Nanosized spinel Cu–Mn mixed oxide catalyst prepared via solvent evaporation for liquid phase oxidation of vanillyl alcohol using air and H₂O₂

Abstract

A highly crystalline and spinel-structured Cu–Mn mixed oxide prepared via a solvent evaporation technique demonstrated superior catalytic activity for the oxidation of vanillyl alcohol to vanillin using different oxygen sources, air and H₂O₂, under optimum reaction conditions. The Cu–Mn mixed oxide catalyst contains large concentrations of structural defects (i.e. surface oxygen vacancy sites, defective oxygen species and grain boundaries) which act as active sites to enrich the transformation of vanillyl alcohol to vanillin. The atomic ratio of Cu and Mn in the metal oxides was observed to influence the catalytic performance significantly under similar reaction conditions. Moreover, reaction conditions such as time, solvent, temperature, and pressure were investigated to achieve suitable reaction parameters for the oxidation of vanillyl alcohol. For the catalytic activity, 94% conversion was measured with 99% selectivity for vanillin using H₂O₂ in the presence of base. Meanwhile, 91% conversion with 81% selectivity for vanillin was obtained by liquid phase aerobic oxidation in base-free conditions. The catalyst also showed high stability for three recycling redox reactions.

---

Introduction

Recently, there has been much interest in the conversion of renewable biomass into green fuel or fuel-based value-added products as a suitable and sustainable way to reduce dependency on petroleum feedstocks.^1–3 Lignocellulosic biomass is a material of interest due to its abundancy. Lignocellulosic biomass consists of three major constituents: cellulose, hemicellulose and lignin. Lignin is 30% carbon based and represents 40% of the energy content of the biomass.⁴ However, lignin is discarded as cellulosic waste in bio-refineries due to the complexity associated with its conversion. Hence, the valorization of lignin into fine chemicals is economically valuable in the bio-refinery process.

Vanillyl alcohol represents the B-0-4 linkage of lignin.⁵ Vanillyl alcohol has been widely studied as a phenolic model compound of lignin.^6–8 Vanillin, the selective oxidative product of vanillyl alcohol, is the major flavor constituent of vanilla. Also, vanillin has great economic value in the pharmaceutical industry, as it is used as an important platform chemical.^9–11 The large-scale consumption of vanillin requires an economically viable production protocol to satisfy demand. 85% of the world supply of vanillin is produced via the chemical route of petroleum-derived guaicol transformation, and 15% of vanillin production is based on lignin feedstock.¹² Thus, the selective oxidation of lignin is a vital alternative and environmentally feasible method to synthesize vanillin.

Alcohol oxidation is commonly performed using homogenous organometallic complexes. Although this process offers high conversion, it has some serious drawbacks, such as a considerable amount of waste, contamination of the product, separation protocols, poor selectivity and fast deactivation of the catalyst.^5,13–17 Thus, heterogeneous catalysts can be a suitable replacement because they possess many advantages compared to their homogeneous counterparts. Extensive studies have been reported of wet air aerobic or liquid phase oxidation of lignin by novel metal-based catalysts.^18–21 However, the use of novel metal-based catalysts for these applications raises the questions of economic sustainability and health issues, as food industries also benefit from the flavor properties of vanillin. Thus, from the environmental and economical points of view, designing a heterogeneous transition metal-based catalyst is highly desirable.

Mixed oxide catalysts have been explored in many oxidation reactions, as they show superior catalytic activity due to their synergistic interactions and electronic properties.²² An isolated Co₃O₄ catalyst has shown promising catalytic activity for vanillyl alcohol oxidation; however, the usage of base (i.e. NaOH) in this oxidation reaction results in a large amount of inorganic waste.²³ Various mixed oxides, such as CoTiO₃,²⁴ Co–Mn mixed oxide,²⁵ and graphene promoted Co–Mn oxide²⁶ catalysts, have been explored for the oxidation of vanillyl alcohol to vanillin. Although some catalysts have exhibited moderate conversion or selectivity, extensive research is still required to substantially enrich the conversion and selectivity under milder reaction conditions for an economically feasible process.

Cu–Mn-based spinel catalyst has also shown promising catalytic activity in many applications, such as steam reforming of methanol, low temperature oxidation of CO and gas phase oxidation of hydrocarbons.^27–30 Cu–Mn mixed oxide catalyst has been prepared via different routes, such as sol–gel,³¹ co-precipitation,^32,33 formate decomposition,³⁴ solid state reaction³⁵ and redox reaction.^36,37 The most frequently used preparation method is co-precipitation of metal precursors using alkaline NaHCO₃ or NaOH. However, the presence of trace metal (i.e. sodium, Na) in the catalyst after the washing procedure poisons the catalytic activity of the catalyst.³⁸ In this work, we synthesized a spinel-structured, highly crystalline, mesoporous Cu–Mn mixed oxide catalyst via a simplified and facile solution method without using alkaline solution. To our knowledge, the catalytic activity of the Cu–Mn mixed oxide catalyst prepared via the solution method was tested for the first time in the oxidation reaction of vanillyl alcohol using environmentally friendly oxidants, such as air and H₂O₂. Moreover, various analytical tools, such as XRD, HRTEM, SAED, XPS, FESEM, TG/DTG, EDX, H₂-TPR, and O₂-TPD, were employed to establish the intimate relationships of the properties of the catalyst, such as crystallinity, phase detection, morphology, thermal stability, surface area and redox properties, with its catalytic activity. Furthermore, the influence of reaction parameters such as catalyst loading, H₂O₂ concentration, time, and the nature of the solvent were explored for vanillyl alcohol oxidation over novel Cu–Mn mixed oxide catalyst. In addition, the stability of the Cu–Mn mixed oxide catalyst was also investigated.

Experimental

All the chemicals used in this study were purchased from various commercial sources and were used without further purification. Copper acetate (Sigma Aldrich, 98%), manganese acetate hydrate (Sigma-Aldrich, 98%), vanillyl alcohol (Sigma-Aldrich, 98%), vanillin (Sigma Aldrich, 97%), vanillic acid (Sigma Aldrich, 97%), guaicol (Sigma Aldrich, 98%), hydrogen peroxide (Sigma-Aldrich, 30% solution in water), acetic acid (Merck, 96%), isopropanol (Sigma Aldrich, 99.8%), and NaOH (Merck) were used.

Catalyst preparation

Different catalysts with various ratios of Cu and Mn loading were prepared in this investigation. In a typical synthesis, 4 mmol copper acetate were completely dissolved in 30 ml methanol inside a 100 ml Schlenk flask connected with a vacuum line adaptor. An equimolar ratio of manganese acetate was added to the solution. The mixture was placed under vacuum for 4 h until the solution became deep green. The solvent was extracted under low pressure until a dry powder was obtained. The powder was ground finely prior to calcination. The calcination was performed in a blast furnace under air at 500 °C using a heating rate of 2 °C min⁻¹. The catalyst was labelled as A. Catalyst B (Cu : 2Mn) was prepared by changing the molar ratio of manganese acetate to 2, whereas the copper acetate ratio was 2 in the case of catalyst C (2Cu : Mn).

Catalyst characterization

X-ray diffractograms of the prepared catalysts were acquired on a Bruker D8 Advance X-ray diffractometer using nickel filtered Cu Kα radiation (d = 1.54 Å) in the Bragg angle range from 10° to 90° with a step size of 0.026 min⁻¹. The current density and the voltage were kept constant at 40 mA and 40 kV, respectively, at ambient temperature during the analysis. The thermal analysis was performed in a ceramic crucible on a Perkin Elmer TGA 4000 Thermogravimetric Analyzer connected with a computer interface. Pure air with a flow rate of 30 cm³ min⁻¹ was passed continuously at a heating rate of 10 °C min⁻¹ under atmospheric pressure. The surface morphology analysis was performed using field emission scanning microscopy (FESEM Quanta FEI 200F), and the elemental composition analysis was performed using energy dispersive X-ray spectroscopy (INCA Software) attached to the FESEM.

Temperature-programmed reduction and oxidation were carried out on a Micrometrics Chemisorb 1100 Series instrument. 0.2 g sample was treated at 180 °C for 1 hour with N₂ gas (20 ml min⁻¹) in a quartz tube for removal of the surface moisture prior to analysis. Thereafter, the temperature was elevated to 900 °C under gas flow of a mixture of 5% H₂ with 95% N₂ at a heating rate of 10 °C min⁻¹ and was maintained for 10 min. The reduced sample was also analyzed to determine the re-oxidation process of the catalyst with the same temperature programme.

Catalytic activity measurement

The performance test of the catalyst was carried out in a three neck round bottom flask attached to a reflux condenser which was placed on a magnetic hotplate. A thermometer was dipped in the reaction mixture through one neck, and another neck was closed with a stopper for sampling at every 0.5 h time interval. In a typical experiment, the reaction mixture contained 1 mmol vanillyl alcohol (0.15 g) with 3 mmol H₂O₂ (0.3 ml) and 0.0037 g cm⁻³ catalyst in 20 ml isopropanol. The temperature of the reaction was maintained at 85 °C. The liquid phase aerobic oxidation was performed in a high pressure 200 ml autoclave reactor supplied by Parr Co. In a typical liquid phase aerobic oxidation experiment, 3 mmol vanillyl alcohol was added to 60 ml acetonitrile. The reaction temperature and pressure were maintained at 120 °C and 21 bar, respectively. The products from the oxidation reactions were collected and analyzed using an Agilent Technology HPLC Chromatography 1100 series connected with a UV detector and computer interface. A mixture of solvent (85% water + 15% acetonitrile) with 1% acetic acid was used as the mobile phase, and an RP-C18 column by Zobrax was used. The products and substrate were successfully detected by a UV detector at λ_max = 270 nm. The column temperature was stable at 28 °C with a flow rate of 1 ml min⁻¹. The conversion and selectivity were measured as follows:

Results and discussion

Thermogravimetric analysis

Thermogravimetric analysis was carried out for the uncalcined catalyst A (Cu:Mn) in order to determine the stability of the catalyst in the temperature range of 60 °C to 700 °C (Fig. 1.) Four distinct weight losses of 3.18%, 3.49%, 29.83% and 24.44% were detected at temperature maxima of 110 °C, 190 °C, 245 °C and 310 °C, respectively. The first and second weight losses of 3.18% and 3.49% correspond to the withdrawal of adsorbed moisture from the surface and the bulk catalyst. Furthermore, the elimination of crystallite water present on the precursors [Cu(CH₃COO)₂·xH₂O and Mn(CH₃COO)₃·2H₂O] could be ascribed to the third weight loss of 29.83% at 245 °C. Moreover, the final weight loss of 24.44% at 310 °C was attributed to the elimination of organic groups from the metal salt precursors. The catalyst was observed to reach thermal stability at 450 °C.

Fig. 1 Thermogravimetric analysis of catalyst A (Cu:Mn).

XRD analysis

The XRD data of all the synthesized catalysts were evaluated using Highscoreplus Panlytical software in terms of peak position at the Bragg angle and the lattice spacing (d) (Fig. 2). The XRD pattern displayed in Fig. 2 shows peaks at Bragg angle 2theta = 18.70, 30.62, 36.06, 37.17, 43.79, 49.15, 54.33, 55.45, 57.89, 63.56, 66.13, 68.08, 75.26 with d spacing (Å) 4.74, 2.91, 2.49, 2.38, 2.06, 1.85, 1.68, 1.65, 1.59, 1.46, 1.441, 1.37, 1.26 which corresponds to Miller indices (111), (220), (311), (222), (400), (420), (422), (430), (511), (440), (530), (442), (533) of cubic spinel Cu_1.5Mn_1.5O₄ phase, respectively, according to JCPDS no. ICSD-01-070-0260. Also, the peaks positioned at Bragg angle 2theta = 33.21 and 38.87 were detected as the minor phase (222) and (111) planes of Mn₂O₃ and CuO, respectively, in accordance with JCPDS no. ICSD-00-041-1442 and ICSD-00-041-0254. No other impurities or phases were observed in the recorded XRD patterns. Cu_1.5Mn_1.5O₄ phase was referred to as the nonstoichiometric form of the cubic spinel Cu_xMn_3−xO₄ structure.

Fig. 2 XRD analysis of the synthesized catalysts.

The Cu_1.5Mn_1.5O₄ phase after calcination at 500 °C observed in the XRD pattern is in harmony with previous literature reports of the phase diagram of Cu–Mn–O.³⁹ the cubic spinel shaped Cu_1.5Mn_1.5O₄ phase detected in this synthesis protocol was reported to be stable at low temperature in the phase diagram. It is worthwhile to mention that this simple and facile synthesis protocol and calcination at 500 °C was very effective and successful for the synthesis of spinel-structured, highly crystalline and stable mixed oxide catalyst. Moreover, the spent catalyst was also analyzed by XRD to detect any changes in the phase and crystallinity. As expected, there was no change in the XRD pattern; it showed a similar pattern at the same Bragg angle as the fresh catalyst. This suggested that the crystallinity and the Cu_1.5Mn_1.5O₄ phase of the catalyst are well preserved. Also, similar reaction conditions was observed to not affect the crystal phase in previous literature reports of CoTiO₃ catalyst in vanillyl alcohol oxidation.²⁴ However, the intensity of the Mn₃O₄ phase was slightly reduced after the catalytic reaction, probably due to amorphization and re-distribution of the corresponding species.⁴⁰

Oxidation state of Cu–Mn mixed oxide catalyst

XPS analysis was conducted to obtain surface chemistry information along with the chemical bonding state of the elements for catalyst A (Cu:Mn) (Fig. 3). The wide XPS spectra revealed that there was no trace of other metals as impurities apart from the presence of Cu, Mn and O (Fig. 3a). The surface elemental composition is shown in Table 1. Cu spin–orbit splitting indicated the presence of two Cu species in the oxide cage (Fig. 3b). The major peaks at the binding energies for 2p_3/2 at 933.1 eV and 2p_1/2 at 953.1 eV correspond to Cu²⁺ species.⁴¹ Also, the peak at a relatively lower binding energy at 931.07 eV along with a weak satellite peak at 945 eV was attributed to Cu⁺ species on the catalyst framework.^42,43 It should be mentioned that Cu²⁺ was predominantly present in the catalyst according to the peak intensities and FWMH of Cu²⁺ and Cu⁺. The presence of Cu⁺ originated from the redox resonance of Cu²⁺ + Mn³⁺ → Cu⁺ + Mn⁴⁺, which is expected to occur on the catalyst surface. The identification of the Mn oxidation state based on the Mn 2p XPS spectra was disputable due to the high fraction of Mn₂O₃. Different valences of Mn showed peaks in the same region of 2p. Therefore, the comprehensive approach to identify the oxidation state of Mn in the prepared catalyst was to analyze the Mn 3s spectra. The spin–orbit splitting of Mn 3s showed a 5.5 eV binding energy difference between the main peak and the satellite peaks in the XPS spectra (Fig. 3d). This confirmed the presence of Mn³⁺ species and disproved the existence of Mn²⁺ and Mn⁴⁺ on the surface of the catalyst, which was in good agreement with results previously reported in the literature.⁴⁴ These observations suggested that the synthesized Cu_1.5Mn_1.5O₄ catalyst contains only Mn³⁺ species, with the co-existence of Cu⁺ as well as Cu²⁺ species. Previously, it was reported in the literature that the ideal structure of cubic spinel (Cu₂)[Mn₃Cu]O₈ contains two types of Cu species, which is supported in the current work.⁴⁵

Fig. 3 Oxidation state analysis of catalyst A (Cu:Mn): [a] wide spectra, [b] Cu 2p spectra, [c] O 1s spectra, [d] Mn 3s spectra.

Table 1 The surface composition of the catalyst from XPS analysis

Cu (%)	Mn (%)	O (%)
5.82	20.57	73.65

---

However, some shake-up peaks due to charge transfer were also observed at the binding energies of 942 eV and 962 eV. This further confirmed the presence of Cu²⁺ species in the Cu–Mn crystal cage.⁴⁶ In addition, the O 1s XPS spectrum is displayed in Fig. 3c. The asymmetric peak of the O 1s spectrum was further resolved into three components by Gaussian fitting. The peak at the lowest binding energy of 529 eV corresponds to the lattice oxygen species (O_L) (O²⁻) of Cu–O and Mn–O. The peak at the binding energy of 531.2 eV could be attributed to surface oxygen vacancies as per previously reported literature.^47,48 Surface oxygen vacancies (O_V) as surface defects can be caused by the non-stoichiometry of the metal ions in the nanocrystalline domain. Also, the peaks at higher binding energies in the range of 531.8 to 532.9 eV have been reported as chemisorbed (O_C) or dissociative species (O₂⁻, O₂⁻ or O⁻) and OH⁻.⁴⁹ The atomic percentages of the three types of oxygen species were calculated from the XPS data and are shown in Table 2.

Table 2 The percentage of oxygen species from the O 1s spectra

Lattice oxygen (OL)	Surface oxygen vacancy (OV)	Defective oxygen (OD)
46.23%	38.36%	15.40%

---

Morphologies of Cu–Mn catalysts

FESEM images of the synthesized Cu–Mn catalysts after calcination are shown in Fig. 4. The images clearly indicate that the nanoparticles grew in an agglomerated state in a 3D frame network to form interconnected clusters. Also, it can be seen that the catalysts possess uneven and jagged pores. The randomly distributed pores originated from the liberation of organic residues during the heat treatment process. Furthermore, the rough and stony surface of the catalyst is believed to facilitate the adsorption of the substrate to the catalyst and hence enrich the catalytic activity in oxidation reactions. The higher resolution pictures of the well-dispersed catalyst show particle shapes with defined edges. The spent catalyst was also observed by microscopic analysis to study the expected changes in the morphology and the surface of the catalyst (Fig. 4d). The results indicated that the particles were segregated after the reaction. A careful interpretation of the image of the used catalyst shows that the surface roughness was reduced to some degree, as expected.

Fig. 4 Morphologies of the catalysts. [a] Catalyst A (Cu:Mn), [b] catalyst B (Cu:2Mn), [c] catalyst C (2Cu:Mn), [d] spent catalyst A.

HRTEM images were captured to further confirm the particle size and the formation of Cu_1.5Mn_1.5O₄ phase in terms of the d spacing of the phase (Fig. 5). The d spacing was measured as 4.74 Å, which confirmed the (111) plane of Cu_1.5Mn_1.5O₄ (Fig. 5b). Also, the crystallite size was measured in the range of 8 to 13 nm. A high resolution insight into the interface of the plane revealed the presence of surface defects in the form of numerous grain boundaries associated by partial dislocations. As shown in Fig. 5c, partial dislocation occurred at the interface of two (111) planes of Cu_1.5Mn_1.5O₄; this was labelled as T. The observed grain boundaries notably contribute to the high catalytic activity of the Cu_1.5Mn_1.5O₄ catalyst. Furthermore, a selective area electron diffractogram (SAED) was also obtained to confirm the crystallinity and the formation of Cu_1.5Mn_1.5O₄ phase in the synthesized catalyst (Fig. 5d). The diffusive ring observed in the SAED pattern confirmed that the material was polycrystalline. The bright spots noted in the SAED pattern confirmed each plane of Cu_1.5Mn_1.5O₄. Fig. 5d shows the formation of the (111) and (220) planes of Cu_1.5Mn_1.5O₄ phase in terms of the measured d spacings of 4.74 Å and 2.91 Å.

Fig. 5 HRTEM images of catalyst A (Cu:Mn). [a] Overview image, [b] lattice parameter, [c] grain boundary, [d] SAED pattern.

Moreover, the elemental composition of the prepared material was determined by EDX analysis (Fig. 6). The atomic ratios observed in the EDX spectra of the three catalysts tentatively matched the primary ratio of metal loading in the synthesis protocol. Also, the area mappings clearly show the high degree of homogeneity of both metals in the catalyst (Fig. 6d).

Fig. 6 EDX spectra of the synthesized catalysts. [a] Catalyst A, [b] catalyst B, [c] catalyst C, [d] homogeneity of the catalyst, [e] Cu atoms, [f] Mn atoms, [g] O atoms.

Redox properties

The quantitative elucidation of the reduction profile of the Cu–Mn catalyst could not be determined precisely, owing to the existence of multiple valences for both Cu and Mn (Fig. 7). However, the correlation of reduction temperature with the catalytic activity was established, as the oxidation reaction involves redox coupling.^50,51 The profile of the synthesized catalysts shows that H₂ consumption starts at 220 °C, with two distinguishable peak maxima at 340 °C and 430 °C for all three catalysts. Isolated oxide phases showed reduction at higher temperatures, as per the literature.^27,30 CuO was reduced at around 300 °C, whereas Mn³⁺ increased slightly at the range of 350 °C to 400 °C. However, the reduction profiles of the catalysts synthesized by this solution method appeared at low reduction temperatures. This indicated that the redox ability of the catalyst prepared by the solution route was strongly improved by the synergy of Cu and Mn. It was postulated that the initial H₂ consumption occurred at the surface oxygen sites. Continuous reduction of the catalyst was observed up to a temperature of 480 °C. The large reduction was mainly due to few possible steps of reduction. The lattice oxygen (Cu–O and Mn–O) was reduced at the high reduction temperature relative to the surface oxygen sites. The XPS spectra confirmed that Mn was present in the catalyst in the Mn³⁺ oxidation state. Therefore, the reduction of Mn may have occurred in the following steps: Mn₂O₃–Mn₃O₄–MnO. Complete reduction of MnO to metallic Mn was not expected even up to 1223 K, due to the more negative reduction potential.^52,53 Moreover, it has been reported that previously reduced Cu⁰ can act as an H₂ activation site to decrease the reduction temperature of Mn species.³⁰ This could contribute to the reduction of spillover species, as suggested by Ferrandon et al.⁵⁴

Fig. 7 H₂-TPR study of the synthesized catalysts.

Moreover, a comprehensive study of the TPR profiles of the synthesized catalysts revealed that the catalyst containing an equimolar ratio of Cu and Mn had the first peak maxima at the lowest reduction temperature. The catalyst A (Cu:Mn) reduction started at 275 °C, while those of the other two catalysts started at 350 °C and 360 °C, respectively. One explanation of the low reduction temperature of catalyst A is the presence of a large concentration of structural defects associated with defective oxygen species; this was also suggested by the O 1s XPS spectra as well as the small particle size. Another possible reason was the formation of a high degree of Cu–O–Mn species, which causes an electronic interaction between Cu and Mn.

In order to study the oxygen vacancies in the catalyst, temperature programme oxygen desorption was used (Fig. 8). In the TPD profile, four distinctive oxygen desorption peaks were noted at temperature maxima around 230 °C, 340 °C, 540 °C and 680 °C. The first oxygen desorption peak at low temperature was attributed to surface oxygen vacancies. Desorption of oxygen at the lattice oxygen vacant sites of Cu occurred at a temperature of 340 °C, which was indicated by the prominent peak of catalyst C with a high loading of Cu. A similar trend was observed in the case of catalyst B with high content of Mn at a temperature of 540 °C. Therefore, the third oxygen desorption peak corresponds to the oxidation of Mn²⁺ to Mn³⁺. A previous literature study reported the presence of defective oxygen species in the lattice at high temperatures of 796 K to 863 K for supported MnO_x species.⁵⁵ A similar observation was made by another group of researchers for Al₂O₃ and SiO₂-promoted MnO_x species.⁵⁶ Therefore, in our work, the oxygen desorption peak at high temperature around 680 °C originated from structural defects associated with defective oxygen species. The most intense peak at high temperature for catalyst A indicated the presence of a high degree of defective oxygen species among all three catalysts. Thus, the reduction temperature of catalyst A (Cu:Mn) was noticeably lower, as observed in the TPR profile.

Fig. 8 O₂-TPD study of the prepared catalysts.

Catalytic activity

As described above, the findings from XRD, HRTEM and SAED confirmed the formation of Cu_1.5Mn_1.5O₄ phase in all synthesized catalysts. Also, XPS revealed the co-existence of two oxidation states of Cu²⁺ and Cu⁺ with Mn³⁺ species. In addition, the presence of surface oxygen vacancies was confirmed by the O 1s spectra and O₂-TPD studies. It was postulated that high catalytic activity could be achieved by the enrichment of oxygen surface defects, which are surface-reactive oxygen species that facilitate the oxidation process. Moreover, grain boundaries, a form of surface defect, as additional active sites were marked in the interface of the plane. The reducing temperature of the prepared Cu_1.5Mn_1.5O₄ catalyst phase was significantly lower than that of the corresponding isolated phases because of the synergistic combination of the metals and also due to large concentrations of structural defective oxygen species. These structural and chemical properties of the synthesized Cu_1.5Mn_1.5O₄ catalyst enhanced the robustness of the catalyst for the liquid phase oxidation of vanillyl alcohol.

To evaluate the catalytic activity, catalyst A (Cu:Mn) was considered to optimize the reaction variables. No conversion was observed in acetonitrile in the presence or absence of catalyst. However, in the presence of equimolar NaOH base relative to vanillyl alcohol in isopropanol, the oxidation reaction occurred. Interestingly, surprising catalytic activity was achieved. To determine the role of the base, the reaction without catalyst in the presence of NaOH and H₂O₂ in isopropanol solvent was also considered. No conversion of vanillyl alcohol was found. Therefore, it should be mentioned that the presence of base in the liquid phase oxidation of vanillyl alcohol by Cu_1.5Mn_1.5O₄ is crucial, as it probably deprotonates the phenoxy ion and favors coordination with the catalyst. It is a well-known fact that H₂O₂ coupled with NaOH favors lignin valorization over metal oxides.^23,57,58

To monitor the progress of the oxidation reaction, HPLC analysis was performed after every 0.5 h interval of reaction time (reaction conditions: 1 mmol vanillyl alcohol, 1 mmol NaOH (0.15 g), 3 mmol H₂O₂, 20 ml isopropanol, 0.0037 g cm⁻³ catalyst) (Fig. 9). A steeply progressive trend for the amount of conversion was observed. After two and a half hours of reaction time, the conversion reached 94% with 99% selectivity for vanillin. With increasing reaction time, the conversion remained almost constant (99%). Based on these results, 2.5 h was chosen as the optimum time for vanillyl alcohol oxidation by Cu_1.5Mn_1.5O₄ catalyst using H₂O₂ oxidant.

Fig. 9 Reaction progress with time. Reaction conditions: 1 mmol vanillyl alcohol, 3 mmol H₂O₂, 1 mmol NaOH, 85 °C, 20 ml isopropanol, catalyst A (Cu:Mn) mass 0.0037 g cm⁻³.

To determine the optimum loading of catalyst in the reaction, four different loadings, 0.0012 g cm⁻³, 0.0025 g cm⁻³, 0.0037 g cm⁻³ and 0.0050 g cm⁻³, were considered (reaction conditions: 1 mmol vanillyl alcohol, 1 mmol NaOH (0.15 g), 3 mmol H₂O₂, 20 ml isopropanol, 85 °C, 2.5 h) (Fig. 10). A sharp increase to 94% from 56% was observed in the amount of conversion when the catalyst loading was changed from 0.0012 g cm⁻³ to 0.0037 g cm⁻³, owing to a significant mass transfer effect. However, in the case of 0.0050 g cm⁻³ catalyst loading, the conversion was further enriched by 2%. This was possibly due to insignificant mass transfer.⁵⁹ Surprisingly, the selectivity towards vanillin remained constant at 99% for all loadings of the catalyst. Thus, 0.0037 g cm⁻³ was the best catalyst mass for liquid phase selective oxidation of vanillyl alcohol to vanillin using H₂O₂ oxidant.

Fig. 10 Effects of catalyst A (Cu:Mn) mass on conversion and selectivity. Reaction conditions: 1 mmol vanillyl alcohol, 3 mmol H₂O₂, 1 mmol NaOH, 85 °C, 20 ml isopropanol, 2.5 h.

Four different molar ratios of H₂O₂ concentration to vanillyl alcohol, 1 mmol, 2 mmol, 3 mmol and 4 mmol, were injected in order to scrutinize the effect of oxidant concentration in the reaction (reaction conditions: 1 mmol vanillyl alcohol, 1 mmol NaOH (0.15 g), 20 ml isopropanol, 2 h, 85 °C, 0.0037 g cm⁻³ catalyst) (Fig. 11). The conversion markedly improved when the concentration of H₂O₂ was increased. The conversion reached a maximum at 94% in the case of 3 mmol. An increased amount of H₂O₂ accelerates the rate of reaction. However, 4 mmol showed a slight decline in the conversion (92%). This was possibly because the degradation of H₂O₂ became predominant over the oxidation of vanillyl alcohol.¹⁷ Notably, the selectivity was not affected for the abovementioned ratios of H₂O₂. Based on the experimental data, 3 mmol was chosen as the best condition for liquid phase oxidation of vanillyl alcohol by H₂O₂.

Fig. 11 Influence of H₂O₂ concentration on conversion and selectivity. Reaction conditions: 1 mmol vanillyl alcohol, 1 mmol NaOH, 85 °C, 20 ml isopropanol, catalyst A (Cu:Mn) mass 0.0037 g cm⁻³, 2.5 h.

The oxidation reaction was carried out at four different temperatures, 65 °C, 75 °C, 85 °C and 90 °C, under optimized reaction conditions to investigate the probable response (reaction conditions: 1 mmol vanillyl alcohol, 1 mmol NaOH (0.15 g), 3 mmol H₂O₂, 20 ml isopropanol, 2 h, catalyst 0.0037 g cm⁻³) (Fig. 12). When the temperature was increased to 85 °C from 65 °C, the catalytic conversion improved to 94% from 56%, as expected. However, at a temperature of 90 °C, the conversion was slightly reduced to 92%. In a previous literature study, it was reported that H₂O₂ degradation was higher above a temperature of 353 K, thus resulting in lower conversion.⁶⁰ A similar observation was made in our study. Thus, 85 °C was considered to be the best temperature for the maximum amount of conversion in the liquid phase oxidation of vanillyl alcohol using H₂O₂.

Fig. 12 Effects of temperature on conversion and selectivity. Reaction conditions: 1 mmol vanillyl alcohol, 3 mmol H₂O₂, 1 mmol NaOH, 20 ml isopropanol, catalyst A (Cu:Mn) mass 0.0037 g cm⁻³, 2.5 h.

Two solvents, acetic acid and isopropanol, were considered to study the influence of the nature of the solvent, as NaOH is highly soluble in both media (reaction conditions: 1 mmol vanillyl alcohol, 1 mmol NaOH (0.15 g), 3 mmol H₂O₂, 2 h, 85 °C, catalyst 0.0037 g cm⁻³) (Fig. 13). For isopropanol, the highest catalytic activity of 94% conversion was obtained, with 99% selectivity for vanillin. In the case of acetic acid, there was a marginal decrement in the catalytic conversion of 3%. However, the selectivity was remarkably increased to 61%. This was probably due to both the polarity and the pH of the reaction. At lower pH, the formation of peracetic acid was expected, which could further generate peroxy (⁻OOH) radicals.⁶¹ This results in further oxidation of vanillin to vanillic acid in acetic acid medium. Therefore, isopropanol is the best solvent to obtain maximum conversion with excellent selectivity for vanillin.

Fig. 13 Influence of the nature of the solvent on conversion and selectivity. Reaction conditions: 1 mmol vanillyl alcohol, 3 mmol H₂O₂, 1 mmol NaOH, 85 °C, catalyst A (Cu:Mn) mass 0.0037 g cm⁻³, 2.5 h.

The catalyst composition is also a vital parameter, and its effects on the catalytic conversion were surveyed (reaction conditions: 1 mmol vanillyl alcohol, 1 mmol NaOH (0.15 g), 3 mmol H₂O₂, 2 h, 85 °C, 20 ml isopropanol, 0.0037 g cm⁻³ catalyst) (Fig. 14). The catalyst with equivalent loadings of Cu and Mn was found to be the most effective in the oxidation of vanillyl alcohol using H₂O₂. Catalytic conversion of 94% with 99% selectivity in the presence of NaOH base was obtained for catalyst A under the optimized reaction conditions. Inferior catalytic conversions were noted for catalysts B and C (67% and 36%, respectively). Notably, the selectivity was not affected by the ratio of the loaded metals. This suggested that the reaction pathways were similar for all the catalysts, as Cu_1.5Mn_1.5O₄ phase was formed in all the catalysts. The above experimental data imply that the catalytic activity was highly influenced by the chemical composition of the metal loading in the synthesis protocol. The metal loading in the bulk phase of the catalyst could significantly assist in the alteration of the concentration of surface reactive oxygen species. Thus, it could be concluded that the formation of the Cu–O–Mn linkage in catalyst A was higher, as was expected due to the key active sites in the catalyst. The superiority of catalyst A was further confirmed by the reducing properties of the catalyst in the H₂-TPR profile. Hence, the control of the surface composition in Cu–Mn mixed oxide is vital to achieve improved catalytic activity in the oxidation of vanillyl alcohol. In addition, it is a well-recognized fact that the origin of the excellent catalytic activity of Cu–Mn mixed oxide catalysts is due to the presence of two Jahn–Teller ions, Cu²⁺ and Mn³⁺, and their solid state charge transfer redox cycle.^62,63

Fig. 14 Effects of catalyst composition on conversion and selectivity. Reaction conditions: 1 mmol vanillyl alcohol, 3 mmol H₂O₂, 1 mmol NaOH, 85 °C, 20 ml isopropanol, catalyst A (Cu:Mn) mass 0.0037 g cm⁻³, 2.5 h.

The best catalyst in terms of conversion, A (Cu:Mn), was further considered for aerobic oxidation of vanillyl alcohol to avoid the use of NaOH base. Acetonitrile was chosen as the solvent, as it was found to be a suitable solvent as per a previous report.²⁵ Furthermore, the reaction variables were optimized using air. A blank reaction in the absence of catalyst was performed for the aerobic oxidation of vanillyl alcohol. No conversion was noted. To inspect the progress, the aerobic oxidation reaction in the presence of catalyst was carried out (reaction conditions: 3 mmol vanillyl alcohol, T = 120 °C, P = 21 bar air, 60 ml acetonitrile, 0.0025 g cm⁻³ catalyst) (Fig. 15). The substrate conversion was sharply enriched with increasing reaction time. It reached a maximum at 97% conversion with selectivity for vanillin after 2 h of reaction time. After that, the reaction did not proceed with appreciable improvement of the conversion. Also, the selectivity decreased owing to overoxidation of vanillin to the corresponding acid. Hence, 2 h was the best reaction time to achieve optimum catalytic activity for liquid phase aerobic oxidation in base-free conditions.

Fig. 15 Influence of reaction time concentration on conversion and selectivity. Reaction conditions: 3 mmol vanillyl alcohol, 120 °C, 21 bar air pressure, 60 ml acetonitrile, catalyst A (Cu:Mn) mass 0.0025 g cm⁻³, 2.5 h.

Four different loadings of catalyst, 0.0006 g cm⁻³, 0.0012 g cm⁻³, 0.0025 g cm⁻³ and 0.0037 g cm⁻³, were used to study the apparent effects of the catalyst on the aerobic oxidation of vanillyl alcohol (reaction conditions: 3 mmol vanillyl alcohol, T = 120 °C, P = 21 bar air, 60 ml acetonitrile) (Fig. 16). The best catalytic activity was obtained for the catalyst loading of 0.0025 g cm⁻³ due to significant mass transfer. The worst catalytic activity (48%) was found for the loading of 0.0006 g cm⁻³. In the case of 0.0037 g cm⁻³ loading, there was no significant increase in the conversion, possibly owing to insignificant mass transfer.⁵⁹ It should be mentioned that the selectivity was not changed by changing the loading of the catalyst.

Fig. 16 Influence of catalyst mass A (Cu:Mn) on conversion and selectivity. Reaction conditions: 3 mmol vanillyl alcohol, 120 °C, 21 bar air pressure, 60 ml acetonitrile, 2 h.

The temperature in the range of 85 °C to 140 °C was found to have a vital role in the catalytic activity for the liquid phase aerobic oxidation of vanillyl alcohol (reaction conditions: 3 mmol vanillyl alcohol, P = 21 bar air, 60 ml acetonitrile, 0.0025 g cm⁻³ catalyst) (Fig. 17). At 85 °C, the lowest catalytic activity was measured, with 64% conversion. As the temperature increased to 120 °C, excellent catalytic activity, 91% conversion with 81% selectivity for vanillin, was obtained. A further increase in the temperature to 140 °C showed a slight increment in the conversion; however, the selectivity was reduced. This was possibly due to the overoxidation of vanillin to vanillic acid at higher temperature. However, char formation was also observed at 140 °C. Thus, it should be noted that 120 °C is a suitable temperature for the liquid phase aerobic oxidation of vanillyl alcohol.

Fig. 17 Effects of temperature on conversion and selectivity. Reaction conditions: 3 mmol vanillyl alcohol, 21 bar air pressure, 60 ml acetonitrile, catalyst A (Cu:Mn) mass 0.0025 g cm⁻³, 2 h.

To investigate the influence of air pressure, the reaction was carried out at four different pressures: 7 bar, 14 bar, 21 bar and 28 bar (reaction conditions: 3 mmol vanillyl alcohol, 120 °C, 60 ml acetonitrile, 0.0025 g cm⁻³ catalyst) (Fig. 18). The catalytic conversion was found to increase with increasing pressure in the order of 7 bar < 14 bar < 21 bar < 28 bar, whereas the selectivity followed a descending order. The selectivity decreased due to the presence of more desorbed oxygen in the reaction, which caused overoxidation of the product vanillin. The optimum catalytic conversion of 91% conversion with 81% selectivity for vanillin was attained for 21 bar air pressure, while the worst conversion (61%) was measured at 7 bar air pressure. Based on the observed results, one can conclude that 21 bar air pressure is the optimum condition for the aerobic oxidation of vanillyl alcohol.

Fig. 18 Effects of air pressure on conversion and selectivity. Reaction conditions: 3 mmol vanillyl alcohol, 120 °C, 60 ml acetonitrile, catalyst A (Cu:Mn) mass 0.0025 g cm⁻³, 2 h.

Furthermore, to detect any leaching of the metals in the reaction, leaching tests were also performed. After 30 min reaction time, the conversion was measured at 28%. After that, the catalyst was filtered and the reaction was carried out in the absence of the catalyst. No further conversion was noticed. This confirms that no metal leached from the catalyst during the catalytic conversion.

Recyclability

The catalyst was recovered after the reaction and washed thoroughly with solvent to remove any desorbed products. Thereafter, it was dried for four hours at 60 °C. An oxidation reaction using spent catalyst was carried out under the optimized conditions for fresh catalyst (reaction conditions: 3 mmol vanillyl alcohol, T = 120 °C, P = 21 bar air, 60 ml acetonitrile, 0.0025 g cm⁻³ catalyst mass, 2 h) (Fig. 19). Similarly, three cycles of the reaction were performed, and a negligible change was observed in the catalytic activity after the third run. The slightly lower catalytic activity was possibly due to some degree of decrement in the roughness of the surface. This was possibly due to the metastability of the rough surface. A similar observation of a reduction in the roughness of the catalyst surface after the catalytic reaction in vanillyl alcohol oxidation was made in a previous report.²⁴ Also, there was a reasonable expectation of blockage of the few active sites by the residual un-desorbed products after washing. Notably, the selectivity remained almost constant. This suggested that the reaction pathway was not changed, as the catalyst phase remained unchanged after the oxidation reaction.

Fig. 19 Recyclability study on conversion and selectivity. Reaction conditions: 3 mmol vanillyl alcohol, 120 °C, 21 bar air pressure, 60 ml acetonitrile, catalyst A (Cu:Mn) mass 0.0025 g cm⁻³, 2 h.

Conclusion

A very effective and simple solution method was used to successfully synthesize a spinel-structured and highly crystalline Cu–Mn mixed oxide in the absence of base. The superior physico-chemical properties of the prepared mixed oxide were confirmed by XRD, HRTEM and SAED data. The catalytic activity was evaluated as 94% conversion with 99% selectivity using H₂O₂ (reaction conditions: 1 mmol vanillyl alcohol, 3 mmol H₂O₂, 1 mmol NaOH, 85 °C, 20 ml isopropanol, 0.0037 g cm⁻³ catalyst mass, 2.5 h), whereas 91% conversion with 81% selectivity to vanillin was observed by aerobic oxidation under optimized reaction conditions (3 mmol vanillyl alcohol, T = 120 °C, P = 21 bar air, 60 ml acetonitrile, 0.0025 g cm⁻³ catalyst mass, 2 h). This indicated that the catalyst was outstanding in the liquid phase aerobic oxidation of vanillyl alcohol in base-free conditions. This work summarizes the importance of the chemical composition of a catalyst to enrich its reducing ability with respect to the transformation of vanillyl alcohol to vanillin, complemented by H₂-TPR. The high catalytic activity was also significantly influenced by the surface properties of the catalyst, such as a large concentration of surface oxygen vacancies as well as defects in the form of grain boundaries, justified by O₂-TPD, XPS and HRTEM analysis. Moreover, the catalyst showed high stability for three subsequent oxidation reactions.

Acknowledgements

The authors are grateful for the financial support from University Malaya Research Grant (GC001A-14AET), which fully supported this work.

References

1. M. L. Besson, P. Gallezot and C. Pinel, Chem. Rev., 2013, 114, 1827–1870.
2. B. Peng, Y. Yao, C. Zhao and J. A. Lercher, Angew. Chem., 2012, 124, 2114–2117.
3. E. J. Steen, Y. Kang, G. Bokinsky, Z. Hu, A. Schirmer, A. McClure, S. B. Del Cardayre and J. D. Keasling, Nature, 2010, 463, 559–562.
4. R. D. Perlack, L. L. Wright, A. F. Turhollow, R. L. Graham, B. J. Stokes and D. C. Erbach, Biomass as feedstock for a bioenergy and bioproducts industry: the technical feasibility of a billion-ton annual supply, DTIC Document, 2005.
5. R. Behling, S. Valange and G. Chatel, Green Chem., 2016, 18, 1839–1854.
6. R. Kondo, H. Yamagami and K. Sakai, Appl. Environ. Microbiol., 1993, 59, 438–441.
7. J. Hemmingson and G. Leary, Aust. J. Chem., 1980, 33, 917–925.
8. J. Pan, J. Fu and X. Lu, Energy Fuels, 2015, 29, 4503–4509.
9. G. Ravishankar, B. Suresh, P. Giridhar, S. R. Rao and T. S. Johnson, Capsicum: The genus Capsicum, 2003, pp. 96–128.
10. H. Priefert, J. Rabenhorst and A. Steinbüchel, Appl. Microbiol. Biotechnol., 2001, 56, 296–314.
11. A. Sheldrake, Nature, 1972, 238, 352–353.
12. T. V. Capital, Study into the establishment of an aroma and fragrance fine chemicals value chain in South Africa (Final Report), FRIDGE, NEDLAC, South Africa, 2004, http://www.nedlac.org.za.
13. V. D. Makwana, Y.-C. Son, A. R. Howell and S. L. Suib, J. Catal., 2002, 210, 46–52.
14. A. Garade, N. Biradar, S. Joshi, V. Kshirsagar, R. Jha and C. Rode, Appl. Clay Sci., 2011, 53, 157–163.
15. V. S. Kshirsagar, A. C. Garade, K. R. Patil, M. Shirai and C. V. Rode, Top. Catal., 2009, 52, 784–788.
16. G. C. Behera and K. Parida, Appl. Catal., A, 2012, 413, 245–253.
17. R. Ma, Y. Xu and X. Zhang, ChemSusChem, 2015, 8, 24–51.
18. W. Deng, H. Zhang, X. Wu, R. Li, Q. Zhang and Y. Wang, Green Chem., 2015, 17, 5009–5018.
19. A. Tarasov, L. Kustov, V. Isaeva, A. Kalenchuk, I. Mishin, G. Kapustin and V. Bogdan, Kinet. Catal., 2011, 52, 273.
20. A. Tarasov, L. Kustov, A. Bogolyubov, A. Kiselyov and V. Semenov, Appl. Catal., A, 2009, 366, 227–231.
21. F. G. Sales, L. C. Maranhão, N. M. Lima Filho and C. A. Abreu, Chem. Eng. Sci., 2007, 62, 5386–5391.
22. E. W. McFarland and H. Metiu, Chem. Rev., 2013, 113, 4391–4427.
23. A. Jha and C. V. Rode, New J. Chem., 2013, 37, 2669–2674.
24. M. Shilpy, M. A. Ehsan, T. H. Ali, S. B. A. Hamid and M. E. Ali, RSC Adv., 2015, 5, 79644–79653.
25. A. Jha, K. R. Patil and C. V. Rode, ChemPlusChem, 2013, 78, 1384–1392.
26. A. Jha, D. Mhamane, A. Suryawanshi, S. M. Joshi, P. Shaikh, N. Biradar, S. Ogale and C. V. Rode, Catal. Sci. Technol., 2014, 4, 1771–1778.
27. M. R. Morales, B. P. Barbero and L. E. Cadús, Appl. Catal., B, 2006, 67, 229–236.
28. T. Fukunaga, N. Ryumon, N. Ichikuni and S. Shimazu, Catal. Commun., 2009, 10, 1800–1803.
29. A. Marinoiu, M. Raceanu, C. Cobzaru, C. Teodorescu, D. Marinescu, A. Soare and M. Varlam, React. Kinet., Mech. Catal., 2014, 112, 37–50.
30. H. Einaga, A. Kiya, S. Yoshioka and Y. Teraoka, Catal. Sci. Technol., 2014, 4, 3713–3722.
31. M. Krämer, T. Schmidt, K. Stöwe and W. Maier, Appl. Catal., A, 2006, 302, 257–263.
32. P. A. Wright, S. Natarajan, J. M. Thomas and P. L. Gai-Boyes, Chem. Mater., 1992, 4, 1053–1065.
33. P. Porta, G. Moretti, M. Musicanti and A. Nardella, Solid State Ionics, 1993, 63, 257–267.
34. V. Koleva, D. Stoilova and D. Mehandjiev, J. Solid State Chem., 1997, 133, 416–422.
35. D. P. Shoemaker, J. Li and R. Seshadri, J. Am. Chem. Soc., 2009, 131, 11450–11457.
36. E. C. Njagi, C.-H. Chen, H. Genuino, H. Galindo, H. Huang and S. L. Suib, Appl. Catal., B, 2010, 99, 103–110.
37. E. C. Njagi, H. C. Genuino, C. K. King'ondu, C.-H. Chen, D. Horvath and S. L. Suib, Int. J. Hydrogen Energy, 2011, 36, 6768–6779.
38. A. A. Mirzaei, H. R. Shaterian, R. W. Joyner, M. Stockenhuber, S. H. Taylor and G. J. Hutchings, Catal. Commun., 2003, 4, 17–20.
39. P. Wei, M. Bieringer, L. M. Cranswick and A. Petric, J. Mater. Sci., 2010, 45, 1056–1064.
40. E. M. Samsudin, S. B. A. Hamid, J. C. Juan, W. J. Basirun and G. Centi, Chem. Eng. J., 2015, 280, 330–343.
41. M. Hernandez, J. Fernández-Bertrán, M. Farias and J. Diaz, Surf. Interface Anal., 2007, 39, 434–437.
42. B. Yang, S. Chan, W. Chang and Y. Chen, J. Catal., 1991, 130, 52–61.
43. S. Angelov, E. Zhecheva, K. Petrov and D. Menandjiev, Mater. Res. Bull., 1982, 17, 235–240.
44. G. Wertheim, S. Hüfner and H. Guggenheim, Phys. Rev. B: Condens. Matter Mater. Phys., 1973, 7, 556.
45. R. Vandenberghe, E. Legrand, D. Scheerlinck and V. Brabers, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1976, 32, 2796–2798.
46. T. Mathew, N. Shiju, K. Sreekumar, B. S. Rao and C. S. Gopinath, J. Catal., 2002, 210, 405–417.
47. M. Roberts, Chem. Soc. Rev., 1989, 18, 451–475.
48. M. O'Connell, A. Norman, C. Hüttermann and M. Morris, Catal. Today, 1999, 47, 123–132.
49. Y. Wang, Y. Lü, W. Zhan, Z. Xie, Q. Kuang and L. Zheng, J. Mater. Chem. A, 2015, 3, 12796–12803.
50. S. A. Kondrat, T. E. Davies, Z. Zu, P. Boldrin, J. K. Bartley, A. F. Carley, S. H. Taylor, M. J. Rosseinsky and G. J. Hutchings, J. Catal., 2011, 281, 279–289.
51. K. Morgan, K. J. Cole, A. Goguet, C. Hardacre, G. J. Hutchings, N. Maguire, S. O. Shekhtman and S. H. Taylor, J. Catal., 2010, 276, 38–48.
52. J. Carnö, M. Ferrandon, E. Björnbom and S. Järås, Appl. Catal., A, 1997, 155, 265–281.
53. F. Kapteijn, L. Singoredjo, A. Andreini and J. Moulijn, Appl. Catal., B, 1994, 3, 173–189.
54. M. Ferrandon, J. Carnö, S. Järås and E. Björnbom, Appl. Catal., A, 1999, 180, 141–151.
55. J. Trawczyński, B. Bielak and W. Miśta, Appl. Catal., B, 2005, 55, 277–285.
56. R. Radhakrishnan, S. T. Oyama, J. G. Chen and K. Asakura, J. Phys. Chem. B, 2001, 105, 4245–4253.
57. G. H. Tomlinson 2nd and H. Hibbert, J. Am. Chem. Soc., 1936, 58, 348–353.
58. V. Tarabanko, D. Petukhov and G. Selyutin, Kinet. Catal., 2004, 45, 569–577.
59. R. Klaewkla, M. Arend and W. F. Hoelderich, A review of mass transfer controlling the reaction rate in heterogeneous catalytic systems, INTECH Open Access Publisher, 2011.
60. J. Andas, F. Adam, I. A. Rahman and Y. H. Taufiq-Yap, Chem. Eng. J., 2014, 252, 382–392.
61. M. Fujita and L. Que, Adv. Synth. Catal., 2004, 346, 190–194.
62. L. S. Puckhaber, H. Cheung, D. L. Cocke and A. Clearfield, Solid State Ionics, 1989, 32, 206–213.
63. D. Cocke and S. Vepřek, Solid State Commun., 1986, 57, 745–748.

---