- Email: [email protected]
Preparation of highly stable Ca-Zn-Al oxide catalyst and its catalytic performance for one-pot synthesis of dimethyl carbonate
Catalysis Today xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today journal homepage: www.elsevier.com/locate/cattod
Preparation of highly stable Ca-Zn-Al oxide catalyst and its catalytic performance for one-pot synthesis of dimethyl carbonate ⁎
Hualiang An, Guangjie Zhang, Xinqiang Zhao , Yanji Wang Hebei Provincial Key Lab of Green Chemical Technology and High Efficient Energy Saving, School of Chemical Engineering and Technology, Hebei University of Technology, Tianjin, 300130, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Urea Methanol One-pot method Dimethyl carbonate Ca-Zn-Al catalyst Response surface methodology
Ca-Zn-Al oxide was prepared by co-precipitation method and the effect of preparation conditions on its catalytic performance was investigated. The Ca-Zn-Al oxide exhibited an excellent catalytic activity and stability for onepot synthesis of dimethyl carbonate (DMC) from urea, 1,2-propylene glycol (PG), and methanol. XRD characterization and activity evaluation results indicated that new crystalline phases Ca3Al4ZnO10 and ZnAl2O4 could create a synergistic effect with CaO and ZnO, being responsible for the excellent catalytic performance of Ca-ZnAl oxide. Response surface methodology (RSM) was used to investigate the effect of operating variables and to optimize conditions for the first-step reaction of PG and urea to propylene carbonate. Then the effect of reaction conditions for the transesterification of propylene carbonate with methanol to DMC was also investigated. As a result, a DMC yield of 84.7% was obtained under the appropriate reaction conditions. The Ca-Zn-Al oxide could be reused for four times without a significant change in its catalytic activity.
1. Introduction Dimethyl carbonate (DMC) is considered as an green chemical and is widely used in a variety of fields owing to its versatile chemical reactivity and unique physical properties, especially as a safe substitute for phosgene, dimethyl sulfate or methyl halide [1–3]. In addition, it is a potential additive to gasoline because of its high oxygen content, low toxicity, and good biodegradability [4]. The methods for the synthesis of DMC include phosgenation process [5], transesterification of cyclic carbonates with methanol [6], alcoholysis of urea [7], reaction of methanol with CO2 [8], and oxidative carbonylation of methanol [9]. Among these routes, transesterification of propylene carbonate (PC) with methanol is an attractive route due to its several distinct advantages such as high DMC yield, mild reaction conditions and no corrosion to equipments. However, PC is derived from depleting petroleum and a great deal of 1,2-propylene glycol (PG) is co-produced as a byproduct in the production process, leading to a low utilization of the feed. To solve the problems mentioned above, we proposed an idea to transform the by-product PG back into PC by the reaction of PG with urea in 2004 [10]. According to this idea, an integrated reaction process can be described as the synthesis of DMC from urea and methanol using PG as a recycle agent (see Scheme 1). In this way, not only can the problem of PG utilization be solved, but also the reliance of PC supply on the petrochemical industry can be decreased.
⁎
In addition, the released ammonia can be recycled to produce urea by the reaction with CO2. So this process meets the principles of green chemistry. However, both the above patent and other studies only related to one or two separate reactions for DMC synthesis [11–17]. Wang et al. [18] used zinc-yttrium oxide to catalyze the synthesis of DMC by a two-step process starting from urea, ethylene glycol, and methanol, and focused on the catalytic performance of zinc-yttrium oxide for the two individual steps. In the first step, ethylene carbonate with a yield of 94% was synthesized from urea and ethylene glycol. In the second step, the transesterification of ethylene carbonate with methanol to DMC was performed, and the yield of DMC was 71%. The integration of multistep reactions is a matter of great importance for solving the problems separation and purification operation [19]. For this reason, we prepared Ca-Zn-Al oxide catalyst by introducing calcium hydroxide into Zn-Al hydroxide, and successfully performed one-pot reaction of urea, PG and methanol to DMC in 2015. The effect of reaction conditions was investigated by single factor experiment which cannot show the interactions among the factors. Furthermore, the stability of that Ca-Zn-Al oxide catalyst was poor; the yield of DMC decreased from 83.9% to 69.9% at the second run of catalyst[20]. Previous studies on the transesterification route and one-pot synthesis of DMC were performed using traditional one‐factor‐at‐a‐time approaches to optimize the operating parameters. Such experimental methods cannot consider interactions among the process variables. Response surface methodology
Corresponding author. E-mail addresses: [email protected], [email protected] (X. Zhao).
https://doi.org/10.1016/j.cattod.2018.03.006 Received 29 November 2017; Received in revised form 3 February 2018; Accepted 6 March 2018 0920-5861/ © 2018 Elsevier B.V. All rights reserved.
Please cite this article as: An, H., Catalysis Today (2018), https://doi.org/10.1016/j.cattod.2018.03.006
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
(CO2-TPD). The measurements were performed on a Micromeritics AutoChem II 2920 Chemisorption Analyzer (Micromeritics Instrument Corp., USA). Prior to the test, 0.2 g of the sample was placed in a quartz sample tube and then heated to a temperature not higher than the calcination temperature of the sample and maintained for 1 h in an atmosphere of helium. When the temperature was decreased to 110 °C, CO2 was introduced to attain adsorption saturation. Then, the sample was purged by helium for 1 h to remove the physically absorbed CO2. Finally, the TPD experiment started with a heating rate of 10 °C/min, and the CO2 desorption signal was detected by a thermal conductivity detector (TCD). The elemental composition of the Ca-Zn-Al oxide samples was determined by inductively coupled plasma-optical emission spectrometry (ICP-OES) on an Optima 7300 V spectrometer (PerkinElmer, USA). The morphological structure and element distribution of Ca-Zn-Al oxides were characterized with a Quanta 450 FEG Scanning Electron Microscope (SEM, FEI, USA) equipped with an Octane Plus Energy Dispersive Spectrometer (EDS, AMETEK Inc., USA).
Scheme 1. Route for DMC synthesis using PG as a recycle agent.
(RSM) is a mixture of mathematical and statistical techniques and can be used to evaluate the effects of process variables and their interactions on response variables [21,22]. Hence, RSM has been extensively applied with good results on chemistry and chemical engineering, biochemical engineering, food engineering, and so on [21–24]. Consequently, we prepared a Ca-Zn-Al oxide catalyst by co-precipitation method and used it to catalyze the one-pot synthesis of DMC from urea, PG and methanol in the present study. The catalyst preparation conditions and reaction conditions were optimized to improve DMC yield and Ca-Zn-Al oxide stability.
2.4. One-pot synthesis of DMC One-pot synthesis of DMC was conducted in two sequential reactions including the reaction of urea with PG to PC and the tranesterification of PC with methanol to DMC. Urea, PG, and Ca-Zn-Al oxide catalyst were first added to a four-necked flask and then heated to the reaction temperature under stirring. During this period of reaction, a flow of pure N2 was introduced into the flask to remove the byproduct of ammonia from the reaction system quickly. After the completion of the first-step reaction, a Vigreux column was fixed to the four-necked flask to drive off the products of the upcoming tranesterification reaction. During the period of tranesterification reaction, a flow of methanol was continuously pumped into the four-necked flask by a metering pump to keep the volume of the reaction liquid steady in the flask. After the completion of reaction, the reaction system was cooled to room temperature. Finally, the distillate obtained from the Vigreux column and the residue left in the flask were collected separately for quantitative analysis.
2. Experimental 2.1. Materials and reagents All of the chemical reagents used in this paper, namely, urea (AR, Tianjin Chemical Reagent Factory, China), 1,2-propylene glycol (AR, Tianjin Chemical Reagent Co. Ltd., China), anhydrous methanol (AR, Rionlon Tianjin Chemical Co. Ltd., China), calcium nitrate (AR, Sinopharm Chemical Reagent Co., Ltd, China), zinc nitrate, aluminium nitrate, sodium carbonate and sodium hydroxide (AR, Tianjin Fengchuan Chemical Reagent Technology Co. Ltd., China), were used as received without further purification. 2.2. Catalyst preparation
2.5. Product analysis Ca-Zn-Al oxide catalysts were prepared using co-precipitation methods. Take the Ca-Zn-Al oxide with n(Ca2+):n(Zn2+):n (Al3+) = 1.6:3:1 as an example: A mixture of Ca(NO3)2·4H2O (12.59 g), Zn(NO3)2·6H2O (29.75 g) and Al(NO3)3·9H2O (12.50 g) was dissolved in 133 mL of distilled water to attain an aqueous solution designated as A. An aqueous solution of NaOH and Na2CO3 with a molar ratio of 4 was prepared and designated as B. A and B were dropwise added into a beaker with 50 mL of distilled water under vigorously stirring, and the rate of solution A was maintained at 2 mL/min. The pH value of the mixed solution in the beaker was adjusted to about 9.5 by controlling the dropping rate of solution B. After aged at 40 °C for 24 h, the mixture was filtrated and the cake was washed several times with distilled water until the filtrate was neutral. Then the cake was dried at 110 °C for 12 h and calcined at 900 °C for 4 h to form Ca-Zn-Al oxide catalyst.
The reaction products were quantitatively analyzed on a SP3420A gas chromatograph (Beijing Beifen Ruili Analytical Instrument co. Ltd, China) with a PEG–20 M capillary column. Nitrogen was used as carrier gas at a flow rate of 30 mL/min. For the analysis of the distillate, the temperature of both injection port and FID was controlled at 180 °C. The column temperature was controlled according to the following program: an initial temperature of 50 °C and held for 3 min and then increased with a rate of 15 °C/min to 200 °C. An internal standard method was employed using n-propyl alcohol as the internal standard. For the analysis of the residue, the temperature of both injection port and FID was controlled at 220 °C. The column temperature was kept at 100 °C for 2 min, raised to 220 °C at a rate of 10 °C/min, and then held for 10 min. n-Butanol was used as the internal standard.
2.3. Catalyst characterization
3. Results and discussion
X-ray diffraction (XRD) patterns were recorded with a Rigaku D/ max-2550 diffractometer (Rigaku International Corp., Japan) using Cu Kα radiation. The scan range covered from 5° to 90° at a rate of 8°/min. The specific surface areas of the catalyst were obtained from N2 adsorption-desorption isotherm at 77 K with a Micromeritics ASAP 2020 Surface Area and Porosity Analyzer (Micromeritics Instrument Corp., USA), after degassing at 200 °C and 10 μm Hg for 4 h. Multipoint BET analysis method was used to calculate the specific surface area. The basicities of the Ca-Zn-Al oxide samples were measured by temperature-programmed desorption using CO2 as the probe molecule
3.1. Effect of preparation conditions on the catalytic performance of Ca-ZnAl oxide Ca-Zn-Al oxide catalysts were prepared using co-precipitation methods with calcium nitrate, zinc nitrate and aluminium nitrate as the precursors, and with NaOH and Na2CO3 aqueous as the precipitant. The effect of pH value, Zn/Al molar ratio, Ca/Al molar ratio, NaOH/Na2CO3 molar ratio, calcination temperature, and calcination time on the catalytic performance was investigated. The suitable reaction conditions obtained in our previous work were determined for the activity 2
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
Table 1 Effect of pH value on the catalytic performance of Ca-Zn-Al oxides.
Table 3 Catalytic performance and specific surface area of Ca-Zn-Al oxides with different Ca/Al molar ratios.
pH value
YDMC/%
YPC/%
YDMC+PC/%
8.5 9.5 10.5 11.5
82.5 82.4 79.3 79.5
1.8 3.5 4.1 2.5
84.3 85.9 83.4 82.0
Ca/Al molar ratio
YDMC/%
YPC/%
YDMC+PC/%
Specific surface area(m2/g)
0.6:1 1.6:1 2.6:1
80.9 82.4 79.8
4.9 3.5 1.8
85.8 85.9 81.6
13.9 12.7 10.8
Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
evaluation [20]: a PG/urea molar ratio of 2, a catalyst weight percentage of 2.7%, a reaction temperature of 170 °C for the first step and 73 °C for the second step, a reaction time of 2 h for the first step and 8 h for the second step, and a methanol flow rate of 0.6 mL/min.
base sites may be responsible for the slight decrease of DMC yield at the Zn/Al molar ratio of 4:1. Thus 3:1 was selected as the appropriate Zn/Al molar ratio. 3.1.3. Effect of Ca/Al molar ratio The effect of Ca/Al molar ratio on the specific surface area and catalytic performance of Ca-Zn-Al oxide was studied at a fixed Zn/Al molar ratio of 3 and the results are reported in Table 3. As can be seen from the table, when the Ca/Al molar ratio was increased from 0.6:1 to 2.6:1, the yield of DMC first increased and then decreased. When the Ca/Al molar ratio was 1.6:1, the yield of DMC and the total yield of both DMC and PC reached their highest values, 82.4% and 85.9%, respectively. Compared with the results for a Ca/Al molar ratio of 0.6:1, an almost identical total yield of both DMC and PC was achieved whereas the yield of DMC was slightly higher at a Ca/Al molar ratio of 1.6:1, indicating that the catalytic active centers of CaO were insufficient at a Ca/Al molar ratio of 0.6:1. When the Ca/Al molar ratio increased to 2.6:1, the total yield of both DMC and PC decreased evidently, possibly because excess CaO led to an unfavorable influence on the reaction of urea and PG [20]. In addition, the specific surface area of Ca-Zn-Al oxide also decreased with increasing the Ca/Al molar ratio, indicating that the specific surface area may be another factor affecting the reaction of urea and PG. Therefore, the appropriate Ca/Al molar ratio was 1.6:1. The morphological structure and element distribution of Ca-Zn-Al oxides with different Ca/Al molar ratio were characterized by SEM and EDS mapping and the results are shown in Fig. S1 (see Supplementary material). It can be seen from the SEM images that Ca-Zn-Al oxide shows a small particle size and the particle size grows with increasing the Ca/Al molar ratio. The elemental mappings show that the element of Ca, Zn and Al are homogeneously distributed in the Ca-Zn-Al oxide catalysts.
3.1.1. Effect of pH value The influence of pH value on the catalytic performance of Ca-Zn-Al oxides was investigated and the results are reported in Table 1. When the pH value was controlled at 8.5 and 9.5, the Ca-Zn-Al oxide exhibited similar catalytic performance; the yield of DMC was 82.5% and 82.4%, respectively. When the pH value was higher than 9.5, the DMC yield decreased obviously. A higher pH value could lead the decrease of Al2O3 content in Ca-Zn-Al oxide due to the dissolution of Al(OH)3. Since Al2O3 mainly played the role of dispersing and stabilizing active components [25,26], the decrease of Al2O3 content would result in the decline in the activity of Ca-Zn-Al oxide. The total yield of both DMC and PC obtained the highest value when the pH value was 9.5, and a higher yield of DMC might be expected upon a further optimization of other preparation conditions and reaction conditions. Therefore, the appropriate pH value was 9.5. 3.1.2. Effect of Zn/Al molar ratio The influence of Zn/Al molar ratio on the catalytic performance and specific surface area of Ca-Zn-Al oxide was studied at a fixed Ca/Al molar ratio of 1.6:1. Additionally, the basicities of Ca-Zn-Al oxides with different Zn/Al molar ratios were also measured by CO2-TPD and the results are listed in Table 2. With increasing the Zn/Al molar ratio, the yield of DMC and the total yield of both DMC and PC first increased and then decreased, whereas the specific surface area and the amount of both weak base sites and strong base sites decreased monotonically. When the Zn/Al molar ratio was 3:1, the yield of DMC and the total yield of both DMC and PC reached their highest values, 82.4% and 85.9%, respectively. ZnO is a good catalytic component for the reaction of urea and PG to PC [14,26], so it is expected that the variation of Zn/ Al molar ratio mainly affects the first-step reaction of PG and urea (1 mol DMC is formed from 1 mol PC, so the total yield of both DMC and PC essentially reflects the results of the first-step reaction). When the Zn/Al molar ratio was less than 3:1, the reaction may be restrained by an insufficient number of catalytic active centers. When the Zn/Al molar ratio was larger than 3:1, the number of catalytic centers was enough for the reaction, however, the catalytic activity decreased slightly. Since the transesterification of PC with methanol to DMC is an alkali-catalyzed reaction, it is inferred that the inadequate amount of
3.1.4. Effect of Na2CO3/NaOH molar ratio The XRD patterns of the Ca-Zn-Al oxides prepared with different Na2CO3/NaOH molar ratio are shown in Fig. 1, and the composition of crystal phase of Ca-Zn-Al oxides was listed in Table S1 (see Supplementary material). The sample using bare NaOH as the precipitator shows only the characteristic diffraction peaks of ZnO and ZnAl2O4, indicating that Ca2+ can not be effectively precipitated by NaOH. When the Na2CO3/NaOH molar ratio increased to 1:4, the characteristic diffraction peaks of CaO in addition to ZnO and ZnAl2O4 were observed obviously. With increasing the Na2CO3/NaOH molar ratio, the peak of ZnAl2O4 decreased gradually while a new crystalline phase of Ca3Al4ZnO10 appeared and increased gradually. The sample prepared with Na2CO3/NaOH molar ratio of 1:1 contained ZnO, CaO and Ca3Al4ZnO10 whereas the characteristic diffraction peaks of ZnAl2O4 disappeared. Unlike using NaOH as the precipitator, the phases of ZnO and CaO appeared in the sample with bare Na2CO3 as the precipitator. The catalytic performance of the Ca-Zn-Al oxides prepared with different Na2CO3/NaOH molar ratio was evaluated and the results are shown in Table 4. The total yield of both DMC and PC was 87.1% whereas the DMC yield was only 36.3% over the Ca-Zn-Al oxide prepared using NaOH only, presumably due to the lack of CaO in the catalyst. The Ca-Zn-Al oxides prepared using the mixed aqueous
Table 2 Catalytic performance, specific surface area and CO2-TPD measurement results of Ca-ZnAl oxides with different Zn/Al molar ratios. Zn/Al molar ratio
2:1 3:1 4:1
YDMC/%
80.1 82.4 80.8
YPC/%
2.8 3.5 2.2
YDMC+PC/%
82.9 85.9 83.0
Specific surface area (m2/g)
14.4 12.7 11.1
Desorption amount of CO2 (mmol/g) weak
strong
0.076 0.053 0.047
0.021 0.016 0.012
Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
3
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
Fig. 1. XRD patterns of the Ca-Zn-Al oxides prepared with different molar ratio of Na2CO3/NaOH. (a) XRD patterns with 2θ = 5 ∼90°, (b) XRD patterns with 2θ = 30 ∼39°.
Table 4 Catalytic performance of Ca-Zn-Al oxides with different Na2CO3/NaOH molar ratio.
Table 5 Effect of calcination parameters on the catalytic performance of Ca-Zn-Al oxides.
Na2CO3/NaOH molar ratio
YDMC/%
YPC/%
YDMC+PC/%
Calcination parameters
YDMC/%
YPC/%
YDMC+PC/%
0:1 1:4 1:3 1:1 1:0
36.3 82.4 82.7 82.0 67.5
50.8 3.5 1.7 1.5 5.5
87.1 85.9 84.4 83.5 73.0
800 °C/4 h 850 °C/4 h 900 °C/4 h 950 °C/4 h 1000 °C/4 h 950 °C/3 h 950 °C/5 h
80.1 79.9 82.4 84.7 83.3 78.3 77.0
3.4 2.6 3.5 1.9 2.1 2.4 1.9
83.5 82.5 85.9 86.6 85.4 80.7 78.9
Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
solution of Na2CO3/NaOH as precipitator exhibited an excellent catalytic performance. When Na2CO3/NaOH molar ratio was 1:4, the yield of DMC and the total yield of both DMC and PC were 82.4% and 85.9%, respectively. Using Na2CO3 alone as precipitator, the yield of DMC and the total yield of both DMC and PC were lower, 67.5% and 73.0%, respectively. The absence of Ca3Al4ZnO10 and (or) ZnAl2O4 might be responsible for the poor catalytic performance of Ca-Zn-Al oxide. Although a slightly lower DMC yield was obtained at a Na2CO3/NaOH molar ratio of 1:4 than that of 1:3, the total yield of both DMC and PC was relatively high at the Na2CO3/NaOH molar ratio of 1:4. The high total yield of both DMC and PC means a higher yield of DMC may be expected by a further optimization. All things considered, 1:4 was determined as the appropriate Na2CO3/NaOH molar ratio. According to the XRD characterization and activity evaluation results, Ca-Zn-Al oxides with the higher catalytic performance have one thing in common: there exist new crystalline phase Ca3Al4ZnO10 and/or ZnAl2O4 besides CaO and ZnO. The results suggest that Ca3Al4ZnO10 and/or ZnAl2O4 can create a synergistic effect with CaO and ZnO for the catalytic performance of Ca-Zn-Al oxide.
peaks of Ca3Al4ZnO10 appeared, meanwhile the characteristic diffraction peaks of ZnAl2O4 disappeared, indicating that Ca3Al4ZnO10 might be formed by the reaction of ZnAl2O4 and CaO at a higher temperature. The yield of DMC was obviously enhanced when the Ca3Al4ZnO10 was formed at a calcination temperature of 950 °C. The calcination time also had an obvious effect on the catalytic performance of the Ca-Zn-Al oxide. When the calcination time increased from 3 to 5 h, the yield of DMC and the total yield of both DMC and PC all first increased and then decreased. The lower catalytic performance of the Ca-Zn-Al oxide calcined for 3 h was possibly due to incomplete decomposition of the Ca-Zn-Al compounds. On the other hand, calcination for 5 h would lead to the sintering of Ca-Zn-Al oxide and then a decrease of the catalytic performance. As mentioned above, the appropriate conditions for the preparation of Ca-Zn-Al oxide were determined as follows: Ca(NO3)2, Zn(NO3)2 and Al(NO3)3 as precursors (n(Ca2+):n(Zn2+):n(Al3+) = 1.6:3:1), NaOH and Na2CO3 aqueous solution (n(OH−):n(CO32−) = 4:1) as the precipitant, using parallel-flow precipitation method, controlling the pH value of the mixed solution at 9.5, aging at 40 °C for 24 h, and calcination at 950 °C for 4 h. To investigate the reproducibility of Ca-Zn-Al oxide catalyst, Ca-Zn-Al oxide was prepared for four times under the appropriate conditions and their catalytic performance was evaluated in one-pot synthesis of DMC. The results listed in Table S3 (see Supplementary material) show that the preparation of Ca-Zn-Al oxide is repeatable. The Ca-Zn-Al oxide prepared under these conditions was then used in optimization of reaction conditions.
3.1.5. Effect of calcination parameters Under the conditions of a Ca/Zn/Al molar ratio of 1.6:3:1, a Na2CO3/NaOH molar ratio of 1:4 and a pH value of 9.5, the effect of calcination temperature and calcination time on the catalytic performance of Ca-Zn-Al was studied and the results are listed in Table 5. The highest DMC yield of 84.7% was achieved over the Ca-Zn-Al oxide catalyst calcined at 950 °C for 4 h, and the total yield of PC and DMC got the highest value as well over the catalyst prepared at the same conditions. The XRD patterns of the catalysts calcinated at different temperature for 4 h are illustrated in Fig. 2, and the composition of crystal phase of Ca-Zn-Al oxides was listed in Table S2 (see Supplementary material). Only the characteristic diffraction peaks of ZnO and CaO existed in the Ca-Zn-Al oxide calcined at 800 °C. When calcination temperature increased to 900 °C, ZnAl2O4 was formed in the sample. With continuing to increase calcination temperature, some new characteristic diffraction
3.2. Optimization of reaction process for PC synthesis from PG and urea The optimization of reaction conditions for PC synthesis from PG and urea was conducted using Box-Behnken design (BBD) coupled with RSM. Box-Behnken design (BBD) is an independent, rotatable or nearly rotatable second-order design based on three-level incomplete factorial designs. BBD is more efficient compared with other response surface 4
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
Fig. 2. XRD patterns of Ca-Zn-Al oxides calcinated at different temperature. (a) XRD patterns with 2θ = 5 ∼90°, (b) XRD patterns with 2θ = 30 ∼39°.
Table 6 Box-Behnken test design factors and levels for the reaction of urea and PG. Variables
Factor
PG/urea molar ratio reaction temperature /°C reaction time /h catalyst weight percentage/%
X1 X2 X3 X4
Table 7 Experimental design and results. Run
Coded levels of variables −1
0
+1
1:1 160 1 2.2
2:1 170 2 2.7
3:1 180 3 3.2
X1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
designs, such as central composite designs. Therefore, BBD is one of the best quadratic models for RSM and has been widely used in many fields [27]. In order to evaluate the influence of operating parameters on the reaction of PG and urea, four main factors of the first step were chosen: PG/urea molar ratio (X1), reaction temperature (X2), reaction time (X3) and catalyst weight percentage (X4) as shown in Table 6. The low, middle, and high levels of each variable were appointed as −1, 0, and +1. Since 1 mol DMC is formed from 1 mol PC, and the total yield of both DMC and PC (YDMC+PC) in one-pot synthesis of DMC essentially reflects the reaction results of PC and urea. YDMC+PC was selected as the response. In the optimization process the response can be related to the chosen variables by a quadratic model as follows: n
n
Y = β0 +
∑ i=1
βi χi +
∑ i=1
βii χi2 +
∑ i
βij χi χj
(1)
where, Y is a predicted response, β0 is a constant coefficient, βi is a linear coefficient, βii is a quadratic coefficient, βii is an interaction coefficient, Xi is the variable attributed to factor i, and Xj is the variable attributed to factor j. For statistical calculations, the levels of independent variables were coded as
Xi =
xi − x 0 Δxi
Variables
1:1 1:1 2:1 2:1 3:1 1:1 2:1 2:1 1:1 2:1 2:1 2:1 1:1 2:1 1:1 3:1 2:1 3:1 3:1 3:1 2:1 2:1 2:1 3:1 2:1 2:1 2:1 2:1 2:1
(−1) (−1) (0) (0) (1) (−1) (0) (0) (−1) (0) (0) (0) (−1) (0) (−1) (1) (0) (1) (1) (1) (0) (0) (0) (1) (0) (0) (0) (0) (0)
Response (YDMC+PC, %) X2/°C
X3/h
X4/%
Experimental
Predicted
180 170 170 160 180 170 170 160 160 170 180 160 170 180 170 160 170 170 170 170 180 180 160 170 170 170 170 170 170
2(0) 1(−1) 1(−1) 1(−1) 2(0) 2(0) 3(1) 2(0) 2(0) 3(1) 3(1) 2(0) 2(0) 1(−1) 3(1) 2(0) 1(−1) 2(0) 1(−1) 2(0) 2(0) 2(0) 3(1) 3(1) 2(0) 2(0) 2(0) 2(0) 2(0)
2.7 2.7 2.2 2.7 2.7 3.2 2.2 3.2 2.7 3.2 2.7 2.2 2.2 2.7 2.7 2.7 3.2 2.2 2.7 3.2 2.2 3.2 2.7 2.7 2.7 2.7 2.7 2.7 2.7
72.0 63.9 85.2 77.3 54.4 73.8 84.7 86.5 61.7 77.2 69.0 83.2 68.7 81.1 75.1 81.1 79.6 76.8 85.5 76.2 78.5 78.6 84.5 76.9 85.1 86.6 85.9 85.5 85.7
72.01 65.65 83.89 77.36 59.46 72.28 83.98 83.25 60.29 82.16 69.71 82.51 70.30 80.23 74.68 84.74 83.98 79.10 85.76 75.38 77.33 74.86 86.15 70.73 85.76 85.76 85.76 85.76 85.76
(1) (0) (0) (−1) (1) (0) (0) (−1) (−1) (0) (1) (−1) (0) (1) (0) (−1) (0) (0) (0) (0) (1) (1) (−1) (0) (0) (0) (0) (0) (0)
(0) (0) (−1) (0) (0) (1) (−1) (1) (0) (1) (0) (−1) (−1) (0) (0) (0) (1) (−1) (0) (1) (−1) (1) (0) (0) (0) (0) (0) (0) (0)
3.2.1. Examining the adequacy of model Analysis of variance (ANOVA) is important in determining the adequacy and significance of the quadratic model. The summary of ANOVA is shown in Table 8. The quadratic regression model is highly significant, which is evident from the very low P-value (< 0.0001) [28]. The significance of each coefficient of Eq. (3) is also determined by their P-values as listed in Table 8. The P-values less than 0.0500 indicate that model terms are significant. In this case, X1, X2, X1X2, X1X3, X2X3, X12 and X22 are significant model terms, i.e., they have significant effect on YDMC+PC. The correlation coefficient (R2) is often used to evaluate the good or bad of the fit between the predicted and experimental data. The high coefficient of determination (R2 = 0.9006 for YDMC+PC) indicates that there is good agreement between the predicted and experimental YDMC+PC. “Adequate precision” measures the signal to noise ratio and a ratio greater than 4 is desirable [29]. In the
(2)
where, xi is the actual value of variable, x0 is the value of xi at the center point, △xi is the step change in xi, i = 1, 2, 3, 4. The design matrix of the variables, and the predicted and experimental values of the response are presented in Table 7. The statistical model developed by applying multiple regression analysis methods using the experimental data can be described as follows: Y(DMC+PC) = 85.76 + 2.97X1 − 3.39X2 − 0.43X3 − 0.43X4 − 9.25 X1X2 − 4.95X1X3 − 1.42X1X4 − 4.82X2X3 − 0.80X2X4 + 0.47X3X4 −10.93X12 − 5.70X22 − 1.69X32 − 0.57X42 (3) 5
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
the present study, the interaction effects between PG/urea molar ratio and reaction temperature, PG/urea molar ratio and reaction time, and reaction temperature and reaction time are significant whereas the interaction effects between PG/urea molar ratio and catalyst weight percentage, reaction temperature and catalyst weight percentage, and reaction time and catalyst weight percentage are negligible. These results agreed well with that obtained from Table 8.
Table 8 Analysis of variance (ANOVA) for response surface quadratic model of the total yield of both DMC and PC. Source
Sum of squares
df
Mean squares
F-value
P-value (Prob > F)
Remark
Model X1 X2 X3 X4 X1X2 X1X3 X1X4 X2X3 X2X4 X3X4 X12 X22 X32 X42 Residual
1690.60 106.21 138.04 2.25 2.25 342.25 98.01 8.12 93.12 2.56 0.90 774.91 211.12 18.58 2.09 186.55
14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 14
120.76 106.21 138.04 2.25 2.25 342.25 98.01 8.12 93.12 2.56 0.90 774.91 211.12 18.58 2.09 13.32
9.06 7.97 10.36 0.17 0.17 25.68 7.36 0.61 6.99 0.19 0.068 58.15 15.84 1.39 1.16
< 0.0001 0.0135 0.0062 0.6871 0.6871 0.0002 0.0169 0.4480 0.0193 0.6678 0.7985 < 0.0001 0.0014 0.2573 0.6981
** * *
*
2
3.2.3. Optimization of reaction conditions It is predicted from the model that the highest total yield of both DMC and PC could reach 87.6%. Considering the existing highest experimental value (86.6% in Table 7) was close to the predicted optimum value, the optimal experiment point in Table 7 was determined as the optimum one. Therefore, the optimized conditions for the reaction of PG and urea were summarized as follows: a PG/urea molar ratio of 2, a catalyst weight percentage of 2.7%, a reaction temperature of 170 °C, and a reaction time of 2 h.
* * *
** *
Significant at 5% (P-value), **Significant at 0.1% (P-value), R = 0.9006, R just) = 0.8012.
2
3.3. Effect of the second-step reaction conditions on the one-pot synthesis of DMC (ad-
3.3.1. Effect of methanol flow rate Because a mixture of methanol and DMC was continuously distilled from the reactor in the tranesterification process, the flow rate of methanol introduced was adjusted to match the distillation rate to keep the volume of the reaction liquid steady in the reactor. The effect of methanol flow rate on the one-pot synthesis of DMC was studied and the results are reported in Table 9. When the flow rate of methanol was 0.4 mL/min, the yield of DMC was 80.2%. The DMC yield increased with increasing flow rate of methanol and reached 84.7% at a flow rate of methanol = 0.6 mL/min. It can be inferred that the formation rate of DMC increases with increasing the distillation rate. Thus, the equilibrium of the transesterification reaction was shifted in the forward direction. With a further increase of the flow rate of methanol, the yield of DMC exhibited no significant change, probably because the distillation rate of DMC was greater than the rate of transesterification. Therefore, 0.6 mL/min was determined as the appropriate flow rate of methanol.
present case, a ratio of 10.165 indicates an adequate signal. So this model can be used to navigate the design space. To understand whether the experimental data is normality, normal probability plot was developed. A normal probability plot is drawn by plotting the residuals of the experimental data against the corresponding residuals of a standard normal distribution. If the plot shows a straight line, it is reasonable to assume that the values come from a normal distribution. The normal probability plot drawn in this case suggests that the response data is normally distributed. Fig. 3 shows that the residuals are equal probability distributed in both sides of the straight line, indicating that the errors are normally distributed. 3.2.2. Interaction of parameters The response surface plot was generated by the above model to investigate the interaction of the parameters required for the reaction of PG and urea. The plots which depict the interactions between two variables by keeping the other variables at their zero levels are shown in Fig. 4. The shape of the contour plots (circle or elliptical) indicates whether there are mutual interactions between the two variables. For example, a circle contour plot indicates that the interaction between the two variables is negligible. In contrast, an elliptical contour plot indicates that the interaction between the two variables is significant. In
3.3.2. Effect of reaction time The second-step reaction temperature was held at about 73 °C for the mixture of DMC and methanol to be distilled out continuously. The effect of the second-step reaction time on the one-pot synthesis of DMC was investigated and the results are shown in Fig. 5. The reaction rate of the transesterification was high in the initial stage because of the high concentration of PC. The yield of DMC reached 74.2% for only 4 h. The transesterification reaction rate decreased gradually with the decrease of the PC concentration. As a result, the increased rate of DMC yield gradually decreased. The DMC yield increased to 84.7% at a reaction time of 8 h. While the reaction time was prolonged from 8 to 10 h, the yield of DMC increased by only 0.7%. Therefore, the appropriate reaction time for the second-step reaction was determined to be 8 h. 3.4. Reusability of Ca-Zn-Al oxide The Ca-Zn-Al oxide catalyst was recovered by filtrating after the completion of reaction and washed with methanol and dried in vacuum. Then the recovered Ca-Zn-Al oxide was used in the next run and the activity evaluation results are illustrated in Table 10. The catalyst could be reused four times without a significant change in its catalytic activity, and the DMC yield and the total yield of both DMC and PC remained at about 84% and 85%, respectively. Because the yield of DMC decreased from 82.9% to 69.9% at the second run over the Ca-Zn-Al oxide prepared by introducing calcium hydroxide into Zn-Al hydroxide [20], the Ca-Zn-Al oxide reported here shows much better reusability. The atomic compositions of the fresh and the fourth recovered Ca-
Fig. 3. Normal probability plot of residuals.
6
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
Fig. 4. Two-dimensional contour plots indicating the interaction effects: (1) PG/urea molar ratio and reaction temperature, (2) PG/urea molar ratio and reaction time, (3) reaction temperature and reaction time, (4) PG/urea molar ratio and catalyst weight percentage, (5) reaction temperature and catalyst weight percentage, (6) reaction time and catalyst weight percentage.
4. Conclusions
Table 9 Effect of methanol flow rate on the one-pot synthesis of DMC. Flow rate of methanol/mL/min
YDMC/%
YPC/%
YDMC+PC/%
0.4 0.6 0.8
80.2 84.7 84.2
4.8 1.9 2.0
85.0 86.6 86.2
(1) Ca-Zn-Al oxide catalyst was prepared by co-precipitation method and exhibited an excellent catalytic activity for the one-pot synthesis of DMC from urea, PG, and methanol. XRD characterization and activity evaluation results indicate that the new crystalline phase Ca3Al4ZnO10 and/or ZnAl2O4 could creat a synergistic effect with CaO and ZnO, being responsible for the good catalytic performance of Ca-Zn-Al oxide. In addition, the Ca-Zn-Al oxide also shows much better reusability, and the catalyst could be reused four times without a significant change in its catalytic activity. (2) The reaction conditions of the first-step reaction of PG and urea was optimized by the response surface methodology (RSM). The high regression coefficient (R2), namely 0.9006, implied good agreement between the predicted values and the experimental results. The RSM results demonstrated significant effects of reaction temperature and PG/urea molar ratio, and significant interactions separately between PG/urea molar ratio and reaction temperature, PG/
Reaction conditions: PG/urea molar ratio of 2, catalyst weight percentage of 2.7%, reaction temperature of 170 °C for the first step and 73 °C for the second step, reaction time of 2 h for the first step and 8 h for the second step. Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
Zn-Al oxide were analyzed by ICP-OES, and the Ca/Zn/Al atomic ratios are 1.51:2.65:1 and 1.45:2.31:1, respectively. Compared with the fresh catalyst, there is a small loss of Zn atom in the fourth recovered catalyst, which may lead to the slight decrease of PC yield.
7
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
Foundation of China (Grant No. 21506046, 21476058), Natural Science Foundation of Tianjin City (Grant No. 16JCQNJC06100), and Science and Technology Research Project of Colleges and Universities in Hebei Province (Grant No. QN2014144). The authors are gratefully appreciative of their contributions. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.cattod.2018.03.006. References [1] Y. Ono, Appl. Catal. A: Gen. 155 (1997) 133. [2] J. Tian, J. Wang, J. Chen, J. Fan, F. Cai, L. He, Appl. Catal. A: Gen. 301 (2006) 215. [3] X. Ding, X. Dong, D. Kuang, S. Wang, X. Zhao, Y. Wang, Chem. Eng. J. 240 (2014) 221. [4] M.A. Pacheco, C.L. Marshall, Energy Fuels 11 (1997) 2. [5] H. Babad, A.G. Zeiler, Chem. Rev. 73 (1973) 75. [6] J.F. Knifton, US patent 4661609 (1987). [7] R.Y. Saleh, R.C. Michaelson, E.N. Suciu, B. Kuhlmann, US patent 5565603 (1996). [8] J. Choi, L.N. He, H. Yasuda, T. Sakakura, Green Chem. 4 (2002) 230. [9] U. Romano, R. Tesel, M.M. Mauri, P. Rebora, Ind. Eng. Chem. Prod. Res. Dev. 19 (1980) 396. [10] X. Zhao, Y. Zhang, Y. Chen, Y. Wang, CN. Patent 200410019175.3 (2007). [11] Z. Gao, S. Wang, C. Xia, Chin. Chem. Lett. 20 (2) (2009) 131–135. [12] D. Wu, Y. Guo, S. Geng, Y. Xia, Ind. Eng. Chem. Res. 52 (2013) 1216. [13] X. Zhao, Z. Jia, Y. Wang, J. Chem. Technol. Biotechnol. 81 (2006) 794. [14] Q. Li, N. Zhao, W. Wei, Y. Sun, J. Mol. Catal. A: Chem. 270 (2007) 44. [15] H. Wang, S. Liu, W. Zhang, N. Zhao, W. Wei, Y. Sun, Acta Chim. Sinica 64 (2006) 2409. [16] P. Kumar, V.C. Srivastava, I.M. Mishra, Catal. Commun. 60 (2015) 27. [17] J. Holtbruegge, M. Leimbrink, P. Lutze, A. Górak, Chem. Eng. Sci. 104 (2013) 347. [18] P. Wang, S. Liu, F. Zhou, B. Yang, A.S. Alshammari, L. Lu, Y. Deng, Fuel Process. Technol. 126 (2014) 359. [19] Y. Wang, J. Hu, W. Xue, X. Zhao, CIESC J. 58 (2007) 2689. [20] G. Zhang, H. An, X. Zhao, Y. Wang, Ind. Eng. Chem. Res. 54 (2015) 3515. [21] M.A. Olutoye, B.H. Hameed, Appl. Catal. A: Gen. 371 (2009) 191. [22] M. Jourshabani, A. Badiei, N. Lashgari, G.M. Ziarani, Chin. J. Catal. 36 (2015) 2020. [23] J. Bourquin, H. Schmidli, P. van Hoogevest, H. Leuenberger, Eur. J. Pharm. Sci. 6 (1998) 287. [24] K.K. Kandimalla, N. Kanikkannan, M. Singh, J. Control. Release 61 (1999) 71. [25] J. Yang, Catalytic Hydrogenolysis of Glycerol to 1,2-Propanediol over Cu-based Catalysts Derived from Hydrotalcite-Type Precursors, Tianjin University, China, 2010 (M.S. thesis). [26] H. An, Y. Ma, X. Zhao, Y. Wang, Catal. Today 264 (2016) 136. [27] J. Song, C. Qiao, S. Li, Y. Zhou, M. Hsieh, H. Xu, J. Chromatogr. A 1216 (2009) 7007. [28] B.Y. Tak, B.S. Tak, Y.J. Kim, Y.J. Park, Y.H. Yoon, G.H. Min, J. Ind. Eng. Chem. 28 (2015) 307. [29] W. Guo, Z. Meng, N. Ren, Z. Zhang, F. Cui, Int. J. Hydrogen Energy 36 (2011) 5843.
Fig. 5. Effect of transesterification reaction time on the one-pot synthesis of DMC. Reaction conditions: PG/urea molar ratio of 2, catalyst weight percentage of 2.7%, reaction temperature of 170 °C for the first step and 73 °C for the second step, reaction time of 2 h for the first step, methanol flow rate of 0.6 mL/min.
Table 10 Reusability of Ca-Zn-Al oxide. Reused times
YDMC/%
YPC/%
YDMC+PC/%
1 2 3 4
84.7 83.8 84.0 84.9
1.9 1.3 1.0 0.9
86.6 85.1 85.0 85.8
Reaction conditions: PG/urea molar ratio of 2, reaction temperature of 170 °C for the first step and 73 °C for the second step, reaction time of 2 h for the first step and 8 h for the second step, the methanol flow rate of 0.6 mL/min. Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
urea molar ratio and reaction time, and reaction temperature and reaction time on the first-step reaction of PG and urea. Then the reaction conditions of the second-step reaction of PC and methanol to DMC was also investigated by single-factor experiments. As a result, the yield of DMC could reach 84.7% under the appropriate reaction conditions of a PG/urea molar ratio = 2, a reaction temperature of 170 °C for the first step and 73 °C for the second step, a reaction time of 2 h for the first step and 8 h for the second step, and a methanol flow rate of 0.6 mL/min. Acknowledgements This work was financially supported by National Natural Science
8
Contents lists available at ScienceDirect
Catalysis Today journal homepage: www.elsevier.com/locate/cattod
Preparation of highly stable Ca-Zn-Al oxide catalyst and its catalytic performance for one-pot synthesis of dimethyl carbonate ⁎
Hualiang An, Guangjie Zhang, Xinqiang Zhao , Yanji Wang Hebei Provincial Key Lab of Green Chemical Technology and High Efficient Energy Saving, School of Chemical Engineering and Technology, Hebei University of Technology, Tianjin, 300130, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Urea Methanol One-pot method Dimethyl carbonate Ca-Zn-Al catalyst Response surface methodology
Ca-Zn-Al oxide was prepared by co-precipitation method and the effect of preparation conditions on its catalytic performance was investigated. The Ca-Zn-Al oxide exhibited an excellent catalytic activity and stability for onepot synthesis of dimethyl carbonate (DMC) from urea, 1,2-propylene glycol (PG), and methanol. XRD characterization and activity evaluation results indicated that new crystalline phases Ca3Al4ZnO10 and ZnAl2O4 could create a synergistic effect with CaO and ZnO, being responsible for the excellent catalytic performance of Ca-ZnAl oxide. Response surface methodology (RSM) was used to investigate the effect of operating variables and to optimize conditions for the first-step reaction of PG and urea to propylene carbonate. Then the effect of reaction conditions for the transesterification of propylene carbonate with methanol to DMC was also investigated. As a result, a DMC yield of 84.7% was obtained under the appropriate reaction conditions. The Ca-Zn-Al oxide could be reused for four times without a significant change in its catalytic activity.
1. Introduction Dimethyl carbonate (DMC) is considered as an green chemical and is widely used in a variety of fields owing to its versatile chemical reactivity and unique physical properties, especially as a safe substitute for phosgene, dimethyl sulfate or methyl halide [1–3]. In addition, it is a potential additive to gasoline because of its high oxygen content, low toxicity, and good biodegradability [4]. The methods for the synthesis of DMC include phosgenation process [5], transesterification of cyclic carbonates with methanol [6], alcoholysis of urea [7], reaction of methanol with CO2 [8], and oxidative carbonylation of methanol [9]. Among these routes, transesterification of propylene carbonate (PC) with methanol is an attractive route due to its several distinct advantages such as high DMC yield, mild reaction conditions and no corrosion to equipments. However, PC is derived from depleting petroleum and a great deal of 1,2-propylene glycol (PG) is co-produced as a byproduct in the production process, leading to a low utilization of the feed. To solve the problems mentioned above, we proposed an idea to transform the by-product PG back into PC by the reaction of PG with urea in 2004 [10]. According to this idea, an integrated reaction process can be described as the synthesis of DMC from urea and methanol using PG as a recycle agent (see Scheme 1). In this way, not only can the problem of PG utilization be solved, but also the reliance of PC supply on the petrochemical industry can be decreased.
⁎
In addition, the released ammonia can be recycled to produce urea by the reaction with CO2. So this process meets the principles of green chemistry. However, both the above patent and other studies only related to one or two separate reactions for DMC synthesis [11–17]. Wang et al. [18] used zinc-yttrium oxide to catalyze the synthesis of DMC by a two-step process starting from urea, ethylene glycol, and methanol, and focused on the catalytic performance of zinc-yttrium oxide for the two individual steps. In the first step, ethylene carbonate with a yield of 94% was synthesized from urea and ethylene glycol. In the second step, the transesterification of ethylene carbonate with methanol to DMC was performed, and the yield of DMC was 71%. The integration of multistep reactions is a matter of great importance for solving the problems separation and purification operation [19]. For this reason, we prepared Ca-Zn-Al oxide catalyst by introducing calcium hydroxide into Zn-Al hydroxide, and successfully performed one-pot reaction of urea, PG and methanol to DMC in 2015. The effect of reaction conditions was investigated by single factor experiment which cannot show the interactions among the factors. Furthermore, the stability of that Ca-Zn-Al oxide catalyst was poor; the yield of DMC decreased from 83.9% to 69.9% at the second run of catalyst[20]. Previous studies on the transesterification route and one-pot synthesis of DMC were performed using traditional one‐factor‐at‐a‐time approaches to optimize the operating parameters. Such experimental methods cannot consider interactions among the process variables. Response surface methodology
Corresponding author. E-mail addresses: [email protected], [email protected] (X. Zhao).
https://doi.org/10.1016/j.cattod.2018.03.006 Received 29 November 2017; Received in revised form 3 February 2018; Accepted 6 March 2018 0920-5861/ © 2018 Elsevier B.V. All rights reserved.
Please cite this article as: An, H., Catalysis Today (2018), https://doi.org/10.1016/j.cattod.2018.03.006
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
(CO2-TPD). The measurements were performed on a Micromeritics AutoChem II 2920 Chemisorption Analyzer (Micromeritics Instrument Corp., USA). Prior to the test, 0.2 g of the sample was placed in a quartz sample tube and then heated to a temperature not higher than the calcination temperature of the sample and maintained for 1 h in an atmosphere of helium. When the temperature was decreased to 110 °C, CO2 was introduced to attain adsorption saturation. Then, the sample was purged by helium for 1 h to remove the physically absorbed CO2. Finally, the TPD experiment started with a heating rate of 10 °C/min, and the CO2 desorption signal was detected by a thermal conductivity detector (TCD). The elemental composition of the Ca-Zn-Al oxide samples was determined by inductively coupled plasma-optical emission spectrometry (ICP-OES) on an Optima 7300 V spectrometer (PerkinElmer, USA). The morphological structure and element distribution of Ca-Zn-Al oxides were characterized with a Quanta 450 FEG Scanning Electron Microscope (SEM, FEI, USA) equipped with an Octane Plus Energy Dispersive Spectrometer (EDS, AMETEK Inc., USA).
Scheme 1. Route for DMC synthesis using PG as a recycle agent.
(RSM) is a mixture of mathematical and statistical techniques and can be used to evaluate the effects of process variables and their interactions on response variables [21,22]. Hence, RSM has been extensively applied with good results on chemistry and chemical engineering, biochemical engineering, food engineering, and so on [21–24]. Consequently, we prepared a Ca-Zn-Al oxide catalyst by co-precipitation method and used it to catalyze the one-pot synthesis of DMC from urea, PG and methanol in the present study. The catalyst preparation conditions and reaction conditions were optimized to improve DMC yield and Ca-Zn-Al oxide stability.
2.4. One-pot synthesis of DMC One-pot synthesis of DMC was conducted in two sequential reactions including the reaction of urea with PG to PC and the tranesterification of PC with methanol to DMC. Urea, PG, and Ca-Zn-Al oxide catalyst were first added to a four-necked flask and then heated to the reaction temperature under stirring. During this period of reaction, a flow of pure N2 was introduced into the flask to remove the byproduct of ammonia from the reaction system quickly. After the completion of the first-step reaction, a Vigreux column was fixed to the four-necked flask to drive off the products of the upcoming tranesterification reaction. During the period of tranesterification reaction, a flow of methanol was continuously pumped into the four-necked flask by a metering pump to keep the volume of the reaction liquid steady in the flask. After the completion of reaction, the reaction system was cooled to room temperature. Finally, the distillate obtained from the Vigreux column and the residue left in the flask were collected separately for quantitative analysis.
2. Experimental 2.1. Materials and reagents All of the chemical reagents used in this paper, namely, urea (AR, Tianjin Chemical Reagent Factory, China), 1,2-propylene glycol (AR, Tianjin Chemical Reagent Co. Ltd., China), anhydrous methanol (AR, Rionlon Tianjin Chemical Co. Ltd., China), calcium nitrate (AR, Sinopharm Chemical Reagent Co., Ltd, China), zinc nitrate, aluminium nitrate, sodium carbonate and sodium hydroxide (AR, Tianjin Fengchuan Chemical Reagent Technology Co. Ltd., China), were used as received without further purification. 2.2. Catalyst preparation
2.5. Product analysis Ca-Zn-Al oxide catalysts were prepared using co-precipitation methods. Take the Ca-Zn-Al oxide with n(Ca2+):n(Zn2+):n (Al3+) = 1.6:3:1 as an example: A mixture of Ca(NO3)2·4H2O (12.59 g), Zn(NO3)2·6H2O (29.75 g) and Al(NO3)3·9H2O (12.50 g) was dissolved in 133 mL of distilled water to attain an aqueous solution designated as A. An aqueous solution of NaOH and Na2CO3 with a molar ratio of 4 was prepared and designated as B. A and B were dropwise added into a beaker with 50 mL of distilled water under vigorously stirring, and the rate of solution A was maintained at 2 mL/min. The pH value of the mixed solution in the beaker was adjusted to about 9.5 by controlling the dropping rate of solution B. After aged at 40 °C for 24 h, the mixture was filtrated and the cake was washed several times with distilled water until the filtrate was neutral. Then the cake was dried at 110 °C for 12 h and calcined at 900 °C for 4 h to form Ca-Zn-Al oxide catalyst.
The reaction products were quantitatively analyzed on a SP3420A gas chromatograph (Beijing Beifen Ruili Analytical Instrument co. Ltd, China) with a PEG–20 M capillary column. Nitrogen was used as carrier gas at a flow rate of 30 mL/min. For the analysis of the distillate, the temperature of both injection port and FID was controlled at 180 °C. The column temperature was controlled according to the following program: an initial temperature of 50 °C and held for 3 min and then increased with a rate of 15 °C/min to 200 °C. An internal standard method was employed using n-propyl alcohol as the internal standard. For the analysis of the residue, the temperature of both injection port and FID was controlled at 220 °C. The column temperature was kept at 100 °C for 2 min, raised to 220 °C at a rate of 10 °C/min, and then held for 10 min. n-Butanol was used as the internal standard.
2.3. Catalyst characterization
3. Results and discussion
X-ray diffraction (XRD) patterns were recorded with a Rigaku D/ max-2550 diffractometer (Rigaku International Corp., Japan) using Cu Kα radiation. The scan range covered from 5° to 90° at a rate of 8°/min. The specific surface areas of the catalyst were obtained from N2 adsorption-desorption isotherm at 77 K with a Micromeritics ASAP 2020 Surface Area and Porosity Analyzer (Micromeritics Instrument Corp., USA), after degassing at 200 °C and 10 μm Hg for 4 h. Multipoint BET analysis method was used to calculate the specific surface area. The basicities of the Ca-Zn-Al oxide samples were measured by temperature-programmed desorption using CO2 as the probe molecule
3.1. Effect of preparation conditions on the catalytic performance of Ca-ZnAl oxide Ca-Zn-Al oxide catalysts were prepared using co-precipitation methods with calcium nitrate, zinc nitrate and aluminium nitrate as the precursors, and with NaOH and Na2CO3 aqueous as the precipitant. The effect of pH value, Zn/Al molar ratio, Ca/Al molar ratio, NaOH/Na2CO3 molar ratio, calcination temperature, and calcination time on the catalytic performance was investigated. The suitable reaction conditions obtained in our previous work were determined for the activity 2
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
Table 1 Effect of pH value on the catalytic performance of Ca-Zn-Al oxides.
Table 3 Catalytic performance and specific surface area of Ca-Zn-Al oxides with different Ca/Al molar ratios.
pH value
YDMC/%
YPC/%
YDMC+PC/%
8.5 9.5 10.5 11.5
82.5 82.4 79.3 79.5
1.8 3.5 4.1 2.5
84.3 85.9 83.4 82.0
Ca/Al molar ratio
YDMC/%
YPC/%
YDMC+PC/%
Specific surface area(m2/g)
0.6:1 1.6:1 2.6:1
80.9 82.4 79.8
4.9 3.5 1.8
85.8 85.9 81.6
13.9 12.7 10.8
Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
evaluation [20]: a PG/urea molar ratio of 2, a catalyst weight percentage of 2.7%, a reaction temperature of 170 °C for the first step and 73 °C for the second step, a reaction time of 2 h for the first step and 8 h for the second step, and a methanol flow rate of 0.6 mL/min.
base sites may be responsible for the slight decrease of DMC yield at the Zn/Al molar ratio of 4:1. Thus 3:1 was selected as the appropriate Zn/Al molar ratio. 3.1.3. Effect of Ca/Al molar ratio The effect of Ca/Al molar ratio on the specific surface area and catalytic performance of Ca-Zn-Al oxide was studied at a fixed Zn/Al molar ratio of 3 and the results are reported in Table 3. As can be seen from the table, when the Ca/Al molar ratio was increased from 0.6:1 to 2.6:1, the yield of DMC first increased and then decreased. When the Ca/Al molar ratio was 1.6:1, the yield of DMC and the total yield of both DMC and PC reached their highest values, 82.4% and 85.9%, respectively. Compared with the results for a Ca/Al molar ratio of 0.6:1, an almost identical total yield of both DMC and PC was achieved whereas the yield of DMC was slightly higher at a Ca/Al molar ratio of 1.6:1, indicating that the catalytic active centers of CaO were insufficient at a Ca/Al molar ratio of 0.6:1. When the Ca/Al molar ratio increased to 2.6:1, the total yield of both DMC and PC decreased evidently, possibly because excess CaO led to an unfavorable influence on the reaction of urea and PG [20]. In addition, the specific surface area of Ca-Zn-Al oxide also decreased with increasing the Ca/Al molar ratio, indicating that the specific surface area may be another factor affecting the reaction of urea and PG. Therefore, the appropriate Ca/Al molar ratio was 1.6:1. The morphological structure and element distribution of Ca-Zn-Al oxides with different Ca/Al molar ratio were characterized by SEM and EDS mapping and the results are shown in Fig. S1 (see Supplementary material). It can be seen from the SEM images that Ca-Zn-Al oxide shows a small particle size and the particle size grows with increasing the Ca/Al molar ratio. The elemental mappings show that the element of Ca, Zn and Al are homogeneously distributed in the Ca-Zn-Al oxide catalysts.
3.1.1. Effect of pH value The influence of pH value on the catalytic performance of Ca-Zn-Al oxides was investigated and the results are reported in Table 1. When the pH value was controlled at 8.5 and 9.5, the Ca-Zn-Al oxide exhibited similar catalytic performance; the yield of DMC was 82.5% and 82.4%, respectively. When the pH value was higher than 9.5, the DMC yield decreased obviously. A higher pH value could lead the decrease of Al2O3 content in Ca-Zn-Al oxide due to the dissolution of Al(OH)3. Since Al2O3 mainly played the role of dispersing and stabilizing active components [25,26], the decrease of Al2O3 content would result in the decline in the activity of Ca-Zn-Al oxide. The total yield of both DMC and PC obtained the highest value when the pH value was 9.5, and a higher yield of DMC might be expected upon a further optimization of other preparation conditions and reaction conditions. Therefore, the appropriate pH value was 9.5. 3.1.2. Effect of Zn/Al molar ratio The influence of Zn/Al molar ratio on the catalytic performance and specific surface area of Ca-Zn-Al oxide was studied at a fixed Ca/Al molar ratio of 1.6:1. Additionally, the basicities of Ca-Zn-Al oxides with different Zn/Al molar ratios were also measured by CO2-TPD and the results are listed in Table 2. With increasing the Zn/Al molar ratio, the yield of DMC and the total yield of both DMC and PC first increased and then decreased, whereas the specific surface area and the amount of both weak base sites and strong base sites decreased monotonically. When the Zn/Al molar ratio was 3:1, the yield of DMC and the total yield of both DMC and PC reached their highest values, 82.4% and 85.9%, respectively. ZnO is a good catalytic component for the reaction of urea and PG to PC [14,26], so it is expected that the variation of Zn/ Al molar ratio mainly affects the first-step reaction of PG and urea (1 mol DMC is formed from 1 mol PC, so the total yield of both DMC and PC essentially reflects the results of the first-step reaction). When the Zn/Al molar ratio was less than 3:1, the reaction may be restrained by an insufficient number of catalytic active centers. When the Zn/Al molar ratio was larger than 3:1, the number of catalytic centers was enough for the reaction, however, the catalytic activity decreased slightly. Since the transesterification of PC with methanol to DMC is an alkali-catalyzed reaction, it is inferred that the inadequate amount of
3.1.4. Effect of Na2CO3/NaOH molar ratio The XRD patterns of the Ca-Zn-Al oxides prepared with different Na2CO3/NaOH molar ratio are shown in Fig. 1, and the composition of crystal phase of Ca-Zn-Al oxides was listed in Table S1 (see Supplementary material). The sample using bare NaOH as the precipitator shows only the characteristic diffraction peaks of ZnO and ZnAl2O4, indicating that Ca2+ can not be effectively precipitated by NaOH. When the Na2CO3/NaOH molar ratio increased to 1:4, the characteristic diffraction peaks of CaO in addition to ZnO and ZnAl2O4 were observed obviously. With increasing the Na2CO3/NaOH molar ratio, the peak of ZnAl2O4 decreased gradually while a new crystalline phase of Ca3Al4ZnO10 appeared and increased gradually. The sample prepared with Na2CO3/NaOH molar ratio of 1:1 contained ZnO, CaO and Ca3Al4ZnO10 whereas the characteristic diffraction peaks of ZnAl2O4 disappeared. Unlike using NaOH as the precipitator, the phases of ZnO and CaO appeared in the sample with bare Na2CO3 as the precipitator. The catalytic performance of the Ca-Zn-Al oxides prepared with different Na2CO3/NaOH molar ratio was evaluated and the results are shown in Table 4. The total yield of both DMC and PC was 87.1% whereas the DMC yield was only 36.3% over the Ca-Zn-Al oxide prepared using NaOH only, presumably due to the lack of CaO in the catalyst. The Ca-Zn-Al oxides prepared using the mixed aqueous
Table 2 Catalytic performance, specific surface area and CO2-TPD measurement results of Ca-ZnAl oxides with different Zn/Al molar ratios. Zn/Al molar ratio
2:1 3:1 4:1
YDMC/%
80.1 82.4 80.8
YPC/%
2.8 3.5 2.2
YDMC+PC/%
82.9 85.9 83.0
Specific surface area (m2/g)
14.4 12.7 11.1
Desorption amount of CO2 (mmol/g) weak
strong
0.076 0.053 0.047
0.021 0.016 0.012
Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
3
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
Fig. 1. XRD patterns of the Ca-Zn-Al oxides prepared with different molar ratio of Na2CO3/NaOH. (a) XRD patterns with 2θ = 5 ∼90°, (b) XRD patterns with 2θ = 30 ∼39°.
Table 4 Catalytic performance of Ca-Zn-Al oxides with different Na2CO3/NaOH molar ratio.
Table 5 Effect of calcination parameters on the catalytic performance of Ca-Zn-Al oxides.
Na2CO3/NaOH molar ratio
YDMC/%
YPC/%
YDMC+PC/%
Calcination parameters
YDMC/%
YPC/%
YDMC+PC/%
0:1 1:4 1:3 1:1 1:0
36.3 82.4 82.7 82.0 67.5
50.8 3.5 1.7 1.5 5.5
87.1 85.9 84.4 83.5 73.0
800 °C/4 h 850 °C/4 h 900 °C/4 h 950 °C/4 h 1000 °C/4 h 950 °C/3 h 950 °C/5 h
80.1 79.9 82.4 84.7 83.3 78.3 77.0
3.4 2.6 3.5 1.9 2.1 2.4 1.9
83.5 82.5 85.9 86.6 85.4 80.7 78.9
Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
solution of Na2CO3/NaOH as precipitator exhibited an excellent catalytic performance. When Na2CO3/NaOH molar ratio was 1:4, the yield of DMC and the total yield of both DMC and PC were 82.4% and 85.9%, respectively. Using Na2CO3 alone as precipitator, the yield of DMC and the total yield of both DMC and PC were lower, 67.5% and 73.0%, respectively. The absence of Ca3Al4ZnO10 and (or) ZnAl2O4 might be responsible for the poor catalytic performance of Ca-Zn-Al oxide. Although a slightly lower DMC yield was obtained at a Na2CO3/NaOH molar ratio of 1:4 than that of 1:3, the total yield of both DMC and PC was relatively high at the Na2CO3/NaOH molar ratio of 1:4. The high total yield of both DMC and PC means a higher yield of DMC may be expected by a further optimization. All things considered, 1:4 was determined as the appropriate Na2CO3/NaOH molar ratio. According to the XRD characterization and activity evaluation results, Ca-Zn-Al oxides with the higher catalytic performance have one thing in common: there exist new crystalline phase Ca3Al4ZnO10 and/or ZnAl2O4 besides CaO and ZnO. The results suggest that Ca3Al4ZnO10 and/or ZnAl2O4 can create a synergistic effect with CaO and ZnO for the catalytic performance of Ca-Zn-Al oxide.
peaks of Ca3Al4ZnO10 appeared, meanwhile the characteristic diffraction peaks of ZnAl2O4 disappeared, indicating that Ca3Al4ZnO10 might be formed by the reaction of ZnAl2O4 and CaO at a higher temperature. The yield of DMC was obviously enhanced when the Ca3Al4ZnO10 was formed at a calcination temperature of 950 °C. The calcination time also had an obvious effect on the catalytic performance of the Ca-Zn-Al oxide. When the calcination time increased from 3 to 5 h, the yield of DMC and the total yield of both DMC and PC all first increased and then decreased. The lower catalytic performance of the Ca-Zn-Al oxide calcined for 3 h was possibly due to incomplete decomposition of the Ca-Zn-Al compounds. On the other hand, calcination for 5 h would lead to the sintering of Ca-Zn-Al oxide and then a decrease of the catalytic performance. As mentioned above, the appropriate conditions for the preparation of Ca-Zn-Al oxide were determined as follows: Ca(NO3)2, Zn(NO3)2 and Al(NO3)3 as precursors (n(Ca2+):n(Zn2+):n(Al3+) = 1.6:3:1), NaOH and Na2CO3 aqueous solution (n(OH−):n(CO32−) = 4:1) as the precipitant, using parallel-flow precipitation method, controlling the pH value of the mixed solution at 9.5, aging at 40 °C for 24 h, and calcination at 950 °C for 4 h. To investigate the reproducibility of Ca-Zn-Al oxide catalyst, Ca-Zn-Al oxide was prepared for four times under the appropriate conditions and their catalytic performance was evaluated in one-pot synthesis of DMC. The results listed in Table S3 (see Supplementary material) show that the preparation of Ca-Zn-Al oxide is repeatable. The Ca-Zn-Al oxide prepared under these conditions was then used in optimization of reaction conditions.
3.1.5. Effect of calcination parameters Under the conditions of a Ca/Zn/Al molar ratio of 1.6:3:1, a Na2CO3/NaOH molar ratio of 1:4 and a pH value of 9.5, the effect of calcination temperature and calcination time on the catalytic performance of Ca-Zn-Al was studied and the results are listed in Table 5. The highest DMC yield of 84.7% was achieved over the Ca-Zn-Al oxide catalyst calcined at 950 °C for 4 h, and the total yield of PC and DMC got the highest value as well over the catalyst prepared at the same conditions. The XRD patterns of the catalysts calcinated at different temperature for 4 h are illustrated in Fig. 2, and the composition of crystal phase of Ca-Zn-Al oxides was listed in Table S2 (see Supplementary material). Only the characteristic diffraction peaks of ZnO and CaO existed in the Ca-Zn-Al oxide calcined at 800 °C. When calcination temperature increased to 900 °C, ZnAl2O4 was formed in the sample. With continuing to increase calcination temperature, some new characteristic diffraction
3.2. Optimization of reaction process for PC synthesis from PG and urea The optimization of reaction conditions for PC synthesis from PG and urea was conducted using Box-Behnken design (BBD) coupled with RSM. Box-Behnken design (BBD) is an independent, rotatable or nearly rotatable second-order design based on three-level incomplete factorial designs. BBD is more efficient compared with other response surface 4
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
Fig. 2. XRD patterns of Ca-Zn-Al oxides calcinated at different temperature. (a) XRD patterns with 2θ = 5 ∼90°, (b) XRD patterns with 2θ = 30 ∼39°.
Table 6 Box-Behnken test design factors and levels for the reaction of urea and PG. Variables
Factor
PG/urea molar ratio reaction temperature /°C reaction time /h catalyst weight percentage/%
X1 X2 X3 X4
Table 7 Experimental design and results. Run
Coded levels of variables −1
0
+1
1:1 160 1 2.2
2:1 170 2 2.7
3:1 180 3 3.2
X1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
designs, such as central composite designs. Therefore, BBD is one of the best quadratic models for RSM and has been widely used in many fields [27]. In order to evaluate the influence of operating parameters on the reaction of PG and urea, four main factors of the first step were chosen: PG/urea molar ratio (X1), reaction temperature (X2), reaction time (X3) and catalyst weight percentage (X4) as shown in Table 6. The low, middle, and high levels of each variable were appointed as −1, 0, and +1. Since 1 mol DMC is formed from 1 mol PC, and the total yield of both DMC and PC (YDMC+PC) in one-pot synthesis of DMC essentially reflects the reaction results of PC and urea. YDMC+PC was selected as the response. In the optimization process the response can be related to the chosen variables by a quadratic model as follows: n
n
Y = β0 +
∑ i=1
βi χi +
∑ i=1
βii χi2 +
∑ i
βij χi χj
(1)
where, Y is a predicted response, β0 is a constant coefficient, βi is a linear coefficient, βii is a quadratic coefficient, βii is an interaction coefficient, Xi is the variable attributed to factor i, and Xj is the variable attributed to factor j. For statistical calculations, the levels of independent variables were coded as
Xi =
xi − x 0 Δxi
Variables
1:1 1:1 2:1 2:1 3:1 1:1 2:1 2:1 1:1 2:1 2:1 2:1 1:1 2:1 1:1 3:1 2:1 3:1 3:1 3:1 2:1 2:1 2:1 3:1 2:1 2:1 2:1 2:1 2:1
(−1) (−1) (0) (0) (1) (−1) (0) (0) (−1) (0) (0) (0) (−1) (0) (−1) (1) (0) (1) (1) (1) (0) (0) (0) (1) (0) (0) (0) (0) (0)
Response (YDMC+PC, %) X2/°C
X3/h
X4/%
Experimental
Predicted
180 170 170 160 180 170 170 160 160 170 180 160 170 180 170 160 170 170 170 170 180 180 160 170 170 170 170 170 170
2(0) 1(−1) 1(−1) 1(−1) 2(0) 2(0) 3(1) 2(0) 2(0) 3(1) 3(1) 2(0) 2(0) 1(−1) 3(1) 2(0) 1(−1) 2(0) 1(−1) 2(0) 2(0) 2(0) 3(1) 3(1) 2(0) 2(0) 2(0) 2(0) 2(0)
2.7 2.7 2.2 2.7 2.7 3.2 2.2 3.2 2.7 3.2 2.7 2.2 2.2 2.7 2.7 2.7 3.2 2.2 2.7 3.2 2.2 3.2 2.7 2.7 2.7 2.7 2.7 2.7 2.7
72.0 63.9 85.2 77.3 54.4 73.8 84.7 86.5 61.7 77.2 69.0 83.2 68.7 81.1 75.1 81.1 79.6 76.8 85.5 76.2 78.5 78.6 84.5 76.9 85.1 86.6 85.9 85.5 85.7
72.01 65.65 83.89 77.36 59.46 72.28 83.98 83.25 60.29 82.16 69.71 82.51 70.30 80.23 74.68 84.74 83.98 79.10 85.76 75.38 77.33 74.86 86.15 70.73 85.76 85.76 85.76 85.76 85.76
(1) (0) (0) (−1) (1) (0) (0) (−1) (−1) (0) (1) (−1) (0) (1) (0) (−1) (0) (0) (0) (0) (1) (1) (−1) (0) (0) (0) (0) (0) (0)
(0) (0) (−1) (0) (0) (1) (−1) (1) (0) (1) (0) (−1) (−1) (0) (0) (0) (1) (−1) (0) (1) (−1) (1) (0) (0) (0) (0) (0) (0) (0)
3.2.1. Examining the adequacy of model Analysis of variance (ANOVA) is important in determining the adequacy and significance of the quadratic model. The summary of ANOVA is shown in Table 8. The quadratic regression model is highly significant, which is evident from the very low P-value (< 0.0001) [28]. The significance of each coefficient of Eq. (3) is also determined by their P-values as listed in Table 8. The P-values less than 0.0500 indicate that model terms are significant. In this case, X1, X2, X1X2, X1X3, X2X3, X12 and X22 are significant model terms, i.e., they have significant effect on YDMC+PC. The correlation coefficient (R2) is often used to evaluate the good or bad of the fit between the predicted and experimental data. The high coefficient of determination (R2 = 0.9006 for YDMC+PC) indicates that there is good agreement between the predicted and experimental YDMC+PC. “Adequate precision” measures the signal to noise ratio and a ratio greater than 4 is desirable [29]. In the
(2)
where, xi is the actual value of variable, x0 is the value of xi at the center point, △xi is the step change in xi, i = 1, 2, 3, 4. The design matrix of the variables, and the predicted and experimental values of the response are presented in Table 7. The statistical model developed by applying multiple regression analysis methods using the experimental data can be described as follows: Y(DMC+PC) = 85.76 + 2.97X1 − 3.39X2 − 0.43X3 − 0.43X4 − 9.25 X1X2 − 4.95X1X3 − 1.42X1X4 − 4.82X2X3 − 0.80X2X4 + 0.47X3X4 −10.93X12 − 5.70X22 − 1.69X32 − 0.57X42 (3) 5
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
the present study, the interaction effects between PG/urea molar ratio and reaction temperature, PG/urea molar ratio and reaction time, and reaction temperature and reaction time are significant whereas the interaction effects between PG/urea molar ratio and catalyst weight percentage, reaction temperature and catalyst weight percentage, and reaction time and catalyst weight percentage are negligible. These results agreed well with that obtained from Table 8.
Table 8 Analysis of variance (ANOVA) for response surface quadratic model of the total yield of both DMC and PC. Source
Sum of squares
df
Mean squares
F-value
P-value (Prob > F)
Remark
Model X1 X2 X3 X4 X1X2 X1X3 X1X4 X2X3 X2X4 X3X4 X12 X22 X32 X42 Residual
1690.60 106.21 138.04 2.25 2.25 342.25 98.01 8.12 93.12 2.56 0.90 774.91 211.12 18.58 2.09 186.55
14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 14
120.76 106.21 138.04 2.25 2.25 342.25 98.01 8.12 93.12 2.56 0.90 774.91 211.12 18.58 2.09 13.32
9.06 7.97 10.36 0.17 0.17 25.68 7.36 0.61 6.99 0.19 0.068 58.15 15.84 1.39 1.16
< 0.0001 0.0135 0.0062 0.6871 0.6871 0.0002 0.0169 0.4480 0.0193 0.6678 0.7985 < 0.0001 0.0014 0.2573 0.6981
** * *
*
2
3.2.3. Optimization of reaction conditions It is predicted from the model that the highest total yield of both DMC and PC could reach 87.6%. Considering the existing highest experimental value (86.6% in Table 7) was close to the predicted optimum value, the optimal experiment point in Table 7 was determined as the optimum one. Therefore, the optimized conditions for the reaction of PG and urea were summarized as follows: a PG/urea molar ratio of 2, a catalyst weight percentage of 2.7%, a reaction temperature of 170 °C, and a reaction time of 2 h.
* * *
** *
Significant at 5% (P-value), **Significant at 0.1% (P-value), R = 0.9006, R just) = 0.8012.
2
3.3. Effect of the second-step reaction conditions on the one-pot synthesis of DMC (ad-
3.3.1. Effect of methanol flow rate Because a mixture of methanol and DMC was continuously distilled from the reactor in the tranesterification process, the flow rate of methanol introduced was adjusted to match the distillation rate to keep the volume of the reaction liquid steady in the reactor. The effect of methanol flow rate on the one-pot synthesis of DMC was studied and the results are reported in Table 9. When the flow rate of methanol was 0.4 mL/min, the yield of DMC was 80.2%. The DMC yield increased with increasing flow rate of methanol and reached 84.7% at a flow rate of methanol = 0.6 mL/min. It can be inferred that the formation rate of DMC increases with increasing the distillation rate. Thus, the equilibrium of the transesterification reaction was shifted in the forward direction. With a further increase of the flow rate of methanol, the yield of DMC exhibited no significant change, probably because the distillation rate of DMC was greater than the rate of transesterification. Therefore, 0.6 mL/min was determined as the appropriate flow rate of methanol.
present case, a ratio of 10.165 indicates an adequate signal. So this model can be used to navigate the design space. To understand whether the experimental data is normality, normal probability plot was developed. A normal probability plot is drawn by plotting the residuals of the experimental data against the corresponding residuals of a standard normal distribution. If the plot shows a straight line, it is reasonable to assume that the values come from a normal distribution. The normal probability plot drawn in this case suggests that the response data is normally distributed. Fig. 3 shows that the residuals are equal probability distributed in both sides of the straight line, indicating that the errors are normally distributed. 3.2.2. Interaction of parameters The response surface plot was generated by the above model to investigate the interaction of the parameters required for the reaction of PG and urea. The plots which depict the interactions between two variables by keeping the other variables at their zero levels are shown in Fig. 4. The shape of the contour plots (circle or elliptical) indicates whether there are mutual interactions between the two variables. For example, a circle contour plot indicates that the interaction between the two variables is negligible. In contrast, an elliptical contour plot indicates that the interaction between the two variables is significant. In
3.3.2. Effect of reaction time The second-step reaction temperature was held at about 73 °C for the mixture of DMC and methanol to be distilled out continuously. The effect of the second-step reaction time on the one-pot synthesis of DMC was investigated and the results are shown in Fig. 5. The reaction rate of the transesterification was high in the initial stage because of the high concentration of PC. The yield of DMC reached 74.2% for only 4 h. The transesterification reaction rate decreased gradually with the decrease of the PC concentration. As a result, the increased rate of DMC yield gradually decreased. The DMC yield increased to 84.7% at a reaction time of 8 h. While the reaction time was prolonged from 8 to 10 h, the yield of DMC increased by only 0.7%. Therefore, the appropriate reaction time for the second-step reaction was determined to be 8 h. 3.4. Reusability of Ca-Zn-Al oxide The Ca-Zn-Al oxide catalyst was recovered by filtrating after the completion of reaction and washed with methanol and dried in vacuum. Then the recovered Ca-Zn-Al oxide was used in the next run and the activity evaluation results are illustrated in Table 10. The catalyst could be reused four times without a significant change in its catalytic activity, and the DMC yield and the total yield of both DMC and PC remained at about 84% and 85%, respectively. Because the yield of DMC decreased from 82.9% to 69.9% at the second run over the Ca-Zn-Al oxide prepared by introducing calcium hydroxide into Zn-Al hydroxide [20], the Ca-Zn-Al oxide reported here shows much better reusability. The atomic compositions of the fresh and the fourth recovered Ca-
Fig. 3. Normal probability plot of residuals.
6
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
Fig. 4. Two-dimensional contour plots indicating the interaction effects: (1) PG/urea molar ratio and reaction temperature, (2) PG/urea molar ratio and reaction time, (3) reaction temperature and reaction time, (4) PG/urea molar ratio and catalyst weight percentage, (5) reaction temperature and catalyst weight percentage, (6) reaction time and catalyst weight percentage.
4. Conclusions
Table 9 Effect of methanol flow rate on the one-pot synthesis of DMC. Flow rate of methanol/mL/min
YDMC/%
YPC/%
YDMC+PC/%
0.4 0.6 0.8
80.2 84.7 84.2
4.8 1.9 2.0
85.0 86.6 86.2
(1) Ca-Zn-Al oxide catalyst was prepared by co-precipitation method and exhibited an excellent catalytic activity for the one-pot synthesis of DMC from urea, PG, and methanol. XRD characterization and activity evaluation results indicate that the new crystalline phase Ca3Al4ZnO10 and/or ZnAl2O4 could creat a synergistic effect with CaO and ZnO, being responsible for the good catalytic performance of Ca-Zn-Al oxide. In addition, the Ca-Zn-Al oxide also shows much better reusability, and the catalyst could be reused four times without a significant change in its catalytic activity. (2) The reaction conditions of the first-step reaction of PG and urea was optimized by the response surface methodology (RSM). The high regression coefficient (R2), namely 0.9006, implied good agreement between the predicted values and the experimental results. The RSM results demonstrated significant effects of reaction temperature and PG/urea molar ratio, and significant interactions separately between PG/urea molar ratio and reaction temperature, PG/
Reaction conditions: PG/urea molar ratio of 2, catalyst weight percentage of 2.7%, reaction temperature of 170 °C for the first step and 73 °C for the second step, reaction time of 2 h for the first step and 8 h for the second step. Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
Zn-Al oxide were analyzed by ICP-OES, and the Ca/Zn/Al atomic ratios are 1.51:2.65:1 and 1.45:2.31:1, respectively. Compared with the fresh catalyst, there is a small loss of Zn atom in the fourth recovered catalyst, which may lead to the slight decrease of PC yield.
7
Catalysis Today xxx (xxxx) xxx–xxx
H. An et al.
Foundation of China (Grant No. 21506046, 21476058), Natural Science Foundation of Tianjin City (Grant No. 16JCQNJC06100), and Science and Technology Research Project of Colleges and Universities in Hebei Province (Grant No. QN2014144). The authors are gratefully appreciative of their contributions. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.cattod.2018.03.006. References [1] Y. Ono, Appl. Catal. A: Gen. 155 (1997) 133. [2] J. Tian, J. Wang, J. Chen, J. Fan, F. Cai, L. He, Appl. Catal. A: Gen. 301 (2006) 215. [3] X. Ding, X. Dong, D. Kuang, S. Wang, X. Zhao, Y. Wang, Chem. Eng. J. 240 (2014) 221. [4] M.A. Pacheco, C.L. Marshall, Energy Fuels 11 (1997) 2. [5] H. Babad, A.G. Zeiler, Chem. Rev. 73 (1973) 75. [6] J.F. Knifton, US patent 4661609 (1987). [7] R.Y. Saleh, R.C. Michaelson, E.N. Suciu, B. Kuhlmann, US patent 5565603 (1996). [8] J. Choi, L.N. He, H. Yasuda, T. Sakakura, Green Chem. 4 (2002) 230. [9] U. Romano, R. Tesel, M.M. Mauri, P. Rebora, Ind. Eng. Chem. Prod. Res. Dev. 19 (1980) 396. [10] X. Zhao, Y. Zhang, Y. Chen, Y. Wang, CN. Patent 200410019175.3 (2007). [11] Z. Gao, S. Wang, C. Xia, Chin. Chem. Lett. 20 (2) (2009) 131–135. [12] D. Wu, Y. Guo, S. Geng, Y. Xia, Ind. Eng. Chem. Res. 52 (2013) 1216. [13] X. Zhao, Z. Jia, Y. Wang, J. Chem. Technol. Biotechnol. 81 (2006) 794. [14] Q. Li, N. Zhao, W. Wei, Y. Sun, J. Mol. Catal. A: Chem. 270 (2007) 44. [15] H. Wang, S. Liu, W. Zhang, N. Zhao, W. Wei, Y. Sun, Acta Chim. Sinica 64 (2006) 2409. [16] P. Kumar, V.C. Srivastava, I.M. Mishra, Catal. Commun. 60 (2015) 27. [17] J. Holtbruegge, M. Leimbrink, P. Lutze, A. Górak, Chem. Eng. Sci. 104 (2013) 347. [18] P. Wang, S. Liu, F. Zhou, B. Yang, A.S. Alshammari, L. Lu, Y. Deng, Fuel Process. Technol. 126 (2014) 359. [19] Y. Wang, J. Hu, W. Xue, X. Zhao, CIESC J. 58 (2007) 2689. [20] G. Zhang, H. An, X. Zhao, Y. Wang, Ind. Eng. Chem. Res. 54 (2015) 3515. [21] M.A. Olutoye, B.H. Hameed, Appl. Catal. A: Gen. 371 (2009) 191. [22] M. Jourshabani, A. Badiei, N. Lashgari, G.M. Ziarani, Chin. J. Catal. 36 (2015) 2020. [23] J. Bourquin, H. Schmidli, P. van Hoogevest, H. Leuenberger, Eur. J. Pharm. Sci. 6 (1998) 287. [24] K.K. Kandimalla, N. Kanikkannan, M. Singh, J. Control. Release 61 (1999) 71. [25] J. Yang, Catalytic Hydrogenolysis of Glycerol to 1,2-Propanediol over Cu-based Catalysts Derived from Hydrotalcite-Type Precursors, Tianjin University, China, 2010 (M.S. thesis). [26] H. An, Y. Ma, X. Zhao, Y. Wang, Catal. Today 264 (2016) 136. [27] J. Song, C. Qiao, S. Li, Y. Zhou, M. Hsieh, H. Xu, J. Chromatogr. A 1216 (2009) 7007. [28] B.Y. Tak, B.S. Tak, Y.J. Kim, Y.J. Park, Y.H. Yoon, G.H. Min, J. Ind. Eng. Chem. 28 (2015) 307. [29] W. Guo, Z. Meng, N. Ren, Z. Zhang, F. Cui, Int. J. Hydrogen Energy 36 (2011) 5843.
Fig. 5. Effect of transesterification reaction time on the one-pot synthesis of DMC. Reaction conditions: PG/urea molar ratio of 2, catalyst weight percentage of 2.7%, reaction temperature of 170 °C for the first step and 73 °C for the second step, reaction time of 2 h for the first step, methanol flow rate of 0.6 mL/min.
Table 10 Reusability of Ca-Zn-Al oxide. Reused times
YDMC/%
YPC/%
YDMC+PC/%
1 2 3 4
84.7 83.8 84.0 84.9
1.9 1.3 1.0 0.9
86.6 85.1 85.0 85.8
Reaction conditions: PG/urea molar ratio of 2, reaction temperature of 170 °C for the first step and 73 °C for the second step, reaction time of 2 h for the first step and 8 h for the second step, the methanol flow rate of 0.6 mL/min. Y: yield, DMC: dimethyl carbonate, PC: propylene carbonate.
urea molar ratio and reaction time, and reaction temperature and reaction time on the first-step reaction of PG and urea. Then the reaction conditions of the second-step reaction of PC and methanol to DMC was also investigated by single-factor experiments. As a result, the yield of DMC could reach 84.7% under the appropriate reaction conditions of a PG/urea molar ratio = 2, a reaction temperature of 170 °C for the first step and 73 °C for the second step, a reaction time of 2 h for the first step and 8 h for the second step, and a methanol flow rate of 0.6 mL/min. Acknowledgements This work was financially supported by National Natural Science
8