Bulk Mixed Oxide Catalysts

Strength (FLS) empirical approach are discussed in Section 3 as methods for determining the molecular structures of metal-oxide species from their Raman spectra. The state-of-the-art in Raman instrumentation as well as new instrumental developments are discussed in Section 4. Sampling techniques typically employed in Raman spectroscopy experiments, ambient as well as in situ, are reviewed in Section S. The application of Raman spectroscopy to problems in heterogeneous catalysis (bulk mixed-oxide catalysts, supported metal-oxide catalysts, zeolites, and chemisorption studies) is discussed in depth in Section 6 by selecting a few recent examples from the literature. The future potential of Raman spectroscopy in heterogeneous catalysis is discussed in the fmal section. 

Bulk Mixed Oxide Catalysts. - Raman spectroscopy of bulk transition metal oxides encompasses a vast and well-established area of knowledge. Hie fundamental vibrational modes for many of the transitional metal oxide complexes have already been assigned and tabulated for systems in the solid and solution phases. Perhaps the most well-known and established of the metal oxides are the tungsten and molybdenum oxides because of their excellent Raman signals and applications in hydrotreating and oxidation catalysis. Examples of these two very important metal-oxide systems are presented below for bulk bismuth molybdate catalysts, in this section, and surface (two-dimensional) tungstate species in a later section. 

Courty, P. Marcilly, C. A Scientific Approach to the Preparation of Bulk Mixed Oxide Catalyst. In Proceedings of the Third International Symposium Scientific Bases for the Preparation of Heterogeneous Catalysts Poncelet, G., Grange, P., Jacobs, P.A., Eds. Elsevier Amsterdam, 1983 485-519. 

Figure 2. Different methods of synthesis of bulk mixed oxide catalysts. ...

Alternatively, we have attempted the molecular design of mixed-oxide catalysts by using crystalline mixed oxides whose bulk structures are known and whose potential for practical use is good. Heteropoly compounds, perovskites, and zeolites are the candidate catalysts. 

The first Raman spectra of bulk metal oxide catalysts were reported in 1971 by Leroy et al. (1971), who characterized the mixed metal oxide Fe2(MoC>4)3. In subsequent years, the Raman spectra of numerous pure and mixed bulk metal oxides were reported a summary in chronological order can be found in the 2002 review by Wachs (Wachs, 2002). Bulk metal oxide phases are readily observed by Raman spectroscopy, in both the unsupported and supported forms. Investigations of the effects of moisture on the molecular structures of supported transition metal oxides have provided insights into the structural dynamics of these catalysts. It is important to know the molecular states of a catalyst as they depend on the conditions, such as the reactive environment. 

The future of Raman spectroscopy in the research and the development of catalysts appears to be extremely promising. The recent revolution in Raman instrumentation has dramatically increased the ability to detect weak Raman signals and to collect the data in very short times. Thus, it is now possible to perform real-time Raman analysis and to study many catal) c systems that give rise to unusually weak Raman signals. The enormous strides in Raman instrumentation now allow for the characterization of a wide range of catalytic materials bulk mixed oxides, supported metal oxides, zeolites, supported metal systems, metal foils, as well as single crystal surfaces. Few Raman studies have been reported for sulfides, nitrides, or carbides, but these catalytic materials also give rise... 

Other mechanisms that could also explain the detrimental interaction between the two catalysts include 1) the solid state reaction between y-alumina and ZnO from the methanol catalyst to form bulk mixed oxides, and 2) the formation of physical agglomerates from the fines of the two catalysts. These are under investigation. 

In principle it would seem reasonable that the bulk structure and surface properties of a solid would influence the catalytic performance. Verification of this view and an assessment of its importance may be more significant than first appears since it incorporates an implication that catalytic preparation should be designed to achieve the bulk structure and surface properties that give the optimized catalytic performance. Materials which have been shown to catalyze the conversion of propylene to acrolein have included metal oxides, mixed oxides, and more lately the multicomponent catalysts. A consideration of all these solids would require the assessment of numerous data and speculation. However, the mixed oxide catalysts have been associated with many of the more recent investigations of the course of catalytic oxidation, and these catalysts therefore seem to be worthy of detailed consideration. 

Although several mixed oxide catalysts have been developed commercially for the selective oxidation of propylene, the investigation of their fundamental physical and chemical properties has resulted in only a slow and steady accumulation of information. It also appears that attempts to correlate data from different investigations have frequently resulted in unsatisfactory interpretations. It seems that some of this uncertainty arises from correlations between results obtained from different catalysts subjected to different pretreatments and assessed under different evaluation conditions. Hence, the comprehensive description of the bulk and surface properties of a single catalyst, their interdependence, and their influence on catalytic performance is in most cases quite unclear. 

The industrial catalyst for n-butane oxidation to maleic anhydride (MA) is a vanadium/phosphoras mixed oxide, in which bulk vanadyl pyrophosphate (VPP) (VO)2P207 is the main component. The nature of the active surface in VPP has been studied by several authors, often with the use of in situ techniques (1-3). While in all cases bulk VPP is assumed to constitute the core of the active phase, the different hypotheses concern the nature of the first atomic layers that are in direct contact with the gas phase. Either the development of surface amorphous layers, which play a direct role in the reaction, is invoked (4), or the participation of specific planes contributing to the reaction pattern is assumed (2,5), the redox process occurring reversibly between VPP and VOPO4. 

Elemental and Structural Characterization Many oxidation reactions occur on mixed oxides of complex composition, such as SbSn(Fe)0, VPO, FePO, heteropolycompounds, etc. Very often the active surfaces are not simple terminations of the three dimensional structure of the bulk phases. There is need to extensively apply structural characterization techniques to the study of catalysts, if possible in their working state. 

AHt can be calculated, in principle, from thermochemical data. It is then necessary to take into account the variable valency of most metals and to fix the different oxidation states which occur during stationary or non-stationary reaction conditions. Some difficulties with this method are th scarcity of data for mixed oxides, the difference in conditions between those on the surface of the catalyst and those in the bulk and the inaccuracy of a number of data obtained by measuring differences in AH. 

Johnson and co-workers (62) have come to the conclusion that interaction of lead with Pt crystallites results in the formation of an inactive phase in which the Pt atoms are ionized and soluble in HC1. These data were derived from engine tests, in which the catalysts were exposed to fuels with 0.03-0.1 g Pb/gal. The amount of crystalline Pt in these catalysts was smaller than in catalysts run on lead-free fuels. The authors indicate that noncrystalline forms of Pt are present on A1203 supports under certain conditions, and that lead stabilizes such forms. The question whether the noncrystalline, ionic Pt is a surface or a bulk phase remains unanswered. Bulk mixed Pt-Pb oxides have been described (98, 99), but, again, the dispersed forms of noble metals supported on A1203, which lead (and other elements) may stabilize, are known to be associated with the surface only. Palladium can be expected to form such noncrystalline dispersed phases to a still greater extent since it is more easily oxidized than Pt. 

---