Synthesis of Diethyl Carbonate from Carbon Dioxide, Propylene Oxide and Ethanol over KNO₃-CeO₂ and KBr-KNO₃-CeO₂ Catalysts

Abstract

One-pot syntheses of diethyl carbonate (DEC) from CO₂, propylene oxide and ethanol were carried out using different solid catalysts. The supercritical CO₂ extraction method was used to separate the liquid products and reactants from the catalysts after reaction. The KNO₃-CeO₂ and KBr-KNO₃-CeO₂ were found to be active for the reaction after calcinations. The catalyst was also reusable. The thermodynamic properties of the reaction were also evaluated. The effects of various conditions, such as reaction time, amount of catalysts, molar ratio of the reactants, the composition and calcination temperature of the catalysts on the conversion and yields, were investigated, and the yield of DEC was about 13.0% with a selectivity of 38.5% over KBr-KNO₃-CeO₂. The yield of DEC was improved about 10-fold by using KBr-KNO₃-CeO₂ catalyst compared to CeO₂.

1. Introduction

The synthesis of chemicals using CO₂ as a raw material is characterized at present by increasing industrial and academic efforts to use this carbon renewable [1]. The production of carbonates [2,3], carbamates [4], methanol [5], formic acid and its derivatives could be synthesized from CO₂ [6,7]. Diethyl carbonate (DEC) is one of the most important green chemicals among carbonate esters. It is an excellent solvent and an intermediate for various pharmaceuticals, such as antibiotics and phenobarbital [8]. DEC has also been proposed as a replacement for MTBE as an attractive oxygen-containing fuel additive for its high oxygen content (40.6 wt. %) compared to MTBE (18.2 wt. %) [9,10].

Since the conventional methodologies for the DEC synthesis, including ethanol phosgenation [11], ethanol oxidative carbonylation [12] and the reaction of ethanol with urea [13], have many problems, such as the toxicity of phosgene, corrosion and low production rates [14], the novel technology for DEC synthesis starting from CO₂ and ethanol is a promising route. However, the reaction hardly occurs spontaneously, even under harsh conditions, due to the thermodynamic limitations (yield of less than 0.5%) [8]. To address this issue, chemical dehydration reagent was usually involved to shift the reaction forward to the carbonate side. In the similar reaction of the synthesis dimethyl carbonate (DMC), acetals [15] and orthoesters [16] were used as the organic dehydrants, respectively, and both the DMC yields could be effectively improved above 20-fold. However, the high cost of acetals and orthoesters makes them difficult for industrial production. Acetonitrile [17] and amines [18] were also reported as being used as the dehydrants for the DEC synthesis, but their co-products were complex. Recently, butylene oxide was also used as the dehydrant for the direct synthesis of the DEC over CeO₂ catalyst [8,19]. According to the results, the yield of DEC had a nine-fold enhancement compared to that over CeO₂ without dehydrant, but it was still not high enough (only 1.5%) in this system and needed to be improved. Besides, CeO₂ with 2-cyanopyridine was also used as the carboxylation/hydration cascade catalyst by Tomishige group [20] for the propylene carbonate synthesis from CO₂ and 1,2-propanediol, and the yield was much higher (>99%), which might be a landmark in carbonate synthesis using CeO₂ catalysis.

In addition, the one-pot synthesis of DEC from carbon dioxide, ethylene oxide (EO) and ethanol on the KI and sodium ethoxide binary homogeneous catalyst was also researched by Wang et al. [21], and the yield of DEC was improved. However, KI and sodium ethoxide are dissolved in ethanol and cannot be separated easily.

On the other hand, the synthesis of cyclic carbonate from epoxide and CO₂ was well established in industrial manufacturing. Furthermore, the transesterification of cyclic carbonate with ethanol to produce DEC was also proven to be feasible [22] However, from views of energy consumption, productivity and investment, the one-pot reaction directly from CO₂ was undoubtedly superior to the two-step separate reaction. Thus, the development of a more effective one-pot reaction to improve the productivity of DEC directly from CO₂ is highly desired.

2. Results and Discussion

The one-pot reaction in DEC synthesis might be composed of two steps, the cycloaddition reaction and subsequent transesterification reaction. The mechanisms of the reaction have been studied and proven by many researchers [21,23]. As analyzed by the GC-MS method in this work, the main products in the one-pot reaction from CO₂, ethanol and PO were 1,2-propanediol (PG), DEC and propylene carbonate (PC) with the side-product 1-ethoxy-2-propanol (EP). The EP might be formed from propylene oxide by ethanolysis [23,24] in the basic catalytic environment. The possible equations of these reactions are presented as follows (Scheme 1).

2.1. The One-Pot Synthesis of DEC over KNO₃-CeO₂ Catalyst

As reviewed in the literature [8], the heterogeneous catalyst CeO₂ has catalytic activity for the one-pot reaction in DEC synthesis. In order to improve the catalytic activity in DEC synthesis, the strong base of KOH and several typical alkali and alkaline-earth metal salts, which might be necessary to meet the requirement for catalyzing the transesterification reaction combined with CeO₂, were researched. Several metal oxides, such as γ-Al₂O₃, ZrO₂, SiO₂ and La₂O₃, were also tested. In the typical reaction, the molar ratio of CO₂, ethanol and PO was fixed as 0.29:0.17:0.14; the reaction temperature was 150 °C; the initial pressure was 5 MPa; and during the reaction, the pressure could reach 9 MPa, which is higher than the critical pressure of CO₂. The results are summarized in

Table 1. Effects of different catalysts on the one-pot reaction. DEC, diethyl carbonate; PG, 1,2-propanediol; EP, 1-ethoxy-2-propanol.

Run	Catalyst	Specific surface area/m2/g	Ethanol conversion/%	DEC yield/%	Selectivity/%
					DEC	PG	EP
1	CeO2	-	1.5	1.3	84.0	-	-
2	NaNO3-CeO2	-	12.6	3.6	28.3	24.1	19.3
3	Ba(NO3)2-CeO2	-	8.1	0.9	12.6	10.2	16.6
4	Mg(NO3)2-CeO2	-	15.9	2.2	13.0	15.9	17.4
5	Ca(NO3)2-CeO2	-	6.6	1.3	22.3	16.7	19.8
6	KOH-CeO2	-	13.4	3.6	28.0	20.1	33.6
7	K2CO3-CeO2	-	9.1	1.8	19.0	17.6	25.1
8	KNO3-CeO2	30.5	34.0	11.2	33.3	23.7	23.5
9	KNO3-γ-Al2O3	102	46.8	11.5	24.5	23.1	16.5
10	KNO3-ZrO2	22.9	23.7	6.3	27	24.4	22.9
11	KNO3-SiO2	78.1	12.3	3.1	25.3	26.1	18.8
12	KNO3-La2O3	31.8	0.6	0.4	69.1	20.7	10.9
13	KI-CeO2 A	-	18.0	0.9	6.3	6.4	10.3
14	KI-CeO2 B	-	9.0	1.3	14.4	17.4	12.8
15	KI-KNO3-CeO2 A	-	17.1	0.9	5.8	4.9	7.7
16	KI-KNO3-CeO2 B	-	9.6	0.4	7.5	7.9	7.2
17	KCl-KNO3-CeO2 A	-	14.4	2.7	17.4	14.4	5.7
18	KCl-KNO3-CeO2 B	-	39.9	10.3	25.9	22.3	5.9
19	KBr-KNO3-CeO2 A	-	23.1	9.4	40.8	30.5	7.6
20	KBr-KNO3-CeO2 B	-	33.9	13.0	38.5	27.7	7.5

Both of the volumes of ethanol and PO were 10 mL; the reaction temperature was 150 °C; the initial pressure was 5 MPa. ^A Catalysts prepared by the impregnation method; ^B catalysts prepared by the solid mixed method.

, Runs 1–12.

As seen in Table 1, when KNO₃-CeO₂ was used as the catalyst, the yield of DEC was effectively improved about 10-fold, compared to that using CeO₂ catalyst. The KNO₃-CeO₂ had better catalytic activity than KOH and other alkali and alkaline-earth metal salts loading on CeO₂, such as KOH-CeO₂, K₂CO₃-CeO₂, NaNO₃-CeO₂, Ba(NO₃)₂-CeO₂, Mg(NO₃)₂-CeO₂ and Ca(NO₃)₂-CeO₂. Then, KNO₃ with other metal oxides, including γ-Al₂O₃, ZrO₂, La₂O₃ and SiO₂, was evaluated. The results indicate that the oxides with acid-base properties, especially ZrO₂, γ-Al₂O₃ and CeO₂ combined with KNO₃, are more active for DEC synthesis. The basicity of the catalysts KNO₃-γ-Al₂O₃, KNO₃-CeO₂ and CeO₂ were analyzed by the CO₂-TPD method (Figure 1). The desorption peaks at about 280 °C–340 °C could be observed for KNO₃-γ-Al₂O₃ and KNO₃-CeO₂, which had better catalytic activities. It is indicated that the addition of a small amount of a moderate base is more effective for enhancing the activity of the catalyst. Considering the KNO₃-CeO₂ catalyst obtaining the better yield and selectivity for DEC, it was finally selected for the following reactions. The XRD spectrums of KNO₃-CeO₂ and CeO₂ are shown in Figure 2. By comparison, the characteristic peaks of KNO₃ cannot be found, and the diffraction peaks of KNO₃-CeO₂ are stronger than CeO₂. It is indicated that the KNO₃ might be well dispersed and caused no considerable distortion in the structure of CeO₂ [25].

The preparation conditions for KNO₃-CeO₂ including the load of KNO₃ and the calcination temperature were optimized. The results are shown in Figure 3a,b. It is indicated that the KNO₃-CeO₂ with n(Ce)/n(K) = 1:0.4 has better catalytic activity (Figure 3a). And the DEC yield reaches higher level (Figure 3b) at the calcination temperature of 500 °C. According to the TG-DTG analysis of KNO₃ (Figure 4), the decomposition temperature of KNO₃ is 520 °C. When the temperature is higher than 520 °C, the KNO₃ will be decomposed to K₂O, and the catalytic activity will decrease.

Then, the reaction conditions, including the amount of catalyst, reaction time and volume ratio of ethanol and PO, were studied. The results are shown in Figure 3c–e. As seen in Figure 3c, the DEC yield first increases and then decreases with the increase of catalyst amounts. The DEC yield reaches the peak value when the amount of KNO₃-CeO₂ is 0.3 g (Figure 4c). Figure 4d shows the dependence of DEC yield on reaction time. The reaction reaches equilibrium in 2 h. Figure 3e shows that the DEC yield reaches the higher level, when both of the volumes of ethanol and PO are 10 mL (0.17 mol and 0.14 mol, respectively) and the CO₂ amount is fixed as 0.25 mol.

2.2. The One-Pot Synthesis of DEC over KBr-KNO₃-CeO₂ Catalyst

As potassium halides (KI, KBr and KCl) were proven to be conducive to the cycloaddition reaction [15]; the KI, KBr and KCl were added in the KNO₃-CeO₂ catalyst for the one-pot synthesis of DEC. The addition methods for potassium halides, including solid mixed and impregnation methods, were compared.

The results are shown in Table 1, Runs 13–20. As seen in Table 1, when KI-CeO₂ is used as the catalyst, the ethanol conversion increases, but the yield of DEC is not improved compared to CeO₂ alone. In addition, when KBr or KCl is added to KNO₃-CeO₂, the ethanol conversion decreases, but the selectivity of DEC is effectively improved. Additionally, the catalysts prepared by the solid mixed method give a higher DEC yield and selectivity. However, when KI-KNO₃-CeO₂ is used as the catalyst, the DEC yield seriously decreases. This might be because that KI, with a stronger reduction property, is oxidized by KNO₃, which causes deactivation.

The molar ratio of KBr and KNO₃ was evaluated with fixed n(CeO₂)/n(KNO₃) = 1:0.4. The results are shown in Figure 5. The XRD patterns of KBr-KNO₃-CeO₂ are presented in Figure 4, and the characteristic peaks of KBr are labeled. The yield and the selectivity of DEC are improved with the addition of KBr into the catalyst. When the molar ration is n(KBr)/n(KNO₃) = 6:4, that is n(CeO₂)/n(KNO₃)/n(KBr) = 1:0.4:0.6, the yield of DEC reaches 13.0% with a selectivity of 38.5% on the ethanol basis. Both the yield and selectivity of DEC are much higher than reported in Leino’s research [8] by using butylene oxide as the dehydration agent and CeO₂ as the catalyst. In their results, the highest obtained yield of DEC was 1.5%, and selectivity to DEC was 10% on the ethanol basis.

The formation kinetics of DEC were also studied, and the dependence of DEC yield on reaction time at 150 °C in 180 min is shown in Figure 6. It is also indicated that the reaction can reach equilibrium in 100 min. The reaction time was selected as 120 min.

2.3. Recycling Experiments

The recycling experiment results of the KBr-KNO₃-CeO₂ catalyst are listed in Figure 7. After three times reuse, the catalyst still keeps good catalytic activity, and the DEC yield is above 90%, as in the primary reaction. The XRD patterns of fresh KBr-KNO₃-CeO₂ and the catalyst after using three times were also compared. The calcination temperature KBr-KNO₃-CeO₂ was also set as 500 °C. Additionally, the KBr was not decomposed according to the TG-DTG analysis (Figure 4). As seen in Figure 7, the characteristic peaks do not change after three runs. This indicates that the active species of KBr and KNO₃ do not leach from the catalyst. The reason might be that CO₂ was the main reactant in this reaction, and the reactants and products were extracted by Sc-CO₂ after the reaction, while the catalyst was not soluble in Sc-CO₂ and left in the reactor, which avoided the loss of catalyst. This is one of the advantages of separating the products by Sc-CO₂.

2.4. Thermodynamic Evaluation of the One-Pot Synthesis of DEC

In order to perform the thermodynamics evaluations, which are important in seeking novel synthesis ideas, the thermodynamic data of various substances, such as ethanol, CO₂, PO, DEC and PG, involved in the reaction are tabulated in

Table 2. Thermodynamic data of various substances in the reaction [26].

Substance	${\mathbf{\Delta }}_f{\mathit{H}}_{\mathbf{298}k}^{\theta }$/kJ·mol−1	$S_{\mathbf{298}k}^{\theta }$/J·mol−1·K−1	${\mathit{C}}_{\mathit{p} \mathbf{298}k}^{id}$/J·mol−1·K−1
C2H5OH	−235.10	282.70	112.6
CO2	−393.509	213.74	37.13
DEC	−637.9	412.94 a	211
H2O	−241.82	188.82	33.58
PO	−94.68	281.15	72.55
PG	−421.29	288 a	189.9

^a Calculated by the Constantinous-Gani ( CG) group contribution method [27].

.

The enthalpy and the entropy of the reaction at 298 K estimated from the ${\Delta }_{\mathrm{f}}{H}_{298\mathrm{k}}^{\mathsf{\theta }}$ and ${\mathrm{S}}_{298\mathrm{k}}^{\mathsf{\theta }}$ values amounted to ${\Delta }_{\mathrm{r}}{H}_{298\mathrm{k}}^{\mathsf{\theta }}=-101.4 \text{KJ}·{\text{mol}}^{-1}$ and ${\Delta }_{\mathrm{r}}{S}_{298\mathrm{k}}^{\mathsf{\theta }}=-357.22 \text{J}·{\text{mol}}^{-1}·{\mathrm{K}}^{-1}$. The Gibbs energy at 298 K and 100 KPa could be calculated by Equation (1) and has a value ${\Delta }_{\mathrm{r}}{G}_{298\mathrm{k}}^{\mathsf{\theta }}=5.10 \text{KJ}·{\text{mol}}^{-1}$.

$${\Delta }_{\mathrm{r}}{G}_{298\mathrm{k}}^{\mathsf{\theta }}={\Delta }_{\mathrm{r}}{H}_{298\mathrm{k}}^{\mathsf{\theta }}+T{\Delta }_{\mathrm{r}}{S}_{298\mathrm{k}}^{\mathsf{\theta }}$$

(1)

Based on the obtained values, it can be concluded that the reaction is exothermic (${\Delta }_{\mathrm{r}}{H}_{298\mathrm{k}}^{\mathsf{\theta }}=-101.4\text{KJ}·{\text{mol}}^{-1}<0$) and does not occur spontaneously at room temperature ${\Delta }_{\mathrm{r}}{G}_{298\mathrm{k}}^{\mathsf{\theta }}=5.10 \text{KJ}·{\text{mol}}^{-1}>0$.

The relative pattern of the reaction heat with the temperature is expressed by Kirchhoff’s law (Equation (2)), whereas the Gibbs energy of the reaction, at different temperatures, can be given by the Gibbs–Helmholtz equation (Equation (3)) [8]. Gibbs energy values of this reaction at different temperatures were calculated. The results are shown in Figure 8. The value of ${\Delta }_{\mathrm{r}}{G}_{\mathrm{T}}^{\mathsf{\theta }}$ increases with the reaction temperature, and the increase in the temperature is disadvantageous to the formation of DEC.

$${\Delta }_{\mathrm{r}}{H}_{\mathrm{T}}^{\mathsf{\theta }}={\Delta }_{\mathrm{r}}{H}_{298\mathrm{k}}^{\mathsf{\theta }}+\Delta C_{\mathrm{p}}(\mathrm{T}-298)$$

(2)

$$d\left(\frac{{\Delta }_{\mathrm{r}}{G}_{\mathrm{T}}^{\mathsf{\theta }}}{T}\right)=\left(-\frac{{\Delta }_{\mathrm{r}}{H}_{\mathrm{T}}^{\mathsf{\theta }}}{T^2}\right)d\text{T}$$

(3)

When the temperature is fixed at 423 K, the Gibbs energy of the reaction is a function of pressure and can be written as follows [8]:

$${\Delta }_{\mathrm{r}}{G}_{\mathrm{p}}={\Delta }_{\mathrm{r}}{G}^{\mathsf{\theta }}-RT\text{ln}\left(\frac{p}{p^{\mathsf{\theta }}}\right)$$

(4)

The equilibrium constant K was determined from the Gibbs energy according to Equation (5):

$${\Delta }_{\mathrm{r}}{G}_{\mathrm{T}}=-RT\text{ln}K$$

(5)

$$K=\frac{\left[\text{DEC}\right]\left[\text{PG}\right]}{{\left[{\mathrm{C}}_2{\mathrm{H}}_5\text{OH}\right]}^2\left[\text{PO}\right]\left[{\text{CO}}_2\right]}$$

(6)

In this work, the reaction temperature and pressure are 423 K and 9 MPa; the ${\Delta }_{\mathrm{r}}{G}_{\mathrm{p}}$ calculated by Equation (4) is 36.07 kJ·mol⁻¹ > 0; and the equilibrium constant K calculated by Equation (5) is 3.5 × 10⁻⁵. When the amounts of CO₂, ethanol and PO are 0.29, 0.17 and 0.14 mol, respectively, and assuming that the reaction is in the gas phase, the equilibrium yield of DEC at 423 K and 9 MPa calculated by Equation (6) is about 14.2% based on ethanol. It is indicated that the yield of DEC in the experiment (13.0%) is close to the equilibrium yield calculated.

3. Experimental Section

3.1. Material

Ethanol (99.8%, analytical grade) was dehydrated by adding spherical 3A molecular sieves before the reactions. Propylene oxide (PO), n-propanol, Ce(NO₃)₃·6H₂O, KNO₃, KOH, NaNO₃, Mg(NO₃)₂·6H₂O, Ba(NO₃)₂, Ca(NO₃)₂·4H₂O, KBr, KCl, KI, ammonium hydroxide (NH₃·H₂O, 28%) and spherical 3A molecular sieves were purchased from Jiangtian Chemical Reagent Co., Ltd., Tianjin, China. All of the chemicals were of analytical grade and used without further purification. CO₂ (99.99%) was purchased from Lianbo Gas Co., Ltd., Tianjin, China.

3.2. Catalyst Preparation and Characterization

CeO₂ was prepared by the precipitation method. The 6.5 g of cerium (III) nitrate hexahydrate (Ce(NO₃)₃·6H₂O) were firstly dissolved in 50 mL water. Then, the precipitating solution of ammonium hydroxide of 50% (v/v) was slowly added into the well-stirred cerium nitrate aqueous solution. The pH of the solution was controlled at 10 throughout the synthesis process. The resulting precipitate was filtrated, washed with deionized water and then dried overnight at 100 °C and calcined at 500 °C for 2 h in air.

KNO₃-CeO₂ was prepared by the incipient impregnation method. The slurry (n(Ce)/n(K) = 1:0.2, 1:0.4, 1:0.6, 1:0.8 or 1:1) was kept in a static state for 24 h and dried at 110 °C for 5 h. The dried samples were calcined at 500 °C for 3 h. Similar methods were used for the preparation of NaNO₃-CeO₂, KOH-CeO₂, Ca(NO₃)₂-CeO₂, Mg(NO₃)₂-CeO₂ and Ba(NO₃)₂-CeO₂.

KBr-KNO₃-CeO₂ was prepared by both the impregnation method and solid mixed method. The impregnation method was performed as in the KNO₃-CeO₂ preparation. For the solid mixed method, the dried solids of KBr and KNO₃-CeO₂ were mixed evenly and then calcined at 500 °C for 3 h.

The specific surface area was measured by using nitrogen adsorption (Sorptometer 1900, Carlo Erba Instruments, Milan, Italy). The sample was out gassed at 150 °C for 3 h prior to the measurement of the surface area. For calculating the surface area, the BET equation was used.

The structural properties of the catalysts were investigated by X-ray diffraction (XRD) (X’Pert Pro MPD, PANalytical, Lelyweg, The Netherlands) using CuKα (40 kV, 50 mA) radiation with 2θ ranging from 0° to 80° at a scanning speed 0.04°/3 s.

Thermogravimetric (TG) analysis was carried out on a thermogravimetric analyzer (TG209, Netzsch Co., Wittelsbacherstr, Germany) under a N₂ stream of 40 mL min⁻¹. The temperature was raised at a heating rate of 10 °C min⁻¹.

Temperature-programmed desorption (TPD) of CO₂ was carried out using the Micromeritics Instrument (Auto Chem 2910, Micromeritics Instrument Co., Norcross, GA, USA). Zero-point-one gram of catalyst was placed in a quartz-made U-shaped tube. The catalyst was firstly treated in a flow of helium (50 mL·min⁻¹) at 200 °C for 2 h to remove any physisorbed organic molecules. The sample was then saturated by adsorption of CO₂ (50 mL·min⁻¹) at 60 °C for 30 min. Physisorbed CO₂ was flushed at 150 °C for 1 h. After the sample was cooled, the furnace temperature was increased from 60 °C to 800 °C at a heating rate of 5 °C·min⁻¹ under a flow of helium (30 mL·min⁻¹). The desorbed CO₂ was detected by using a TCD (thermal conductivity detector).

3.3. Catalytic Test

All experiments were carried out in a laboratory-scale stainless steel autoclave (Weihai Chemical Machinery Co., Ltd., Weihai, China) with an inner volume of 100 mL equipped with a stirrer and an electric heater. In a standard procedure, 0.3–0.5 g of catalyst, 170 mmol (10 mL) of ethanol and 140 mmol (10 mL) propylene oxide were introduced into the autoclave. The reactor was purged with CO₂ at room temperature. Then, the reaction system was pressurized to the predetermined initial pressure with CO₂ by a pump (Model SFC-24, SSI/LabAlliance, State College, PA, USA). The amount of CO₂ was calculated from measuring the weight difference of the CO₂ cylinder before and after charging the CO₂. After the reaction, the reactor was cooled to about 50 °C. The CO₂ was released slowly through a receiving flask with methanol as the absorbent. Then, the liquid chemicals in the reactor were extracted in situ by supercritical CO₂ at 12 MPa and 50 °C. All of the reactants and products were collected using methanol in a flask. The catalyst was left in the reactor. About 240 g CO₂ were required to extract all liquid chemicals. For the recycling experiments, the catalyst was used directly after the extraction process by repeating the procedures above.

$$\text{Yield} \text{of} \text{product} Y_{\mathrm{i}}(\%)=\frac{\text{mole of product i}}{0.5\times \text{mole of ethanol charged }}\times 100\%$$

$$\text{Selectivity }{S}_{\mathrm{i}}(\%)=\frac{\text{mole of product i}}{0.5\times \text{mole of ethanol convered}}\times 100\%$$

3.4. Analysis Method

The products were analyzed by gas chromatography (GC) with an FID detector (6890N, Agilent Co., Santa Clara, CA, USA) and GC–MS (HP 6890 gas chromatograph coupled with mass detector HP 5973N, Agilent Co., Santa Clara, CA, USA) using an HP-5MS (30 m × 0.25 mm × 0.25 μm, Agilent Co., Santa Clara, CA, USA) fused silica capillary column. n-Propanol was used as the internal standard. The chromatography conditions were the same as our previous studies [14].

4. Conclusions

The direct synthesis of DEC from carbon dioxide, ethanol and PO was developed over KNO₃-CeO₂ and KBr-KNO₃-CeO₂ catalyst. Supercritical CO₂ was used to extract the products and reactants in site, and the catalysts could be reused. In comparison with the direct synthesis of DEC from CO₂ and ethanol, the involvement of PO could improve the formation of DEC, simultaneously with the important chemicals of glycol coproduced. The yield of DEC was improved about 10-fold compared to that using single CeO₂ as the catalyst. The yield of DEC was about 13.0% with a selectivity of 38.5% under the optimized conditions with KBr-KNO₃-CeO₂ as the catalyst, which was close to the equilibrium yield calculated in this work.