into 10Cs/SiO₂, all catalysts displayed an obvious desorption peak in the range of 120 °C to 250 °C, corresponding to the appearance of weak acid sites offered by ZrO₂ species. Meanwhile, it was obvious that using Zr(acac)₄ as zirconium source to modify SiO₂ could provide more acid and base sites than that of using Zr(O-n-Bu)₄ and Zr(NO₃)₄. Therefore, it was deduced that modification of highly dispersed ZrO₂ could tune the acid − base property and provide more weak acid and base sites, which is more conducive to the adsorption and activation of reactants during the aldol condensation reaction.

Figure 6. CO₂ (a) and NH₃-TPD (b) profiles of prepared 10Cs/SiO₂, Zr(A)-10Cs/SiO₂, Zr(B)-10Cs/SiO₂, and Zr(N)-10Cs/SiO₂.

2.2. Aldol condensation evaluation

2.2.1. Effect of ZrO₂ modification

The representative ZrO₂-modified Cs/SiO₂ catalysts with different zirconium source were evaluated for the aldol condensation of MAc and FA. The feed mixture of FA, MAc and CH₃OH (molar ratio = 1:1:2) was fed with a LHSV of 0.6 h⁻¹, and the reaction was conducted at 360 °C. CH₃OH was used to prevent the hydrolysis of MAc and MA. Except for MA, various byproducts such as acetic acid, acrylic acid and acrolein could also be detected in the liquid-phase products according to the GC analysis results, as shown in Table S2. Very few gas-phase products could be separated out, which were mainly composed of CO, CO₂, ethylene and hydrogen.

The activity of the 10Cs/SiO₂ and three kinds of ZrO₂ modified Cs/SiO₂ catalysts is recorded in Figure 7. ZrO₂ modified 10Cs/SiO₂ catalysts exhibited higher conversion of MAc and selectivity of MA than 10Cs/SiO₂. In particular, Zr(B)-10Cs/SiO₂ demonstrated the highest conversion of MAc (34.16%), and Zr(A)-10Cs/SiO₂ also had similar performance (32.16%). Zr(A)-10Cs/SiO₂ possessed the highest selectivity of MA (88.97%) among all catalysts. The difference in catalytic activities can be attributed to the balance of basic and acid sites in the catalysts. Based on the mechanism of aldol condensation, the deprotonation of α-H of MAc will proceed on the base sites to form the intermediate enol species, while the FA will be activated on the acid sites (Wang et al. 2022; Bao et al. 2017). The acid and base site densities of all catalysts are summarized in Table 2. After the modification with ZrO₂, more base and acid sites would appear on the Cs/SiO₂ sample. Compared to the 10Cs/SiO₂, three types of ZrO₂-modified Cs/SiO₂ catalysts show a higher density of weak base sites, which is more favourable for the activation and transformation of MAc (Hattori 2015). For the Zr(B)-10Cs/SiO₂ catalyst with the highest concentration of base sites (0.33 mmol/g_cat), the numerous weak base sites would contribute to activation of MAc, resulting in the highest conversion among all the samples (Ma et al. 2021). Beyond that, with ZrO₂ modification, the total acid sites amount of Zr(A)-10Cs/SiO₂ catalyst reached maximum values of 1.03 mmol/g_cat and the percentage of weak acid sites reached 62.82%. An appropriate number of weak acid sites in Zr(A)-10Cs/SiO₂ catalyst was beneficial for the formation of activated FA, which further converted the α-carbanion into MA, resulting in the higher selectivity among all the samples. A small amount of medium and strong acid sites over the surfaces of ZrO₂-modified Cs/SiO₂ could suppress accumulation of excess surface activated FA which can be involved in the aggregation to form carbon deposits (Wu et al. 2022), "thus resulting in the highest selectivity of MA among all the catalysts. Taking an overview, these results revealed that Zr(A)-10Cs/SiO₂ sample displayed the best performance in selectivity and yield of MA, further illustrating using Zr(acac)₄ as zirconium source to modify Cs/SiO₂ could be considered as optimal for aldol condensation reaction of MAc with FA.

Figure 7. Catalytic activities of prepared 10Cs/SiO₂, Zr(A)-10Cs/SiO₂, Zr(B)-10Cs/SiO₂, and Zr(N)-10Cs/SiO₂.

Table 2. Densities of acid and base sites on prepared 10Cs/SiO₂, Zr(A)-10Cs/SiO₂, Zr(B)-10Cs/SiO₂, and Zr(N)-10Cs/SiO₂.

Catalyst	Total acid sites (mmol/gcat)	Percentage of acid sites %			Total base sites (mmol/gcat)
		Weak	Medium	Strong
10Cs/SiO2	0.55	35.08	54.06	10.85	0.21
Zr(N)-10Cs/SiO2	0.86	50.96	43.42	5.62	0.28
Zr(B)-10Cs/SiO2	0.85	57.97	35.76	6.27	0.33
Zr(A)-10Cs/SiO2	1.03	62.82	32.43	4.75	0.30

2.2.2. Effect of Cs₂O and ZrO₂ loading

Since the Zr(A)-10Cs/SiO₂ sample showed best catalytic activity, this catalyst was chosen for the optimizing experiments of the loading of the ZrO₂ and Cs₂O, as shown in Figure 8. With the enhancement of ZrO₂ loading, the conversion of MAc increased accordingly from 21.60% to 37.38%, while the selectivity of MA increased from 62.37% to 88.97% and then declined to 78.16%. As shown previously, the modification using ZrO₂ could promote the dispersion of Cs species and increase the specific surface area of Cs/SiO₂. In addition, the increase of Zr content would favour the formation of active weak acid and base sites, as shown in Figure S2 and Table S3, thereby improving the conversion of MAc and selectivity of MA. Once excessive loading of the ZrO₂ in Cs/SiO₂ catalyst, the more acid sites will be introduced in catalyst, resulting in the lower selectivity of MA.

Figure 8. Catalytic performance of Zr(A)-Cs/SiO₂ catalysts with various ZrO₂ (a) and Cs₂O (b) loading.

The catalytic performance of ZrO₂ modified Cs/SiO₂ catalysts with various Cs₂O loading is shown in Figure 8(b). As the Cs₂O loading increases, the selectivity and yield of MA demonstrated the typical “volcanic curve” reaching its peak at 10 wt.% Cs loading, while the conversion of MAc significantly increased from 25.54% to 60.84%. It was noteworthy that a suitable amount of Cs promoted the activation of MAc and the formation of MA, which attributed to the formation of the Si − O−Cs species. On the other hand, when the loading of Cs₂O in SiO₂ was excessive, the CsNO₃ could not be completely transformed to Si − O−Cs species, causing a reduction in the specific surface area and forming the excessive medium and strong base sites, as shown in Figure S2. This would cause undesired hydrolysis of MAc and reduce the selectivity of MA.

All these results revealed that the content of ZrO₂ and Cs₂O affected the specific surface area and acid–base balance of the ZrO₂ modified Cs/SiO₂ simultaneously, and determine the catalytic activity for aldol condensation reaction. The optimal ZrO₂ modified Cs/SiO₂ catalysts can be achieved with 5% ZrO₂ loading and 10% Cs₂O loading.

2.2.3. Effect of reaction conditions

The aldol condensation catalytic activity of Zr(A)-Cs/SiO₂ was tested at different reaction temperatures from 340 °C to 380 °C, as shown in Figure 9(a). As the temperature increased, the conversion of MAc gradually elevated from 26.80% to 53.46% and the selectivity of MA first increased from 77.79% to 88.97% and then decreased to 61.38%. Further increasing the temperature, the pyrolyzation and polymerization of reactants turned to be more competitive than the aldol condensation, which resulting in the formation of precursors of coke, gaseous products (CO_x) and other byproducts, and reducing the selectivity of MA. Thus, 360 °C was regarded as the suitable reaction temperature for Zr(A)-Cs/SiO₂.

Figure 9. Effect of reaction temperatures (a) and LHSV (b) on the aldol condensation over a Zr(A)-10Cs/SiO₂ catalyst.

The effect of LHSV on the catalytic performance was explored, as shown in Figure 9(b). The LHSV was determined by the feed flow rate and catalyst loading volume. As the LHSV was increased from 0.2 h⁻¹ to 2 h⁻¹, the marked decline of the conversion of MAc was detected from 40.15% to 15.21%, while the selectivity of MA rosed at first and kept in a stable higher level gradually later. Essentially the extending of the residence time, is in favor of conversion of MAc and formation of byproducts. On this basis, A maximum yield of MA was explored at an LHSV of 0.6 h⁻¹, which was suitable for aldol condensation on Zr(A)-Cs/SiO₂.

2.2.4. Stability and reusability

The long-term stability of prepared Zr(A)-10Cs/SiO₂ samples was investigated by the reaction and regeneration cycle experiments of the aldol condensation. After a reaction cycle, the catalyst was regenerated by calcining in the air atmosphere at 400 °C for 10 h. As shown in Figure 10, the conversion of MAc decreased from 32.18% to 17.61% after 200 h continuous reaction. After 3 regeneration cycles, the Zr(A)-10Cs/SiO₂ was successively employed for 600 h, indicating thar prepared Zr(A)-10Cs/SiO₂ catalyst displayed good stability and reusability.

Figure 10. Stability of the Zr(A)-10Cs/SiO₂ in the aldol condensation.

Conclusions

Highly dispersed ZrO₂ modification with ligand–assisted strategy was applied to improve the catalytic performance of Cs/SiO₂-based catalyst for the aldol condensation of MAc with FA. The effect of ZrO₂ modification had quite important influence to the existence state of Cs species, specific surface area, and surface acid–base sites, thus further affected the catalytic performance of aldol condensation. Compared to traditional impregnation methods with using inorganic Zr(NO₃)₄, ligand–assisted impregnation strategy with using organic Zr(acac)₄ and Zr(O-n-Bu)₄ possessed better performance in the process of ZrO₂ modification. On the one hand, the higher dispersion of ZrO₂ on the surface catalyst was beneficial for the formation Cs–O–Si species which could promote the formation of MA. On the other hand, ZrO₂ modification could improve pore structure and tune acid–base sites balance of Cs/SiO₂. The optimal ZrO₂ modified Cs/SiO₂ catalyst had 5% ZrO₂ loading and 10% Cs₂O loading, which could achieve the highest selectivity being 88.97% and the yield being 28.63% of MA under optimal reaction conditions. Besides, the optimal catalyst exhibited good reusability in 600 h recycling experiments. Using ligand–assisted strategy during ZrO₂ modification is a very effective approach for enhancing catalytic activity of the Cs/SiO₂ catalyst for the aldol condensation, which has a good prospect for industrial application.

Additional information

Funding

This work was supported by National Natural Science Fund for Distinguished Young Scholars (Grant 22025803), the Joint Fund of the Yulin University and the Dalian National Laboratory for Clean Energy (Grant YLU-DNL Fund 2021018) and Key R&D Program of Xinjiang (2022B01035).