Indian Journal of Chemistry Vol. 49A, September 2010, pp. 1182-1188

Transesterification of propylene with using Mg-Al-CO 3 hydrotalcite as solid base catalyst

C Murugan & H C Bajaj* Discipline of Inorganic Materials and Catalysis, Central Salt and Marine Chemicals Research Institute Council of Scientific and Industrial Research (CSIR), G B Marg, Bhavnagar 364 021, Gujarat, India Email: [email protected] Received 18 June 2010; revised and accepted 17 August 2010 The catalytic activity of synthetic hydrotalcites of varying Mg/Al molar ratio from 2 to 5 has been investigated for the transesterification of with methanol. The hydrotalcite with Mg/Al molar ratio of 5 at different calcination temperature has been characterized using FT-IR, P-XRD and surface area analyses. Under optimized conditions, the catalyst with Mg/Al molar ratio 5 shows highest propylene carbonate conversion (72 %) with 97 % selectivity. The catalyst shows the highest TON (280 mmol DMC/g cat ) ever reported with heterogeneous catalysts. A plausible reaction mechanism has been proposed based on the results obtained. Keywords : Catalysts, Transesterification, Hydrotalcites, , Magnesium, Aluminum IPC Code: Int. Cl. 9 B01J21/02; B01J21/10; B01J27/232; C01F5/00; C01F7/00; C07B41/14

Dimethyl carbonate (DMC) finds diversified DMC using hydrotalcite. The ionic species from the applications in the chemical industries 1,2 . Though one catalyst were found to play an important role in step synthesis of DMC from CO 2 and CH 3OH is overall reactivity of the reaction. The anion present in attractive, yield is relatively low due to high the hydrotalcite gallery plays an important role in thermodynamic stability and kinetic inertness of CO 2. determining the basic behavior of the materials Also, it leads to the deactivation of catalyst by water calcined below the decomposition temperature 18 . We formed during the process 3. Recently, cyclo-addition have studied the effect of N,N ′-dimethylformamide in of to epoxides under ambient cycloaddition reaction of carbon dioxide to propylene condition has been reported 4,5 and needs more oxide (step I in Scheme 1) using Mg/Al hydrotalcite 19 . emphasis on the transesterification step (Scheme 1) to In the present study, we report the efficiency of utilize CO 2 directly in a one pot synthesis. Mg-Al-CO 3 hydrotalcite in transesterification of Several researchers have reported the formation of propylene carbonate with methanol in atmospheric DMC from cyclic carbonates using different pressure. The effect of calcination and reaction catalysts 6-15 . The equilibrium yield of DMC from temperatures, reactant ratio and catalyst amount has propylene carbonate (PC) is found to be three times also been studied. lower than that from carbonate (EC) in the 6 presence of Na 3PO 4 as solid base catalyst . This has been explained in terms of steric factor though both the reactions have same activation energy 6 (29 ± 2 kJ/mol). Synthesis of dimethyl carbonate from at atmospheric pressure 9 and dialkyl carbonates from cyclic carbonates at higher pressure (25-30 bar) 11 have been reported using Mg-Al hydrotalcite as solid base catalyst. We have reported the synthesis of dimethyl carbonate from propylene carbonate and methanol using KF/Al 2O3 as efficient base catlyst 16 . Takagaki et al. 17 have recently reported the synthesis of glycerol carbonate from glycerol and MURUGAN & BAJAJ: CATALYTIC ACTIVITY OF Mg-Al-CO 3 HYDROTALCITES 1183

Materials and Methods indicator, were measured separately and the results Propylene carbonate (PC), ethylene carbonate (EC) are given under three different strengths. and biphenyl were purchased from Acros Organics, Catalytic reactions Belgium. Dimethyl carbonate, , In a typical catalytic experiment, the oven dried methanol, Mg(NO 3)2·6H 2O, Al(NO 3)3·9H 2O, Na 2CO 3, 25 mL round bottom flask was fed with 2.04 g and NaOH were procured from SD Fine Chemicals, (20 mmol) propylene carbonate, 6.4 g (200 mmol) India. Methanol was dried using standard procedure methanol, 80 mg biphenyl (internal standard) and and stored under activated molecular sieve 3A. Milli- 50 mg catalyst. The reaction was carried out under N 2 pore deionized water was used for the synthesis of atmosphere using N 2 filled balloon and long spiral hydrotalcite catalyst. condenser setup. The reaction setup was placed in a preheated oil bath (±2 °C temperature variation) and Synthesis of the catalyst the reaction was initiated by stirring (500 rpm). After The hydrotalcite samples with Mg/Al molar ratio 4 h, the product mass was filtered to remove the varying from 2-5 were synthesized by co-precipitation catalyst and analyzed by gas chromatography method 19,20 . The catalysts obtained were powdered, (Shimadzu 17A) equipped with a flame ionization dried at 120 °C for 4 h and stored in a vacuum detector (FID). The products were further confirmed desiccator to avoid contamination by atmospheric by GC–MS (Shimadzu GCMS, QP 2010); 5 % moisture. The catalysts were coded as HT-X with X diphenyl- and 95 % dimethylsiloxane universal representing the molar ratio of Mg/Al salts used for capillary column (60 m length, 0.25 mm diameter) synthesis. The catalyst (HT-5) was calcined separately was employed in both GC and GC-MS analyses. at 200 °C and 450 °C for 4 h to evaluate the effect of Propylene glycol was co-produced in all the reactions calcination temperature on the title reaction. equimolar to the DMC formed and thus it has been

Characterization of the catalyst exempted from the discussion. The conversion and Powder X-ray diffraction (P-XRD) patterns of the selectivity were calculated using the following relationship in response to the internal standard: hydrotalcite samples were recorded with Phillips X’Pert MPD system equipped with XRK 900 reaction % Conversion of PC = chamber, using Ni-filtered Cu-Kα radiation initial moles of PC − final moles of PC λ θ ×100 ( = 1.5405 Å) over a 2 range of 2–70°. The Fourier initial moles of PC transform infra red (FT-IR) spectra of the hydrotalcite moles of DMC formed samples were recorded with Perkin–Elmer (Spectrum % Selectivit y of DMC = ×100 moles of PC reacted GX) FT-IR system using KBr pellets. The surface area was computed using N 2 adsorption data Results and Discussion measured at 77 K on a Micromeritics, ASAP 2010 Catalyst characterization USA system. The BET isotherm model was used to The X-ray powder diffraction patterns of the calculate surface area while the HK model was used synthesized samples with the reflection planes (003), to calculate pore volume and pore radius. The samples (006), (012), (015), (018), (110), (113) and (116) were degassed at 120 °C for 4 h under vacuum matched perfectly with the literature report 22 . No −2 (5 × 10 mm Hg) prior to N 2 adsorption additional impurity phase was observed in any of the measurement. Hammett indicator method was used to synthesized catalysts (see Supplementary Data). The 21 determine the basic strength of the catalyst . basal spacing d(003) of HT-CO 3 at ~7.65 Å was in Phenolphthalein (H ≥ 9.8), bromothymol (H ≥ 7.2) good agreement with the literature value and and methyl red (H ≥ 4.2) were used as indicators for increased with decrease in the Al content (Table 1) as different basic strengths. In a typical measurement, also reported by Miyata 23 . The sample calcined at 50 mg of the catalyst in 5 mL dry methanol was 200 °C showed decreased intensity of the HT peaks stirred for 2 h under nitrogen atmosphere and the (003 and 006 reflections) and that at 450 °C showed resultant solution was titrated against standard acid the complete decomposition of HT phase into Mg-Al (0.02 mol/L benzene carboxylic acid in dry methanol) mixed oxide (Fig. 1). The intensity of 003 reflection solution. The end points, i.e., pink to colorless with of the reused catalyst after the 4 th cycle was found to phenolphthalein indicator, blue to yellow with decrease as compared to the fresh catalyst, showing bromothymol and yellow to red with methyl red some crystallinity loss.

1184 INDIAN J CHEM, SEC A, SEPTEMBER 2010

Table 1 – Physical characteristic values of HT samples with varying Mg/Al ratio

Catalyst d spacing of BET surface Pore Pore (003) plane surface area vol. radius (Å) (m 2/g) (cm 3/g) (nm) HT-2.0 7.60 86 0.45 10.6 HT-3.0 7.69 65 0.38 11.7 HT-4.0 7.92 90 0.39 8.6 HT-5.0 7.95 93 0.46 9.9 HT-5.0 a 7.55 82 0.40 9.6 HT-5.0 b - 229 0.72 6.3 aCalcined at 200 oC. bCalcined at 450 oC.

Fig. 2 ― FT-IR profile of HT-5 catalyst [(a) as synthesized; (b) calcined at 200 °C; (c) after first cycle; (d) after fourth cycle].

The BET surface, pore volume and pore radius values decreased on calcination at 200 °C as compared to those of the as synthesized catalyst. This may be due to the loss of interlayer water followed by decreased d-spacing of (003) plane (7.95 Å to 7.55 Å). On calcination at 450 °C, the catalyst completely loss its interlayer anions and surface hydroxyl groups resulting in the formation of Mg-Al mixed oxides. This resulted in increased surface area and pore volume but decreased pore radius. The hydrotalcite samples were found to have different basicity at different basic strengths

(Table 2). The basicity values in the basic strength Fig. 1 ― P-XRD patterns of HT-5 catalyst. [(a) as synthesized; between 7.2 and 9.8 increased with increase in Mg/Al (b) calcined at 200 °C; (c) calcined at 450 °C; (d) after first cycle; (e) after fourth cycle]. ratio with as synthesized catalysts and decreased with calcination temperature of HT-5. The basicity of ≥ The FT-IR spectra of hydrotalcite samples matched stronger basic sites (H 9.8) decreased with increase well with the literature data 24,25 . The FT-IR peak in Mg/Al ratio and increased with calcination at ~1650 cm −1 showed the presence of monodentate temperature of HT-5. The total number of basic sites presented in the calcined catalysts may not be the carbonate species with δO-C-OH asymmetric mode in as synthesized catalyst. The sample calcined at actual one, since the adsorption depends on the 200 °C showed the splitting of carbonate peak number of accessible sites which in turn depends on −1 the pore size of the catalyst. into two peaks at around 1530 and 1380 cm (Fig. 2). This is assigned to the change in the coordination of Screening of the catalyst the carbonate ion upon loss of interlayer The catalytic activity of hydrotalcite samples dried water molecules, as also noted by Cabrera et al .26 at 120 °C was studied for the transesterification Similar observation, though less pronounced was reaction and the data are given in Table 2. There was noticed in the reused catalyst ( d in Fig. 2) indicating no conversion of propylene carbonate at the studied some loss (~ 11.7 %) in catalytic activity after reaction temperature (130 °C) in the absence of four cycles. catalyst. The conversion of PC increased with MURUGAN & BAJAJ: CATALYTIC ACTIVITY OF Mg-Al-CO 3 HYDROTALCITES 1185

Table 2 – Effect of Mg/Al molar ratio on transesterification of PC with methanol. {React. cond.: Methanol/PC ratio 10:1 (mol/mol); Catalyst amt: 50 mg; Temp.: 130 °C ; Time 4 h}

a Catalyst Conv. of PC Selectivity of DMC Basicity (µmol/g cat ) (%) (%) 4.2

HT-2.0 61.3 90.2 44 126 14 184 HT-3.0 63.8 95.0 46 136 14 196 HT-4.0 67.8 97.9 42 159 11 212 HT-5.0 72.2 97.1 36 190 10 236 HT-5.0 b 54.5 95.4 40 144 24 208 HT-5.0 c 52.8 76.1 20 104 44 168 aValues are within ±5 % error limit. bCalcined at 200 °C. cCalcined at 450 °C. increase in Mg/Al molar ratio reaching a maximum of synthesis of dimethyl carbonate. Based on this 72.2 % with HT-5. The selectivity of DMC also concept and the selectivity data obtained with the increased with increase in Mg/Al molar ratio and samples calcined at high temperature, one can say that reached a maximum of 97 % with HT-5. Considering PC can be activated at the surface of the catalyst and the catalytic efficiency of the catalyst in terms of both methanol by the hydrogen bonded carbonate ions conversion and selectivity, HT-5 was selected for present in the interlayer of the catalyst. Thus, the further detailed investigations. overall reactivity of Mg-Al-CO 3 hydrotalcite samples in this transesterification reaction was assigned to the Effect of calcination temperature of the catalyst (HT-5) surface basicity and carbonate coordination with The catalyst HT-5 dried at 120 °C showed higher interlayer water/OH. activity as compared to the samples calcined at higher temperatures (Entries 5 and 6 in Table 2). The sample Effect of temperature and amount of catalyst calcined at 200 °C resulted in 54.5 % conversion with The increase in temperature increased the 95.4 % selectivity. The drop in conversion (~18.5 %) formation of DMC and reached equilibrium at 130 °C may be due to the change in coordination of carbonate within 4 h (Fig. 3a). Reaction temperature >130 °C species from interlayer water to interlayer OH upon resulted in lower selectivity of DMC, shifting the losing the interlayer water (Fig. 2). The presence of equilibrium back to the reactant side. Wei et al. 27 have H-bonding between interlayer OH (Lewis basic sites) reported similar reversing of equilibrium at higher and carbonate ion gave selectivity greater than 95 %, temperature. The cyclic carbonate with no sterically while the absence of interlayer water brought the hindered methyl group, i.e., ethylene carbonate, conversion down to 54.5 %. The presence of attained optimum conversion (entry 11 in Table 3) at interlayer water was found to accelerate the activation 80 °C in 4 h. This is in accordance with the difference of methanol via hydrogen bonded carbonate ion. The in the frequency factor between EC →DMC and weak and medium basicity values of the hydrotalcite PC →DMC reaction as explained by Filippis et al. 6 samples were matched well with the observed The conversion of PC was significantly increased conversion of PC (Table 2). The sample calcined at with increase in the reactant ratio up to 10 and was 450 °C showed lower selectivity (76.1 %) of DMC. less significant beyond the ratio 10 (Fig. 3b). This This might be due to the presence of high stronger might be due to the formation of methanol-DMC basic sites (Brønsted basic sites) present in Mg-Al azeotrope which is subsequently involved in deciding mixed oxide leading to the partial decomposition of the equilibrium shift. The variation of amount of PC. A similar lower selectivity has been observed catalyst shows that the conversion of PC increased with metal oxide catalyst in transesterification of with increase in amount of catalyst, keeping the ethylene carbonate with methanol 15 . Watanabe et al. 9 selectivity greater than 97% (Fig. 4). Above 50 mg, have reported that the ethylene carbonate gets the amount of catalyst was not significant in the activated at the edge of the HT layer during the optimum conversion (72 %) of 20 mmol PC.

1186 INDIAN J CHEM, SEC A, SEPTEMBER 2010

Fig. 3 ― Effect of reaction temperature at methanol/PC ratio 10:1 (mol/mol) (a) and reactant ratio at reaction temperature 130 °C (b). [React. cond.: catalyst (HT-5): 50 mg; time 4 h. 1, Conversion of PC (%); 2, Selectivity of DMC (%)].

Fig. 4 ― Effect of catalyst amount on [React. cond.: methanol/PC ratio: 10:1 (mol/mol); catalyst HT-5; temp. 130 °C; time 4 h. 1, Conversion of PC (%); 2, Selectivity of DMC (%)].

Efficiency of the catalyst in terms of TON

The catalytic efficiency of Mg-Al-CO used in this 3 remove the various adsorbed molecules such as study was compared with the data for catalysts 8,14,15,23 internal standard, propylene carbonate propylene reported in literature (Table 3). Since the glycol, etc. The washed catalyst was dried at 120 °C basicity values are not available for the reported for 4 h before use in the consecutive runs. The catalysts, TON (TON = mmol of product formed per catalyst showed 60.5 % conversion of PC with 92.4 % gram catalyst) was employed as a unit of comparison. DNC selectivity after the fourth cycle. This slight The efficiency of various heterogeneous catalysts on drop in activity (11.7 % in conversion and 4.7 % in the title reaction was compared with the presently selectivity) is due to the structural changes, i.e., loss studied catalyst (Table 3). The TON obtained in the crystallinity observed in the used catalyst as (280 mmol DMC/g ) was much higher than values cat seen by the P-XRD (e in Fig. 1) and FT-IR studies reported in literature. The TON of Mg-Al mixed (d in Fig. 2). oxide (HT-5 calcined at 450 °C) was also comparably higher (160 mmol DMC/g cat ) than that of metal oxides Plausible reaction mechanism reported in the literature 14,15,27 . Based on the above results, a possible reaction Reusability of the catalyst HT-5 mechanism is proposed (Scheme 2). In the first step, The catalyst once used in the fresh run was the H-bonded carbonate ions present in the basic collected and thoroughly washed with methanol to gallery of the catalyst abstract proton from methanol MURUGAN & BAJAJ: CATALYTIC ACTIVITY OF Mg-Al-CO 3 HYDROTALCITES 1187

Table 3 – Comparative efficiency of Mg-Al-CO 3 in transesterification of PC with methanol

Run Catalyst React. temp. React. time Conv. of PC Selectivity of DMC TON c Ref. (°C) (h) (%) (%)

1a HT-5.0 130 4 72.2 97.1 280.4 This work

2 Au/CeO 2 (1.5 wt%) 140 6 63 55 30.1 14 3 Fe-Zn double metal 170 8 - 86.6 36.4 8 cyanide 4 MgO 160 2 ~ 40 ~93 ~21 23 5 MgO 150 4 28 100 14 15

6 ZrO 2 150 4 11.8 100 5.9 15

7 CeO 2 150 4 32.8 98.8 16.2 15 8 ZnO 150 4 23 100 12.5 15 9 CaO 150 4 25.6 100 12.8 15 10 a Mg/Al mixed oxide 130 4 52.8 76.1 160.7 This work 11 b HT-5.0 80 4 82.9 98.6 326.9 This work aReact. cond.: Methanol/PC ratio: 10:1 (mol/mol); catalyst amt.: 50 mg. bEthylene carbonate was used instead of propylene carbonate. cYield of DMC in mmol/g catalyst.

- o forming methoxy species (CH 3O ) and the Lewis acid 200 C resulted in slightly lower conversion with site (active metal centre) present on the surface of HT nearly the same selectivity while that at 450 °C interacts with carbonyl oxygen of PC to generate resulted in decreased conversion and selectivity. The relatively charged carbonyl carbon. In the second increase in methanol to propylene carbonate ratio step, the methoxy ions attack the carbonyl carbon of increased the conversion of propylene carbonate, PC at the edge of the HT structure to produce DMC. shifting the equilibrium to the product side. The Simultaneously, the relative positive charge is catalyst showed very high TON with moderate transferred into the gallery to release proton, which is recyclability even after four cycles. subsequently involved in the formation of propylene glycol. The water molecule present in the gallery is Supplementary Data proposed to be involved in easier activation of P-XRD patterns of HT samples of varying Mg/Al methanol followed by the proton transfer during ratio may be obtained from the authors on request. simultaneous formation of DMC and PG. In the reaction with Mg-Al mixed oxide, both the reactants Acknowledgement are to be activated on the surface of the catalyst where Authors are thankful to CSIR for financial support the transfer of methoxy species and proton is under CSIR Network Programme NWP-0010. supposed to be slower due to the strong basic nature of the surface. This leads to the decomposition of PC References into CO and propylene oxide leading to decreased 1 Delledonne D, Rivetti F & Romano U, Appl Catal A: Gen, 2 221 (2001) 241. selectivity of DMC. 2 Shaikh A G & Sivaram S, Chem Rev, 96 (3) (1996) 951. 3 Wu X L, Meng Y Z, Xiao M & Lu Y X, J Mol Catal A: Conclusions Chem, 249 (2006) 93.

The hydrotalcite materials, Mg-Al-CO 3 with 4 Mele´ndez J, North M & Villuendas P, Chem Commun, different Mg/Al molar ratio from 2-5 have been (2009) 2577. synthesized and characterized with FT-IR and P-XRD 5 Yin S & Shimada S, Chem Commun, (2009) 1136. 6 Filippis P D, Scarsella M, Borgianni C & Pochetti F, Energy techniques. The catalyst HT-5 dried at 120 °C showed Fuels, 20 (1) (2006) 17. optimum activity at 130 °C. The catalyst calcined at 7 Knifton J F & Duranleau R G, J Mol Catal, 67 (1991) 389.

1188 INDIAN J CHEM, SEC A, SEPTEMBER 2010

8 Srivastava R, Srinivas D & Ratnasamy P, J Catal, 241 18 Constantino V R L & Pinnavaia T J, Inorg Chem, 34 (1995) (2006) 34. 883. 9 Watanabe Y & Tatsumi T, Micropor Mesopor Mater, 22 19 Murugan C, Sharma S K, Jasra R V & Bajaj H C, Indian J (1998) 399. Chem , 49A (2010) 288. 10 Sankar M, Nair C M, Murty K V G K & Manikandan P, Appl 20 Sharma S K, Parikh P A & Jasra R V, J Mol Catal A: Chem, Catal A: Gen, 312 (2006) 108. 278 (2007) 135.

11 Jonge J P D, Lange J &. Pello D H L, US Patent 6953864 B2 21 Yan S, Kim M, Salley S O &. Simon Ng K Y, Appl Catal A: (2005) , to Shell Oil Company. Gen, 360 (2009) 163. 12 Jagtap S R, Bhor M D & Bhanage B M, Catal Commun, 9 22 Chibwe K & Jones W J Chem Soc Chem Commun, (1989) (2008) 1928. 926. 13 Williams D B G, Sibiya M S, van Heerden P S, Kirk M & Harris R, J Mol Catal A: Chem, 304 (2009) 147. 23 Miyata S, Clays Clay Miner, 31 (1983) 305.

14 Ju´arez R, Corma A & García H, Green Chem, 2009, 11, 24 Kloprogge J T & Frost R L, J Solid State Chem, 146 (1999) 949. 506. 15 Bhanage B M, Fujita S & Ikushima Y, Arai M, Appl Catal A: 25 Pagano M A, Forano C & Besse J P, J Mater Chem, 13 Gen, 219 (2001) 259. (2003) 1988. 16 Murugan C, Bajaj H C & Jasra R V, Catal Lett, 137 (2010) 26 Cabrera I, Granados M L & Fierro J L G, Phys Chem Chem 224. Phys, 4 (2002) 3122. 17 Takagaki A, Iwatani K, Nishimura S & Ebitani K, Green 27 Wei T, Wang M, Wei W, Sun Y & Zhong B, Fuel Process Chem, 12 (2010) 578. Technol, 83 (1–3) (2003) 175.