Possible Effects of Magnetic Fields on the Quenching of Triplet Excited States of Polynuclear Hydrocarbons by Oxygen

Abstract

A theory of the magnetic field effect on the rate constant γ of the quenching of aromatic hydrocarbon triplets by molecular oxygen is developed. Based on theoretical considerations of Kawaoka, Khan, and Kearns and on experimental evidence obtained by other workers, it is assumed that the final state of the quenching reaction gives rise to oxygen in a singlet excited state. The conclusions should, however, be applicable to other cases in which the initial state involves two triplets and the final state two singlets. The theory takes the intramolecular spin-spin interaction and the charge-transfer interaction in the bimolecular intermediate complex state into account. The distribution of singlet character among the nine possible collision complex states is considered. The theory predicts that as long as the splitting among the singlet, quintet and triplet like pair states due to charge-transfer interaction is negligible compared to the spin-spin interaction, a negative effect on γ in a quantizing magnetic field is expected. If the charge-transfer induced splitting ΔETS between singlet and triplet like pair states and ΔEQS between quintets and singlets is large compared to the zero-field spin-spin interaction and also the Zeeman energy, no effect of magnetic field on γ will be observed. Both positive and negative field effects on γ are predicted if ΔETS and ΔEQS are of the order of the Zeeman energy or if ΔETS∼0, but ΔEQS is larger than the spin-spin interaction. The magnetic field dependence of γ may then display minima and maxima, which may be very broad if ΔEQS and ΔETS are a function of the collision complex geometry which may assume many configurations. Experiments on partially quenched phosphorescent samples of chrysene, a,h-dibenzanthracene and coronene adsorbed on a polystyrene matrix were carried out at a field of 145 kG. Within the experimental error of 1%–2%, no effect of magnetic field on γ was detected. The experimentally observed null effect is interpreted assuming either (1) that the ratio ε of the transition rate to final singlet states to the rate of dissociation of the collision complex is small, or (2) that the effect of charge transfer mixing in the complex is significant. It is tentatively concluded that the results are better explained in terms of the latter hypothesis. It is estimated that the energy spread due to charge-transfer interaction between singlet and triplet-like complex states is greater than 30 cm−1.