The Effect of Using H4P2O7 as Phosphorus Source for Synthesizing Vanadyl Pyrophosphate Catalysts

Vanadyl pyrophosphate (VO)2P2O7 catalysts synthesized via VOPO4·2H2O were investigated by using BET surface area measurement, X-ray Diffraction (XRD), Scanning Electron Microscope (SEM), and Temperature-Programmed Techniques (TPD and TPRS). H3PO4 and H4P2O7 were used as the phosphorus source. Only pyrophosphate phase was observed for both final catalysts after 75 hours of calcination in a reaction flow of n-butane/air mixture (0.75% n-butane/air). However, catalyst derived from H4P2O7 based preparation (denoted VPDpyro) exhibit better crystallinity and slightly higher BET surface area compared to the H3PO4 based preparation (denoted VPDortho). The nature of the oxidants for both catalysts was investigated by O2-TPD. For VPDpyro, TPD showed an oxygen peak maximum at 986 K and a shoulder at 1003 K, whereas for VPDortho, the oxygen was desorbed as two peaks maxima at 966 and 994 K. The total amount of oxygen desorbed thermally from VPDpyro (3.60×1020 atom×g-1) is higher than that obtained for VPDortho (3.07×1020 atom×g-1). VPDpyro displayed a slightly improved activity and selectivity for n-butane oxidation. A proper amount of V5+ species may have an effect on the enhancement of the catalytic activity.


Introduction
Vanadyl pyrophosphate, (VO) 2 P 2 O 7 catalysts currently are the most active phase to transform the only alkane used commercially in selective oxidation process from n-butane to maleic anhydride (MA). Maleic anhydride is an important material for manufacturing unsaturated polyester resins, agricultural chemical, food additives, lubricating oil additives and pharmaceuticals [1].
Most of the studies reported that (VO) 2 P 2 O 7 was obtained from the precursor VOHPO 4 ·0.5H 2 O, by calcination in a reaction condition at 673 K. This precursor played an important role in the transformation of the final catalyst [2]. Preparation of the precursor is the key to obtain a high performance catalyst [3,4,5,6]. The phosphorus sources employed for the synthesis of VOHPO 4 ·0.5H 2 O also reported to give a strong influence in the synthesis of vanadium phosphate catalysts [3].
The purpose of this paper is to investigate the physico-chemical properties of the vanadyl pyrophosphate, (VO) 2 P 2 O 7 catalyst prepared by using two different sources of phosphorus i.e.: H 3 PO 4 and H 4 P 2 O 7 . The catalytic activity of these catalysts will also be reported.

Catalysts Preparation
Via VOPO 4 ·2H 2 O and H 4 P 2 O 7 V 2 O 5 (15.0 g) and H 4 P 2 O 7 (69.9 g) mixture was refluxed in water for 4 hours. Then the resulting mixture (yellowish coloured) was recovered by centrifuged, following by filtration and washed with cold water. The solid obtained was then dried in air for 24 hours at 373 K and confirmed by XRD as VOPO 4  These resulting precursors were then calcined in a flow of n-butane/air mixture (0.75% n-butane in air) for 75 hours at 673 K. The final catalysts were denoted VPD pyro (using H 4 P 2 O 7 ) and VPD ortho (using H 3 PO 4 ).

Catalysts Characterization
The BET surface areas of the catalysts were measured by using nitrogen adsorption at 77 K. This was done by using ThermoFinnigan Sorptomatic Instrument model 1990.
The bulk average oxidation states of vanadium in the catalysts were determined by redox titration following the method of Niwa and Murakawi [7].
The X-ray diffraction (XRD) analyses were carried out using a Shimadzu diffractometer model XRD 6000 employing Cu K α radiation to generate diffraction patterns from powder crystalline samples at ambient temperature.
SEM was done using a Quanta 200 ESEM FEG electron microscope.
Temperature Programmed Desorption (TPD) of O 2 analysis was done by using a ThermoFinnigan TPDRO 1110 apparatus utilizing a thermal conductivity detector (TCD).
Temperature Programmed Reaction Spectroscopy (TPRS) profile was obtained by passing an n-butane/ He stream (2.0%, 101 kPa, 25 cm 3 ×min -1 ) over the catalyst (0.5 g) and raising the temperature from ambient to 1173 K at 5 K×min -1 . This was done by using ThermoFinnigan TPDRO 1100 apparatus connected via a heated capillary to a quadrupole MS (Pfeiffer).

BET surface area measurement and redox titration
The BET surface area for VPD ortho and VPD pyro are 23.18 and 25.40 m 2 ×g -1 , respectively. These values are higher compared to 12.0 m 2 ×g -1 as reported by Bartley et al. [3]. The average oxidation numbers of vanadium for both catalysts are given in Table 1. By using H 4 P 2 O 7 as the phosphorus source was found to decrease the average oxidation number from 4.26 to 4.18 i.e. a reduction of V 5+ oxidation state from 26 to only 18%. This observation is similar to those reported by Bartley et al. [3].

X-ray Diffraction
The XRD pattern of the precursor, Ortho pre ( Fig. 1) showed the characteristic peaks that matched perfectly with VOHPO 4 ·0.5H 2 O with peaks appeared at 2θ = 15.6°, 27.0° and 30.4° which correspond to (001), (121) and (220) reflections, respectively. However, the usage of H 4 P 2 O 7 leads to the domination of reflection at 2θ = 30° which indexed to the (220) plane. This observation is similar to the previous report [3]. The XRD patterns for VPD ortho and VPD pyro catalysts shown in Fig. 2 were matched perfectly with the patterns of a pyrophosphate phase. The main peaks observed at 2θ = 23.0°, 28.5° and 29.9° which correspond to (020), (204) and (221) Table 2 shows the line width of the (020) and (204) plane reflections. The parameter used to determined the crystallite size is the half width of the (020) peak. The line width increases with the decreasing size of the crystallite. The particle size of VPD ortho at (020) and (204) were calculated by using Debye-Scherrer equation [8] as 151.9 and 97.7 Å, respectively. However, the particle size for VPD pyro catalyst was decreased to 105.3 and 55.8 Å, respectively. raised to 1173 K following the conductivity of oxygen by a thermal conductivity detector (TCD).
VPD ortho gave two overlapped peaks at 966 and 994 K, whereas VPD pyro also gave two desorption peaks maxima at 986 and 1003 K. There are no desorption peaks for oxygen in the temperature range from 300 to 850 K which correspond to the specific site for chemisorbed molecular oxygen. The total amount of oxygen desorbed from the VPD ortho is 3.07×10 20 atom×g -1 ( Table 3). The usage of H 4 P 2 O 7 slightly increased the amount to 3.62×10 20 atom×g -1 i.e., an increment of ~15% of the total oxygen desorbed for VPD pyro . These are the lattice oxygen which also has been observed earlier and showed 100% selectivity towards partial oxidation of C 4 hydrocarbons [9].

Scanning Electron Microscope
The surface morphologies of VPD ortho and VPD pyro are shown in Fig. 3. Both catalysts show a platelets structure which closely packed together forming a rosette-shape cluster. VPD ortho showed a rosette type structure with a crystallite of uniformed sized however VPD pyro catalyst showed an improper arrangement of rosette cluster with more isolated platelet.

Temperature-Programmed Desorption of Oxygen
The oxygen desorption spectra shown in Fig. 4 were obtained by pretreating the catalyst by heating them to 673 K in oxygen flow (1 bar, 25 cm 3 ×min -1 ), holding them under that stream for 1 hour before cooling them to ambient temperature. Then the flow was switched to helium and the temperature was

Catalytic activity
Both samples exhibit catalytic activity with the production of intermediates for n-butane selective oxidation to maleic anhydride (MA), i.e. butene, butadiene and furan. The temperature profiles of Fig. 5 indicate that by replacement of H 3 PO 4 to H 4 P 2 O 7 exhibit minimal influence on the evolution of the catalytic function. VPD ortho developed activity at 674 K (as shown in the onset of furan production in Fig. 6 whereas VPD pyro requires a lower temperature at 588 K. A higher rate of butene production and a lower temperature peak maximum of butene (Fig. 5) shows that VPD pyro gave higher activity as compared to VPD ortho . This is in an agreement with the earlier finding [3]. Total amount of selective products for both catalysts as shown in Table 4 indicate that VPD pyro has a better selectivity. The most significant observation in the present study is that furan was observed as one of the in- Furan termediate found for this anaerobic partial oxidation of n-butane. This was not observed previously by using the same catalyst prepared using ''classical'' organic method [10]. The total amounts of CO 2 produced by these two catalysts are similar. However VPD ortho produced higher amount (24% more) of CO 2 at lower temperature peak (783 K) as compared to  Table 4 Percentage of selective and unselective products VPD pyro which may due to the increment of the V 5+ oxidation state in the catalyst. Previous study has shown that the presence of V 5+ phases can decrease the catalytic performances [11,12]. However, according to Abon et al. [13], the existence of V 5+ sites in the catalyst leads to the enhanced catalytic performances for n-butane oxidation. The vanadium oxidation state of VPD pyro in this study is 4.18 indicates that the presence of a small amount of V 5+ is essential for a higher selectivity.

Conclusions
The effect of using H 4 P 2 O 7 in replacement of H 3 PO 4 in the synthesised of (VO) 2 P 2 O 7 are: • A decreased of vanadium oxidation state from 4.26 to 4.18. • An increasing of the lattice oxygen desorbed thermally. • The morphology of VPD pyro showed the increasing of isolated platelet for each rosette shape clusters. • Higher activity and selectivity obtained from nbutane oxidation.