A SEiMKLASSKAL JUSTIFICATION FOR THE USE OF NON-SPREADING WAVEPACKEI-S IN DYNAMICS CALCULATiONS

This manuscript considers the spreading or non-spreadin, 0 of wavepackets. On one hand molecular dynamics simulations have given useful results for many properties of bulk systems These simulations are based upon classical mechanics and would seem to indicate that some Iocahzed wavepacket description for the pat-tide wavefunctions must be applicable on the time scales of interest. Otherwise the use of classical mechanics wouhi appear to be inappropriateOn the other hand a simple estimate of the packet spreading for a one-dimensional argon atom for instance, indicates that the locahzed packet picture is applicable only for very short times For a rough estimate we use the web-known result for a free particle gaussian wavepacket [l]_ Taking the initial wavefunction to be