Synthesis and Characterization of Iron Oxide Nanoparticles Supported on Ziconia and Its Application in the Gas-Phase Oxidation of Cyclohexanol to Cyclohexanone ()
1. Introduction
Since the early 1900’s, iron-based dispersed catalysts have been used for the liquification of coals. Continuous efforts have been made to reduce size of the particles with a simultaneous focus on preserving and enhancing the dispersion, in order to improve the quality and affordability of these iron-based catalysts. Several attempts have been made to prepare iron oxides and explore their catalytic activities in petrochemical industries [1-9]. Research has established the enhancing effect of promoters on the catalytic activity of iron-based catalysts, for example, molybdenum promotes the catalytic activity of sulfated hematite while tungsten has additive effect when used in combination with molybdenum. Similarly, nickelcobalt has synergetic effect while used with molybdenumetc. [10-12]. Recently, researchers have diverted attention towards nano-materials and their applications in the field of catalysis. In this scenario, iron oxide nanoparticle is a potent candidate to be investigated as a catalyst in various industrially important reactions, including the synthesis of NH3, water shift reaction, desulfurization of natural gas, dehydrogenation of ethyl benzene, oxidation of alcohol, and manufacture of butadiene [13-18].
In this work, we prepared iron oxide nanoparticles from iron nitrate with precipitating agents, i.e., ammonium hydroxide and ammonium acetate, without the aid of any surfactant. Iron oxide particles were found to be in the range of 8 - 10 nm as estimated from XRD of pure iron oxide nanoparticles. Following the same procedure, iron oxide nanoparticles supported on zirconia were prepared by adding deionized water and monoclinic zirconia to iron hydroxide prepared with ammonium hydroxide, or an excess of acetone and monoclinic zirconia to Fe2(CHOO)6 sol prepared with ammonium acetate. The prepared catalysts were characterized by various techniques, and tested for catalytic activities in the gas phase conversion of cyclohexanol to cyclohexanone, cyclohexene and benzene, keeping in view the fact that hematite and magnetite are semiconductors and can catalyze oxidation/reduction reactions [4].
2. Experimental
2.1. General
Chemicals were of high purity grade and were used as such without further purification. The catalyst was prepared in the laboratory from its precursor compounds. Nitrogen, oxygen and hydrogen were supplied by BOC Pakistan limited.
2.2. Catalyst Preparation
The catalyst was prepared in two steps.
2.2.1. Monoclinic Zirconia
Monoclinic zirconia was used as a support for the catalyst which was prepared by ammonolysis of ZrOCl2∙8H2O (0.5 M aqueous solution) [19]. The precipitate was washed, dried, grinded, and then calcined at 1203 K in furnace at a rate of 1˚C/min.
2.2.2. Fe2O3/Monoclinic ZrO2
Fe2O3 nanoparticles were prepared by drop wise addition of 1) 0.5 M ammonium acetate solution to 0.1 M iron nitrate solution under vigorous stirring to obtain Fe2(CHOO)6 sol [16]. After the reaction, an excess of acetone and monoclinic zirconia were added and centrifuged to precipitate Fe2(OH)6/ZrO2.
2) 0.5 M ammonium hydroxide solution to 0.1 M iron nitrate solution under vigorous stirring to obtain brownish precipitate. The precipitate was washed twice with triple distilled water and centrifuged. To the purified precipitates, deionized water and monoclinic zirconia were added and stirred, followed by centrifugation at 3000 rpm. Iron hydroxide/zirconia was then dried and calcined at 823 K which resulted in the nanoparticles of iron oxide supported on ZrO2.
2.2.3. Reduction of the Catalyst
The prepared catalyst was reduced at 553 K in tube furnace for 2 h. The nitrogen gas was continuously passed through the reactor until the desired temperature (i.e., 553 K) was achieved, and then a mixture of hydrogen and nitrogen (1:1) was passed for two hours at 40 mL/min, with a subsequent cooling in the nitrogen atmosphere.
2.3. Characterization of the Catalyst
Modern techniques such as XRD (X-ray differactrometer Rigaku D/Max-II, Cu tube, Japan), SEM (JSM 5910, JEOL, Japan), TGA/DTA (Diamond Series PerkinElmer, USA), U.K), FT-IR (Shimadzu prestige-21), Surface area and pore size analyzer (Quantachrome) were used for characterization of the catalyst.
2.4. Catalytic Test
100 mg of iron oxide nanoparticles supported on monoclinic ZrO2 catalysts were used for the gas phase oxidation of cyclohexanol to cyclohexanone in a fixed-bed flow type Pyrex glass reactor at 433 - 463 K in a tubular furnace attached to temperature controller. Cyclohexanol vapors were fed from saturators, using N2 as a carrier gas with a fixed flow rate of 40 mL/min. Reaction mixtures of 0.5 mL were injected at specified time intervals with six-port gas sampling valve to GC (PerkinElmer Clarus 580) with column (rtx@-Wax 30 m, 0.5 mm ID, 0.5 nm) and FID.
3. Results and Discussion
3.1. Characterization of the Catalyst
XRD of the catalysts, i.e., one prepared with ammonium acetate and the other prepared with ammonium hydroxide as a precipitating agent, was carried out. XRD pattern of the former catalyst presented with peaks at 2θ = 28.5 and 31.8 which point towards a retention of the monoclinic phase of zirconia. However, the monoclinic phase of zirconia was lost in the latter catalyst which could probably be due to the presence of water. Furthermore, peaks at 2θ = 35.4 and 45.3 were recorded for both samples prior to being reduced, which account for the crystallized nanoparticles of iron oxide (αFe2O3/γFe2O3). In case of reduced samples, a peak was recorded at 2θ = 69.2 as shown in Figure 1 which owes for the presence of magnetite in the samples [12,20,21].
The average particle size was estimated (8 - 10 nm) by using the Debye-Scherrer equation from peak width broadening in the XRD data of pure iron oxide (Figure 2). This finding is consistent with earlier studies of iron oxide nanoparticles [20,22].
BET surface area of the catalysts was found to be in the range of 150 - 300 m2g−1. FT-IR spectra (Figure 3) of freshly prepared sample exhibited peaks at 3300, 1642 and 1300 cm−1 corresponding to the O-H group, carboxyl group and 
ion, respectively [20,23]. But none of these peaks could be recorded for the unused calcined sample. However, peaks at 470 cm−1 and 570 cm−1 confirmed the presence of hematite particles in the unused calcined sample.
Figure 1. XRD patterns of the iron oxide nanoparticles supported on monoclinic zirconia: (a) Prepared with ammonium hydroxide as precipitating agent, and calcination at 823 K; (b) Prepared with ammonium acetate as precipitating agent, and calcination at 823 K; (c) Prepared with ammonium acetate as precipitating agent and calcination at 823 K, followed by reduction at 553 K in hydrogen flow for 2 hours.
Figure 2. XRD patterns of the iron oxide nanoparticles: (a) Prepared with ammonium acetate as precipitating agent and calcination at 823 K; (b) Prepared with ammonium hydroxide as precipitating agent and calcination at 823 K, followed by reduction at 553 K in hydrogen flow for 2 hours.
SEM images (Figure 4) show that iron oxide nanoparticles prepared by using NH4OH as precipitating agent are clustered together and bear a rough overall morphology, whereas those prepared by ammonium acetate as precipitating agent are well dispersed and have a smooth morphological presentation. SEM images clearly show that the morphological smoothness of the iron oxide nanoparticles in the latter sample enhances with calcination and reduction. After reduction in hydrogen flow, the nanoparticles appear as spherical balls. TGA/DTA study of freshly prepared dried catalysts, reflects a three stage reduction of iron oxide, and thereby strongly supports the procedure we have adopted for the preparation of hematite nanoparticles on monoclinic zirconia by calcination at 823
Figure 3. FT-IR of the iron oxide nanoparticles supported on monoclinic zirconia: (a) Before calcination, (b) After calcination.
Figure 4. SEM images of iron nanoparticles on ZrO2: (a) Prepared with ammonium hydroxide as precipitating agent; (b) Prepared with ammonium acetate as precipitating agent; (c) Prepared with ammonium acetate as precipitating agent and calcined; (d) Prepared with ammonium acetate as precipitating agent, calcined and reduced in hydrogen flow.
K, and for the preparation of magnetite nanoparticles on monoclinic zirconia by calcination at 823 K 4 hours followed by reduction at 553 K in hydrogen flow for 2 hours.
3.2. Oxidation/Dehydrogenation
Oxidation/dehydrogenation was demonstrated by comparing the data obtained with and without oxygen under identical experimental conditions using magnetite nanoparticles supported on zirconia as a catalyst as shown in Figure 5. A noticeable effect was observed on the rate of reaction, when oxygen (95 Torr) was introduced in the gas feeding system, in comparison to that obtained in pure N2 atmosphere. The dehydrogenation activity falls again to the same negligible level when O2 is removed from the gaseous mixture. This increase in the rate of formation of cyclohexanone in the presence of O2 is common for metal oxide catalysts and is known as oxidation. Most of the industrially important reactions proceed through oxidation pathway in the presence of metal oxide as catalysts.
The FT-IR spectra of the catalyst recovered from the reaction carried out in the oxygen flow using magnetite nanoparticles supported on zirconia showed intense peak for carbonyl group, thus showing that the major product of gas phase oxidation of cyclohexanol is cyclohexanone as shown in Figure 6. The percent yield calculated from FT-IR was in a good agreement with that of GC results.
The same reaction when carried out under identical condition using hematite nanoparticles supported on zirconia gives cyclohexene as a major product by dehydration of cyclohexanol as shown in FT-IR spectra (Figure 7). However with an increase in temperature, benzene was also detected suggesting digression of the reaction towards dehydrogenation as shown in Scheme 1 [25-29].
3.3. Activation Energy
Variation in the rate of reaction with temperature at con-
Figure 5. Activity profile of magnetite nanoparticles/monoclinic ZrO2 for the conversion of cyclohexanol to cyclohexanone in an atmosphere of oxygen and nitrogen.
Figure 6. FT-IR of the recovered catalyst (magnetite nanoparticles supported on monoclinic zirconia), after being used for conversion of cyclohexanol.
Figure 7. FT-IR of the recovered catalyst (hematite nanoparticles supported on monoclinic zirconia), after being used for conversion of cyclohexanol.
Scheme 1. Major products of cyclohexanol while using magnetite nanoparticles supported on monoclinic zirconia (i) or hematite nanoparticles supported on monoclinic zirconia (ii & iii).
stant pressures ranging from 12 Torr to 33 Torr was observed, using magnetite nanoparticles supported on monoclinic zirconia. Activation energies were calculated using Equation (4). Activation energies at pressures 12, 15, 20, 26, 33Torr are 17.48, 18.2, 17.73, 19.27, 19.73 kJ·mol−1, respectively. The activation energies for all these pressures are in the range where reaction is diffusion controlled. The rate of a reaction is given by following equation.
(1)
According to Arrhenius equation
(2)
Putting the value of k from Equation (2) in Equation (1), we get
(3)
ln rate = (lnA + nlnP)
(4)
Plot ln rate vs. 1/T will give slope = 
while intercept = lnA + nlnP as shown in Figure 8. Different values of intercept as obtained from figure were plotted against lnP (Figure 8) where intercept = lnA and slope = n represent the order of reaction. The exposure of the catalyst in presence of oxygen was carried out for 480 minutes as shown in Figure 9. It is clear from the Figure 9 that there is no noticeable decline in the catalyst activity.
4. Conclusion
The catalytic activity of hematite and magnetite nanoparticles supported on monoclinic zirconia was investigated for the gas-phase reaction of cyclohexanol. It was found that the catalyst was more active when reduced and selective for cyclohexanol oxidation to cyclohexanone, but it showed activity and selectivity for dehydration (i.e., conversion of cyclohexanol to cyclohexene) when unreduced. Furthermore, the catalyst effectuated dehydration/ dehydrogenation of cyclohexanol to benzene at higher temperatures. The model reaction, i.e., the gas-phase oxidation of cyclohexanol to cyclohexanone using magnetite nanoparticles supported on monoclinic zirconia, in
Figure 8. Plot of ln rate vs. 1/T (slope = 
and intercept = lnA + nlnPROH) and plot of lnPROH against calculated intercept.
Figure 9. Continuous exposure of magnetite nanoparticles supported on monoclinic zirconia, used for conversion of cyclohexanol to cyclohexanone at 463 K.
flow reactor, was found to be in diffusion control regime. In addition, the iron oxide nanoparticles supported on monoclinic zirconia could be easily recovered and used several times without significant loss of catalytic activity.
Acknowledgements
The authors gratefully acknowledge financial support of Higher Education Commission of Pakistan (Project No: 20-1604/R&D/092198) and Pakistan Science Foundation (Project No.PSF/Res/F-UM/Chem-434).