A Study of Propane and Propene Ammoxidation over the Mo-V-Nb-Te-Oxide System

Abstract: The ammoxidation of propane is commonly regarded as being the future route for producing acrylonitrile, and the Mo-V-Nb-Te-oxide system studied here has thus far been seen as the most promising candidate for achieving this. Yields of up to 62% have been presented very elegantly in the patent literature. A great disadvantage has been, however, that the crystalline quality of the catalyst tends to be highly dependent on the synthesis conditions. The system taken up here consists of two phases, termed M1 and M2. It was shown to be highly important when aiming for the M1 phase in the structure to keep the mixing and drying time short, whereas samples involving solely the M2 phase are more easily synthesised. It was also shown that in order to stabilize tellurium and exclude the rutile phase, it is important to calcine the precursors in air at 275°C prior to calcination in an inert gas. In order to achieve the highest yield possible it is important that both phases are present in the catalyst. Ammoxidation of propane was shown to be a consecutive reaction passing over propene. It was also found that only the M1 phase is capable of converting propane to propene, whereas the M2 phase is more selective in the ammoxidation of propene than the M1 phase is. The symbiosis between the two phases when physical mixtures of the M1 and M2 phase are prepared was demonstrated. It was concluded, however, that for this symbiosis to take place, the two phases have to be intimately mixed, since the propene formed cannot diffuse over any great distance without readsorbing on the M1 phase. Regarding the various elements in the phases, it is well established that vanadium is the propane-activating element, the present results concurring with this. In addition, results obtained for the M2 catalysts strongly indicate molybdenum and tellurium to be the active species in propene ammoxidation. It was also shown that tungsten could fully replace molybdenum in the M2 phase, whereas only 50% of the vanadium and 30% of the tellurium could be replaced by titanium and by cerium, respectively. Niobium could substitute for both molybdenum and vanadium. In addition iron could replace vanadium. Niobium and titanium were found to enhance the activity, but without affecting the selectivites. Both the substitution of cerium and especially that of tungsten were found to improve the selectivity to acrylonitrile at the expense of acrolein. (Mo0.3W0.7)6V3.23Te3.52Ox was found to be the best catalyst, having a selectivity to acrylonitrile from propene of 80%. Catalysts were prepared in sol-diluted and impregnated form, using SiO2, Al2O3 and TiO2 as supports. Dry-impregnation was found to give neither M1 nor M2 nor any active or selective catalysts. The sol-diluted catalysts were found to perform better. Silica dilution gave catalysts with pure M1 and M2 phases. Dilution with Al2O3 did not yield these phases, but rather Al2(MoO4)3, AlVO4, TeMo5O16 and NbO2. Titania dilution of the nominal M1 composition produced Mo5-x(V/Nb/Te)xO14, whereas the nominal M2 and M1/M2 nominal compositions yielded phases of the correct type. The catalysts diluted by SiO2 were similar in performance to the neat catalysts but were less active, whereas the Al2O3-diluted samples were inferior to the neat catalysts. The selectivities of the samples diluted with TiO2 and consisting of the correct phases were comparable to the corresponding neat samples, but not very active for propane.