Effects of solvent on TMP photophysics. Transition from no barrier to barrier case, induced by solvent properties

Publisher Website
Google Scholar
Abstract
The dynamics of the radiationless relaxation of triphenylmethane (TPM) molecules in the n‐alcohols methanol to octadecanol have been studied with picosecond absorption recovery techniques as a function of viscosity, temperature, and wavelength of the exciting and analyzing light. It is shown that the solvent dependence of the relaxation rate in TPM molecules cannot, as is usually done, be described by a viscosity dependence only. In going through the n‐alcohol series the excited state potential surface changes from a practically flat one (E0=0) in methanol to a surface having a definitive potential barrier (E0≊15 kJ mol−1) in the higher alcohols. This variation of the potential surface implies that the relaxation is largely controlled by rotational diffusion in methanol and ethanol whereas it is mainly controlled by the potential barrier in the higher alcohols. When the viscosity dependence of the relaxation rate is obtained, by compensating for the contribution from the potential barrier, a turnover behavior such as that predicted by activated barrier crossing theories, is observed. We have fitted the expression of Skinner and Wolynes [J. Chem. Phys. 6 9, 2143 (1978)] to our experimental results in the solvents where a barrier exists, and obtain a frequency of 550 cm−1 at both the well and the top of the potential. The fit is consistent with a friction that is lower than that corresponding to hydrodynamic slip boundary conditions. A decrease in solute–solvent friction relative the hydrodynamic value is also observed in going from small to large solvent molecules. This effect is attributed to a decreasing solute/solvent molecular volume ratio and to the structure of the TPM molecule which could prevent large solvent molecules from coming into close contact with the relaxing groups. In addition to these features we discuss the wavelength dependence and previously suggested change from single‐ to bi‐ and multiexponential decay of the relaxation kinetics. The experimental observations are also discussed in relation to the recent theory of Bagchi et al. [J. Chem. Phys. 7 8, 7375 (1983)].