Search

Current models of sputtered excited atom emission are compared and it is concluded that a direct atom-atom collision, model affords a more realistic description than plasma or quasi plasma processes. However, although a formalism is derived for excited atom yields, quantification is not attempted because of present absence of knowledge of (1) collision induced excitation functions, (2) bound-free state excitation and dissociation promoted excitation processes during atomic ejection, and (3) the nature and magnitude of competing de-excitation processes. The influence of quasi thermal processes in producing surface atomic emission via sublimation during space-time extension of the initial cascade is also examined briefly. A formalism is developed to describe how this process may contribute to excited atom emission.

Scattering of H3+ ions from a clean Ni(111) surface is observed at energies below 1 keV and grazing incidence in a time-of-flight system. The interaction with the surface leads to neutral particles mainly which are identified from the energy distributions as H and H2. The neutralisation/dissociation mechanism is probably a resonant charge transfer of Ni electrons into the non-binding ground state of H3.

Equivalent relative populations of Znl levels of similar lifetime and small cascade contribution in arc, sputtering, and single collision sources are demonstrated, and imply that the physical basis of the Andersen and Hinthorne theory is only one possible description of the excitation mechanism operative during sputtering.

Single atomic collision and thermodynamic equilibrium level populations are compared. The population distributions for relatively cascade free levels with similar lifetimes are shown to be easily described by a Boltzmann relation even if the excitation mechanism is non-thermal.

A rigorous method of calculating the photon yield from sputtered excited atoms in terms of the level population distribution, which in turn is related to the velocity and angular dependent excitation function appropriate to the particular sputtering conditions employed, is derived.

The method of calculating relative initial level populations of excited states of sputtered atoms is developed in principle and compared with those in current use. The reason that the latter, although mathematically different, have generally led to similar population distributions is outlined.

The concept of LTE and its existence criteria are applied to the sputtering problem. After presenting arguments against the existence of LTE within the solid, expressions are derived which allow a calculation of the minimum secondary electron yields and sputtering conditions required for the establishment of partial and complete LTE within the hemisphere or hemispherical shell immediately above the point of impact of an incident, heavy projectile. It is found by application of the theory of 50 keV Ar+ bombardment of zinc that secondary electron yields per incident projectile of greater than 10s are required before either complete, or partial, LTE can be established during sputtering. Further refinements of the model are not expected to reduce this figure the four or more orders of magnitude required for agreement with experimental yields. In the event that this were possible, statistical arguments cannot then validly be applied to such small numbers of particles. The physical basis of Anderson and Hinthorne's theory of excited and ionized state production during sputtering can, therefore, have no validity.

The ability of relevant current theories of secondary ion emission to explain experimental ion energy spectra obtained from polycrystalline Al, Sc, Ti, V, Cr, Fe, Ni, Cu, Hf, W and Au is analysed. It is shown that our present understanding or interpretation of the physics of the emission of secondary ions seems incapable of accounting for the observed changes in the energy spectra as a function of surface contamination or oxidation.