Genetic Algorithm in the EPR Study of Powder Spectra of Charge-Transfer tm-p-PD:Chloranil Complex

. The temperature dependence of the electron paramagnetic resonance (EPR) powder spectra of tm-p-PD:Chloranil has been studied. A qualitative analysis of the spectra suggests that the line width of an unpaired spin center varies with temperature, g parameters may vary, too, but generally there is no simple way to assign parameters to the spectra unless the EPR data are confronted with an appropriate synthetic model. As a model for the detailed simulation and analysis, an orthorhombic spin Hamiltonian is adopted and an anisotropic line width of the paramagnetic charge-transfer center is assumed. The results are the best-fit values g xx , g yy , g zz and the three line width parameters, respectively. A genetic algorithm was designed for the EPR spectra best-fit search in the six-parameter space and the results were obtained automatically without operator intervention. The numerical procedure was accelerated with a computing cluster.