Synthesis and characterization studies of γ-Al2O3 -supported, V-Zr mixed oxide catalysts prepared from OV(OEt)3 and Zr(OnBu)4

Arindom Saha and Darrell P. Eyman *
Department of Chemistry, University of Iowa, Iowa City, IA 52242-1294, USA. E-mail: darrell-eyman@uiowa.edu

Received 21st January 2012 , Accepted 28th March 2012

First published on 19th April 2012


Abstract

Steam reforming of hydrocarbons is one of the methods widely used to produce clean hydrogen for fuel cells. In previous studies in this group, a nickel-containing, V2Ox-ZrO2 co-grafted catalyst on γ-Al2O3, prepared from a solution of OV(OiPr)3 and Zr(OnBu)4 (precursor feed of 1 V nm−2[thin space (1/6-em)]:[thin space (1/6-em)]4 Zr nm−2) was found to generate significant turnover rates in the production of hydrogen from steam-reforming of alkanes and light alcohols. To investigate the influence of change in vanadium precursor ligand and other reaction conditions including time, temperature, concentration of co-grafting solution etc. on the surface optimization of these supported mixed oxides a compositional range of V2Ox-ZrO2 samples were synthesized using co-grafting solutions of OV(OEt)3 and Zr(OnBu)4. Temperature Programmed reduction (TPR) and X-ray Photoelectron Spectroscopy (XPS) studies furnished further insight into the red-ox behavior of the catalysts under steam-reforming conditions. It was observed that a V5+/V4+ red-ox couple was involved in this catalytic process. Zirconium was observed to be present only in the +4 oxidation state irrespective of temperature and other reduction conditions. Based on this premise it was concluded that the refractory nature of the Zr4+ species on the surface was responsible in preventing the vanadyl groups from sintering at temperatures above 600 °C, thereby stabilizing the red-ox active species during the steam-reforming process.


1. Introduction

Fuel cells convert the chemical energy in fuels such as hydrogen, alcohols, and hydrocarbons into usable electricity at efficiencies higher than those obtained by conventional thermal combustion. Such conversion is accomplished without the production of pollutants such as SOx, NOx and carbon soot. However, optimal efficiencies in fuel cells are obtained when H2 is used as the fuel and so there is a widespread interest to develop and refine means of generating H2 from conventional fuels. Steam reforming of hydrocarbons is a very effective way of producing clean hydrogen. It has been observed that materials containing nickel and the mixed transition metal oxides, V2Ox-ZrO2, deposited on a γ-Al2O3 support phase through co-grafting techniques, can be very efficient catalysts in generation of hydrogen through steam reforming of hydrocarbons and light alcohols.1

The technique of co-grafting refers to simultaneously grafting two or more metal derivatives on a surface. This is achieved by reacting, a solution containing two different metal oxide precursors with the surface functional groups of the support which are usually hydroxyls. Basset et al., were one of the foremost researchers in the area of co-grafted bimetallic catalysis.2–6 They used surface organometallic chemistry for the preparation of novel selective bimetallic catalysts by reacting tetra-n-butyl tin with silica supported Rh, Ru and Ni. These bimetallic catalysts were observed to be highly active and selective in hydrogenolysis reactions of various alkanes and esters. Bimetallic Sn-Pt complexes synthesized in the supercages of NaY zeolite using co-grafting techniques also showed excellent catalytic properties in selective hydrogenation of furfural to furfuryl alcohol.7 Silica-supported Pd-Ni bimetallic catalysts were also prepared using surface organometallic chemistry by reacting Ni(C5H5)2 with a reduced Pd/SiO2 support. A comparison between the X-ray methods and a careful analysis of the particle composition by energy–dispersive X-ray mapping in the scanning transmission electron microscope (EDX-STEM) showed that through this process nearly homogeneous alloys could be synthesized.8 Bimetallic Pt-Sn catalysts prepared through grafting processes also provided better activity and selectivity in dehydrogenation of isobutane to isobutene than when prepared through successive impregnation.9

To understand how change in vanadium ligand precursor and other controlled factors like time, concentration of grafting solution, temperature etc. affect the optimization of the support surface, a range of V2Ox-ZrO2 samples were synthesized on γ-Al2O3 using co-grafting solutions of OV(OEt)3 and Zr(OnBu)4 and studied. TPR and XPS studies furnished further insight into the red-ox behavior of the catalysts under steam-reforming conditions.

2. Experimental section

All gases used for TPR studies and for maintaining an inert atmosphere in Schlenk lines and glove boxes were purchased from Air Products. Drisolv@ Tetrahydrofuran (THF) with special mesh caps on the bottles purchased through VWR to prevent entry of moisture was used as a solvent to perform all the grafting reactions. Distilled water, nitric acid (50–70%, Fischer Scientific) and sulfuric acid (98%, Fischer Scientific) were used to dissolve samples for elemental analysis. The vanadium and zirconium precursors used for these co-grafting studies were OV(OEt)3 (95–98%, Alfa Aesar) and Zr(OnBu)4 (98%, Aldrich Chemicals) respectively. Vanadium and zirconium metal precursors were stored and used in an inert glove box environment. The V-ICP standard and the Zr-ICP standard used for calibration of vanadium and zirconium respectively in ICP-OES studies, were purchased from Aldrich Chemicals and stored ambiently. All reagents and solvents were used as received without any further purification or processing.

The γ-Al2O3 pellets used for the grafting studies were spherical in shape (∼1 mm diameter) and were generously donated by Saint-Gobain Norpro Corporation. The samples were dried at 110 °C for at least 24 h before use. A Micromeritics ASAP 2000 was used to measure surface area by adsorption and desorption of N2 at liquid nitrogen temperature. The surface area was determined from the acquired data using the BET (Brunaeur-Emmett-Teller) equation. Thermogravimetric Analysis (TGA) was used to determine the surface -OH density of the γ-Al2O3 pellets.

The physical properties of the γ-Al2O3 used for these studies were as follows: surface area ∼ 219 m2 g−1, pore volume ∼ 0.59 cm3 g−1 and average pore diameter of 6.7 nm. From TGA the surface -OH density was observed to be ∼ 13.02 ± 0.31 OH nm−2.

2.1 Sample synthesis and nomenclature

In all studies except the one depicted in Fig. 3, a co-grafting solution feed of 1 V nm−2[thin space (1/6-em)]:[thin space (1/6-em)]4 Zr nm−2 was used for sample synthesis. The co-grafted V2Ox-ZrO2 samples on γ-Al2O3 were synthesized by taking appropriate amounts of the metal precursors in THF in a two-necked round bottom flask and attaching it to a Schlenk-line under an inert nitrogen atmosphere. After the initial synthesis the samples were calcined at 500 °C for 24 h to get the final co-grafted V2Ox-ZrO2 catalysts (See ESI for detailed synthetic procedures). Four series of co-grafted V2Ox-ZrO2 mixed oxides t-V2Ox-ZrO2/Al2O3, f-V2Ox(fixed)-ZrO2/Al2O3, f-V2Ox-ZrO2(fixed)/Al2O3 and c-T-V2Ox-ZrO2/Al2O3 were synthesized using OV(OEt)3 and Zr(OnBu)4 as the precursors. In these sample groups “t” stands for time of co-grafting, “f” for metal alkoxide feed in co-grafting solution, “c” stands for concentration of co-grafting solution and “T” for temperature of co-grafting. “x” is used to denote the number of oxygen groups in the final grafted vanadium species. An effort has been made to maintain a consistent nomenclature for these co-grafted samples. For example t(24h)-V2Ox-ZrO2/Al2O3, indicates a time sample synthesized by co-grafting for 24 h.

2.2 Catalyst characterization and testing

2.2.1 Elemental analysis. The surface densities of vanadium and zirconium on the co-grafted samples were calculated from the surface area of the γ-Al2O3 substrate and the amount of metals determined through quantitative elemental analysis. These analyses of various samples were performed using a Varian 720-ES Inductively Coupled Plasma-Optical Emission Spectrometry (ICP-OES). The V2Ox-ZrO2/Al2O3 samples were first powdered, then weighed out quantitatively in an analytical balance and the masses were recorded. The samples were then dissolved in a H2SO4/HNO3 acid mixture (80[thin space (1/6-em)]:[thin space (1/6-em)]20 v/v) and transferred carefully to a 25 mL volumetric flask. The volume was made up with deionized water. Standard curves were generated each time with known concentrations of commercially available vanadium and zirconium ICP standards.
2.2.2 X-ray diffraction studies. The X-ray diffraction (XRD) patterns were measured on powdered samples using a Siemens D5000 powder X-ray diffractometer equipped with Cu-Kα radiation. The diffractograms were obtained using Cu-Kα radiation at a wavelength of 0.1542 nm.
2.2.3 H2/He TPR and XPS studies. Reduction studies of co-grafted V2Ox-ZrO2 samples were performed in a specially constructed apparatus (Fig. 1). A few pellets of the samples were loaded on a ceramic boat which was then inserted into a quartz tube (1” external diameter). The tube was fitted at both ends with Swagelok fittings containing graphite ferrules to make the setup withstand very high temperatures. It was then placed in a calcining oven where the reduction temperature was fixed and monitored. The inlet point of the quartz tube was connected to an assembly of stainless steel tubes controlled through valves so that the He and H2 gases coming out from the mass-flow controllers could be mixed and directed towards the calcining oven for the reduction to take place. The reduction was performed with two distinct gas compositions: a) 10% H2/He mixture, and b) 20% H2/He mixture. Samples were reduced for 6 h at temperatures of 500 °C, 600 °C, 700 °C and 750 °C using these two gas compositions. They were then carefully transferred and stored in the glove box under inert conditions before oxidation state determinations were made using XPS.
Schematic diagram of the TPR apparatus used for reduction studies.
Fig. 1 Schematic diagram of the TPR apparatus used for reduction studies.

3. Results and discussion

3.1 Studying the effects of various reaction conditions on surface loading with the two metal oxides

All the surface density values reported here were taken as an average of three measurements. The errors in the loadings of vanadium and zirconium obtained from all the time studies discussed were observed to be in a range of ± 2.2%–3.8%, in the precursor loading studies were observed to be in a range of ± 1.9%–4.2% and in the concentration studies were observed to be in a range of ± 1.3%–4.2%.
3.1.1 Variation of loading with time. Five different time samples were synthesized by varying the co-grafting time from 12 h to 72 h with an increment of 12 h between samples. All the co-grafting studies discussed here proceeded slowly with time. With this precursor combination OV(OEt)3 and Zr(OnBu)4, the highest loading of 0.97 ± 0.02 V nm−2 for vanadium was observed after 48 h of co-grafting and the highest loading for zirconium, 0.51 ± 0.01 Zr nm−2 was observed after 36 h respectively (Fig. 2). A significant increase in total metal loading (sum of the V and the Zr loadings in a particular time sample) was observed from the 24 h grafted sample (0.75 ± 0.01 M nm−2) to the 36 h grafted sample (1.39 ± 0.01 M nm−2) as compared to the increase observed from the 12 h grafted sample (0.54 ± 0.01 M nm−2) to the 24 h grafted sample (0.75 ± 0.01 M nm−2). The total metal loading increased steadily with time until 36 h when the loading of 1.39 ± 0.01 M nm−2 (0.88 V nm−2 + 0.51 Zr nm−2 = 1.39 M nm−2) was observed. Pursuing this co-grafting reaction for higher time periods indicated that the total metal loading remained constant at this value. However, it was also observed that the individual vanadium loading increased and the corresponding zirconium loading decreased further till 48 h of grafting, following which the respective loadings were more or less constant and a grafting-degrafting equilibrium was established on the surface. This grafting-degrafting equilibrium involves forming and breaking of the metal-oxygen bond(s) which anchor(s) the metal to the catalyst surface.
Change in loading as a function of time for (a) vanadium, (b) zirconium, and (c) total metal using 1 V nm−2 : 4 Zr nm−2 precursor ratio in the co-grafting solution.
Fig. 2 Change in loading as a function of time for (a) vanadium, (b) zirconium, and (c) total metal using 1 V nm−2[thin space (1/6-em)]:[thin space (1/6-em)]4 Zr nm−2 precursor ratio in the co-grafting solution.

An interesting aspect observed in this study was that the vanadium precursor grafted at a faster rate than the corresponding zirconium precursor. As mentioned earlier, a factor of four times greater precursor feed was used for zirconium compared to vanadium. Surprisingly, very low zirconium loading was observed on the catalyst surface and in most cases it was even lower than the corresponding vanadium loading. This observation is very important as it provides insight into the difference in the rate of grafting of the two metal oxide precursors.

3.1.2 Variation of loading with precursor feed when vanadium feed is fixed. The next study pursued was from the series f-V2Ox(fixed)-ZrO2/Al2O3 and involved varying the sample precursor feed of zirconium from 1.0–5.0 Zr nm−2 in the co-grafting solution with an increment of 1.0 Zr nm−2 feed between consecutive samples. The vanadium feed was kept fixed at 1.0 V nm−2. All the syntheses were carried out at room temperature for 24 h using 0.08 M (25 mL THF) co-grafting solution.

With OV(OEt)3 precursor, vanadium loading remained consistent at ∼ 0.97 V nm−2 for all feeds. The corresponding zirconium loading kept increasing with increase in total metal feed until a loading of 0.45 ± 0.02 Zr nm−2 was observed using a feed of 5.0 M nm−2. With higher feeds of the co-grafting solution, it was observed that the zirconium loading remained constant at this value. The total metal loading with this precursor increased with increase in feed until a loading of 1.42 ± 0.05 M nm−2 was observed using a feed of 5.0 M nm−2 (Fig. 3). This loading was very similar in value to that obtained at 36 h from the time study with this precursor, 1.39 ± 0.01 M nm−2. As this loading at 36 h was considered to be the one at equilibrium, we can conclude that in this study equilibrium was reached using a feed of 5.0 M nm−2. In this study too, similar low loadings of zirconium were observed compared to those of vanadium even on using a molar feed ratio of 1 V nm−2[thin space (1/6-em)]:[thin space (1/6-em)]4 Zr nm−2. This again confirmed the observation from the time studies that the vanadium precursor grafted at a faster rate than the corresponding zirconium precursor.


Change in surface loading as a function of precursor feed when vanadium feed is kept constant at 1.0 V nm−2 and zirconium feed is varied for (a) vanadium (b) zirconium and (c) total metal.
Fig. 3 Change in surface loading as a function of precursor feed when vanadium feed is kept constant at 1.0 V nm−2 and zirconium feed is varied for (a) vanadium (b) zirconium and (c) total metal.
3.1.3 Variation of loading with precursor feed when zirconium feed is fixed. In these studies, the vanadium feed was varied from 1.0–5.0 V nm−2 while the zirconium feed was kept constant at 1.0 Zr nm−2. As in the previous study, all syntheses were performed at room temperature using 0.08 M (25 mL THF) co-grafting solution.

In this study it was observed that the vanadium loading increased consistently with increase in precursor feed and reached a maximum loading of 3.16 ± 0.10 V nm−2 at a total metal feed of 6.0 M nm−2 (Fig. 4). In the same study, the zirconium loading remained more or less constant (0.48 ± 0.01 to 0.58 ± 0.01 Zr nm−2) with increase in total metal feed. The highest total metal loading (sum of individual V and Zr loading) of 3.50 ± 0.10 M nm−2 was observed at a total metal feed of 6.0 M nm−2.


Change in loading as a function of precursor feed when zirconium feed is kept constant at 1.0 V nm−2 and vanadium feed is varied for (a) vanadium (b) zirconium and (c) total metal.
Fig. 4 Change in loading as a function of precursor feed when zirconium feed is kept constant at 1.0 V nm−2 and vanadium feed is varied for (a) vanadium (b) zirconium and (c) total metal.

This experiment again indicates that vanadium grafting occurs faster than zirconium and the high vanadium loadings were consistently observed to corroborate this.

3.1.4 Variation of loading with precursor concentration and temperature. The next study in this series was pursued with the OV(OEt)3 as the vanadium precursor and Zr(OnBu)4 as the zirconium precursor. Precursor feeds of 1.0 V nm−2 for vanadium and a 4.0 Zr nm−2 for zirconium were used in the co-grafting solution and each sample was grafted for 24 h. The co-grafting solution volume was changed from 5 mL–25 mL in 5mL increments while the amount of precursors taken remained constant. This resulted in a concentration range of the co-grafting solution from 0.08 M–0.4 M. The four different temperatures of study chosen were namely 25 °C, 40 °C, 60 °C, and 80 °C. The elemental analyses on these samples are shown in Fig. 5.
Change in loadings of vanadium, zirconium and total metal as a function of concentration of the co-grafting solution at: (a) 25 °C, (b) 40 °C, (c) 60 °C, and (d) 80 °C.
Fig. 5 Change in loadings of vanadium, zirconium and total metal as a function of concentration of the co-grafting solution at: (a) 25 °C, (b) 40 °C, (c) 60 °C, and (d) 80 °C.

Comparing zirconium loadings at four different temperatures, it was observed that the highest loading achieved decreased with increase in temperature. For instance, at 25 °C the highest loading of zirconium obtained was 1.90 ± 0.11 Zr nm−2 whereas at 80 °C it was 1.00 ± 0.12 Zr nm−2. At 25 °C and at 40 °C, high vanadium loadings of 0.95 ± 0.02 V nm−2 and 0.96 ± 0.02 V nm−2 were observed, respectively, even at low concentrations of 0.1 M of the co-grafting solution. Further increase of concentrations up to 0.4 M did not result in significant changes in the loadings. At 60 °C and 80 °C the loadings showed gradual increase with increase in concentration of the co-grafting solution and highest loading values of 0.90 ± 0.03 V nm−2 and 0.59 ± 0.04 V nm−2, were observed, respectively. The highest total metal loadings (sum of individual V and Zr loadings) observed with these precursors at 25 °C and 80 °C were 2.90 ± 0.12 M nm−2 and 1.89 ± 0.12 M nm−2, respectively.

3.2 Powder XRD studies

Powder X-ray diffractograms were recorded for the time sample t(12h)-V2Ox-ZrO2/Al2O3. If it is assumed that the grafting technique produces homogeneously dispersed single sites on the support surface, the surface V and Zr oxide species should appear amorphous in nature. The diffractograms observed for powdered γ-Al2O3 and this sample were identical suggesting that the material produced by the grafting process displayed an amorphous nature (Fig. 6). However, due to the detection limit of XRD, these observations do not exclude the possibility of crystallites with sizes smaller than 4.0 nm being present on the surface.
Powder XRD spectra of bare γ-Al2O3 sample and a grafted time sample, t(12h)-V2Ox-ZrO2/Al2O3.
Fig. 6 Powder XRD spectra of bare γ-Al2O3 sample and a grafted time sample, t(12h)-V2Ox-ZrO2/Al2O3.

3.3 H2/He TPR and XPS studies

TPR and XPS were used together to identify the vanadium and zirconium oxidation states present under conditions comparable to those used for catalytic production of hydrogen from steam reforming of hydrocarbons and alcohols. The experimental procedures were devised to determine the oxidation states of vanadium and zirconium present following oxidizing (calcined) or reducing procedures.

All the prepared V2Ox-ZrO2 catalysts appeared yellowish white following calcination. The intensity of color is a very helpful optical indicator of the loading and agrees qualitatively with the data obtained through elemental analysis.

3.3.1 XPS studies of calcined samples. A previous report on vanadium grafting indicated that after calcining the catalyst samples, tetrahedral “single sites” were formed on the surface.10 TPR studies were not necessary in identifying the oxidation state of the calcined samples (calcined at 500 °C) because they were expected to be entirely in the highest oxidized states (V5+ and Zr4+) and were therefore subjected to XPS studies directly. One sample from each of the time, precursor feed, and concentration studies were chosen. The time sample, the precursor feed sample, and the concentration sample chosen for this study were t(24h)-V2Ox-ZrO2/Al2O3, f-V2Ox(fixed)-(2.0)ZrO2/Al2O3 and c(25mL)-T(25 °C)-V2Ox-ZrO2/Al2O3, respectively. CasaXPS software was used for peak fitting and all the peaks and the standard deviations in the peak positions were generated using Monte Carlo simulations. All the sample peaks in this study lay in the B.E. range of 517.00 eV–518.80 eV and could be peak-fitted entirely with only one Shirley fit. This confirmed the presence of only the V5+ species in all the three samples. The corresponding zirconium peaks also confirmed the presence of only the Zr 4+ species (Fig. 7). The percentage concentration of the V5+, Zr4+ and Al3+ (from the alumina substrate) species in all the three samples have been compiled in Table S1 (See ESI). From and Table S1 is clearly evident that the Al3+ species from the support material is the predominant species and depending on which sample is studied, the Zr4+ and V5+ conc. vary. It can also be predicted that if other samples from each group (time, feed, conc.) were calcined and characterized with XPS, the same V5+ and Zr4+ species would be obtained but the percentage composition of the respective species would be different depending on the metal loadings of the samples.
XPS Spectra of the three calcined samples (a) V2p spectra of the time sample, (b) V2p spectra of the precursor feed sample, (c) V2p spectra of conc. sample, (d) Zr 3d spectra of the time sample, (e) Zr 3d spectra of the precursor feed sample, and (f) Zr-3d spectra of the concentration sample.
Fig. 7 XPS Spectra of the three calcined samples (a) V2p spectra of the time sample, (b) V2p spectra of the precursor feed sample, (c) V2p spectra of conc. sample, (d) Zr 3d spectra of the time sample, (e) Zr 3d spectra of the precursor feed sample, and (f) Zr-3d spectra of the concentration sample.
3.3.2 TPR and XPS studies of reduced samples. For the reduction studies of samples synthesized using OV(OEt)3 and Zr(OnBu)4, the three samples mentioned above: t(24h)-V2Ox-ZrO2/Al2O3, f-V2Ox(fixed)-(2.0)ZrO2/Al2O3 and c(25mL)-T(25 °C)-V2Ox-ZrO2/Al2O3, were chosen again. The XPS spectra of all three samples reduced at 600 °C using 10% H2/He gas mixture are presented in Fig. 8. After data collection, asymmetry was observed in the vanadium peaks indicating the presence of multiple oxidation states. When these peaks were fit using CasaXPS software, a mixture of the oxidation states V5+ and V4+ was observed. However, no asymmetry was observed on the zirconium spectra and each of them could be fitted using a single peak fitting which indicated the presence of the Zr4+ species. The percentage concentration of the V5+, V4+, Zr4+ and Al3+ (from the alumina substrate) species in all the three samples reduced under different temperatures have been compiled in Table S2 (See ESI). From Table S2 it is clearly evident that the Al3+ species from the support material is the predominant species. It is also observed that the percentage composition of the Zr4+ for one particular sample remains more or less similar irrespective of the temperature of reduction. In most samples, the percentage concentration of the reduced species, V4+ is observed to increase with increase in temperature while the concentration of the V5+ species decreases simultaneously. It can also be predicted that if other samples from each group (time, feed, conc.) were reduced similarly and characterized with XPS, the same V5+, V4+ and Zr4+ species would be obtained and the percentage composition of the reduced species, V4+ would increase with rise in temperature of reduction. This increase in the V4+ percentage concentration with rise in temperature also indicates the possibility of higher hydrogen conversion at elevated temperatures.
XPS Spectra of the three reduced samples (a) V2p spectra of the time sample, (b) V2p spectra of the precursor feed sample, (c) V2p spectra of conc. sample, (d) Zr 3d spectra of the time sample, (e) Zr 3d spectra of the precursor feed sample, and (f) Zr-3d spectra of the concentration sample.
Fig. 8 XPS Spectra of the three reduced samples (a) V2p spectra of the time sample, (b) V2p spectra of the precursor feed sample, (c) V2p spectra of conc. sample, (d) Zr 3d spectra of the time sample, (e) Zr 3d spectra of the precursor feed sample, and (f) Zr-3d spectra of the concentration sample.

4. Conclusions

The syntheses and characterization studies of the V2Ox-ZrO2 mixed oxides using co-grafting solutions of OV(OEt)3 and Zr(OnBu)4 lead to a number of conclusions. In the times studies initially the total metal loading increased steadily with time till 36 h. Pursuing this co-grafting reaction for higher time periods indicated that the total metal loading remained constant at this value and that a grafting-degrafting equilibrium was established on the surface. However, the most interesting aspect observed in this study was that all the vanadium precursors grafted at a faster rate than the corresponding zirconium precursor even though four times more molar feed was used for the latter in the co-grafting solution. Similar observations were also made in the precursor feed studies and these consistent observations provide insight into the difference in the rate of grafting of the two metal oxide precursors.

In the studies performed at different temperatures it was observed that the total metal loadings decreased with increase in temperature. It is possible that this decrease was an effect of some degrafting process taking place on the surface. This decrease of loading with increase in temperature indicates that the co-grafting process has a negative ΔH value and thus is exothermic in nature. These results are consistent with previous studies of zirconium grafting, in which it was reported that zirconium loading decreased with increasing temperature indicating the exothermic nature of the grafting process.11

XRD studies showed that even after co-grafting, the support surface appeared to be amorphous in nature, thereby confirming the success of the co-grafting process. Combined TPR and XPS studies helped to identify the vanadium and zirconium oxidation state(s) in different samples. It was observed that in the calcined co-grafted samples, V5+ was the only observed vanadium species. This observation was anticipated as the calcination was carried out under aerobic conditions. On performing reduction studies, using 10% H2/He mixtures on the same samples at various temperatures, a mixture of V5+ and V4+ species were observed for vanadium. It was also observed that with increase in reduction temperature the percentage composition of the reduced species (V4+) increased while that of V5+ decreased. Irrespective of the fact that the sample was calcined or reduced, only one oxidation species of zirconium, Zr4+, was observed. The observations in this research suggest that zirconium oxides are essential for the successful functioning of these catalysts. Vanadium oxide “single sites” are known to decompose above temperatures of 600 °C and the supported zirconium oxides in all probability play a very significant role in holding the vanadium “single-sites” intact at the typically higher temperatures of catalytic operation resulting in the longevity of the catalyst.

Nomenclature

Co-graftingA catalyst loading technique, which typically utilizes a direct chemical reaction between two or more metal precursors and the surface hydroxyl groups of the support material.
FeedThe theoretical maximum loading possible with the available quantity of grafting precursor.
LoadingThe actual amount of metal precursor deposited on the support surface.

Acknowledgements

The authors gratefully acknowledge the support of the Department of Chemistry, University of Iowa. The assistance of Dr Dale Swenson, Dr Sarah Larsen, and the staff at the Central Microscopic Facility at University of Iowa for equipment use is much appreciated.

References

  1. D. P. Eyman, Int. Patent. Appl. 04/004827, 2004 Search PubMed.
  2. M. Agnelli, P. Louessard, El. Mansour, A. J. P. Candy, J. P. Bournonville and J. M. Basset, Catal. Today, 1989, 6, 63 CrossRef CAS.
  3. F. Z. Bentahar, J. P. Candy, J. M. Basset, F. Le Peltier and B. Didillon, Catal. Today, 2001, 66, 303 CrossRef CAS.
  4. J.-P. Candy, B. Didillon, E. L. Smith, T. B. Shay and J.-M. Basset, J. Mol. Catal., 1994, 86, 179 CrossRef CAS.
  5. F. Humblot, J. P. Candy, F. Le Peltier, B. Didillon and J. M. Basset, J. Catal., 1998, 179, 459 CrossRef CAS.
  6. F. Humblot, F. Lepeltier, J. P. Candy, J. Corker, O. Clause, F. Bayard and J. M. Basset, J. Am. Chem. Soc., 1998, 120, 137 CrossRef CAS.
  7. Y. W. Zheng, X. Fu and K. Wei, Appl. Organomet. Chem., 2007, 21, 836 CrossRef CAS.
  8. J. F. Faudon, F. Senocq, G. Bergeret, B. Moraweck, G. Clugnet, C. Nicot and A. Renouprez, J. Catal., 1993, 144, 460 CrossRef CAS.
  9. G. J. Siri, M. L. Casella, G. F. Santori and O. A. Ferretti, Ind. Eng. Chem. Res., 1997, 36, 4821 CrossRef CAS.
  10. T. Kim and I. E. Wachs, J. Catal., 2008, 255, 197 CrossRef CAS.
  11. L. Xu, PhD. Thesis, Department of Chemistry, University of Iowa, Iowa City, Iowa, 2004 Search PubMed.

Footnotes

Electronic supplementary information (ESI) available: Additional discussion sections, surface density plots and XPS spectra of TPR studies. See DOI: 10.1039/c2cy20039b
Present address: 6E Fenske Laboratories, Department of Chemical Engineering, Pennsylvania State University, State College, PA 16801, USA.

This journal is © The Royal Society of Chemistry 2012
Click here to see how this site uses Cookies. View our privacy policy here.