A combined simulated annealing and quasi-Newton-like conjugate-gradient method for determining the structure of mixed argon-xenon clusters

This paper shows how various limited-memory quasi-Newton large-scale unconstrained minimization methods can be used to speed up the location of global minima of potential energy surfaces related to the structures of mixed Ar-Xe clusters. Both a simulated annealing method and a finitetemperature lattice-based Monte Carlo method are accelerated by the various quasi-Newton limitedmemory methods which are then compared for computational efficiency.