What do we want from computer simulation of SIMS using clusters?

Abstract

Computer simulation of energetic cluster interactions with surfaces has provided much needed insight into some of the complex processes which occur and are responsible for the desirable as well as undesirable effects which make the use of clusters in SIMS both useful and challenging. Simulations have shown how cluster impacts can cause meso-scale motion of the target material which can result in the relatively gentle up-lift of large intact molecules adsorbed on the surface in contrast to the behaviour of single atom impacts which tend to create discrete motion in the surface often ejecting fragments of adsorbed molecules instead. With the insight provided from simulations experimentalists can then improve their equipment to best maximise the desired effects. The past 40 years has seen great progress in simulation techniques and computer equipment. 40 years ago simulations were performed on simple atomic systems of around 300 atoms employing only simple pair-wise interaction potentials to times of several hundred femtoseconds. Currently simulations can be performed on large organic materials employing many body potentials for millions of atoms for times of many picoseconds. These simulations, however, can take several months of computation time. Even with the degree of realism introduced with these long time simulations they are still not perfect are often not capable of being used in a completely predictive way. Computer simulation is reaching a position where by any more effort to increase its realism will make it completely intractable to solution in a reasonable time frame and yet there is an increasing demand from experimentalists for something that can help in a predictive way to help in experiment design and interpretation. This paper will discuss the problems of computer simulation and what might be possible to achieve in the short term, what is unlikely ever to be possible without a major new break through and how we might exploit the meso-scale effects in cluster impacts to simplify the simulations.