Orientation-dependent energy transfer in gas-surface collisions: Scattering of vibrationally excited nitric oxide from Au(111)

Abstract

The work in my thesis is a contribution to the field of chemical dynamics at surfaces. In this field we seek to develop a detailed microscopic understanding of chemical events taking place on surfaces. Progress in this field has been spurred by a fruitful interplay between experimental work and theory. My work continues in this tradition. The starting point was a striking theoretical prediction and the measurements I made test that prediction and also provide many new discoveries which I hope will help stimulate improvements in theory. I focused on a hot topic in this field - the breakdown of the Born Oppenheimer approximation and the role of non-adiabatic electronic energy transfer in surface dynamics. Specifically, I studied the scattering of vibrationally excited nitric oxide (NO) from Au(111), an important and extensively studied model system for non{adiabatic dynamics. Upon collision with the surface, vibrationally excited NO molecules very effciently transfer vibrational energy to electronic degrees of freedom in the metal, a striking case of electronic non-adiabaticity (Born-Oppenheimer breakdown), which is believed to be driven by a transient electron transfer (ET). I measured ro-vibrational state distributions of NO molecules prepared in excited vibrational states (<i>v</i>_i = 3, 11, and 16) after scattering from Au(111) as function the incidence translational energy(<i>E</i>_i^trans= 0.05 - 1 eV) and orientation. The goal was to investigate the influence of these parameters on the ET-driven energy transfer and, on a more general level, to improve our understanding of the rules that govern the dynamics of molecules at metal surfaces. To make these measurements, I contributed to the development of two experimental techniques: 1) a new method to orient polar molecules in the laboratory frame (optical state selection with adiabatic orientation) and 2) a new method to improve the quantum state purity in optical pumping (pump-dump-sweep). The effect of orientation is both dramatic and complicated. For NO with <i>v</i>_i = 3 and 11, vibrational relaxation is significantly enhanced for molecules pointing with the N-atom towards the surface compared to molecules oriented with the O-atom towards the surface. For these states vibrational relaxation is furthermore promoted by incidence translational energy. Interestingly, for NO <i>v</i>_i = 16 neither the orientation or incidence energy have an effect and all molecules relax to lower vibrational states. NO <i>v</i>_i = 16 has no survival probability in its initial vibrational state. Rotational state distributions of surface scattered molecules exhibit pronounced rotational rainbow structure that strongly depends on the incidence translational energy, initial orientation, and final vibrational state. These are the first observation of rotational rainbows for molecules that have undergone vibrational relaxation. The measurements have a complicated dependence on orientation, initial energy and vibrational state. Nonetheless, the trends in the vibrational relaxation probability can be understood in terms of a simple model based on the barrier in the energetics of the underlying electron{transfer reaction. Vibrational relaxation requires overcoming this barrier. The barrier decreases as the initial vibrational state increases. This explains the trend to stronger vibrational relaxation as <i>v</i>_i is increased. The variation in barrier height also explains the fact that translational energy promotes vibrational relaxation for <i>v</i>_i = 3 and 11, but is not required for <i>v</i>_i = 16. The barrier is lowered by a favorable N-atom first orientation, resulting in the dramatic increase in relaxation for N-atom first collisions. The experimental data from this work provides a valuable test for theory. Recent calculations of vibrational relaxation based on electronic friction or independent electron surface hopping fail to explain the final vibrational state distributions and how they vary with incidence energy and orientation. My hope is that these measurements will stimulate further theoretical work and new insight into the dynamics of this important example of non-adiabatic chemical dynamics at surfaces. An independent topic covered in my thesis is the generation of a molecular beam of highly vibrationally excited CO using perturbations. This technique is called pump-pump-perturb and dump and I demonstrate its successful implementation in our laboratory.