Photochemical Exploration of Reaction Dynamics at Catalytic Metal Surfaces: From Ballistics to Statistics

Surfaces form the working template upon which much of the world’s technologically important chemistry occurs.1 Catalysis, corrosion, adhesion, friction/lubrication, and materials growth all depend critically upon interfacial chemistry. This account discusses how photochemical methods can be exploited to explore the fundamental dynamics of elementary catalytic reactions occurring at the gas/metal interface.2 By studying reactions on single crystal surfaces, several rather unique opportunities arise. The first stems from the natural self-organization of adsorbed molecules which become stereochemically arranged by the periodic surface forces. By initiating reactions nonthermally at low temperatures, one may investigate restricted and organized reactive geometries. Second, in contrast to reactions occurring in bulk condensed phases, characterization of nascent product state distributions for surface reactions is readily achieved through gas phase spectroscopy if the products are promptly desorbed. Studies of such surface reactions can provide us with a dynamical window into the world of condensed phase reactions. Finally, the remarkable ability of the scanning tunneling microscope (STM) to image and manipulate atoms and molecules currently permits nonthermal reactions to be examined microscopically at the level of individual reagents adsorbed at specific surface sites.3 In the long run, we must anticipate that surfaces will prove to be among the very best places to study the elementary details of chemical reactions. Much of our dynamical information about surface reactions derives from molecular beam and recombinative thermal desorption studies of reactions carried out on single crystal surfaces prepared in ultrahigh vacuum.4 State analysis of products desorbed into the gas phase has come from time-of-flight (TOF) spectroscopy to a quadrupole mass spectrometer, infrared chemiluminescence, laser induced fluorescence, or resonance enhanced multiphoton ionization. Unfortunately, most catalytic reactions, such as CO oxidation, proceed through the Langmuir-Hinshelwood (L-H) mechanism in which thermalized adsorbed reagents react at the surface temperature (e.g. /2O2(g) + CO(g) f O(ad) + CO(ad) f CO2(g)). The energetic monochromaticity of reagent molecular beams goes largely to waste for such L-H reactions, but this may not be the case for reactions proceeding through the Eley-Rideal (E-R) mechanism in which an impinging gas phase species reacts directly with an adsorbate. Despite their long standing discussion, Eley-Rideal reactions have only recently been definitively observed when radical species such as H atoms were made incident on adsorbate covered surfaces.5 Cross-sections for radicals reacting with closed shell species are generally much greater than those for reactions involving two closed shell species because the reduced Pauli repulsion experienced during reactive approach of a radical allows for much lower activation barriers.6 Just as for gas phase reactions,7 it is only practically feasible to study the dynamics of stateprepared radical/closed shell or radical/radical surface reactions which occur with relatively high probability upon every interreagent collision. In the case of typical L-H reactions, at least one of the closed shell molecular reagents incident from the gas phase is dissociated on the surface to give an adsorbate whose reactivity lies somewhere intermediate between a radical and closed shell species (e.g. O2(g) f 2O(ad) during CO oxidation).1 This reactivity is typically insufficient to generate a L-H reaction upon every interreagent collision on the surface, at least not after the dissociated fragments of the molecular reagent have become thermalized to the surface temperature. However, prior to thermalization, the “hot” molecular fragments may display an enhanced reactivity before their heat of adsorption is dissipated. Reactions of such “hot precursors” to the thermalized L-H reagents through a mechanism intermediate between the E-R and L-H mechanisms were first discussed by Harris and Kasemo8 to explain low-temperature reactions occurring when H2 and O2 gases were adsorbed on a polycrystalline Pt foil. Hot H atom fragments liberated by thermal dissociation of adsorbing H2 were suggested to react with adsorbed O and OH species to form OH and H2O products through a precursor (or H-K) mechanism. Analogous surface photochemical reactions involving hot photofragments produced by adsorbate photofragmentation were first demonstrated on insulators by Polanyi’s group at Toronto9 and on metals by Ho’s group at Cornell.10 Photochemical methods for exploring reaction dynamics on metals have been somewhat delayed in developing because of the popular perception that electronic quenching rates for adsorbates on metals were prohibitively fast for surface photochemistry to be observable.11 It is only over the past decade that adsorbate photochemistry on metals has begun to be identified and examined in detail.12 Adsorbed molecules on a metal may be excited directly by light absorption or indirectly by interaction Ian Harrison was born in Montreal, Canada, in 1959. He joined the faculty at Virginia in 1989 after a Ph.D. with John Polanyi at the University of Toronto and a postdoctoral fellowship with Gabor Somorjai at Berkeley. His current research focuses on the study of surface reaction kinetics and dynamics using laser and STM techniques. Acc. Chem. Res. 1998, 31, 631-639