Catalytic Performance of Solid Base Na-ZrO₂ in Subcritical Water

Abstract

In this study, Na-ZrO₂ was prepared from two types of ZrO₂ as solid base catalyst and the catalytic activity, selectivity, and mechanism of Na-ZrO₂ for base catalyzed reactions in subcritical water were elucidated using aldol condensation between acetophenone and benzaldehyde as a model reaction. Based on the different selectivity for Cannizzaro reaction of benzaldehyde, a side reaction accompanied with the aldol condensation, two plausible mechanisms were proposed. One was that the doped Na₂O could directly act as a base, deprotonating the acetophenone for catalyzing aldol condensation, and the other was that the doped Na₂O could dissociate the water to provide OH⁻ for catalyzing both aldol condensation and Cannizzaro reaction. The dominant mechanism varied depending on the nature of the ZrO₂ support. The effect of the amount of subcritical water was also examined, showing that as reducing the loading of water, the yield of byproduct from the Cannizzaro reaction decreased due to the lower amount of dissociated OH⁻. On the other hand, the presence of water adversely affected the reaction equilibrium, but meanwhile sufficient water in the system also prevented the carbonization of organics and polymerization reaction on the catalyst surface, favoring for a ring balance.

Keywords:

• Subcritical water
• Solid base catalyst
• Modified zirconium oxide
• Aldol condensation
• Cannizzaro reaction

1. Introduction

Subcritical water, defined as water at a temperature above the normal boiling point but below the critical point (374 °C), while maintaining in the liquid state under high pressure, has attracted much attention as an alternative medium for organic synthesis because of the rising concerns about the negative impact which organic solvents have on the environment (Ko et al. 2020). Subcritical water exhibits drastically different physicochemical properties with water in ambient condition (Kruse and Dinjus 2007). The lower dielectric constant and higher ion product (K_W = [H⁺][OH⁻]) of water at a subcritical state facilitates the dissolution of organic compounds and makes it suitable for acid/base catalyzed reactions (Iwamura et al. 2014; Savage 1999). In the past, the feasibility of conducting organic reactions in subcritical water without catalysts was mainly investigated, and major types of reactions had been reviewed by several researchers (Iwamura et al. 2014; Kruse and Dinjus 2007; Kus 2012; Savage 1999). However, the limitation of the acid/base catalytic activity of subcritical water was also distinct.

To realize the enhancement and control of reactions in subcritical water while further reducing the generation of waste salts, the application of solid base catalysts for reactions in subcritical water had been studied and proved to be practical to achieve higher reaction rates, which could not be achieved by utilizing the acid/base catalytic activity of subcritical water alone (Akizuki and Oshima 2018). Investigation of the basicity and catalytic mechanism of solid catalysts was important to control the reactions in water at high temperature and pressure since water with unique properties at such conditions would affect the base catalytic properties on the catalyst surface. Watanabe and his group had investigated the acidity and basicity of some metal oxides (CeO₂, MoO₃, TiO₂, and ZrO₂) in supercritical water at 400 °C and 25–40 MPa using formaldehyde reaction as a test reaction (Watanabe et al. 2003). They elucidated the order of the basicity is CeO₂ > ZrO₂ > MoO₃ > TiO₂(rutile) > TiO₂(anatase) and mentioned that the acidity and basicity on a metal oxide surface in supercritical water are probably due to adsorption of dissociated water molecules on the catalyst surface. Akizuki and his colleague studied the kinetics of ZrO₂ catalyzed amide hydrolysis in subcritical water at 240–360 °C and 10–41 MPa, in which hydrolysis of amide was considerably promoted due to the basicity of ZrO₂ (Akizuki and Oshima 2015). They found that compared with the reaction with NaOH as a homogeneous base catalyst, amino groups might hardly form hydrogen bond with water due to steric hindrance near the ZrO₂ surface, thus affecting the reaction rate. However, these kinds of mechanism research were still limited and further investigation in the basicity and catalytic mechanism particularly for the supported catalyst, which was expected to exhibit higher basicity than simple metal oxide investigated before, would be necessary for controlling organic synthetic reactions.

Therefore, in this study, the basicity of ZrO₂ would be enhanced by incorporating with active sodium compounds and the catalytic performance of such Na-ZrO₂ in subcritical water would be evaluated. Crossed aldol condensation of acetophenone and benzaldehyde (Scheme 1) was used as a model of base catalyzed reaction. Since side reactions such as the Cannizzaro reaction of two benzaldehydes (Scheme 2), which was also base catalyzed, might occur, the selectivity of the reaction was expected to provide a deeper insight of base catalytic properties. We hope to elucidate the catalytic mechanisms of ZrO₂ based catalysts in subcritical water, quantitatively analyze the catalytic activity and selectivity by establishing kinetic models and clarify how the amount of subcritical water affected the reaction.

Scheme 1. Synthesis of chalcone via aldol condensation.

Scheme 2. Cannizzaro reactions of benzaldehydes.

2. Experimental

2.1. Materials

All solvents and chemicals were commercially available and used without further purification. Acetophenone, benzaldehyde, chalcone, benzyl alcohol, sodium nitrate and sodium hydroxide were purchased from FUJIFILM Wako Pure Chemical Industries, Ltd. Zirconium oxide, zirconium hydroxide, and acetonitrile were offered by Sigma-Aldrich. Dihydrochalcone was provided by Tokyo Chemical Industry Co., Ltd. Distillation water was prepared by water distillation apparatus, RFD240NC, produced by ADVANTEC Toyo Kaisha, Ltd.

2.2. Preparation of catalysts

Commercial ZrO₂ and ZrO₂ prepared by calcination of Zr(OH)₄ at 500 °C for 3 hours were named as ZrO₂ (type I) and ZrO₂ (type II), respectively. Na-ZrO₂ (type I and type II) were prepared by wet impregnation from their respective ZrO₂ (type I and type II), which followed a procedure as listed below. The mixture of ZrO₂ (type I or type II) and NaNO₃ aqueous solution was stirred for overnight at ambient condition, followed by condensation with rotary evaporation under reduced pressure at 50 °C and drying at 100 °C for 1 h. The resulting solid was then calcined in a muffle furnace at 500 °C for 3 h under a standard atmosphere. In this research, Na-ZrO₂ was prepared with 5.0 g of ZrO₂ and 0.5 g of NaNO₃.

2.3. Experimental procedures

To A mixture of acetophenone (4.3 mmol) and benzaldehyde (4.3 mmol) with distillation water in a 10 mL of 1/2-inch SS316 stainless steel tube batch reactor (Swagelok Co., tube length: 122 mm, outside diameter: 12.7 mm, thickness: 1.24 mm, fitted with a SS316 1/2-inch cap), 250 mg of solid catalyst was added, purging with nitrogen, preventing the oxidation of benzaldehyde to form benzoic acid with oxygen from the air. The reactor was then sealed and heated to 250 °C in sand bath and reacted for a range of time, from 0.5 h to 6 h. After the reaction, the reactor was cooled down to ambient temperature in a water bath and the reaction mixtures were diluted with acetonitrile, accompanied by the filtration of the catalysts. The filtrate was then subjected to GC analysis for identification of the product and determination of the yield and conversion.

2.4. Analysis

Quantitative analysis of the products was performed by using gas chromatograph-flame ionization detector (GC-2014, Shimadzu Corp.), equipped with a capillary column (GL Sciences, Inc., TC-1701, length: 30 m, film thickness: 0.25 μm, inside diameter: 0.25 mm). The initial temperature was set as 80 °C, holding for 25 min, and then the temperature was increased to 260 °C with a rate of 30 °C per min, holding at 260 °C for another 34 min. Identification of the side products was performed by gas chromatography–mass spectrometry (GC-2010, Shimadzu Corp.), equipped with the same capillary column as described above. Blank experiments were also performed to confirm that aldol condensation between acetophenone and benzaldehyde in acetonitrile would not take place in the GC system.

2.5. Characterization of catalysts

The crystal structure of the different catalysts was analyzed by a Rigaku SmartLab 9 kW X-ray diffractometer (XRD) using Cu-Kα radiation, wavelength of 0.154 nm, in the 2θ range from 10° to 90° with a scanning spend of 4° per minute and a step of 0.02 at room temperature.

The measurements of specific surface areas of the catalysts were performed based on the BET method in NOVA 2200e. Prior to the measurement, the samples were degassed at 200 °C for 2 hours and N₂ adsorption/desorption isotherms were operated at −196 °C.

Temperature-programmed desorption with CO₂, CO₂-TPD, was applied to measure the base strength of the catalysts. 200 mg of catalyst were enclosed in a quartz tube, treated with a flow of 50 mL per min of Helium gas, while heating at a rate of 10 °C per min to 700 °C and holding at 700 °C for 1 h. After cooling down to 50 °C with helium gas flowing, the sample was saturated with 10% CO₂/He mixture gas in a flow rate of 20 mL per min for 1 h. The sample was then purged with helium gas for an additional 1 h to remove physically adsorbed CO₂. Finally, heating the samples to 700 °C with a ramp of 10 °C per min, and the desorbed CO₂ was monitored with a thermal conductivity detector (TCD) and quantified with a TCD calibration curve.

To measure the quantity of sodium cation doped on ZrO₂, the catalyst was treated by acid digestion in 66% H₂SO₄ acid at 200 °C overnight, followed by dilution for inductively coupled plasma-atomic emission spectroscopy (ICP-AES) analysis.

3. Results and Discussion

3.1. Characterization of catalysts

The XRD patterns of Zr(OH)₄, ZrO₂ and Na-ZrO₂ were shown in Figure 1. Figure 1 presented that ZrO₂ (type II), obtaining from calcination of amorphous Zr(OH)₄, contained both monoclinic and tetragonal phases, while ZrO₂ (type I) contained only a monoclinic phase with higher crystallinity. However, a transformation of the tetragonal phase to the monoclinic phase was observed after modification of ZrO₂ (type II), which was in agreement with the previous literature (Eom et al. 2015), while in the case of ZrO₂ (type I), the modification did not lead to any phase changes. The detection of the peak of NaNO₃ implied the existence of bulk-phase NaNO₃ on Na-ZrO₂, but Figure 1 didn’t show the presence of Na₂O, indicating the formation of amorphous Na₂O or the presence of dispersed Na₂O undetectable by XRD (Aguila et al. 2016; Liu et al. 1998).

Figure 1. XRD patterns of catalysts. (a) ZrO₂ (Type I), (b) Na-ZrO₂ (Type I), (c) ZrO₂ (Type II), (d) Na-ZrO₂ (Type II).

The nomenclatures of the catalysts involved in this study and the corresponding physicochemical properties were summarized in Table 1, in which ZrO₂ (Type II) presented much higher specific areas than ZrO₂ (type I). In the case of the Na-ZrO₂, compared with ZrO₂, a significant decrease of the specific area was observed, probably being attributed to that the pores were covered by sodium compounds. To accurately measure how much Na⁺ was doped on the support, the fresh catalysts were rinsed with ambient water to eliminate the extra bulk phase NaNO₃. With the removal of the undecomposed NaNO₃, ICP analysis was performed for the Na-ZrO₂, indicating 2.72 mol% and 2.93 mol% of Na⁺ was doped on the surface of ZrO₂ (type I) and ZrO₂ (type II) respectively.

Table 1. Physicochemical properties of the catalysts.

Catalysts	Specific areas	Amounts of basic sites	Na^+/Zr^4+
	m^2‧g^−1	μmol‧g^−1
ZrO2 (Type I)	5.03	27.0	0
Na-ZrO2 (Type I)	1.84	55.6	2.72%
ZrO2 (Type II)	39.6	71.6	0
Na-ZrO2 (Type II)	15.6	136.6	2.93%

CO₂-TPD was utilized to investigate the amount and the strength of basic sites of the catalysts. To accurately measure the basicity of Na-ZrO₂, the same rinsing treatment was performed to remove the extra bulk phase NaNO₃ covered on the ZrO₂ support, followed by drying at 110 °C overnight before subjecting to CO₂-TPD analysis. The amounts of basic sites were given in Table 1, and the desorption profiles were depicted in Figure 2. According to the desorption temperature of CO₂, the strength of basic sites could be roughly classified into four types, weak basic sites, (< 250 °C, I), medium basic sites (250–400 °C, II), strong basic sites (400 °C < 620 °C, III) and super strong basic sites (> 620 °C, IV) (Li et al. 2021). The profile of ZrO₂ (type I) barely revealed a weak desorption peak below 250 °C, whereas the profile of ZrO₂ (type II) showed that ZrO₂ (type II) possessed medium and strong basic sites, along with a strong desorption peak in the range of weak basics sites. In the case of Na-ZrO₂, compared with ZrO₂, the profile presented prominent desorption peaks in the range of medium and strong basic sites with increased intensity, indicating an increased basicity of ZrO₂ after modification. The generation of strong basic sites on ZrO₂ was assumed to be primarily attributed to the formation of Na₂O on the surface of ZrO₂ during heat treatment (Liu et al. 2007).

Figure 2. CO₂-TPD profiles of catalysts. (a) ZrO₂ (Type I), (b) Na-ZrO₂ (Type I) with rinsing treatment, (c) ZrO₂ (Type II), (d) Na-ZrO₂ (Type II) with rinsing treatment.

3.2. Catalytic activities of ZrO₂ and Na-ZrO₂ in subcritical water

Initially, to investigate the catalytic activities of ZrO₂ and Na-ZrO₂ in subcritical water, the reaction was carried out with 3.0 g of distillation water under a condition mentioned in the procedures. To eliminate the possibility that the reaction was catalyzed by the undecomposed NaNO₃, both the rinsed and fresh Na-ZrO₂ were tested. Na-ZrO₂ with rinsing treatment didn’t show a different catalytic performance with the fresh one (Figure 3).

Figure 3. Yield of chalcone with and without catalysts at 250 °C vs. time. (■) Catalyst free, (●) ZrO₂ (Type I), (▲) ZrO₂ (Type II), (▼) Na-ZrO₂ (Type I), (♦) Na-ZrO₂ (Type II), (★) Na-ZrO₂ (Type I) with rinsing treatment.

Chalcone was the major product produced from aldol condensation, and the yield of chalcone was quantitatively determined for the evaluation of the catalytic activity. Benzyl alcohol and benzoic acid were the major side products, yielded from the Cannizzaro reaction of benzaldehydes exhibited in Scheme 2. The catalytic selectivity for the Cannizzaro reaction was calculated based on the yield of benzyl alcohol, which was primarily monitored for its higher FID sensitivity. One minor side product, 1,3,5-triphenylpentane-1,5-dione, shown in Scheme 3, obtained from the Michael addition of acetophenone to chalcone, was traced by GC-MS. Other minor side products, such as dihydrochalcone, produced from the hydrogenation of double bond between the aliphatic carbons (Comisar and Savage 2004), were only in trace amounts.

Scheme 3. Minor products accompanied with the synthesis of chalcone.

Figures 3 and 4 presented the yield of chalcone and benzyl alcohol with and without catalysts depending on time, respectively. The comprehensive experimental results were summarized in Tables S1–S5 in the supplementary materials. The uncertainty was experimentally determined by calculating the standard deviation. In the case of catalyst free, the yield of chalcone increased slowly with respect to the reaction time, and the benzyl alcohol was only traced. The reaction catalyzed by ZrO₂ (type I) demonstrated that ZrO₂ (type I) did not show clear catalytic activity for either aldol condensation or Cannizzaro reaction. On the other hand, the reaction catalyzed by ZrO₂ (type II) showed a higher yield of both chalcone and benzyl alcohol.

Figure 4. Yield of benzyl alcohol with and without catalysts at 250 °C vs. time. (■) Catalyst free, (●) ZrO₂ (Type I), (▲) ZrO₂ (Type II), (▼) Na-ZrO₂ (Type I), (♦) Na-ZrO₂ (Type II).

The reactions catalyzed by Na-ZrO₂ (type I and II) had a considerable increase in the yield for chalcone. However, Na-ZrO₂ (type II) revealed significantly higher selectivity for Cannizzaro reaction to form benzyl alcohol than Na-ZrO₂ (type I). A close highest yield of chalcone in 3 hours was likely indicating the aldol condensation reached an equilibrium.

To compare the catalytic performance of ZrO₂ based catalysts with that of homogeneous catalyst in subcritical water, the reaction with 0.1 equivalent of NaOH was also carried out. Figure 5 showed the yield and conversion with respect to time. The detailed data was summed up in Table S6 in the supplementary materials. The production of chalcone and benzyl alcohol was considerable, indicating both aldol condensation and Cannizzaro reaction were promoted by strong OH⁻ provided by NaOH.

Figure 5. Yield and conversion with 0.1 eq NaOH at 250 °C vs. time. Yield: (■) Chalcone, (●) Benzyl alcohol, Conversion: (▲) Acetophenone, (▼) Benzaldehyde.

3.3. Kinetics analysis with model and proposed mechanisms of ZrO₂ based catalysts

To investigate the difference in the reaction rate quantitatively, kinetic analysis was conducted for reactions under different catalytic conditions. The reaction network involved in this system was presented in Figure 6. Acetophenone (A) reacted reversibly with Benzaldehyde (B) to form Chalcone (C). Chalcone reacted further to form other products, such as 1,3,5-triphenylpentane-1,5-dione, dihydrochalcone and char. The Cannizzaro reaction of benzaldehyde formed benzyl alcohol (Bol) and benzoic acid. Based on this network with assumptions of first order for each reactant (second order for benzaldehyde in Cannizzaro reaction), the differential equations of concentrations depending on time during the reaction in a constant volume batch reactor were shown from Eqs. (1)–(4), which provided a mathematical model for this chemical reaction system. Since the amount of water did not change appreciably in the system compared with that of reactants and products, the concentration of water was assumed to be constant as 44.4 mol/L during the reaction (the density of subcritical water at 250 °C, saturated, was 799 g/L (Wagner and Pruß 2002). For the reason that 1,3,5-triphenylpentane-1,5-dione only became considerable when the loaded water was below 0.5 g, the rate constant k₄ was supposed to be 0 when the system contained 1.0 g of water or more. To solve the differentia equations and estimate the kinetic rate constants, Python was used to perform the fitting of a system of ordinary differential equations. (1) $$\begin{array}{c}\frac{d{C}_A}{dt}=-k_{1f}{C}_A{C}_B+k_{1r}{C}_C{C}_{H_{2}O}-k_4{C}_A{C}_C\end{array}$$(1) (2) $$\begin{array}{c}\frac{d{C}_B}{dt}=-k_{1f}{C}_A{C}_B+k_{1r}{C}_C{C}_{H_{2}O}-k_2{C}_B^2\end{array}$$(2) (3) $$\begin{array}{c}\frac{d{C}_C}{dt}=k_{1f}{C}_A{C}_B-k_{1r}{C}_C{C}_{H_{2}O}-k_3{C}_C-k_4{C}_A{C}_C\end{array}$$(3) (4) $$\begin{array}{c}\frac{d{C}_{\mathit{Bol}}}{dt}=k_2{C}_B^2\end{array}$$(4)

Figure 6. Reaction network for the synthesis of chalcone via aldol condensations in subcritical water.

The experimental data of concentration with respect to time with models for reactions under different catalytic conditions were shown in Figure 7. The estimated values of rate constants were summarized in Table 2. Since the reactions catalyzed by Na-ZrO₂ better expressed a reaction equilibrium from the figures, the equilibrium constant, k_1f/k_1r, was derived from the reaction catalyzed by Na-ZrO₂ via fitting to be 41.6, and such value was used to substitute into the rate constant for k_1r as k_1f/41.6 to fit the rate equation for other cases.

Figure 7. Comparison of model and experimental data for reaction in 3.0 g of water. (a) Catalyst free, (b) ZrO₂ (Type I), (c) ZrO₂ (Type II), (d) Na-ZrO₂ (Type I), (e) Na-ZrO₂ (Type II), (■) Chalcone, (●) Benzyl alcohol, (▲) Acetophenone, (▼) Benzaldehyde, (—) Model with R-squared values shown on the top-right corner.

Table 2. Estimated values of rate constants for reaction in 3.0 g of water.

Catalyst	k1f	k1r	K1	k2	k3
	mol^−1‧L‧s^−1	mol^−1‧L‧s^−1	[−]	mol^−1‧L‧s^−1	s^−1
No catalyst	3.82 × 10^−2	9.02 × 10^−4	41.6	7.58 × 10^−4	9.13 × 10^−2
ZrO2 (type I)	3.75 × 10^−2	9.02 × 10^−4	41.6	1.50 × 10^−3	3.97 × 10^−2
ZrO2 (type II)	6.66  × 10^−2	1.60 × 10^−3	41.6	1.43  × 10^−2	1.28  × 10^−1
Na-ZrO2 (type I)	1.67  × 10^−1	4.00 × 10^−3	41.6	3.31 × 10^−3	3.74 × 10^−2
Na-ZrO2 (type II)	2.36 × 10^−1	5.68  × 10^−3	41.6	2.52 × 10^−2	8.72  × 10^−2

To explain the discrepancy of catalytic activities, the mechanism of the basic model was proposed as Figure 8. The reaction catalyzed by NaOH (Figure 5) provided evidence that the presence of strong OH⁻ in subcritical water would induce both the aldol condensation and the Cannizzaro reaction. The larger values of k_1f and k₂ in the case of the ZrO₂ (Type II) than ZrO₂ (Type I) indicate that OH⁻ on the ZrO₂ (Type II) surface showed basicity. The CO₂-TPD profiles suggested that the basic sites of ZrO₂ (Type II) were much stronger than those of ZrO₂ (Type I) in the gas phase, thus being more capable of dissociating the water into H⁺ and OH⁻ on the catalyst surface in subcritical water.

Figure 8. Mechanisms of basicity model of ZrO₂ and Na-ZrO₂ in subcritical water.

In the case of Na-ZrO₂ (both type I and II), the rate constant for the aldol condensation (k_1f) was 4 times larger than that catalyzed by respective ZrO₂ (type I and II) while the rate constant for the Cannizzaro reaction (k₂) was 2 times larger than that catalyzed by respective ZrO₂ (type I and II). Because k_1f was increased rather than k₂ by impregnation of Na species, there should be another mechanism to mainly promote aldol condensation. A plausible explanation composed of two different mechanisms of the basicity model in subcritical water are shown in Figure 8. One mechanism was that the doped Na₂O could directly act as a base, deprotonating the acetophenone for catalyzing aldol condensation, and the other was that the OH⁻ provided by water dissociation catalyze both aldol condensation and Cannizzaro reaction. When comparing Na-ZrO₂ (type I) with Na-ZrO₂ (type II), Na-ZrO₂ (type II) considerably promoted the Cannizzaro reaction than Na-ZrO₂ (type I), which could be explained by the fact that ZrO₂ (type II) support itself had high ability to provide OH⁻. Solid base catalyzed reactions in subcritical and supercritical water were considered to be promoted by OH⁻ provided by the water dissociation (Akizuki and Oshima 2015; Watanabe et al. 2003) since the surface of catalysts are surrounded by high density water, but the possibility of contributing another kind of basic site even in subcritical water having high ion product was an interesting finding.

3.4. Effect of amount of water

To further investigate the effect of subcritical water on solid base catalysis, the effect of the amount of solvent water, in the reactor with a volume of 10 mL, was investigated by reducing from 3.0 g (condition used for the experiments above) to 0 (water free). The reaction was catalyzed by Na-ZrO₂ (type I) for 3 h, and the results were depicted in Figure 9.

Figure 9. Yield and conversion after 3.0 h reaction at 250 °C with Na-ZrO₂ (type I) vs. the amount of water in 10 mL reactor. Yield: (■) Chalcone, (●) Benzyl alcohol, Conversion: (▲) Acetophenone, (▼) Benzaldehyde.

As the amount of loaded water decreased from 3.0 g to around 1.0 g, the yield of chalcone gradually increased, while the yield of benzyl alcohol slightly decreased. Figure 10 showed the time dependence of the yield of chalone and benzyl alcohol. The equilibrium yield of the chalcone when the amount of water was 1.0 g was larger than that when the amount of water was 3.0 g (Figure 3). Since the equilibrium of aldol condensation shifted to the product side in low water concentration condition (Scheme 1), more benzaldehyde was available for aldol condensation and the yield of chalcone increased but that of benzyl alcohol decreased. To further confirm this, the kinetic model with the rate constants from Table 2 was applied for this condition. Figure 11 showed the predicted values which generally matched the experimental results, indicating that the reaction mechanism was not changed as the amount of water decreased from 3.0 g to 1.0 g.

Figure 10. Yield of chalone and benzyl alcohol of reaction with 1.0 g of water in 10 mL reactor at 250 °C with Na-ZrO₂ (type I) vs. time. Chalcone: (■) Chalcone, (●) Benzyl alcohol.

Figure 11. Comparison of predicted model and experimental data for reaction with 1.0 g of water in 10 mL reactor. (a) Na-ZrO₂ (Type I), (■) Chalcone, (●) Benzyl alcohol, (▲) Acetophenone, (▼) Benzaldehyde, (—) Model.

On the other hand, as the amount of water decreased from 1.0 g to 0 (water free), the yield of benzyl alcohol continued to decline and reached practically 0 in the case of water free. It could be explained that less water in presence led to a lower amount of dissociated OH⁻ on the catalyst surface, thus inhibiting the Cannizzaro reaction. In the case of water free, there would be no dissociated OH⁻ on the catalyst surface and the catalytic mechanism could be the doped Na₂O on the ZrO₂ support directly function as a base, deprotonating the acetophenone for catalyzing aldol condensation. In addition, under a condition where the amount of water was below 1.0 g, the conversion of acetophenone was higher than benzaldehyde, indication that the formation of 1,3,5-triphenylpentane-1,5-dione and further products became noticeable. Furthermore, from Table S7 in the supplementary material, there is an obvious decreasing trend of the ring balance between the reactant conversion and product yield. Also, Figure S1, given in the supplementary materials, demonstrated a darkening trend of the reaction mixtures as reducing the amount of water involved in the reaction. Further investigation of the mechanism was necessary, but these results suggested that sufficient water could inhibit the polymerization reaction on the catalyst surface.

4. Conclusions

The basicity and catalytic mechanism of ZrO₂ and Na-ZrO₂ in subcritical water at 250 °C were investigated using aldol condensation of acetophenone and benzaldehyde and Cannizzaro reaction of two benzaldehyde as model reactions.

Among the two ZrO₂ (Type I and II) catalyst, ZrO₂ (Type II) showed higher base catalytic activity because stronger basic site on ZrO₂ (Type II) promote water dissociation in subcritical water and produced OH⁻ on the surface showed base catalytic activity.

On the other hand, for the Na-ZrO₂ (Type I and II) catalyst, the existence of two catalytic mechanisms was suggested; one was that the doped Na₂O could directly act as a base, and the other was that OH⁻ provided by the dissociation of water act as a base. Since ZrO₂ (type II) support itself had a high ability to provide OH⁻, Na-ZrO₂ (type II) considerably promoted the Cannizzaro reaction while the selectivity for aldol condensation was high for Na-ZrO₂ (type I) catalyst. The presence of other type of basic site than the OH⁻ provided by the dissociation of water even in subcritical water having high ion product, and that their dominance can be changed depending on the catalyst support, is important from the viewpoint of control in reactions in subcritical water.

The effect of the amount of solvent water suggested that reducing the water loading from 3.0 g to 1.0 g in a 10 mL reactor did not cause a change in the catalytic mechanism. On the other hand, when the amount of water decreased from 1.0 g to 0, less water in presence lead to a lower amount of dissociated OH⁻ on the catalyst surface. In addition, sufficient water might be able to inhibit the polymerization reaction on the catalyst surface.

Acknowledgement

BET measurements were performed at Otomo lab, School of Environment and Society, Tokyo Institute of Technology. XRD and ICP analysis were performed at the Institute for Solid State Physics, The University of Tokyo. The GC-MS analysis was conducted at Kashiwa Branch, Environmental Science Center and Tonokura lab, The University of Tokyo.

No potential conflict of interest was reported by the author(s).

Disclosure statement

Additional information

Funding

This study was supported by JST SPRING, Grant Number [JPMJSP2108].