Bismuth-cerium molybdates

Our recent work on the bismuth-cerium molybdate catalyst system has shown that it can serve as a tractable model for the study of the solid state mechanism of selective olefin oxidation by multicomponent molybdate catalysts. Although compositionally and structurally quite simple compared to other multiphase molybdate catalyst systems, bismuth-cerium molybdate catalysts are extremely effective for the selective ammoxidation of propylene to acrylonitrile (16). In particular, we have found that the addition of cerium to bismuth molybdate significantly enhances its catalytic activity for the selective ammoxidation of propylene to acrylonitrile. Maximum catalytic activity was observed for specific compositions in the single phase and two phase regions of the phase diagram (17). These characteristics of this catalyst system afford the opportunity to understand the physical basis for synergies in multiphase catalysts. In addition to this previously published work, we also include some of our most recent results on the bismuth-cerium molybdate system. As such, the present account represents a summary of our interpretations of the data on this system. 

Bismuth cerium molybdates were prepared by coprecipitation using aqueous solutions of (NH ) Mo 02, (NH,)2Ce(N0 ), and Bi(NO ) 5H2O. The catalysts were supported on oiO (20% by weight) using an ammonium stabilized silica sol. Samples for diffraction analysis were unsupported. Samples were calcined in air at 290 and 425°C for three hours each followed by 16 hours at 500, 550, or 600 C. X-ray powder patterns were obtained using a Rlgaku D/Max-IIA X-ray diffractometer using Cu K radiation. 

Physically, the relationship between catalytic activity and Z f can be understood from a study of single phase bismuth cerium molybdate solid solutions. The results show that maximum activity is achieved when there exists a maximum number and optimal distribution of all the key catalytic components bismuth, molybdenum and cerium in the solid. Therefore, it reasonably follows that the low catalytic activity observed for the two phase compositions where Af Af(min) results from the presence of interfacial regions in the catalysts where the compositional uniformity deviates significantly from the equilibrium distribution of bismuth and cerium cations present in the solid solutions. These compositions may contain areas in the interfacial region which are more bismuth-rich or cerium-rich than the saturated solid solutions. Conversely, at Af(min), the catalyst is similar to an ideal mixture of the two optimal solid solutions. The compositional homogeneity of the interfacial region approaches that of the saturated solid solutions. Therefore, the catalytic behavior of compositions at Af(min) is similar to that of the saturated solid solutions. 

As in the bismuth rich structure discussed above, the Bi atoms in Bi doped cerium molybdate are not randomly distributed on the Ce sites. The compositions of M(l), M(2), and M(3) (see Table II) are 86%Ce and 14%Bi, 92%Ce and 8%Bi, 73%Ce and 27%Bi, respectively. This compositional difference is reflected in the M(3)-0 distances as well. Note that the difference between the maximum and minimum M(3)-0 distances is larger for Bi doped cerium molybdate than lanthanum molybdate. Evidently, a 27% occupation of Bi (with the attendent lone pair) is sufficient for a small but noticable distortion in the average local 0 environment. Summarizing the structural results on Ce incorporation into bismuth... 

The maximum in catalytic activity observed for the multiphase region of the phase diagram necessarily arises from interactions between the separate phases. The bismuth rich and cerium rich solid solutions can readily form coherent interfaces at the phase boundaries due to the structural similarities between the two phases which can permit epitaxial nucleation and growth. A good lattice match exists between the [010] faces of the compounds, this match is displayed in Figure 6. We have also shown that regions of an [010] face of a Ce doped bismuth molybdate crystal resembles cerium molybdate compos tionally. This means that the interface between the two compounds need not have sharp composition gradients. It is structurally possible for the Bi-rich phase to possess a metal stiochiometry at the surface that matches that of the Ce-rich phase. 

Application of Raman spectroscopy to a study of catalyst surfaces is increasing. Until recently, this technique had been limited to observing distortions in adsorbed organic molecules by the appearance of forbidden Raman bands and giant Raman effects of silver surfaces with chemisorbed species. However, the development of laser Raman instrumentation and modern computerization techniques for control and data reduction have expanded these applications to studies of acid sites and oxide structures. For example The oxidation-reduction cycle occurring in bismuth molybdate catalysts for oxidation of ammonia and propylene to acrylonitrile has been studied in situ by this technique. And new and valuable information on the interaction of oxides, such as tungsten oxide and cerium oxide, with the surface of an alumina support, has been obtained. 

The Bi2 Ce Mo-O. two phase system has been examined for its activity in the catalytic oxidation of propylene to acrylonitrile. The two phases have been characterized as a solid solution of Bi in cerium in bismuth molybdate. Results of studies have been correlated with... 

---