Ceria–zirconia solid solution loaded hierarchical MFI zeolite: An efficient catalyst for solvent free oxidation of ethyl benzene

2.1 Synthesis of hierarchical MFI

Studies have shown that a high aluminium content (low Si/Al ratio of about 15) stabilizes the surrounding silicon atoms to a high degree. Consequently, relatively small fraction of materials gets dissolved in alkali solution leading to hardly any or small pores in lower nanometer size. On the other hand, low framework aluminium concentration (high Si/Al ratio of about 1000) allows leaching of large portions for the zeolite crystal thereby inducing large pores. These large pores contribute much less to the required mesoporosity (Groen et al., 2006). That is why MFI zeolite with moderate Si/Al ratio (i.e. 100) is synthesized in the present study. For crystallization of a sample of parent MFI with silica to alumina ratio (SAR) 100, hydrothermal technique was adopted following our previously reported method under autogenous pressure (Kalita and Talukdar, 2009). The calcined product (UZ100) was subjected to alkali treatment with 0.4 M NaOH solution for a period of 60 min. 150 mL aqueous solution of 0.4 M NaOH solution was heated to 343 K in a polyethylene beaker placed in a water bath. To the pre-heated solution 2 g of parent MFI was introduced and set under constant stirring at the same temperature. The slurry was immediately cooled down in an ice-bath to prevent further reactions. The suspension was then filtered and the solid sample was washed several times with deionized water till all the alkali got removed. The product was dried overnight at ambient temperature and then at 383 K in an air oven for 6 h. The resulting desilicated sample is designated as DZ100.

2.2 Preparation of supported solid solution

The effect of initial composition of ceria and zirconia in CZ solid solution has been studied by Kenevey et al. According to their study for a Ce_0.5Zr_0.5O₂ solid solution, during calcination the tetragonal phase undergoes demixing, resulting in the creation of a ceria rich and zirconia rich compositions. Moreover, they have reported that throughout the advancement of the calcination this composition is maintained and no material with intermediate compositions is formed (Kenevey et al., 2001). That is why ceria and zirconia with Ce:Zr = 1:1 M ratio were supported on the modified MFI (DZ100) following deposition–precipitation method using ammonium hydroxide solution as hydrolysing agent. 0.051 g of ammonium ceric nitrate (Rankem) and 0.030 g of zirconyl nitrate (Loba Chemie GR grade) were individually dissolved in deionized water and finally mixed together. To the homogenous solution 1 g of finely powdered DZ100 was added and stirred for 2 h. Aqueous ammonium hydroxide solution was added dropwise with constant stirring until precipitation was complete. The resulting precipitate was filtered, washed several times to remove the excess alkali, dried at 383 K for 6 h and finally calcined at 753 (CZD480), 923 (CZD650) and 1073 K (CZD800) in a muffle furnace. Scheme 1 depicts the schematic representation of catalyst preparation.

Close

Scheme 1

Schematic representation for preparation of the catalyst.

Export to PPT

2.3 Catalytic oxidation of ethyl benzene

Catalytic oxidation of ethyl benzene (EB) was carried out with 70% tert-butyl hydroperoxide (TBHP) in a 25 mL reaction flask equipped with a water condenser. Various parameters such as the reactant molar ratio, temperature, amount of catalyst and reaction time were optimized maintaining the total volume of reactants at 5 mL for each run. The catalyst was preheated at 383 K for 3 h before every use. The progress of the reaction was monitored by withdrawing the reaction mixture after definite intervals of time and analysed by gas chromatography (Bruker 430 GC) fitted with WCOT fused silica 30 m × 0.32 mm column and FID detector. The identities of the products were confirmed by comparing with the authentic sample. The conversion and selectivity were calculated from the gas chromatography data by the following equations:

2.4 Characterization of the samples

Powdered X-ray diffraction patterns were recorded on a Bruker D-8 Advance X-ray diffractometer operated at 40 kV and 40 mA using Ni-filtered Cu Kα (0.15418 nm). The intensity data were collected over a 2θ range 5–80° with step size 0.05° and using counting time of 1 s. The surface morphology of the samples was obtained by a Hitachi, S-3600 N scanning electron microscope. Hitachi 4100 spectrometer equipped with a diffuse reflectance attachment was used to study the diffuse reflectance UV–vis spectra of the samples in the range 200–800 nm. Transmission electron microscopy (TEM) images were recorded on a JEOL 2011 electron microscope. For high resolution electron microscopy a suspension of the material in ethanol was placed in an ultrasound bath. A drop of it was placed on a carbon grid and the solvent was allowed to evaporate. The particles that remained on the walls of the grid were studied. The surface area and pore size measurement were done on a Micromeritics Tristar 3000 surface area and porosity analyser by nitrogen physisorption at liquid nitrogen temperature.

3.1 X-ray diffraction analysis

X-ray diffraction patterns for ceria–zirconia solid solution supported on MFI samples are shown in Fig. 1. In addition to the well resolved peaks and prominent reflections of MFI characteristic lines, diffraction lines due to CZ solid solution were also observed. The prominent peaks at 2θ = 7–9° and 23° correspond to characteristic peaks for MFI indicating formation of pure ZSM-5 phase (Treacy and Higgins, 2001). As reported in our earlier works alkaline treatment of MFI up to 90 min did not lead to any significant changes in the basic structure (Kalita and Talukdar, 2011). The observed XRD line patterns for CZ are found to be similar to those of CeO₂ with only slight shifting of the 2θ values (Reddy and Lakshmanan, 2005). The diffraction peaks at 2θ = 29.8°, 46°, 55.4° and 68.8° can be assigned to (1 1 1), (2 2 0), (3 1 1) and (4 0 0) planes respectively of CZ (Fig. 1) and characteristic cubic phase with fluorite structure (PDF-ICDD 28-0271). With increase in calcination temperature the intensity of lines pertaining to the cubic phases of CZ becomes more prominent. The tetragonal phase (2 2 0) for zirconia at 2θ = 50.3° was preserved in each case which is in agreement with Damayanova et al. that zirconia preserves its tetragonal phase under any composition of CZ depending on the fact that they contain a number of defects due to strong interactions with ceria sites (Damyanova et al., 2008). Crystallite size is a measure of the size of coherently diffracting domain. The crystallite size was calculated using Williamson and Hall plot (Fig. 2) for deconvoluting size and strain broadening. The simplified equation in the Lorentzian peak shape is as follows: $${\beta }_{\mathit{hkl}}\cos \theta =\frac{K}{D}+4\epsilon \mathit{sin}\theta$$ where D is average crystallite size, K is a dimensionless shape factor with a value close to unity, λ is the wavelength of Cu Kα1 radiation (0.15406 nm), β_hkl is the line broadening at half the maximum intensity (FWHM) in radian and θ_hkl is Bragg angle in radian. The additional term ε denotes the average strain. Williamson and Hall plot is obtained by plotting βcos θ against 4sin θ. The crystallite size was calculated from the intercept of the linear fit. Crystallinity of the samples was determined using the following relation: $$\%C=\frac{I_{\mathit{hkl}}}{I_b+I_{\mathit{hkl}}}$$ for a particular crystallographic plane (h k l) where, I_b is the integral background intensity and I_hkl is the integral peak intensity with reference to (1 0 1) and (5 0 1) planes. The results are presented in Table 1.

Close

Figure 1

XRD pattern of (a) CZD480, (b) CZD650, (c) CZD800. Inset shows variation of (a) crystallite size and (b) strain with the calcination temperature.

Export to PPT

Close

Figure 2

Williamson and Hall plot for (a) CZD480, (b) CZD650, (c) CZD800.

Export to PPT

The strain (ε) was obtained from the slope of the plot. It is clearly observed that with the increase in the calcination temperature from 753 to 1073 K the strain associated with the samples decreased with increase in the crystallite size (Fig. 1, inset).

3.2 TEM analysis

To ascertain the observations obtained from XRD analysis and to investigate the structural evolution at atomic scale, TEM analysis was carried out for the three samples. The TEM image of CZD480 (Fig. 3a) shows the particles to be homogenously dispersed on the zeolite surface. Further increase in the temperature to 923 then to 1073 K leads to larger particles with distinct crystallites (Fig. 3b and c). It is clearly observed from the TEM images that the increase in the calcination temperature is accompanied by a change in the morphology of the crystallites from irregular to spherical. The selected area diffraction pattern (SAED) for the different samples is shown in Fig. 3. Increase in temperature generates more ordered ring patterns accompanied with bright spots indicating enhanced crystallinity. The results are in good parity with the XRD data. The d-spacing value of about 0.31 nm for all the samples corresponds to (1 1 1) planes of CZ phase with fluorite structure and is the dominant exposed facets on the surface (Colon et al., 1999).

Close

Figure 3

TEM, SAED image and lattice fringe of (a) CZD480, (b) CZD650, (c) CZD800. The inset in each case represents the histogram of particle size distribution for the respective sample.

Export to PPT

The histogram of particle size distribution at 753 K ranges from 2 to 4.5 nm centred around 4 nm (Fig. 3, inset). Similarly, the distributions at 923 and 1073 K centred about 7 and 11 nm respectively complementing the XRD results that with increase in the calcination temperature there is a significant increase in the particle size. Increase in the calcination temperature might lead to rapid rate of nucleation and increase in the supersaturation of the reaction products might accelerate the crystal core formation reaction (Parra and Haque, 2014).

3.3 SEM analysis

SEM micrograph for CZD650 is shown in Fig. 4. It can be seen that uniform and twinned cuboidal MFI particles were formed. The particles show uniform morphology with particle size in the range 20–25 μm. The elemental composition of the same sample was obtained from energy dispersive X-ray spectroscopy (EDX) analysis which indicates that CZD650 is composed of only Ce and Zr in addition to Si, Al and O (Fig. 4). No evidence of any impurities was observed and the formed solid solution was near to our expected stoichiometry (Table 2).

Close

Figure 4

SEM and EDX analysis of CZD650.

Export to PPT

3.4 Surface area analysis

Fig. 5 illustrates the nitrogen adsorption–desorption isotherm of the parent and modified MFI samples. Prior to analysis all the samples were degassed for 4 h to remove any pre-adsorbed gaseous species. The as-synthesized UZ100 seems to exhibit a type 1 isotherm with high up-taking at relatively low pressure indicating its microporous nature (Perry and Green, 1997). Upon alkali treatment for at least 60 min a typical combination of type-I and type-IV isotherms with H3 hysteresis loop is observed (Caicedo-Realpe and Perez-Ramirez, 2010; Perez-Ramirez et al., 2008). The desilicated MFI and the CZ samples at different calcination temperatures exhibit two hysteresis curves suggesting the presence of two mesopores sizes (Kalita and Talukdar, 2011). The capillary condensation loop at p/p₀ = 0.1–0.20 represents the mesoporosity confined in the framework and the prominent H3 type hysteresis loop in the higher partial pressure range (p/p₀ = 0.42–1.0) is characteristic of considerable amount of secondary mesopores originated from intercrystalline voids in the arrangement of crystals (Chen et al., 1997). The surface area and the pore volume decreased upon loading of CZ and are found to depend on the calcination temperature (Table 3). The decrease in surface area may be due to penetration of the impregnated oxides into the composite pores of the zeolite thereby reducing the pore volume (Reddy et al., 1997). Increase in the calcination temperature leads to bigger crystallites as evidenced from XRD (Table 1) and TEM analysis and the occupancy of these bigger crystallites on the pores might have decreased the pore volume.

Close

Figure 5

Nitrogen adsorption–desorption isotherm of (a) UZ100, (b) DZ100, (c) CZD480, (d) CZD650, (e) CZD800.

Export to PPT

3.5 UV–vis diffused reflectance spectroscopy

UV–vis DR spectroscopy has been used extensively to analyse ceria-based materials and transition metal oxides to gain information of different oxidation states of metal ions and surface co-ordinations by measuring d-d, f-d transitions and oxygen-metal ion charge transfer bands. Fig. 6 depicts the UV–vis DR spectra of CZ samples calcined at different temperatures ranging from 753 to 1073 K. The observed absorption bands are in the range 200–400 nm. The DRS pattern for all the samples can be resolved into four bands 220, 250, 290 and 335 nm. The band at 220 nm may be assigned to 4f-5d transitions due to Ce³⁺ → Zr⁴⁺. This seems to weaken with the increase in the calcination temperature of the samples indicating more involvement of Zr⁴⁺ in the solid solution (Rao and Sahu, 2001). The UV–vis band at 250 and 290 nm may be assigned to Ce³⁺ ← O²⁻ and Ce⁴⁺ ← O²⁻ charge transfer transitions respectively (Bensalem et al., 1995). The broad band in the range 320–350 nm may be assigned to interband transition and also Zr⁴⁺ ← O²⁻ transition of the substituted fluorite structure. The broad band may be the characteristic of the solid solution that replaces two intense DRS features due to ceria and zirconia. This is in agreement with the explanation proposed by Bensalem (Bensalem et al., 1992) and Rao and Sahu (2001) for mechanical mixture of ceria and zirconia.

Close

Figure 6

UV–vis DR spectra of (a) CZD480, (b) CZD650 (shifted by 0.13 units), (c) CZD800 (shifted by 0.24 units). Shifting is done for clarity.

Export to PPT

4.1 Kinetics of the reaction

The kinetics of the reaction was studied for CZD650 catalyst under the optimized condition at different temperature ranges of 323–363 K. A plot of –ln (1- conversion) versus reaction time (Fig. S1) gives a linear relationship with satisfactory R² value as displayed in Table 5 indicating the reaction to follow a pseudo-first order dependence. The rate constant at different temperatures was calculated from the slope of the linear fit.

The activation parameters for the overall reaction were calculated by performing the reaction at 353 K with 0.1 g for 4 h with CZD650 catalyst. The activation energy obtained from the Arrhenius equation (Fig. S2) was found to be 21 kJ mol⁻¹. The enthalpy of activation (ΔH) and entropy of the reaction (ΔS) were calculated from the following relations:

The enthalpy and entropy of activation were calculated to be 13 kJ mol⁻¹ and −264 J mol⁻¹ K⁻¹ respectively. In addition, compared with other reported catalysts, the CZD650 catalyst behaves in a comparable way or even better in most of the cases, under relatively mild reaction conditions (Lu et al., 2010; Jothiramalingam et al., 2006).

Based upon earlier studies and our experimental data, the oxidation of EB with TBHP was proposed to follow via a free radical mechanism (Dhakshinamoorthy et al., 2009; George and Sugunan, 2008). The active sites of the solid solution could co-ordinate with distant oxygen of TBHP, leading to the generation of tert-butylperoxyl radical (t-BuOO^•). This initiated the formation of aryl radical by abstraction of H-atom from EB. Compared with the methyl proton, the methylene proton is more susceptible to oxidation due to the steric reasons of –CH₃ than that in –CH₂–. Moreover the rotation in case of –CH₃ group is more than that of –CH₂–. The activated and co-ordinated oxygen of TBHP reacted with EB to give (1-tert-butoxy ethyl) benzene. The formation of different products follows a competitive pathway. Elimination of tert-butanol/water from (1-tert-butoxy ethyl) benzene led to the formation of acetophenone whereas loss of MeOH yielded benzaldehyde (Jana et al., 2007).

The free radical pathway of the reaction was verified with 2,6-di-tert-butyl-4-methylphenol (BHT). In a particular run under the optimized conditions when the radical scavenger was added, after 4 h the conversion just reached to about 40% (Fig. 11). The GC analysis after prolonged reaction time showed no improvement in the conversion and selectivity. The situation was completely different when compared in the absence of BHT where gradual progress of the reaction was observed with time. All these observed data confirm the reaction to follow free radical mechanism.

Close

Figure 11

Evidence of free radical mechanism for the oxidation of EB over catalyst CZD650. (a) Conversion of EB with addition of BHT after 4 h and (b) selectivity of acetophenone with addition of BHT after 4 h.

Export to PPT

4.2 Leaching and reusability

In order to check leaching of CZD650, a fresh reaction was performed and suddenly stopped after 2 h. The GC analysis records a 20% conversion of EB. The catalyst was then removed from the reaction system by simple filtration using Whatman filter paper No. 42 and reaction was continued with the filtrate under similar conditions. Analysis of the mixture after 6 h showed no improvement in the conversion of EB. This assures the absence of leaching and supports the reaction to be purely heterogeneous. This also indicates the strong interaction between the active phase and hierarchical support matrix thereby proving remarkable stability of the catalyst towards oxidation of EB. Recovery and reusability are the two prominent hallmarks for any heterogeneous catalysts. To ensure the reusability of the catalyst the reaction was carried out for 4 h with CZD650 under optimized conditions. The catalyst was then separated from the reaction mixture, washed with acetone and dried. Without further activation the recovered catalyst was tested for fresh reaction mixture of EB and TBHP under the optimized conditions. It is observed that (Fig. 12) CZD650 could maintain its catalytic performance even after the fifth run.

Close

Figure 12

Reusability of the CZD650 catalyst.

Export to PPT