Monometallic Catalysts

Methods of catalyst preparation continue to receive attention: detailed studies were reported on the interaction of Pt(acac)₂ (1) and of Rh(π-allyl)₃ (2) with common oxidic supports such as alumina, silica and titania, various intermediates in the conversion of these complexes to metallic particles being detected spectroscopically. Koningsberger and colleagues (3) reported a most interesting study of the effect of reduction conditions on the structure and properties of supported metal particles. By the use of EXAFS (Extended X-ray Absorption Fine Structure) they found that at low reduction temperatures there was a layer of chemisorbed hydrogen between the metal (platinum) and the support, and this was irreversibly removed on reduction at 450°C. The capacity of the surface platinum atoms to chemisorb hydrogen was then lowered, because, according to the authors, there was a permanent change in the structure of the metal-support interface, which in turn changed the electronic properties of the metal particles.

Other questions concerning metal-support interactions, and the use of novel and unusual supports, continue to generate interest. Until recently the use of basic supports was not much favoured; it is however known that platinum supported upon magnesia converts alkanes to aromatics with good selectivity, but it is now reported that palladium supported upon magnesia catalysts perform even better (4). This system was also investigated by a “model” technique, that is, condensing palladium vapour onto a magnesium (100) surface (5); the palladium particles had the unique shape of a truncated square pyramid. A first attempt has been made to use buckminsterfullerene as a support for a metal (6); it was first deposited onto a support of higher area, and then Ru(acac)₃ was used as a source of metal. The resulting catalyst was used for hydrogenation of cinnamaldehyde, giving selec-tivities to cinnamyl alcohol of about 80 per cent.

The so-called “strong metal-support interaction” still fascinates catalytic chemists. The modification of titania by incorporating higher-valent cations, such as W⁶⁺, produces a support for a rhodium catalyst which is twenty times more active for carbon monoxide hydrogénation than is the unmodified support (7). A method for preparing mesoporous platinum supported upon titania aerogels from Ti(OⁱBu)₄ + PtCl₄ or Pt(acac)₂ was reported (8); the product contained anatase particles about 8 nm in size, and had an area of 200 m²/g. Further evidence for the encapsulation model was obtained by ion-scattering spectroscopy of rhodium supported on titania model catalysts (9).

Catalyst Characterisation

In addition to the methods alluded to above, a number of comparatively novel methods were described for characterising the structure of metallic catalysts. EXAFS is finding ever wider application, but the theoretical interpretation of the observed oscillations is still not firmly established, and failure to allow for the anharmonicity of the motion of atoms in the surface may lead, especially with lighter elements, to an underestimation of the particle size (10). ¹H NMR of hydrogen adsorbed on supported metal catalysts is another rich source of information; when applied to palladium catalysts in conjunction with measurements of uptake, it is capable of distinguishing between the strongly-chemisorbed + α-hydride phase and β-hydride phase (11). Metal surfaces are known to be labile and to reconstruct to minimise surface free energy; but what is surprising is that the background pressure in the residual vacuum under UHV conditions is sufficient to effect an adsorbate-induced reconstruction observable by Scanning Tunnelling Microscopy (12).

Bimetallic Catalysts

The preparation of small, homogeneous alloy particles on supports suitable for catalytic applications still poses a number of difficulties. One approach is to use as precursor a compound in which the desired elements are linked; for example, the compound (p -tolyl)₄Cu₂Rh₂ was used to prepare supported CuRh particles which showed useful properties in liquid-phase reactions (13). Similarly the binuclear acetate-bridge bridged complex PdCu(OAc)₄ has been employed to make Pd-Cu/Al₂O₃ catalysts (14). Pt-Au/SiO₂ catalysts can be made by reacting a reduced platinum on silica catalyst with HAuCl₄, a process that can lead either to the random deposition of gold atoms on the platinum particles, or to a specific coverage of low-index planes, depending on the conditions used (15). Conventional impregnation of silica with solutions of the chlorides of palladium and gold produces materials, the homogeneity of which is uncertain, but careful X-ray Photoelectron Spectroscopy work reveals the presence of Si° close to the small alloy particles after reduction at only 380°C (16).

The ability of the platinum group metals to form sols of finely divided metal particles has been known for many years, but only recently have bimetallic sols been studied. Bimetallic palladium-copper sols, which are supported on carbon are effective catalysts for hydrogenating alkynic alcohols (17).

Applications as Catalysts

Mention of metal sols as supported metal catalyst precursors leads us naturally to their use in unsupported form; rhodium particles prepared under phase-transfer conditions show high stereoselectivity for reducing ortho -disubstituted aromatic rings to cis -disubstituted cyclohexanes (18). Crown ethers containing benzene rings fused into the macrocycle can be similarly reduced. A most interesting innovation is the use of an osmium tetroxide complex with an alkaloid, such as cinchonidine, incorporated in a polymeric matrix, as a catalyst for enan-tioselective di-hydroxylation of alkenes (19). Chiral palladium complexes containing neutral bidentate optically-active nitrogen ligands effect enantioselective allylic substitutions.

One symposium was devoted to “Catalysis and the Environment”, and not surprisingly the selective reduction of nitrogen oxides featured in a number of papers. Although vanadia-tita-nia catalysts and copper-exchanged ZSM-5 zeolite are effective, attention is now turning to lean-burn conditions, where catalysts based on the platinum group metals have potential (20). There were also several studies reported of the kinetics and mechanism of the nitric oxide-carbon monoxide reaction. On rhodium single crystal surfaces, nitric oxide is strongly chemisorbed and blocks access of carbon monoxide to the surface: their reaction is therefore slower than is the oxidation of carbon monoxide (21). Bimetallic combinations involving palladium were compared with rhodium; the palladium-chromium showed comparable activity, but palladium-silver was less active (22).

Fundamental investigations enlarging our understanding of catalysis were comparatively few in number. The use of transient kinetic methods to study the methane-deuterium equilibration on Rh-Ag/SiO₂ catalysts revealed that the desorption rate of C₁ species was faster than over rhodium supported on silica, and the strength of hydrogen chemisorption less (23). The reasons underlying the dependence of activity on the size of metallic particles have long been questioned by catalytic scientists. It now appears possible that the effect may have an energetic rather than a geometric explanation, for a particle-size dependence of reaction kinetics in alkane hydrogenolysis has been traced to differences in hydrogen chemisorption strength (24).

The concept of using a semi-permeable membrane to overcome equilibrium limitations, for example, in dehydrogenations, is also not new, but no practical application on any scale has yet resulted: attempts to develop usable systems are however continuing. Palladium-silver membranes used for isobutane dehydrogenation are unfortunately quickly poisoned (25), but efforts to construct membranes from platinum supported on alumina are showing promise (26, 27).

Conclusion

It has only been possible to mention a few of the highlights of the papers presented at EUROPACAT-I, and scarcely any of the many posters which featured the platinum group metals. However, it is clear that they continue to dominate the catalytic scene, and that their efficiency in very many applications remains unchallenged. EUROPACAT-II will be held in Maastricht in September 1995: put that date in your diary now.