This paper reports the simultaneous internal state and translational energy resolved associative desorption flux of N 2 from Ru͑0001͒ using two different experimental approaches. Both experiments show that the nascent N 2 is formed with little vibrational excitation and that the total excitation in all N 2 degrees of freedom accounts for only 1 3 of the barrier energy. Roughly 2 3 of the energy necessary to surmount the barrier is lost to the surface in desorption. This behavior, as well as the unusual behavior noted previously in direct measurements of dissociative adsorption, both imply strong vibrational quenching in reactive trajectories passing over the high exit channel ͑vibrational͒ barrier. Adiabatic quasiclassical dynamical calculations based on the ab initio potential energy surface and various models of coupling to the lattice are not qualitatively consistent with N 2 vibrational damping to phonons. However, including a strong nonadiabatic coupling of the vibrational coordinate to electron-hole pairs in the dynamics does yield qualitative agreement between experiments and calculated dynamics, and we suggest this as indirect evidence for strong nonadiabatic coupling. We argue that the nonadiabatic coupling is strong in this case because of the high vibrational excitation necessary to pass over the high exit channel barrier in the reactive processes and the large charge transfer inherent in making or breaking bonds. We believe that the same factors will be important in most activated dissociations of bonded molecules on transition metal surfaces, e.g., for O 2 , NO, N 2 , and CO, and if this scenario is correct then nonadiabaticity should be important in the activated dissociation dynamics of these systems as well.

New molecular beam experiments on the dissociation probability S 0 for N 2 on Ru͑0001͒ are presented. These are in general agreement with prior measurements and exhibit very unusual behavior; a very slow increase of S 0 with incident kinetic energy E and the fact that S 0 is still only ϳ10 Ϫ3 at incident energies considerably above the barrier. A simple dynamical model is developed to describe this unusual sticking behavior. The key aspect is that there is considerable energy loss ⌬ from E upon initial impact with the surface ͑principally to the lattice͒ and only EϪ⌬ is then available to surmount the activation barrier in the exit channel. Using experimentally measured values of ⌬ from scattering experiments gives good qualitative agreement of this model with the measured S 0. One implication of the strong energy loss is that there is an apparent violation of detailed balance when comparing only the reactive fluxes of activated adsorption and associative desorption.

An experimental technique, laser assisted associative desorption ͑LAAD͒, is described for determining adiabatic barriers to activated dissociation at the gas-surface interface, as well as some aspects of the dynamics of associative desorption. The basis of this technique is to use a laser induced temperature jump ͑T-jump͒ at the surface to induce associative desorption and to measure the translational energy distribution of the desorbing molecules. The highest translational energies observed in desorption are a lower bound to the adiabatic barrier and the shapes of the translational energy distributions provide information on the dynamics. Implementation of the experimental technique is described in detail and unique advantages and possible limitations of the technique are discussed. The application of this technique to very high barrier surface processes is described; associative desorption of N 2 from Ru͑0001͒ and CO formed by CϩO and C 2 ϩO on Ru͑0001͒. N 2 barriers to dissociation increases strongly with N coverage and co-adsorbed O, in good agreement with DFT calculations. No isotope effects are seen in the associative desorption, indicating that tunneling is not important. The full energy distributions suggest that very large energy loss to the lattice occurs after recombination at the high barrier and prior to N 2 desorption into the gas phase. The mechanism for this remarkably large energy loss is not well understood, but is likely to be general for other high barrier associative desorption reactions. CO associatively desorbs nearly thermally from both CϩO and C 2 ϩO associative reactions. It is argued that this is due to large energy loss for this system as well, followed by indirect scattering in the deep CO molecular well before final exit into the gas phase.

Detailed measurements of state resolved inelastic scattering of N2 from Ru(0001) are reported for a wide range of initial energies (0–3 eV) and angles of incidence. The ion time-of-flight resonantly enhanced multiphoton ionization (REMPI) detection scheme developed here and used with cw molecular beams simultaneously measures the internal quantum state and translational energy normal to the sample surface. Doppler broadening of the REMPI spectrum of scattered particles yields the dispersion in scattering out of plane. The results are qualitatively similar to inelastic N2 scattering studies for a wide variety of other metal surfaces; i.e., no observable vibrational excitation, weak rotational excitation described as a Boltzmann distribution, strong surface excitation depending upon the incident normal energy, and an anticorrelation between rotational and surface excitation. The absence of any vibrational excitation at E≈3 eV is inconsistent with adiabatic model dynamics based on the ab initio potential-energy surface. It is, however, consistent with a strong nonadiabatic damping of vibration to electron-hole pairs in the region of the barrier. This same suggestion was previously found necessary to rationalize unusual dissociative adsorption and associative desorption of N2 on Ru(0001).

Detailed knowledge of the tip apex structure is necessary for quantitative comparison between theory-based simulations and experimental observations of tip-substrate interactions in scanning probe microscopy (SPM). Here, we discuss field ion microscopy (FIM) techniques to characterize and atomically define SPM tungsten tips. The tip radius can be estimated from field emission data, while FIM imaging allows the full atomic characterization of the tip apex. We find that when FIM is applied to tips with a radius of a few nanometers (as is desirable for high-resolution atomic force microscopy imaging), limitations not apparent with less sharp tips arise; successful resolution of these limitations will extend the utility of FIM. Field evaporation can be used to atomically engineer the apex into a desired atomic configuration. Starting from a W(111) wire, a tip terminating in three atoms can reproducibly be fabricated; due to its geometry and stability, this apex configuration is well suited for application as an atomically defined electrical contact in SPM experiments aimed at understanding contact mechanics at the atomic scale