Catalytic Properties of Alumina-Supported Ruthenium , Platinum , and Cobalt Nanoparticles towards the Oxidation of Cyclohexane to Cyclohexanol and Cyclohexanone

A series of metal-loaded (Ru, Pt, Co) alumina catalysts were evaluated for the catalytic oxidation of cyclohexane using tertbutylhydroperoxide (TBHP) as oxidant and acetonitrile or acetic acid as solvent. These materials were prepared by the impregnation method and then characterized by Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES), H2 chemisorption, Fourier Transformed Infrared Spectroscopy (FTIR), High-Resolution Transmission Electron Microscopy (HRTEM), and X-ray Diffraction (XRD). All the prepared materials acted as efficient catalysts. Among them, Ru/Al2O3 was found to have the best catalytic activity with enhanced cyclohexane conversion of 36 %, selectivity to cyclohexanol and cyclohexanone of 96 % (57.6 mmol), and cyclohexane turnover frequency (TOF) of 288 h-1. Copyright © 2018 BCREC Group. All rights reserved


Introduction
The oxidation of organic substrates represents one of the most important industrial chemical reactions [1].This explains the significant effort invested in research and development of new heterogeneous catalysts with increased activities and selectivities in these types of reactions.In particular, cyclohexane oxidation is one of the main goals in this area.The oxidation of cyclohexane to cyclohexanol and cyclohexanone is a key reaction in the syn-thesis of adipic acid (AA), which is an extremely important commodity chemical [2,3] mainly used in the production of urethanes and nylon-6, 6 (for fibers and resins).The global AA capacity, produced in 2006, was around 2.8 million metric tons and the demand for it has been growing, year after year, in the last decade [4].The contemporary industrial process for manufacturing adipic acids (AA) is a two-step process involving nitric acid oxidation of KA oils, a mixture of cyclohexanol and cyclohexanone (which is produced from cyclohexane) using a homogeneous cobalt-based catalyst [5,6].The main disadvantage of this process is that large amounts Huang et al. [20] conducted a series of studies on cyclohexane by air oxidation; using boehmite ((AlOOH) immobilized cobalt tetra (4 carboxyl) phenylporphyrin catalyst (CoTCPP).Cyclohexane conversion of 19.6 % and selectivity of 73 % were obtained.Lu et al. [21] reported a 9.2 % conversion and 92.7 % selectivity when Co3O4 catalyst was used for the oxidation of cyclohexane with O2, at 130 °C.Ti-Zr-Co alloys were prepared by Hao et al. [2], and then used as catalysts in solvent-free liquid phase oxidation of cyclohexane by oxygen; the conversion obtained was 6.8 % and the selectivity of cyclohexanol and cyclohexanone was 90.4 %.
In our present work, ruthenium, platinum and cobalt nanoparticles supported on alumina were used as catalysts for the aerobic oxidation of cyclohexane to KA oils (i.e., cyclohexanone and cyclohexanol), giving moderate yields and improved catalyst turnover number.Noble metals (Ru, Pt) are much active, stable and selective.They are widely used because of their better performance in many oxidation reactions.Cobalt was chosen due to it prevalent use in many oxidations, including that of cyclohexane commercially.

Synthesis
The Al2O3 support of commercial grade was purchased in the form of nanopowder (in the range 10-50 nm).It required further treatment before impregnation of the metal precursors.The support was mixed with water (200 mL of H2O for 100 g of support) in order to form a paste, which was dried overnight at 120 °C, and then sieved to retain only the particles having a diameter ranging between 0.1 and 0.25 mm.Then the support underwent calcination at 400 °C under oxidative flow (20 % O2, 80 % Ar, 60 mL/min), during 4 h.After activation, Al2O3 showed a BET specific surface area of 95 m 2 g -1 , pore volume of 0.72 cm 3 g -1 and pore size of 31 nm.
Copyright © 2018, BCREC, ISSN 1978ISSN -2993 An aqueous solution of Pt(NH3)2(NO2)2 and Co(NO3)2.6H2Oor an acetone solution of Ru(acac)3 was used for the wetness impregnation of the support, in order to obtain 1 or 5% wt. of metal catalyst.After solvent evaporation, the solids were dried at 120 °C overnight, then calcined at 400 °C (Co and Pt) and 350 °C (Ru), for 4 h, under oxidative atmosphere (Ar: 48 mL/min -O2: 12 mL/min).Finally, the solids were reduced by H2 (60 mL/min) at the same temperatures, for 4 h.These calcination and reduction temperatures for the ruthenium catalyst were selected because Ru(acac)3 decomposes in air in one step, starting at 200 °C and completing at 350 °C.Activation at temperature as high as 350 °C ensures complete removal of acetylacetonate ligands from the ruthenium precursor [22].

Characterization of catalysts
The chemical compositions of the samples were analyzed by inductively coupled plasmaatomic emission spectroscopy (ICP-AES), using an OPTIMA 2000 DV spectrometer.Diffractograms of the catalysts were obtained from XRD experiments, performed on a Siemens D5005 powder diffractometer using Cu-Kα radiation (λ = 0.15186 nm) and a back monochromator.XRD patterns were recorded using a dwell time of 2 s, a step size of 0.04° and a constant divergence slit equal to 1°.The crystalline planes were identified by comparison with PDF standards from ICDD.The N2 adsorptiondesorption measurements of Al2O3 were carried out at -196 °C, using a Micromeritics TriStar device.The surface area of the support was calculated using the Brunauer-Emmett-Teller method.
Electron microscopy studies of the catalysts were carried out on a JEOL 2000 FX instrument, operating at 200 kV.Catalyst specimens for electron microscopy were prepared by grinding the powder samples in an agate mortar, suspending and sonicating them in ethanol, and placing a drop of the suspension on a holey carbon copper grid.After evaporation of the solvent, the specimens were introduced into the microscope column.The diameter distribution of the particles was determined by counting a large number of particles (>500) on the TEM micrographs and plotting ni as a function of di (ni is the number of particles within different intervals with given average diameters di).The mean surface diameter [23] of the particles is then given by Equation ( 1). (1) The measurements were done manually using the ImageJ program.The chemisorption measurements were carried out in a glass volumetric system.Catalyst samples of ~ 0.2 g were placed into a reactor and reduced by pure H2, at 400 °C, for 2 h.A 1/1 H/metal stoichiometry was assumed, in order to calculate the metallic dispersion from which the mean diameter (d) of metal particles could be calculated using a cubic model, with one face in direct contact with the support (Equation (2)). ( where ρM is the metal density, SM surface area occupied per metal atom, and D the dispersion of the metal (Table 1) [24].

Evaluation of the catalytic activity
The catalytic oxidation of cyclohexane, with tertiobutylhydroperoxide (TBHP) as oxidant, was carried out in a glass round-bottom flask with a magnetic stirrer and a reflux condenser.First, commercial TBHP 70 % in H2O (Aldrich) was stirred with cyclohexane in order to perform a phase transfer from water to cyclohexane.In a typical reaction, 60 mmol (6.5 mL) of cyclohexane and 60 mmol (8.5 mL) of oxidant (TBHP) are mixed in a closed Erlenmeyer flask and magnetically stirred for 24 h.The organic phase was then separated from the aqueous phase.In order to control the phase transfer, the concentration of the remaining TBHP, in the aqueous phase, was determined by iodometric titration, and was found to be < 10 %.The solvent (50 mL) was then added to the TBHP-cyclohexane mixture.These reactants and the solvent were introduced in a glass round-bottom flask and heated to 70 °C under vigorous stirring.The catalyst (0.05 g) was subsequently added to the reaction mixture (time zero).The reaction products were identified by comparison with authentic products and the course of reactions was followed by gas chromatography (GC), using a Varian CP-3800 gas chromatograph equipped with a CP-WAX 52 CB column.A flame ionization detector (FID) was used and a quantity of 0.5 μL from the sample was analyzed.Before GC analysis, the remaining TBHP was decomposed by introducing an excess of triphenylphosphine.On the other hand, to control this remaining TBHP, an iodometric titration was performed at the end of the reaction (after 6 h) by testing the organic phase.The catalytic performances were reported in terms of cyclohexane conversion and selectivity towards products.They are calculated following the expressions in Equations (3-4): (3) (4)

Characterization
As measured by ICP-OES, the metal content in all catalysts is either 1 or 5 % (Table 2).Figure 1 shows the XRD patterns of the bare Al2O3 support and the different catalysts.The alumina support shows two crystalline phases with sharper peaks, indicating that alumina is in δ and γ forms.The diffractogram of Ru/Al2O3 confirms that metallic ruthenium is present in catalysts as a result of the reduction operation, and no RuO2 peak is observed.The Ru diffraction peaks (38.4°, 42.2°, 44.0°, 58.3°, and 69.4° 2θ, JCPDS no.6-663) are sufficient for an estimation of the average ruthenium crystallite size, which is 9 nm, with a relative standard deviation of 10 %.Diffraction peaks at 39.76°, 46.24° and 67.45° 2θ (JCPDS no.4-802), characteristic of metallic Pt, were detected on the diffraction patterns of Pt/Al2O3.No platinum oxide was observed.Crystalline Co3O4 and CoO phases were observed in Co/Al2O3.No peaks of metallic Co were detected.The peaks at 19.08°, 31.38°,36.88°,44.88°, 59.38° and 65.28° 2θ were attributed to the cubic phase of Co3O4 with Fd3m space group (JCPDS no.42-1467).In addition, the lines at 42.4° and 34.5° 2θ could be assigned to the cubic structures of CoO (JCPDS no.71-1178).
The metal dispersion of catalysts was obtained by H2 chemisorption.The percentage of metallic dispersion was calculated, assuming an H/M atomic ratio equal to 1 [25,26], as shown in Table 2.The 1 % Ru or Pt catalysts gave a dispersion (20 %) lower than that obtained with the 5 % catalysts.Ruthenium catalysts showed higher dispersion than platinum ones.DRX results of cobalt catalysts showed the formation of cobalt oxides and chemisorption measurements were not possible.
TEM analysis (micrographs and histograms) for Ru/Al2O3, Pt/Al2O3 and Co/Al2O3   2, 3, and 4, respectively.The 5 %Pt/Al2O3 and 5 %Ru/Al2O3 catalyst samples were well dispersed with no indication of agglomeration of the metal particles.All of these materials have a spherical morphology.The ruthenium system showed nanoparticle sizes ranging from 1.7 to 4.6 nm with a mean size of 2.5 nm (Figure 2).In the platinum system, the size distribution ranges between 1.8 and 6 nm, with a mean particle size of 3.2 nm (Figure 3).TEM analysis of Co/Al2O3 (Figure 4) revealed an intensive agglomeration of cobalt particles on these samples and, then, a true reliable distribution could not be given.The average size of these agglomerations was between 8 and 10 nm.Zhou et al. [5] showed that Co3O4 oxide has a platelet morphology forming, in most cases, quite large agglomerations.As shown, a very good correspondence between the mean particle sizes estimated from H2 chemisorption and TEM was obtained for all catalyst samples (Ru and Pt) (Table 2).

Catalytic oxidation of cyclohexane
The activities of the catalysts for the catalytic oxidation of cyclohexane with TBHP at 70 °C, in acetic acid or acetonitrile, are presented in Table 3.The desired products are cyclohexanol (C6H11OH) and cyclohexanone (C6H10O), but other products like cyclohexyl hydroperoxide, adipic acid, ester dicyclohexyl adipate, hexanolactone, and other esters [5], among which is cyclohexyl acetate [27], did also form.In general, cyclohexyl hydroperoxide can be decomposed to cyclohexanol and cyclohexanone in presence of PPh3 (triphenylphosphine) introduced before GC analysis, thus increasing the yield of the target compounds.
The present study focused on the selectivity towards olone only (cyclohexanol and cyclohexanone).The amount of catalyst used was 0.05 g and the cyclohexane to TBHP mole ratio was (1:1).A blank oxidation reaction (i.e.without catalyst) was carried out under typical reaction conditions and no oxidation product was formed.Moreover, to check the impact of the support on the cyclohexane oxidation reaction, Al2O3 was tested for its catalytic properties in the presence of acetic acid as solvent.The support exhibited a conversion less than 15 % with selectivity around 10.3 mmol (17.2 %) (Olone).As expected, these results are found to be close to those of Bellifa et al. [28] and Zhao et al.  [29].Table 3 details the catalytic activities of alumina-supported ruthenium in the presence of acetic acid and acetonitrile.Cyclohexanol and cyclohexanone selectivities increase with metal content [17].The 1 % catalysts showed a modest selectivity increase compared to 5 % ones.So, one can assume that the support participates in the reaction via a functional mechanism (support and metal), as reported by Wangcheng et al. [30].
In the presence of TBHP and acetic acid as solvent, turnover frequencies (TOF) of 1863 h -1 and 288 h -1 were obtained, after 6 h of catalytic oxidation, for 1 % and 5 % Ru/Al2O3 systems, respectively.The TOF values obtained are comparable (1907 h -1 for 1.21 % Au/MCM-41) [31] to or higher (184 h -1 ) [32] than those of CoFe2O4 catalysts reported in the literature; however they are smaller than the highest value which amounts to 11900 h -1 for Au/ HAP (Hydroxyapatite) [33], and 2592 h -1 for 1 % Au/Al2O3.Nevertheless, gold-based catalysts are regarded as expensive catalysts compared to ruthenium based catalyst which is the least expensive among the noble metals.This represents a cyclohexane conversion of 35.6 %, whereas the selectivity towards the formation of cyclohexanol is 10.0 mmol (16.7 %) and towards cyclohexanone 12.4 mmol (20.7 %).By comparing the two solvents, the activities obtained under acidic conditions were found to be higher compared to those obtained in acetonitrile.These activities decreased when the ruthenium content increased.This observation suggests that the acetic acid was oxidized by TBHP to form peracetic acid in situ, which served as a better oxidant [34,36].The amount of ruthenium leached, of our nanostructured material, is determined after reaction and found < 50 ppm.
Platinum catalysts (Table 3), in acetic acid and after 6 h, showed high activity towards cyclohexane oxidation, and gave turnovers of 2830 h -1 and 782 h -1 , for 1 % and 5 % Pt/Al2O3, respectively.These correspond to conversions 14.8 % and 28.4 %, respectively.In the presence of acetonitrile, it is clear that platinum does not promote the formation of cyclohexanone.The cyclohexanol selectivity decreases when the platinum content increases from 1 to 5 % [30].As can be seen from the comparison data, the ruthenium-based catalyst is more active towards oxidation products than the platinum-based catalyst.
After the reaction, the ICP analysis of platinum-based catalyst did not exhibit leaching, which means that it is a heterogeneous catalytic system.For cobalt nanoparticles supported on alumina, the results reveal that these solids are capable of oxidizing the substrate used in this study.After 6 h of reaction time, and in the presence of acetic acid as solvent, the turnovers were found equal to 177 h -1 and 49 h -1 , which correspond to 14.8 % and 20.7 % conversions to cyclohexane, for 1 % and 5 % CoxOy /Al2O3, respectively (Table 3).These results show that cobalt-based catalysts are less active than their ruthenium and platinum counterparts, in the conditions under consideration.It can be seen from Table 3 that a relatively lower conversion was obtained in the case of acetonitrile.The chemical analysis of the catalyst, after reaction, revealed that 20 % of cobalt, which was in the catalyst, passed into the solution.Therefore, it is not possible to check whether the catalysis took place in the heterogeneous system CoxOy/Al2O3, in a mixture of heterogeneous and homogeneous catalysts, or just as a homogeneous system.In summary, one can say that concerning the catalytic activity and selectivity of aluminasupported ruthenium, platinum and cobalt nanoparticle towards the oxidation reaction of cyclohexane, it can be proven that noble and non-noble metals can be active for the oxidation of cyclohexane.The support is active during the cyclohexane oxidation, but the presence of the metal clearly improves the reaction activity which increases with the metal loading.From the literature, it is well established that some transition metal ions can be catalytically activated towards the oxidation reactions of certain organic substrates when H2O2 or RO2H are used as oxidizing agents.
These reactions can be classified in two categories, depending on whether the active species in the oxidation reaction is (i) peroxometal or (ii) oxometal [37].Generally, the last transition metals employ an oxometal intermediate.Some other elements, such as vanadium, can use either route, depending on the substrate.Most transition metals that catalyze oxygen-transfer processes, whether via peroxometal or oxometal, are also able to catalyze processes via free radicals formed from peroxide compounds [38,40].
Thus, as expected in the cyclohexane oxidation reaction, and in the presence of the two solvents, the activity increases as Ru/Al2O3 ˃ Pt/Al2O3 ˃ Co/Al2O3.The acetic acid when used as solvent is better than acetonitrile, as it allows higher conversion and selectivity towards cyclohexanol and cyclohexanone.From the leaching results, it is possible to state that the systems studied have the following order of stability: Pt/Al2O3 ˃ Ru/Al2O3 ˃ CoxOy /Al2O3.

Optimization of reaction conditions
Due to the good performance of the catalyst Ru/Al2O3, this study focused, in the following section, on the influence of various parameters on cyclohexane conversion and selectivity towards cyclohexanone and cyclohexanol over 5 % Ru/Al2O3.The reaction conditions were optimized in acetic acid.The reactions were conducted at the temperatures 25, 45, 60, 65, and 70 °C, as shown in Figure 5.
As expected, the cyclohexane conversion increases as a function of the temperature.The distribution of cyclohexanol and cyclohexanone shows no formation of cyclohexanone at 25 and 45 °C, and a temperature increase clearly favors the production of cyclohexanone.A similar behavior was noted by Zhou et al. [5], and Yuan et al. [41].These results show that when the temperature increased from 25 to 70 °C, the conversion and selectivity towards cyclohexanone and cyclohexanol increased and The second order plots gave a straight line, indicating that the reaction is of the 2 nd order, as shown in Figure 6.The activation energy, calculated from the slope of the straight line (Figure 6), is 31.23 kJ mol -1 , which is lower than the values reported in several previous works [42][43][44][45][46][47].The effect of the mole ratio was examined by varying the cyclohexane/TBHP mole ratios (1:1 and 1:2), with 0.05 g of Ru/Al2O3 catalyst at 70 °C, for 6 h (Figure 7).An increase in the TBHP concentration (1:2) resulted in an increase in the selectivity of the oxidized products, such as: cyclohexanol (from ~10 mmol (16.7 %) to 20 mmol (33.4 %)) and cyclohexanone (from 12.4 mmol (20.7 %) to 37 mmol (61.7 %)); however, the conversion remained constant (35.6 %).This might be explained by the fact that the TBHP used may have increased the conversion of cyclohexanol, which in turn reacted with the acetic acid in excess, which led to an increase in cyclohexanone formation.
A similar behavior was noted on TS-1 [48].On the other hand, TBHP gave cyclohexanone as the major product, owing to its much stronger oxidizing capability.The same observation was reported by Selvam et al. [49] and Kumar et al. [27].According to them, the increase in the cyclohexanone selectivity, with the increase in TBHP, is due to the oxidation of cyclohexanol to cyclohexanone in the presence of excess amount of TBHP.
The effects of the amount of catalyst on the reaction were also studied.The results are presented in Figure 8.One can see that the olone selectivity increased from 22 mmol (~ 37 %) to 57.6 mmol (96 %), while the conversion decreased from 35.6 % to 25 %, as the mass of catalyst rose from 0.05 g to 0.1 g.These results coincide with those of Lu et al. [21].The presence of a high number of active centers eventually causes an increase in the rate of reaction, thereby increasing the olone selectivity.On the other hand, these results indicate that in the system catalyzed by Ru/Al2O3, only a small amount of catalyst was active in the oxidation of cyclohexane [22].This could also be attributed to the competition of interactions between the metal-oxo species in the catalyst and the hydroperoxy species and cyclohexane [50,51].It can therefore be concluded that the amount of 0.1 g of catalyst is optimal for this catalytic system.
The effect of ruthenium loading on the oxidation of cyclohexane was also examined.The cyclohexane conversion, as well as the (cyclohexanone + cyclohexanol) selectivity over different Ru-loaded catalysts, is shown in Figure 9.However, as the ruthenium loadings of our samples increased from 0 to 5.0 %, the cyclohexane conversion and the total selectivity of the oxygenated products increased, while the TOF value decreased.This behavior is closely related to the size of Ru particles.The catalyst 5% Ru/Al2O3 exhibited the smallest and most uniform Ru particles (with an average size of about 2.5 nm), followed by catalysts 3.7 % Ru/Al2O3 (about 4.4 nm in size) and 1 % Ru/Al2O3 (around 5.8 nm).The high activity of the catalyst, which contains small size Ru particles, may be explained as follows.This result was expected, because small particles have a higher density, which gives a more active surface of Ru atoms.This is favorable for the adsorption and activation of the oxidant during the oxidation reaction, resulting in a higher catalytic activity [50,52].It is therefore clear that the catalytic activity depends on the Ru particle size.Another reason for the low catalytic activity, with low ruthenium concentrations, may be due to partial incorporation of Ru into the bulk support [53].
In addition, in order to check the catalyst recyclability, five reaction runs were carried out under typical reaction conditions, using acetic acid as solvent.The results are shown in Figure 10.After each reaction, the catalyst was separated by filtration, and then reused under the same reaction conditions.After deactivation, it was dried at 350 °C, for 2 h, under argon.The regenerated catalyst was examined under the same conditions.
It can be seen that the olone selectivity decreases with the number of reaction runs, and after the fifth run the catalyst is completely deactivated.The selectivity towards cyclohexanol and cyclohexanone on a recycled catalyst was slightly lower than that obtained on a fresh one.The deactivation of the catalyst is likely related to common inhibition processes found in oxidation reactions, such as the adsorption of the oxygenated product species on the catalyst surfaces [54].These results disclose the fact that the catalyst is stable or could be recycled after the reaction.

Conclusions
Various metal-supported alumina catalysts were prepared by the incipient wetness impregnation method.The characterization of the catalysts obtained, confirmed that ruthenium and platinum are in metallic form while cobalt is in the form of oxides Co3O4 and CoO.The liquid-phase oxidation of cyclohexane was performed using tertbutyl hydroperoxide (TBHP) at 70 °C.The support Al2O3 was active during the oxidation of cyclohexane but the presence of a metal clearly improved the activity of the reaction.All the prepared catalysts showed higher catalytic activity and stability for the oxidation of cyclohexane in acetic acid, compared to acetonitrile.All the prepared catalysts exhibited high catalytic activities and could reach a cyclohexane conversion of 35.6 % and selectivity of 57.6 mmol towards the desired products (olone).
Several reaction parameters in the cyclohexane oxidation reaction over 5 %Ru/Al2O3, i.e. the temperature, amount of catalyst and concentration of TBHP in terms of the cyclohexane/TBHP ratio, were optimized to achieve maximum selectivity to cyclohexanol and cyclohexanone.After optimization of these parameters, the best reaction conditions were obtained for 0.1 g of catalyst, T = 70 °C, C6H12/TBHP molar ratio =1:2 and t = 6 h.The catalyst was recycled and reused five times.It exhibited practically the same activity and selectivity as those of fresh catalyst after heat treatment, under argon.

C6H12 = 6 .
5 mL; TBHP = 8.5 mL; solvent = 50 mL; catalyst = 0.05 g; t = 6 h; T = 70 • C; TOF = mole of converted cyclohexane per unit time per mole of dispersed ruthenium or platinum.a TOF : mole of converted cyclohexane per unit time per mole of cobalt or Al2O3.b After 6 h of reaction time

Figure 6 .Figure 5 .
Figure 6.Kinetic data at different temperatures and Arrhenius plot for the oxidation of cyclohexane

Table 1 .
Physico-chemical properties of metals

Table 2 .
Metallic accessibility and crystallite size for various catalysts

Table 3 .
Oxidation of cyclohexane with M/Al2O3 catalysts