No abstract

No abstract

The two-electron, Auger-type transitions which occur when an ion of suSciently large ionization energy is neutralized at the atomically clean surface of a diamond-type semiconductor are discussed. Consideration of the basic elements of the problem leads to a computing program which enables one to calculate the total electron yield and kinetic energy distribution of ejected electrons in terms of a number of parameters. It is possible to account for the experimental results for singly-charged noble gas iona incident on the (111) faces of Si and Ge and the (100) face of Si. The fit of theory to experiment is unique in its principal features yielding numerical results concerning: (1) the state density function for the valence bands of Si and Ge, (2) the energy dependence of the matrix element as it is determined by symmetry of the valence band wave functions, (3) the effective ionization energy near the solid surface, (4) energy broadening, and (5l electron escape over the surface barrier. Over-all width of the valence band is found to be 14 -16 ev for both Si and Ge. Width of the degenerate p bands is 5.1 ev in Si, 4.3 ev in Ge. The matrix element for p-type valence electrons is 0.3 times that for s-type valence electrons. Effective ionization energy is 2.2 ev less than the free-space value for 10-ev He+ ions and decreases linearly with ion velocity. Energy broadening is small for 10-ev ions and increases approximately linearly with ion velocity. Probability of electron escape is several times that predicted for an isotropic distribution of excited electrons incident on a plane surface barrier. A general theory of Auger neutralization is given in which the conclusions of the fit to experiment are interpreted. Investigation of the matrix element as a Coulomb interaction integral involving wave functions whose general characteristics are known but which are not explicitly evaluated leads to an understanding of its principal dependences on energy and angle.

No abstract

%e have established that chemisorption of atomic hydrogen on Si(100) forms a previously unreported dihydride phase, Si(100)1 X 1::2H, as well as the monohydride phase Si(100)2 X 1:H. The interplay of these phases as exposure and temperature are varied casts new light on surface structure and bonding, and on the kinetics of thermal desorption. These studies suggest that the reconstruction inherent in the clean Si(100gX l surface is achieved by the pairing of adjacent surface rows (pairing model) rather than by surface vacancies (vacancy model).Hydrogen chemisorption on silicon surfaces has been studied by a number of investigators as a basic system offering the possibility of both theoretical and experimental understanding of chemisorption. Recently, in the case of Si(111},experiment'"' and theory'4 agree in the conclusion that the interaction of atomic hydrogen with suitable starting surfaces can produce both a monohydride phase, Si(111):H, ' and a trihydride phase, Si(111);SiH, . The observation of this trihydride phase suggested the possibility of formation of a previously unobserved dihydride phase on Si(100).The present work demonstrates for the first time that the interaction of atomic hydrogen with the clean reconstructed Si(100)2&&1 surface produces not only a monohydride phase but also a dihydride phase. %'e represent these phases by the symbols Si(100}2X1: H snd Si(100}1xl::2H, respectively. In each symbol the portion following the low-energy-electron diffraction (LEED) symmetr y designation, 2 && 1 or 1 x 1, indic ates the hydrogen bonding to each surface Si atom. The formation of Si(100)1X1::2H clearly indicates that atomic hydrogen can modify the geometrical structure of surface substrate atoms and leads to a conclusion concerning the nature of the reconstruction inherent in the clean Si(100)2x1 surface. The monohydride and dihydride surfaces exhibit different surface electronic structures as observed by ultraviolet photoemission spectroscopy (UPS). Variation of substrate temperature during adsorption and thermal desorption each yield results which confirm our structural conclusions and illuminate the kinetics of the interplay of the tmo phases.The starting surface in these experiments is the well-ordered clean Si(100)2&&1 surface' obtained by sputter etching with Ne' ions, annealing at 600'C for about 10 min, and cooling to room temperature. This surface exhibits a sharp 2&1 LEED pattern with no impurities distinguishable by Auger electron spectroscopy above noise level (-,~of the principal Si signal at 91 eV). Hydrogenation was achieved by dissociation of H, to 2H at heated W filaments of our sputtering apparatus. e In order to maintain the sample near room temperature it is placed 6 cm distant from the % filaments and is shielded from them. In lieu of the unknomn H arrival rate at the sample surface me specify the constant H, pressure (1.5 X10 ' Torr) used and the constant temperature (1710'C}at which the filaments are held during the exposure.In the UPS measurements employing 21.2-eV He I radiatio...