Solvent-free carbonylation of glycerol with urea using metal loaded MCM-41 catalysts

S. E. Kondawar, A. S. Potdar and C. V. Rode*
Chemical Engineering and Process Development Division, CSIR-National Chemical Laboratory, Dr Homi Bhabha Road, India. E-mail: cv.rode@ncl.res.in; Fax: +912025902621; Tel: +912025902349

Received 1st October 2014 , Accepted 21st January 2015

First published on 23rd January 2015


Abstract

Reacting glycerol with urea is the most attractive option for the production of glycerol carbonate (GC) as it utilizes two inexpensive chemicals readily available in the chemical cycle. The overall result is the chemical fixation of carbon dioxide. A Zn/MCM-41(im) catalyst prepared by a wet impregnation method exhibited excellent activity for the reaction of glycerol and urea with 75% glycerol conversion and 98% selectivity to GC. Such excellent activity of the catalyst is explained based on the presence of both basic and acidic sites on the same catalyst which activates the glycerol and urea molecules, respectively.


Introduction

Today, worldwide commercial production of biodiesel via the transesterification route has resulted in a glut of glycerol that is available for downstream processing to a variety of value added products.1–3 This provides the best choice for excess glycerol utilization as well as a significant economic incentive for biodiesel process economies. On the one hand, selective hydrodeoxygenation of glycerol has been studied extensively to obtain the corresponding hydrogenation and/or dehydration products.4 On the other hand, glycerol to glycerol carbonate (GC) transformation offers a very challenging and interesting example from scientific and industrial points of view. As glycerol carbonate synthesis involves retention of all the three oxygen atoms of glycerol molecule, it represents a typical case of atom economy. In addition, glycerol carbonate is an environmentally benign polar solvent due to its biodegradability, low toxicity and excellent miscibility with water.5 More importantly, glycerol carbonate thus obtained from biorenewable resource, can become a substrate for manufacturing still high value glycidol for the new generation polymers.6 Current market price of the GC is US $3–4 per kg which is much more significant than that of glycerol (US $0.75 per kg). The conventional method reported for the synthesis of GC is from glycerol and CO or phosgene in the presence of solid heterogeneous Cu(I), Cu(II) and soluble PdCl2(1,10-phenanthroline) catalysts.7–10 More interestingly, Doro et al. proposed Waker-type Pd/Mn catalyst for direct reaction of 1,2 diols with CO and O2 to form the corresponding cyclic carbonates.11 However both CO and phosgene being highly toxic, make the entire process environmentally prohibitive. Although direct carboxylation of glycerol using CO2 has been also attempted for GC, CO2 activation requires severe conditions of temperature and pressure. For example, Aresta et al. reported Sn-based catalyst for the reaction of glycerol and CO2, at 180 °C and CO2 pressure of 5 MPa giving only 40% conversion with negligible formation of glycerol carbonate rendering the process unviable for practical purposes.12,13 Third route for glycerol carbonate involves transesterification of glycerol with di-ethyl/methyl carbonate over hydrotalcite catalysts with almost complete conversion and 76% selectivity to GC. Nevertheless, transesterification proceeds with a very high molar ratio of glycerol[thin space (1/6-em)]:[thin space (1/6-em)]diethyl carbonate of 1[thin space (1/6-em)]:[thin space (1/6-em)]17 which becomes a major bottleneck for commercial production.14 Again, the carbonates employed during transesterification are generally prepared using phosgene or by energy intensive process involving carboxylation of epoxide thus, posing the serious drawback of handling highly toxic materials. Use of dimethyl carbonate (DMC) could overcome these drawbacks to some extent but requirement of organic solvents, longer reaction time (8 h) and tedious downstream processing steps still are the major issues to resolve.15 Hence, attempts are being focused on using urea due to its ready availability at cheaper cost and safe handling, as well as the ammonia generated in the process can be recycled to produce urea again. This aspect in addition to use of bioglycerol, makes urea based GC synthesis completely atom efficient and sustainable process. GC synthesis via urea route is being studied extensively over the past decade, using mainly noble metal as well as basic oxide type of catalysts. Hammond et al. have reported gold (2.5% Au/ZnO) catalyst for carbonylation of glycerol using urea with 88% conversion but with only 56% selectivity to GC, along with side products like 2,3-dihydroxylpropyl carbamate, 4-(hydroxymethyl)oxazolidin-2-one, (2-oxo-1,3-dioxolan-4-yl) methyl carbamate.16 Aresta et al. have proposed a new, γ-zirconium phosphate catalyst that showed good activity with 80% conversion of glycerol with almost 100% selectivity to GC.17 With a homogeneous ZnCl2 catalyst almost complete selectivity to GC was reported for 80% conversion of glycerol but it again suffers from catalyst reusability for recycling.18 Noble metal catalysts like gold and Pd supported on MgO were also proved to be good catalysts for glycerol carbonylation with urea.19 Other catalyst variations included oxides of Co and La among which La showed higher activity of 68% conversion due its strongly basic nature, than cobalt catalyst.20,21 Acid–base bifunctional catalysts included hydrotalcite based Zn catalyst that gave the highest activity (82% conversion) and selectivity (88%) to GC.22 However, similar type of Zn-smectite catalyst showed a lower activity (65% conversion) and selectivity (75%) to GC.23 Recently, a homogeneous ZMG [diamminediisocyanatozinc(II)] complex for the GC synthesis via carbonylation of glycerol with urea showed 71% conversion of glycerol and 92% selectivity to GC.24 Thus, glycerol carbonylation with urea reported so far, has shown a wide variation in terms of both GC selectivity in the range of 50–70% and glycerol conversion. As this reaction needs an appropriate balance of acidic and basic sites of solid catalyst the scope therefore, still exists to develop a new catalyst system to achieve consistent high activity and selectivity, with an excellent recyclability (Scheme 1).
image file: c4ra11590b-s1.tif
Scheme 1 Synthesis of glycerol carbonate by carbonylation of glycerol with urea.

Our results presented here, demonstrate glycerol carbonate synthesis from glycerol carbonylation with urea over a mixed metal oxide catalyst supported on MCM-41 possessing an appropriate balance of acid–base sites under solvent-free conditions and with equimolar glycerol to urea ratio of 1[thin space (1/6-em)]:[thin space (1/6-em)]1. To the best of our knowledge, there is no report to date on the application of MCM-41 either as a support or as a catalyst for the synthesis of glycerol carbonate. A series of metals (Cu, Ni and Zn) loaded MCM-41 catalysts were prepared by two different methods, such as in situ and impregnation methods, which showed dramatic difference particularly on the GC selectivity due to variation in acid–base sites distribution on the surface.

Experimental

Glycerol (98% pure) purchased from Merck, India was used after pretreating with molecular sieves at 150 °C to remove moisture. Urea and H2SO4 were purchased from Thomas Baker, India while, cetyltrimethylammonium bromide (CTAB) and tetraethylorthosilicate (TEOS) were purchased from Aldrich Chemicals, Bangalore, India. All chemicals were used as received without further purification.

Characterization of the catalysts

X-ray diffraction (XRD) patterns were recorded on a P Analytical PXRD system (Model X-Pert PRO-1712), using Ni filtered Cu Kα radiation (λ = 0.154 nm) as an X-ray source (current intensity, 30 mA; voltage, 40 kV) and an X-accelerator detector. The samples were scanned in a 2θ range of 0–10° and 10–80° with the step size (2θ) of 0.0080° and scanning rate of 10.944° s−1. The species present in the catalysts were identified by their characteristic 2θ values of the relevant crystalline phases. The software program X-Pert High Score Plus was employed to subtract the contribution of the Cu Kα 2 line prior to data analysis.

The BET surface area, of the catalysts was determined by N2 adsorption–desorption technique using Micromeritics-2720 (Chemisoft TPx) volumetric instrument. NH3-TPD and CO2-TPD experiments were also carried out on a Micromeritics-2720 (Chemisoft TPx) instrument. In order to evaluate the acidity and basicity of the catalysts, ammonia and carbon dioxide TPD measurements were carried out by: (i) pre-treating the samples from room temperature to 200 °C under helium flow rate of 25 mL min−1, (ii) adsorption of ammonia/carbon dioxide at 40 °C and (iii) desorption of ammonia/carbon dioxide with a heating rate of 10 °C min−1 starting from the adsorption temperature to 700 °C.

Transmission electron microscopy (TEM) analysis was performed on a Jeol Model JEM 1200 electron microscope operated at an accelerating voltage of 120 kV. A small amount of specimen was prepared by ultrasonically suspending the powder sample in ethanol, and drops of the suspension were deposited on a carbon coated copper grid dried at room temperature before analysis.

X-Ray photoelectron spectroscopy (XPS) data were collected on a VG Scientific ESCA-3000 spectrometer using a non-monochromatised Mg Kα radiation (1253.6 eV) at a pressure of about 1 × 10−9 Torr (pass energy of 50 eV, electron takeoff angle 55) and overall resolution ∼0.7 eV determined from the full width at half maximum of the 4f7/2 core level of the gold surface. The error in the binding energy values were within 0.1 eV. The binding energy values were charge-corrected to the C1s signal (285.0 eV).

Catalyst preparation

Synthesis of catalyst by in situ metal loading. MCM-41 support was prepared by a hydrothermal synthesis method as described earlier.25 In a typical synthesis method, 2.46 g of NaOH was dissolved in 146 mL of deionised water and stirred for 5 min. To this solution, 5.96 g of cetyl trimethylammonium bromide (CTAB) was added, and the mixture was stirred for 15 min. A 14 g portion of tetraethyl orthosilicate (TEOS) was added drop wise, and the pH of solution was maintained in a range of 9–10 by adding dilute H2SO4. The resulting reaction mixture was stirred for 5 h at room temperature. Thereafter, the entire reaction mixture along with mother liquor was transferred to teflon lined autoclave and heated under static condition at 100 °C for 48 h. The resulting solid product was filtered and washed several time with water, dried at 100 °C overnight, and calcined at 540 °C for 6 h.

In a in situ preparation method, calculated amount of each of Ni, Cu and Zn solution was added before the addition of dilute H2SO4 for maintaining the pH. The prepared catalysts were nominated as Cu/MCM-41(in), Ni/MCM-41(in) and Zn/MCM-41(in), respectively.

Synthesis of catalyst by wet impregnation method. Ni, Cu and Zn (5 wt%) on MCM-41 catalysts were prepared by wet impregnation method. In a typical procedure, mixture of 0.4 g of each of Cu(NO3)2, NiCl2·6H2O and Zn(NO3)2·6H2O dissolved in 10 mL of water was stirred for 5 min, followed by the slow addition of MCM-41 support to the solution. The solution was kept for stirring for 6 h at room temperature. The solvent was evaporated on a rotavap and then the catalysts were calcined at 813 K for 3 h. The prepared catalysts were nominated as Cu/MCM-41(im), Ni/MCM-41(im) and Zn/MCM-41(im), respectively.
Synthesis of ZnO by chemical precipitation method. According to this method, two separate solutions zinc sulfate (1.5 mol L−1) and sodium bicarbonate (2.5 mol L−1) were simultaneously mixed by drop wise addition under continuous stirring at room temperature. The precipitate formed was filtered and washed with deionised water and ethanol. Then it was dried at 50 °C for 24 h followed by calcination at 500 °C for 1 h. The prepared catalyst was nominated as ZnO (PPT).

Reaction procedure

The activity evaluation experiments were performed in a glass batch reactor equipped with a reflux condenser. In a typical reaction, respective catalyst (0.23 g), glycerol (50 mmol), and urea (50 mmol) were charged into the reactor. The reaction mixture was heated in a silicon oil bath at 140 °C in absence of solvent under nitrogen flow. After completion of the reaction, 10 mL of MeOH was added in the reaction crude followed by filtration to remove the catalyst. Repeated washing with MeOH was given to the solid catalyst, in order to dissolve any adhered impurities on the catalyst surface. Liquid samples were withdrawn from the reactor and analyzed by gas chromatography to monitor the progress of the reaction. The gas chromatograph (Shimatzu 2025) was equipped with HPFFAP column (30 m length × 0.53 mm ID × 1 μm film thickness) and a flame ionization detector.

Results and discussion

The catalytic carbonylation of glycerol with urea to GC needs the balanced strength of both acidic as well as basic sites in a solid catalyst. In order to develop such a catalyst, we chose high surface area (1050 m2 g−1) MCM-41 having acidic sites, as a matrix for dispersion of Cu, Ni and Zn oxides possessing basic characteristics due to M2+O2− ion pairs. The extent of their basicity would depend on the individual co-ordination numbers of Cu, Ni and Zn.

The surface areas of the parent MCM-41 along with the catalysts prepared by in situ and impregnation methods are presented in Table 1. Cu, Ni and Zn loaded MCM-41 catalysts irrespective of their method of preparation, showed substantially lower surface area than that of the parent MCM-41, confirming the loading of the respective metal into the matrix.26 The surface area of Ni impregnated on the MCM-41, was lower than that of the Ni/MCM-41 prepared by the in situ protocol. This could be explained as the impregnation protocol would allow the surface of MCM-41 also covered by NiO while, in situ preparation preferably involves NiO occupying pores of the matrix. Whereas in case of CuO loaded MCM-41, both impregnated and in situ catalysts showed 798 m2 g−1 and 788 m2 g−1 surface area, respectively. The specific surface area of the MCM-41 samples was found to decrease on loading the ZnO which indicates the insertion of ZnO inside the mesopores.

Table 1 Crystallite size and surface area of prepared catalysts
Sr. no. Catalysts Crystallite size by XRD (nm) BET surface area (m2 g−1)
1 Cu/MCM-41(in) 43 788
2 Cu/MCM-41(im) 45 798
3 Ni/MCM-41(in) 51 968
4 Ni/MCM-41(im) 44 736
5 Zn/MCM-41(in) 753
6 Zn/MCM-41(im) 692
7 MCM-41 cal 550 °C 1050


XRD patterns of various (Cu, Ni and Zn)/MCM-41 are shown in Fig. 1. All the catalysts, showed a strong reflection at 2θ value of 2.3° due to (1 0 0) plane and two weak peaks at 2θ = 4.1° and 4.9° due to (1 1 0) and (2 0 0) planes, respectively which is consistent with the characteristic diffraction patterns of mesoporous MCM-41 confirming the retention of typical hexagonal structure of MCM-41 in all the samples. The intensity of the diffraction peak (1 0 0) of the Cu/MCM-41(im) was strong as compared with that of all the other samples, suggesting that the mesoporous ordering of Cu/MCM-41(im) was higher than that of the other samples.


image file: c4ra11590b-f1.tif
Fig. 1 XRD patterns of the (a) MCM-41 cal 550 °C, (b) Cu/MCM-41(in), (c) Cu/MCM-41(im), (d) Ni/MCM-41(im), (e) Ni/MCM-41(in), (f) Zn/MCM-41(in), (g) Zn/MCM-41(im).

Wide angle XRD pattern of various M/MCM-41 catalysts are shown in Fig. 2A. In all the samples, a broad peak at 2θ = 23.3° (4 1 1) was assigned to SiO2 (orthorhombic) [JCPDF File no. 82-1395]. In case of Cu/MCM-41 sample, peaks at 2θ = 35.2° (1 1 1), 38.6° (1 1 1) and 48.7° (2 0 2) were assigned to CuO phase [JCPDF File no. 80-1917] but this was of weak intensity. In Ni/MCM-41(im) sample, three peaks at 2θ = 37.2° (1 1 1), 43.2° (2 0 0) and 75.4° (3 1 1) were assigned to NiO phase [JCPDF File no. 86-0681]. However, these peaks were not observed in case of Ni/MCM-41 sample prepared by in situ method. The reason might be uniform distribution of NiO within the sheets of MCM-41 matrix and XRD being surface technique, peaks corresponding to NiO phase could not be visible. In case of both impregnated and in situ prepared samples of Zn-MCM-41, no any reflections corresponding to ZnO phase was observed which clearly indicated that ZnO nanoparticles were uniformly distributed inside the silica framework which is in accordance with the earlier reports on similar systems.27 The formation of CuO, NiO and ZnO basic phases in the respective catalysts were responsible for their basicity required in glycerol carbonylation with urea. On the other hand, wide angle XRD pattern of bare ZnO prepared by precipitation method in Fig. 2B showed strong peaks at 2θ = 31.7° (1 0 0), 34.4° (0 0 2), 36.2° (1 0 1), 47.6° (1 0 2), 56.7° (1 1 0), 66.4° (2 0 0) and 72.6° (0 0 4) which were assigned to ZnO phases [JCPDF no. 800075].


image file: c4ra11590b-f2.tif
Fig. 2 (A) Wide angle XRD patterns of the (a) MCM-41 cal 550 °C, (b) Cu/MCM-41(in), (c) Cu/MCM-41(im), (d) Ni/MCM-41(im), (e) Ni/MCM-41(in), (f) Zn/MCM-41(in), (g) Zn/MCM-41(im). (B) Wide angle XRD pattern of Bare ZnO cal 500 °C.

The acidity of the catalyst is also necessary in glycerol carbonylation reaction for the activation of carbonyl group of urea molecule to facilitate its interaction with glycerol.22 Hence, the strength and nature of acid sites of our catalysts were determined by NH3-TPD and the results are shown in Table 2. Although NH3-TPD gave the total acid strength without distinguishing Lewis and Brǿnsted acidity, the distribution of acid sites was observed in low (50–200 °C), medium (200–400 °C) and high (400–700 °C) temperature regions. MCM-41 support without loading of any metal, showed the lowest total concentration of acidic sites in terms of 0.01628 mmol g−1 NH3 desorbed (Fig. 3). All the catalysts prepared by wet impregnation method invariably showed higher acid strength than those prepared by in situ hydrothermal method. Cu-MCM-41 catalyst prepared by in situ protocol showed 0.02281 mmol g−1 total acidity which was lower as compared to the impregnated catalyst (0.02883 mmol g−1 NH3) where as Ni-MCM-41 catalyst showed total acidity in the lower temperature region (50–200 °C) with 0.02432 mmol g−1 of NH3 desorbed. Both impregnated and in situ prepared Zn-MCM-41 catalyst samples showed total acid strength of 0.0285 mmol g−1 and 0.0205 mmol g−1, respectively, with the highest concentration in the range of 100–200 °C relevant to the reaction temperature. Where as, bare ZnO prepared by precipitation method showed the highest total acidity of 0.0477 mmol g−1 of NH3 desorbed. Glycerol carbonylation with urea is also mediated by basic sites for activation of glycerol molecule which then becomes prone to attack the carbonyl carbon of urea molecule, as illustrated in Scheme 2b. Hence, the basic strength of the catalyst was determined by CO2-TPD and the results are presented in Table 3. A small peak below 100 °C was observed in all the samples due to the physically adsorbed CO2 (Fig. 4(A)). Among all the catalysts, the highest amount of CO2 desorbed, 5.298 mmol g−1 was observed in case of Cu/MCM-41(im) sample with a peak located at 357 °C. While in case of Ni/MCM-41(im) catalyst, a broad peak at 350 °C was observed with 3.218 mmol g−1 of CO2 desorption. Total amount of desorbed CO2 in case of Zn/MCM-41(in) and Zn/MCM-41(im) was 0.385 mmol g−1 and 1.617 mmol g−1 at 189 °C and 334 °C, respectively, which is quiet lower as compared to all other catalysts. The bare ZnO showed basicity of 1.793 mmol g−1 in terms of CO2 desorbed with a sharp peak at 286 °C (Fig. 4(B)).

Table 2 NH3-TPD results of prepared catalystsa
Sr. no. Catalysts Total NH3 desorbed (mmol g−1) Temperature range-wise desorption of NH3 (mmol g−1)
40–200 °C 200–400 °C 400–800 °C
a Case of bare ZnO prepared by precipitation method showed the highest total acidity of 0.0477 mmol g−1 of NH3 desorbed.
1 MCM-41 cal 500 °C 0.01628 0.01174 0.00454
2 Cu/MCM-41(in) 0.02281 0.00915 0.0083 0.00536
3 Cu/MCM-41(im) 0.02883 0.02608 0.00275
4 Ni/MCM-41(in) 0.02432 0.02432
5 Ni/MCM-41(im) 0.02724 0.02724
6 Zn/MCM-41(in) 0.02851 0.01045 0.00666 0.0114
7 Zn/MCM-41(im) 0.02054 0.01572 0.00473
8 Bare ZnO cal 500 °C 0.04778 0.01787 0.017 0.01291



image file: c4ra11590b-f3.tif
Fig. 3 (A) NH3-TPD plots of various catalysts. (B) NH3-TPD plot of bare ZnO cal 500 °C.

image file: c4ra11590b-s2.tif
Scheme 2 (A) Proposed of mechanism of acidic site of MCM-41 and basic site of metal-oxide catalyzed carbonylation of glycerol. (B) Proposed mechanism of Lewis acid–base sites of metal-oxide catalysed carbonylation of glycerol.
Table 3 CO2-TPD results of prepared catalysts
Sr. no. Catalysts Total CO2 desorbed (mmol g−1) Temperature range-wise desorption of CO2 (mmol g−1)
40–200 °C 200–400 °C 400–800 °C
1 Cu/MCM-41(in) 4.64432 0.32698 0.866080 3.45126
2 Cu/MCM-41(im) 5.29841 1.37886 1.28595 2.6336
3 Ni/MCM-41(in) 4.49588 0.2874 4.20848
4 Ni/MCM-41(im) 3.21815 0.3791 1.2494 1.58965
5 Zn/MCM-41(in) 1.38550 0.03540 1.3468 0.0033
6 Zn/MCM-41(im) 1.61776 0.33142 0.49061 0.79573
7 Bare ZnO cal 500 °C 1.79352 1.0684 0.72512



image file: c4ra11590b-f4.tif
Fig. 4 (A) CO2-TPD plots of various catalysts (a) Cu/MCM-41(in), (b) Cu/MCM-41(im), (c) Ni/MCM-41(in), (d) Ni/MCM-41(im), (e) Zn/MCM-41(in), (f) Zn/MCM-41(im). (B) CO2-TPD of bare ZnO cal 500 °C.

A more detailed structural characterization and metal particle size of freshly prepared catalysts were studied by HR-TEM. The honeycomb feature, typical of MCM-41 is clearly seen for all the samples (Fig. 5(A)a–d), which suggests that hexagonal pore channels were retained even after the addition of various metal components. HR-TEM image of Zn/MCM-41(in) sample taken with beam direction perpendicular to the pores indicates well aligned hexagonal pores. One cannot observe the particles of ZnO in the channels of MCM-41 by TEM method, probably due to the fact that the weak contrast exists between the silica framework of MCM-41 and ZnO particles.28 In line with this observation, absence of ZnO phases in XRD clearly indicates that ZnO was well dispersed in the framework of MCM-41. Nonetheless, strong zinc signals could be clearly observed in the EDX spectrum of Zn/MCM-41 as depicted in Fig. 5(B), confirming the presence of ZnO in MCM-41. On the other hand, Fig. 5(d) shows the HR-TEM of Ni/MCM-41(im) sample with well dispersed NiO particles having cube shape morphology and d-spacing of 0.22 nm, matching with the value obtained by XRD. This sample possessed a larger particle size of 44 nm and a wider size distribution (40–60 nm) than both Ni/MCM-41 and Cu/MCM-41 samples prepared by in situ method. These features were due to an excellent dispersion of the metal oxide particles within MCM-41 framework for the samples prepared by in situ method. The peaks for Cu, Ni and Zn were clearly detected in EDX, confirming the presence of Cu, Ni and Zn atoms in the respective catalysts.


image file: c4ra11590b-f5.tif
Fig. 5 (A) TEM images at different resolution of various catalysts (a) Cu/MCM-41(in), (b) Cu/MCM-41(im), (c) Ni/MCM-41(in), (d) Ni/MCM-41(im), (e) Zn/MCM-41(in), (f) Zn/MCM-41(im). (B) EDAX pattern of (a) Zn/MCM-41(in), (b) Zn/MCM-41(im).

The presence of ZnO was confirmed by X-ray photoelectron spectroscopy. The total scan of Zn/MCM-41(im) catalyst suggest that only Si, Zn, O and C elements were present and no other impurity except carbon was present in the sample. All the binding energies were corrected with reference to the carbon 1s binding energy position at 284.4 eV. The detail scans of O1s, Zn2p1/2 and Zn2p3/2 are illustrated in Fig. 6. Deconvolution of O1s spectra resulted into three peaks with binding energies of 532.9, 531.5 and 530.6 eV (Fig. 6(b)). The peaks at 531.5 eV and 530.6 eV were due to O1s for ZnO in accordance with the values reported by Lihitkar et al.29 Additional peak observed at BE = 532.9 eV was due to oxygen of adsorbed –OH over the surface of the catalyst. The binding energies for Zn2p3/2 and 2p1/2 were observed at 1022.2 eV and 1045.3 eV, respectively, with spin orbital splitting of ∼23.1 eV which was quite different from those for metallic Zn (24.6 eV) and bare ZnO (∼22.2 eV).30 This clearly suggests that in our Zn/MCM-41(im) catalyst, ZnO was not pure instead, it was bound to silica surface of MCM-41 support.


image file: c4ra11590b-f6.tif
Fig. 6 XPS spectra of (a) Zn2p of Zn/MCM-41(im), (b) O1s of Zn/MCM-41(im).

Activity measurement

A comparison of activity of various catalysts, expressed in terms of conversion, TOF and selectivity is presented in Table 4. Both Cu catalysts prepared by impregnation and in situ hydrothermal methods, showed almost same performance (65–66% conversion) but GC selectivity was significantly higher (83%) for the impregnated Cu catalyst. In both the cases, glycidol formation up to 5% was observed however, the major difference was the formation of (2-oxo-1,3-dioxolan-4-yl) methyl carbamate to the extent of 27%, sacrificing the GC selectivity, in case of in situ prepared sample. Although glycidol formation is via GC, however being a 3-membered ring its formation and stability is more difficult energetically than that of a 5-membered GC.31 Due to this, probably in the reported literature so far, glycidol has not been observed, while our Cu-catalyst system showed appreciable amount of glycidol. Between the two Ni catalysts, the impregnated Ni/MCM-41, showed the highest conversion of 68% as well as GC selectivity of 98% in 3 h time. In both the Ni catalysts glycidol formation was minimum, in a range of 0.8–1.2%. Where as, Zn catalyst showed highest conversion of 65 and 69% respectively, with almost 98% selectivity towards the glycerol carbonate without further dehydration to form glycidol. Due to excellent performance shown by Zn catalyst, the reaction was studied for extended hours and both in situ and impregnated Zn catalysts gave 73 and 75% conversion within 5 h. The promising performance of this Zn catalyst could be due to the appropriate balance of both acidic and basic sites.32 The presence of ZnO as evidenced in XPS (Fig. 6) was responsible for the basic sites required for the activation of glycerol molecule by abstracting its acidic proton (Scheme 2a). The RO thus formed subsequently attacks the carbonyl of urea (carbo-cation) generated by the Brǿnsted acidic sites formed on MCM-41.22 The extent of basicity desirable for optimal performance of Zn/MCM-41(im) catalyst is highlighted by the observation of lower activity of Cu and Ni containing catalysts as CuO and NiO were more basic than ZnO shown by CO2-TPD results discussed previously (Fig. 4).33 In addition, the amphoteric ZnO possesses appropriate balance of acid–base sites, is also responsible for the highest selectivity to GC whereas, Cu and Ni catalysts due to their higher basicity catalyze the further dehydration step to form glycidol. The critical role of amphoteric ZnO sites could be also understood by few control experiments. The activity and selectivity of only MCM-41 was lower 51% and 92% respectively, than that of Zn/MCM-41(im) as MCM-41 possessed only acidic sites. Similarly, the performance of basic ZnO alone prepared by precipitation also showed lower activity, however selectivity to GC was comparable to that of Zn/MCM-41(im) catalyst.
Table 4 Activity results of prepared catalysts for glycerol carbonylationa
Sr. no. Catalysts Conversion (%) Selectivity (%) TOF (h) % Yield GC Acid–base ratio
Glycerol carbonate Glycidol Other
a Reaction conditions: glycerol (4.6 g), urea (3.06 g), catalyst 0.23 g, temperature 145 °C, N2 flow to remove NH3 formed during reaction, reaction time (a) 3 h, (b) 5 h.
1 Cu/MCM-41(in)a 66 68 4.6 27.4 311 45 0.0049
2 Cu/MCM-41(im)a 65 83 4.1 12.4 306 54 0.005
3 Ni/MCM-41(in)a 60 89 1.2 9.8 340 53 0.0054
4 Ni/MCM-41(im)a 68 98 0.8 1.2 385 67 0.0084
5 MCM-41(cal 550 °C)a 51 92 1.7 6.3 2.28 47
6 Zn/MCM-41(in)a 65 98 2 368 64 0.0205
7 Zn/MCM-41(im)a 69 98 2 391 68 0.0126
8 Zn/MCM-41(in)b 73 98 2 248 71
9 Zn/MCM-41(im)b 75 98 2 255 73
10 ZnO cal 500 °C 68 99 1 231 67 0.0266


In order to investigate the mechanistic pathway, instead of urea, carbonylimidazole was reacted with glycerol to form glycerol carbonate. As shown in Scheme 2, intermediate is formed with the elimination of one mole of imidazole, which further undergoes attack of a secondary –OH group on to the carbonyl carbon to form GC with elimination of a second molecule of imidazole (Scheme 3).


image file: c4ra11590b-s3.tif
Scheme 3 Carbonylation of glycerol using carbonylimidazole to form glycerol carbonate.

Catalyst stability was proved by its recycle experiments in which Zn/MCM-41 catalyst was separated by filtration after its use, washed with methanol followed by calcination at 813 K for 5 h. Fig. 7 clearly reveals that after the third recycle also, 73% conversion of glycerol with 98% selectivity to GC was achieved. Catalyst stability was also confirmed by performing the standard leaching test, in which the catalyst was removed from the reaction by filtration after 60 min, and the reaction was continued with the filtrate devoid of catalyst. The conversion was 58% at the end of 60 min. and no any appreciable increase in the conversion (57%) was observed even after 3 h, (Fig. 8) which confirmed that there is no leaching of the catalyst ruling out possibility of any homogeneous complex formation with zinc. Quantitative elemental analysis of the reaction crude after separation of the catalyst by filtration, was done by inductively coupled plasma optical emission spectroscopy (ICP-OES) on a Spectro Arcos FHS-12 instrument equipped with winlab software and which did not show any presence of Zn. This confirmed that no catalyst dissolution occurred in the glycerol phase under the reaction conditions.


image file: c4ra11590b-f7.tif
Fig. 7 Recycle experiments for the synthesis of glycerol carbonate. Reaction conditions: glycerol (4.6 g), urea (3.06 g), catalyst Zn/MCM-41(im) 0.23 g, atmospheric pressure with nitrogen as passing gas, 140 °C, time 5 h.

image file: c4ra11590b-f8.tif
Fig. 8 Leaching test.

Influence of the reaction time

The graphical representation of time online analysis is shown in Fig. 9. After 3 h, 68% conversion of glycerol was observed with almost 98% selectivity towards the glycerol carbonate. Extended reaction time up to 7 h gave much higher conversion of 92% with 99% selectivity to GC. Further increase in reaction time up to 12 h reaction time led to a marginal increase in conversion to ∼94%. This is because of the fact that the reaction is carried out in pure glycerol which is the limiting reactant hence, as the reaction progresses, the decreasing concentration of glycerol affects the kinetics of the reaction.
image file: c4ra11590b-f9.tif
Fig. 9 Time online analysis for in carbonylation of glycerol over Zn/MCM-41(im) catalyst. Reaction conditions: glycerol (4.6 g), urea (3.06 g), catalyst 0.23 g, temperature 145 °C, N2 flow to remove NH3 formed during reaction, reaction time 12 h.

Conclusions

A series of transition metal loaded (Cu, Ni and Zn) mesoporous MCM-41 catalysts were prepared by two different protocols viz. in situ and wet impregnation method to investigate the effect of preparation method and aicd–base sites on the activity and selectivity of glycerol carbonylation reaction. Among these catalysts, Zn/MCM-41(im) prepared by wet impregnation method proved to be the most active catalyst for glycerol carbonate synthesis with 75% conversion of glycerol and 98% selectivity to GC. The higher activity of the catalyst can be explained by high surface area of about 692 m2 g−1 and excellent dispersion of ZnO phase over the mesoporous framework of MCM-41 support. Well balanced acid (0.020 mmol g−1) and basic sites (1.61 mmol g−1) present in the catalyst were responsible for the good catalytic performance. HRTEM images clearly showed the presence of amphoteric ZnO phase, which is important for the activation of glycerol molecule. The catalyst was thoroughly characterized by XRD, BET, NH3-TPD, TEM and XPS. Easy separation, excellent activity and recyclability of our catalyst make it a candidate for the application in industrial synthesis of GC.

Abbreviations

GCGlycerol carbonate
MeOHMethanol

Acknowledgements

One of the authors, S. E. Kondawar thanks to the CSIR New Delhi, for the award of Junior research fellowship.

References

  1. C. Cannilla, G. Bonura, E. Rombi, F. Arena and F. Frusteri, Appl. Catal., A, 2010, 382, 158–166 CrossRef CAS.
  2. A. Behr, J. Eilting, K. Irawadi, J. Leschinski and F. Lindner, Green Chem., 2008, 10, 13–30 RSC.
  3. M. Pagliaro, R. Ciriminna, H. Kimura, M. Rossi and C. D. Pina, Angew. Chem., Int. Ed., 2007, 46, 4434–4440 CrossRef CAS PubMed.
  4. C. V. Rode, A. A. Ghalwadkar, R. B. Mane, A. M. Hengne, S. T. Jadkar and N. S. Biradar, Org. Process Res. Dev., 2010, 14, 1385–1392 CrossRef CAS.
  5. B. Schaffner, F. Schaffner, S. P. Vervekin and A. Borner, Chem. Rev., 2010, 110, 4554–4581 CrossRef CAS PubMed.
  6. C. L. Bolivar-Diaza, V. Calvino-Casildaa, F. Rubio-Marcosb, J. F. Fernandeza and M. A. Banaresa, Appl. Catal., B, 2013, 129, 575–579 CrossRef.
  7. J. H. Teles, N. Rieber and W. harder, US Pat., 5359094, 1994.
  8. J. Li, Y. Gu, Z. Guan, W. Mo, Y. Ni, T. li and G. Li, Appl. Catal., A, 2010, 386, 188–193 CrossRef.
  9. M. Casiello, A. Monopol, P. Cotugno, A. Milella, M. Michela, F. Ciminale and A. Nacci, J. Mol. Catal. A: Chem., 2014, 381, 99–106 CrossRef CAS.
  10. J. Hu, Y. Gu, Z. Guan, J. Li, W. Mo, T. Li and G. Li, ChemSusChem., 2011, 4, 1767–1772 CrossRef CAS PubMed.
  11. F. Doro, P. Winnertz, W. Leitner, A. Prokofieva and T. E. Muller, Green Chem., 2011, 13, 292 RSC.
  12. M. Aresta, A. Dibenedetto, F. Nocito and C. Pastore, J. Mol. Catal. A: Chem., 2006, 257, 149–153 CrossRef CAS.
  13. J. George, Y. Patel, S. M. Pillai and P. Munshi, J. Mol. Catal. A: Chem., 2009, 304, 1–7 CrossRef CAS.
  14. M. G. Alvarez, M. Pliskova, A. M. Segarra, F. Medina and F. Figueras, Appl. Catal., B, 2012, 113–114, 212–220 CrossRef CAS.
  15. R. Jose, O. Gomez, O. G. J. Aberasturi, C. R. Lopeza and M. Madurgaa, Green Chem., 2012, 14, 3368 RSC.
  16. C. Hammond, J. A. LopezSanchez, M. H. Rahim, N. Dimitratos, R. L. Jenkins, A. F. Carley, Q. He, C. J. Kiely, D. W. Knight and G. J. Hutchings, Dalton Trans., 2011, 40, 3927–3937 RSC.
  17. M. Aresta, A. Dibenedetto, F. Nocito and C. Ferragina, J. Catal., 2009, 268, 106–114 CrossRef CAS.
  18. J. Park, J. Choi, S. K. Woo, S. D. Lee, M. Cheong, H. S. Kim and H. Lee, Appl. Catal., A, 2012, 433–434, 35–40 CrossRef CAS.
  19. M. H. Rahim, Q. He, J. Lopez-Sanchez, C. Hammond, N. Dimitratos, M. Sankar, A. F. Carley, C. J. Kiely, W. David and G. J. Hutchings, Catal. Sci. Technol., 2012, 2, 1914–1924 Search PubMed.
  20. F. Rubio-Marcos, V. Calvino-Casilda, M. A. Banares and J. F. Fernandez, J. Catal., 2010, 275, 288–293 CrossRef CAS.
  21. L. Wang, Y. Ma, Y. Wang, S. Liu and Y. Deng, Catal. Commun., 2011, 12, 1458–1462 CrossRef CAS.
  22. M. J. Climent, A. Corma, P. De Frutos, S. Iborra, M. Noy, A. Velty and P. Concepcion, J. Catal., 2010, 269, 140–149 CrossRef CAS.
  23. S. Fujita, Y. Yamanishi and M. Arai, J. Catal., 2013, 297, 137–141 CrossRef CAS.
  24. T. W. Turney, A. Patti, W. Gates, U. Shaheen and S. Kulasegaram, Green Chem., 2013, 15, 1925–1931 RSC.
  25. J. K. Satyarthi, L. Saikia, D. Srinivas and P. Ratnasamy, Appl. Catal., A, 2007, 330, 145 CrossRef CAS.
  26. C. Wu, L. Wang, P. T. Williams, J. Shi and J. Huang, Appl. Catal., B, 2011, 108–109, 6–13 CrossRef CAS.
  27. Q. Jiang, Z. Y. Wu, Y. M. Wang, Y. Cao, C. F. Zhou and J. H. Zhu, J. Mater. Chem., 2006, 16, 1536–1542 RSC.
  28. W.-H. Zhang, J.-L. Shi, L.-Z. Wang and D.-S. Yan, Chem. Mater., 2000, 12, 1408 CrossRef CAS.
  29. P. B. Lihitkar, et al., Mater. Chem. Phys., 2012, 133, 850–856 CrossRef CAS.
  30. T. A. Carlson, Photoelectron and Auger Spectroscopy, Plenum Publishing Corporation, New York, 1975 Search PubMed.
  31. J. Macmurry, Organic Chemistry, 7th edn, 2007, ch. 04 Search PubMed.
  32. Q. Li, W. Zhang, W. Wei and Y. Sun, Catal. Today, 2006, 115, 111–116 CrossRef CAS.
  33. R. R. Reddy, Y. N. Ahemmed, P. A. Azeem, K. R. Gopal and T. V. R. Rao, J. Non-Cryst. Solids, 2001, 286, 169–180 CrossRef CAS.

This journal is © The Royal Society of Chemistry 2015
Click here to see how this site uses Cookies. View our privacy policy here.