New Perspectives in Surface Crystallography

Abstract

The ideal balance between theoretical and experimental techniques occurs where comparable amounts of time and money are spent on each side of the problem. In LEED structure analysis this is true to within an order of magnitude for the sorts of structures currently being studied, but in the future we can already see a serious imbalance appearing. Experiments scale linearly in time with N, the number of beams (assuming that all are measured) and theory scales as N³. The moral is that we should be working to increase the speed of computation and if possible to devise new methods that scale in time as N, so that a competitive position, once established, remains so for a wide class of experiments. Clearly N must be the optimum scaling because even the printing of all beams scales as N.