Our research in this area is directed toward understanding the dynamics of dissociative adsorption in reactive collisions on clean and adsorbate-covered surfaces. These processes are fundamental to heterogeneous catalysis. Dissociative adsorption can occur either by formation of bonds between the incident molecule and the surface at impact (direct collisional activation) or via trapping into a molecularly adsorbed state which acts as a precursor for dissociation (precursor-mediated). Bond activation in this manner is a critical step in many heterogeneously catalyzed reactions important in the energy field; the activation of CO in methanation on nickel, of alkanes in steam reforming, of N2 in ammonia synthesis, of alkanes in selective oxydehydrogenation, and catalytic combustion are but a few important examples. These processes leading to dissociative adsorption are not well understood.
The long term goal of this research is to develop a fundamental understanding useful for the prediction, or at least accurate estimates, of rates of dissociative adsorption. In recent years progress has been made in understanding activated adsorption on model catalytic surfaces using both molecular beam and more conventional batch reactor techniques in combination with ultrahigh vacuum methods. Both experiment and theory show that all modes of energy in the molecule, including translation, vibration and rotation, and the vibrational modes of the surface can influence the probability of direct activation into the adsorbed state. In general, it has been found that increasing translational energy in the incident molecules leads to a higher dissociation probability via direct collisional activation. For molecules incident with low kinetic energies precursor-mediated activation can become the dominant process. For precursor-mediated processes the maximum rate of reaction is dictated by the probability that an incoming molecule associatively adsorbs. The convolution of these two processes over the Boltzmann distribution of energies and the angle of impingement on the surface determines the reactivity under catalytic conditions.
It is clear that the competing processes of direct and precursor-mediated dissociation differ substantially in detail. Generally, direct collisional activation occurs at or very near the point of impact on the time scale of a single collisional event, and precursor-mediated dissociation occurs via an adsorbed molecular species which samples large regions of the surface, including defects. Thus, direct collisional activation is sensitive to the electronic potential energy surface at the point of impact, whereas trapping into the precursor state is dictated by energy exchange processes governed by the weaker "physical" potential between the gas and the surface, and subsequent reaction of the precursor occurs after it samples a variety of local electronic potential surfaces and molecular arrangments as it traverses the surface. As a result, probabilities for direct and precursor-mediated activation show very different dependencies on the incident energy and angle of incidence of the molecule and the surface temperature, allowing them to be distinguished experimentally in most cases.
In heterogeneously catalyzed reactions surface coverages of reactants and intermediates may vary, and appreciable coverages may be reached under some conditions of temperature and pressure. In many cases unreactive residues may build up. For example, in the hydrogenation of ethylene on both supported and model platinum catalysts, the surface shows appreciable coverage by ethylidyne - which itself is not an intermediate in the reaction. Thus, it is clear that understanding adsorption on both clean and adsorbate-covered surfaces is important to catalysis. The purpose of our work in this area is to advance the understanding of the dynamical factors in adsorption in both clean and adsorbate-covered metal surfaces in order to contribute to the development of a predictive theory of rates of dissociative adsorption. Stochastic trajectory simulations are combined with experiment in order to develop the capabilities for predicting adsorption rates.
