Sun 12 February, 2012

Low Energy Electron Diffraction

Si(111) 7x7 LEED patternFigure 1: A Si(111) 7×7 LEED pattern (click to enlarge)
LEED involves the use of an electron gun to bombard a sample with a beam of electrons, the energy of which can be controlled; typically electrons with energies in the range 30 to 300 eV are used in LEED. Diffracted electrons travel away from the sample and pass through a set of retarding grids, which serve to select elastically scattered electrons, and after being accelerated by a high potential they impinge on a phosphor screen. This causes the screen to glow with an intensity at each point on the screen proportional to the incident electron flux, thus producing a visual map of the electron diffraction pattern. An example of such a pattern, from the Si(111) 7×7 surface, is shown in figure 1. The pattern consists of a number of bright spots associated with particular diffraction beams, and actually images the surface reciprocal lattice. As the primary electron energy is increased, spots move towards the centre of the image and higher order diffracted beams come into view, revealing more of the surface reciprocal lattice. The spacing and positions of the spots may be interpreted to yield the surface real-space lattice vectors.
An example of experiment to simulation comparison of a LEED I-V curveFigure 2: Comparison of experimental and simulated LEED I-V curves for Ge(111)1×1Dy (click to enlarge)

While the LEED pattern can give some immediate information as to the quality of the surface and its reconstruction, to use LEED as an anlaytical technique it is necessary to consider the pattern over a range of energies. The intensity of LEED spots changes as the primary electron energy changes. By tracking the change of intensity of a LEED spot as the primary energy is varied an I-V curve is built up. Whilst this I-V curve contains information about the postitions of surface atoms, retrieving the information is not trivial. In practice it is necessary to compare experimentally obtained I-V curves against those produced by computer simulations for model structures. Comparison to many different model structures continues until the agreement with experimental data is good enough, and it is certain that all plausible models have been considered. Figure 2 shows an example of a comparison between simulated and experimental I-V curves, for Ge(111)1×1-Dy.

Copyright © 2003-2012 University of York
Legal statements
Valid HTML 4.01! Valid CSS!