Catalysis on Well-defined Metal Surfaces and Non-metallic Substrates

BY S. J.THOMSON

1 Introduction

The previous review in this series was written with more emphasis on adsorption and techniques than on catalysis. With the passing of two years the number of papers devoted to catalysis has increased and this is reflected in the material in this review and in its arrangement. The papers chosen illustrate the direction in which the applications are proceeding; with limited space it cannot be exhaustive. It is hoped that the references given to books and to reviews will be of value to readers.

2 Changes in Surfaces Accompanying Adsorption and Catalysis

A surprising variety of changes occur in the surfaces of catalysts. There are intentional changes brought about through the use of modifying agents or poisons and there are inevitable changes which occur during the use of catalysts. Chemical intuition would also lead to the expectation of segregation of surface species, corrosive chemisorption, and adsorption-induced surface enrichments. There is in addition a range of more subtle and unexpected changes which might be collected under the heading of surface rearrangements.

Modifying Agents and Poisons. — Marbrow and Lambert have studied O2 adsorption on Na-modified Ag(110) surfaces. Group I and Group II metals modify catalytic properties of metal surfaces and are particularly important in the selective oxidation of ethylene to ethylene oxide over silver. LEED, Auger, and TPD were used in the study. Na could be adsorbed on Ag(110) to θ ~ 1 to form a (1 x 1) monolayer; dosing to the two-monolayer level produced a (1 x 2) LEED pattern. Thermal desorptions were studied and it was noted that some Na diffused into sub-surface layers. The main feature of the work with relevance to catalysis is that whereas O2 has a low sticking probability on Ag(110), if surface Na is present there is an enormously enhanced probability of O2 adsorption. There was also evidence for diffusion of O into the lattice, in that thermal desorption failed to show release of O2. Na atoms in the (1 x 1) structure were probably in simple register with Ag atoms. When sub-surface Na was present, added Na below the monolayer level produced a (1 x 2) LEED pattern. When Na-dosed specimens were exposed to O2, a (4 x 1) LEED pattern developed where it was likely that both O and Na were present in the surface as an ordered layer; a c(4 x 2) structure was also observed with a Na: O ratio of 1 : 1. What has not yet been resolved is the relationship between these results and the possibility that it is a molecularly adsorbed O2 which is active in ethylene oxidation.

Riassian et aL have used ESCA to investigate the changes in activity of supported Ag for ethylene oxidation; the origin of the change lies in the accumulation of organic impurities on the Ag/SiO2 or Ag/Al2O3 catalysts. Persistence of an unchanged Ag doublet peak meant that there was no change in the Ag: changes in the C 1s signal supported the model of organic deposition.

The next example is also concerned with modification of catalysts, again in an industrially important process. X-ray diffraction, Mössbauer spectroscopy, and SIMS have all been utilized to examine promoted iron catalysts for ammonia synthesis. After reviewing the problems associated with defining the states of the promoters, Ludwiczek et al. proceeded to establish that in Al2O3-promoted Fe catalysts the aluminium may appear, for example, as Al2FeO4 units endo-tactically built into the a-Fe surface; this accounted for the stability of the Fe surface.

A paper of significance in considering catalyst poisoning is that by Fisher, who propounds interesting views. He uses UPS information to interpret bonding of adatoms adsorbed on Ni(100). Taking the surface structures shown in Figure 1, he then distinguishes between two features of the system: (a) atoms may be adsorbed at sites with a number of nearest neighbours, i.e. the site co-ordination Cs, and (6) they may be bonded to only some or all of the neighbouring atoms; this is the bond co-ordination Cb.

In considering this situation the author chooses to examine UPS spectra for adsorbed S, Se, and Te and to compare spectra with solid-state spectra for these chalcogens. He found compounds and elemental solids in which the S, Se, or Te were bonded covalently to two neighbours. The main feature observed from known standards was a double-peaked spectrum, and the general result that the shape of the density of valence states is largely determined by the number of nearest neighbours to which the atom is bonded. Applying these ideas to the Ni(100) one half monolayer c(2 x 2) surface, the author finds the double peaks characteristic of 2-co-ordination. For one quarter coverage of the surface, from his own and other work, the author concludes that bonding is four-fold.

The overall conclusion reached, which is of importance to catalysis, is the possibility that the strong adatom–adatom interactions may alter electronic structure for the catalyst system and that this may be more important than a simple site-blocking mechanism for poisoning.

The subject of ISS has been considered in some detail by Baun. The technique is a powerful one for identification of atoms adsorbed on surfaces. It is best illustrated by an example. In catalysts where ionic solids are doped with foreign ions it must be of importance to know the distribution of the foreign ions. Though not used here in a catalytic study, ISS has been used to establish the way in which Pb concentrates in the surface of AgBr crystals. In the experiment Ne+ ions were used as projectiles with energy E1 scattered (through 90°) with energy E2. E2/E1 = (M2 - M1)/( M2 + M1), where M2 = mass of substrate atom and M1 = mass of neon atom.

As an example of the results it was found that for 1000 p.p.m. Pb impurity the Pb/Br ratio on the surface reached 13. For 300 p.p.m. Pb, the ratio was 4. The author proposes that the concentration of Pb2+ on the surface arises from the negative charge produced there by the movement of Ag+ ions into interstitial positions in the lattice.

Segregation. — In a number of catalytic systems in which a poison is present or in which C deposition takes place on a metal it is of importance to know the surface concentration of these...