1. Introduction

The adsorption of organic molecules on inorganic surfaces has been a continually evolving subject since the early work of Langmuir and other vacuum pioneers. As experimental technologies have improved, the number of adsorbate/substrate systems that could be studied has expanded considerably. Although originally limited to relatively inert surfaces such as Au(111) or surfaces that could be regenerated by cleaving such as those of the alkali halides, advances in ultra-high vacuum techniques have enabled the preparation and long-term clean storage of high-quality, single-crystal transition metal and metal oxide samples. Similarly, advances in characterization technologies such as scanning probe and near-field optical microscopy have enabled researchers to image and to perform spectroscopy on individual adsorbates. Meanwhile, due to the development of surface organic chemistry and the growing technological importance of organic surfaces, the number of interesting adsorbate/substrate combinations has also grown tremendously.

With the growing number and complexity of the systems to be studied, it becomes more and more essential to understand in detail the three basic surface processes: physisorption, chemisorption, and surface chemical reaction. In other words, as the complexity of the molecules increases, it becomes more difficult, but not less necessary, to find those several features of the surface process that apply to a whole family of molecules so that not every system needs to be measured before its surface behavior can become known. In this way, our knowledge keeps its predictive character as opposed to becoming purely observational.

1.1. Surface Interactions

Of course, the simplest interaction between a gas-phase molecule and a surface is no long-term interaction at all. When the energy of the collision is large with respect to the adsorption energy, the molecule will collide with the surface and promptly return to the gas­phase. If the collision is fully elastic, the molecule will reflect off the surface at the specular angle. If the incident molecules are monoenergetic (and not too heavy) and the surface is well-ordered with reasonable corrugation, diffraction will be observed. Alternatively, if some energy is transferred between the degrees of freedom of the molecule or between the molecule and the lattice phonons of the surface, the velocity of the molecule will be altered and inelastic scattering is observed. In each non-reactive case, the residence time of the molecule is short and results in little or no energy transfer between the molecule and the surface.

Physisorption

However, at sufficiently low temperatures, any molecule can be made to physisorb onto any substrate. When a slow-moving molecule approaches a surface, the distribution of its electron density will be altered by the influence of the electrons of the surface. A van der Waals attractive force will arise from the coupling of the instantaneous dipole fluctuations in the surface and those of the molecule. This force will accelerate the molecule towards the surface, resulting in greater energy transfer than would be predicted by the initial velocity of the molecule. Upon collision, energy will be lost to the surface and the resulting velocity of the departing molecule may be less than the velocity required to escape the potential well. In this case, the molecule becomes adsorbed and eventually fully equilibrates with the surface temperature at a distance of typically 3-5 Å above the surface.

Physisorption occurs readily at low surface temperatures and has little dependence on the choice of the adsorbate and substrate. While larger molecules will tend to physisorb more readily than smaller molecules due to a greater polarizability and a greater potential for the transfer of translational energy to internal modes, even helium can be made to condense on a crystal at low enough surface temperatures. If only physisorption is involved, adsorption energies are relatively small and proportional to the polarizability of both the adsorbate and the substrate.

Chemisorption

If the adsorbate is able to share electrons with the substrate, chemisorption is possible. Unlike physisorption which relies on physical forces (primarily between dipoles and induced dipoles) to keep the adsorbate bound to the surface, chemisorption requires the formation of a chemical bond between the adsorbate and the surface. If the redistribution of electrons in the molecule to form a bond with the surface does not empty bonding orbitals or cause antibonding orbitals to become occupied, the molecule will adsorb intact. While the formation of new chemical bonds during adsorption may be exothermic by 1-5 eV, this process is much more system specific than physisorption. Due to the presence of energetically-unfavorable transition states along the reaction coordinate, activation barriers often prevent direct chemisorption at low impingement energies and surface temperatures.

At lower surface temperatures, these activation barriers can alternatively be overcome by increased residence time of the molecule on the surface. While in a physisorbed bound state, the molecule has the opportunity to make repeated attempts to overcome the activation barrier. With an appropriate choice of surface temperature, the rate of desorption from the physisorbed state can be minimized while providing adequate thermal energy to promote chemisorption. Since the rate of chemisorption from the physisorbed state is related to the population of adsorbates and the residence time, the rate of precursor-mediated chemisorption will be seen to decrease when the surface temperature is increased.

1.2. Predictions from Theory

At present, semiempirical models of physisorption have been utilized to connect theory with experiment with varying degrees of success for more complex adsorbate-substrate systems. Through the use of "universal functions" to represent the potential resulting from van der Waals attraction and Pauli repulsion, an extensive set of adsorbate-substrate potentials has been compiled. However, the predictive capability of these universal functions is somewhat limited due to the specific attributes of each adsorbate-substrate pair. Additional corrections to the potential function are required to account for factors such as the elasticity or corrugation of the substrate and the possibility of screening of electrons within the adsorbate.

The theoretical treatment of the process of chemisorption has traditionally examined only the interaction of the frontier orbitals. In this model, a chemical bond can be formed between the molecule and the surface while perturbing only the highest occupied molecular orbital (HOMO) or lowest unoccupied molecular orbital (LUMO). While this simplification allows for the description of the reaction along a limited number of "reaction coordinates", it also requires the adsorption process to be localized on the surface. Although current simulations of adsorption on clusters of substrate atoms are able to determine qualitative energy differences between different adsorption sites, quantitative determination of absolute binding energies remain elusive without the use of a semiempirical treatment. In addition, the presence of impurities, the specific surface structure, and the adsorbate coverage can each strongly influence the binding energy of a particular adsorbate-substrate system.

With the increased experimental resolution of synchrotron x­ray emission spectroscopy (over that of conventional photoemission-based spectroscopic techniques), recent evidence has indicated that this basic bonding model may be an oversimplification of the adsorbate-substrate binding interaction. In x-ray emission experiments performed with N₂

(to examine a predominantly -bonding system) or benzene (-bonding) adsorbed on Ni(100), emission from unexpected hybridized orbitals were observed. This lack of access to quantitative estimates of interaction energies fundamentally hinders the refinement of dynamical theories and leaves the field open to a mostly experimental approach.

1.3. Outline of this Thesis

In response to the need to study systems of increased complexity in a more systematic way and because of the lack of predictive abilities of existing theories of adsorption, we decided to study the adsorption of a series of saturated, unsaturated, and substituted hydrocarbons on a simple metal surface aiming at the observation of both physisorption and chemisorption.

Three adsorbate/substrate systems were studied:

1) The physisorption of hydrocarbons on Au(111).

2) The physisorption and chemisorption of alkanethiols on Au(111).

3) The chemisorption of methane on Pt(111) and Ni(111) with and without vibrational excitation of the molecule.

To determine the required rates of adsorption and desorption, a simple but sensitive technique was used that is based on helium atomic beam reflectivity. As a clean and well­ordered surface reflects a large fraction of a collimated molecular beam of helium while the presence of adatoms scatters the helium atoms off-axis, the intensity of the reflected beam provides direct information about the surface coverage. This low-energy technique can measure adsorption rates down to 10^-7 ML/sec with no risk for damage to the surface. The experimental apparatus, techniques, and analytical protocols for the application of atom reflectivity to specific systems are presented in Chapter 2.

In Chapter 3, the physisorption of alkanes, alkenes, and related cyclic hydrocarbons on Au(111) is explored. Twenty-five hydrocarbon species were deposited individually to determine relations between adsorption behavior and the identity of the molecule. Reflectivity-detected temperature programmed desorption was performed to determine the activation energy for the desorption of each species. In addition, sticking coefficients were calculated from helium atom reflectivity measurements of the change in adsorbate coverage in response to systematic dosing at specific fluxes and surface temperatures. To determine sticking coefficients and the functional relationships between the observed specular intensity signal and the adsorbate coverage, a model was developed to predict coverage with the assumption of Langmuir adsorption.

Capable of both physisorption and chemisorption on Au(111), the adsorption of the alkanethiols is studied in Chapter 4 using many of the same techniques described in Chapter 2 and 3. Both linear and branched alkanethiols were deposited to determine the dependence of the chemisorption energy on the adsorbate structure. Annealing experiments were also performed with linear alkanethiols of varying chain length to investigate precursor-mediated chemisorption in the particularly interesting case where the chemically active thiol group is kept constant but the lifetime of the precursor can be adjusted by altering the chain length.

In Chapter 5, the effect of vibrational excitation on the rate of chemisorption of methane on Pt(111) and Ni(111) surfaces is examined. With an experimental apparatus which is uniquely capable of generating a continuous high flux of vibrationally excited methane (~10¹¹ molecules/sec/cm²), sticking coefficients of methane were measured with and without vibrational excitation for a range of impinging translational energies by helium atom reflectivity. To determine the degree of surface accommodation of vibrational energy, bolometric detection of vibrational energy in the molecular beam both before and after collision with the surface was performed.