Global versus local minimization procedures for the determination of spin Hamiltonian parameters from electron spin resonance spectra

Abstract

The simulated annealing procedure based on the Metropolis Monte Carlo algorithm is presented as a numerical method for the analysis of ESR spectra. It provides a way of finding the global minimum of the function which describes the difference between the theoretically calculated spectrum and the experimentally measured one. Compared with the commonly used iterative procedure, it has the advantage of not getting stuck in local minima of the parameter space. This implies that the procedure can be started with arbitrary initial parameters. The usefulness of the method is illustrated by its application to the analysis of ESR spectra in fast motion conditions. The results are compared with those obtained by different minimization algorithms.