The adsorption and reaction of methanol with iron molybdate catalysts of widely varying Mo/Fe ratio have been investigated using X-ray diffraction (XRD), Raman, temperature-programmed desorption (TPD), and pulsed flow reactor measurements. From these data, we can conclude that Mo is segregated to the surface of all catalysts, even that with only a 1:50 Mo/Fe ratio in the bulk of the sample. There is a very marked difference between the behavior of pure ferric oxide and catalysts with even a small amount of Mo present. The former produces only CO2 as the product of methanol oxidation with no formaldehyde production, yet with only small amounts of Mo in the preparation the sample is already highly selective for formaldehyde. TPD shows the reason for this. For iron oxide, the methanol adsorbs to form a formate intermediate, resulting in CO2 evolution, whereas after addition of Mo the main intermediate seen is the selective one, that is, the methoxy intermediate, which decomposes to produce formaldehyde in the gas phase upon heating. However, the best catalysts for the reaction are still those with a ratio of Mo/Fe of greater than 3:2, even though this is the ratio for single phase ferric molybdate, Fe2(MoO4)3. This is because, although the low Mo catalysts have high selectivity at low temperature, they decline at high temperature, producing mainly CO as the product. This effect is strongly Mo-loading dependent. The surface of all the Mo-loaded catalysts manifest the behavior of MoO3 itself, implying that the surface is greatly enriched in Mo, even for those with very low loadings.

The role of Mo in the selective oxidation reaction is considered in some detail, focusing on the selective oxidation of methanol to formaldehyde. The reaction mechanism and kinetics will be described. It is notable that Mo tends to segregate to the surface of iron molybdate catalysts, proven by scanning transmission electron microscopy and XPS, and so it dominates the surface, even at very low loadings. This is manifest in reaction data too: for instance, the selectivity to formaldehyde for a catalyst with only 20% Mo present is 50% at 50% conversion, whereas for pure iron oxide it is close to zero at all conversions. The active site for the reaction is Mo(VI), which cycles through Mo(IV) during the reaction. Mo(IV) itself is shown to be unselective for the reaction. Lattice oxygen in the material can readily re-oxidise the surface at temperatures above 300°C. Some of the mechanistic behaviour is analogous to the role of Mo in enzymatic processes, such as xanthine oxidation, and the two areas of catalysis by molybdenum are compared and contrasted. The role of different types of oxygen, such as 'lattice' oxygen, 'surface' oxygen, bridging and terminal oxygen species will be defined and clarified, and the modified Mars-van Krevelen description of the heterogeneous reaction will be given.

No abstract

Iron molybdate catalysts are used for the selective oxidation of methanol to formaldehyde. In this paper we have attempted to understand what determines high selectivity in this reaction system by doping haematite with surface layers of Mo by incipient wetness impregnation. This works well and the Mo appears to form finely dispersed layers. Even very low loadings of Mo have a marked effect on improving the selectivity to formaldehyde. Haematite itself is a very poor catalyst with high selectivity to combustion products, whereas, when only 0.25 monolayers of Mo are deposited on the surface, formaldehyde and CO selectivities are greatly enhanced and CO2 production is greatly diminished. However, even with as much as seven monolayers of Mo dosed on to the surface, these materials achieve much less selectivity to formaldehyde at high conversion than do the industrial catalysts. The reason for this is that the Mo forms a 'skin' of ferric molybdate on a core of iron oxide, but does not produce a pure Mo oxide monolayer on the surface, a situation which is essential for very high yields of formaldehyde.

The oxidation of methanol has been measured on MoO 3 and MoO 2 . The properties of these two materials are interchangeable, depending upon the conditions in which the reaction is run. MoO 3 produces high yields of formaldehyde, but MoO 2 does not, due to the importance of the Mo 6+ state for the selective reaction. However, if the MoO 3 material is run in anaerobic conditions it behaves in a very similar way to MoO 2 , due to the presence of Mo 4+ in the surface layers. In complement to this MoO 2 converts to high yield behaviour when run in aerobic conditions, due to the conversion of the material to Mo 6+ at the surface, and, ultimately to MoO 3 in the bulk. In TPD experiments MoO 3 yields formaldehyde, whereas MoO 2 yields CO. In both materials oxygen transport within the lattice becomes appreciable above 300°C, and the reaction proceeds via the Mars-van Krevelen mechanism.