Solvent-dependent singlet oxygen lifetimes: temperature effects implicate tunneling and charge-transfer interactions

The effect of solvent on the lifetime of singlet oxygen, O2(a(1)Δg), particularly the pronounced H/D solvent isotope effect, has drawn the attention of chemists for almost 50 years. The currently accepted model for this phenomenon is built on a foundation in which the electronic excitation energy of O2(a(1)Δg) is transferred to vibrational modes in a solvent molecule, with oxygen returning to its ground electronic state, O2(X(3)Σg(-)). This model of electronic-to-vibrational (e-to-v) energy transfer specifically focusses on the solvent as a "sink" for the excitation energy of O2(a(1)Δg). On the basis of temperature-dependent changes in the solvent-mediated O2(a(1)Δg) lifetime, we demonstrate that this energy-sink-based model has limitations and needs to be re-formulated. We now show that the effect of solvent on the O2(a(1)Δg) lifetime is more reasonably interpreted by considering an activation barrier that reflects the extent to which a solvent molecule perturbs the forbidden O2(a(1)Δg) → O2(X(3)Σg(-)) transition. For a given solvent molecule, this barrier reflects contributions from (a) the oxygen-solvent charge transfer state that mediates nonradiative coupling between the O2(a(1)Δg) and O2(X(3)Σg(-)) states, and (b) vibrations of specific bonds in the solvent molecule. The latter establishes connectivity to the desirable features of the energy-sink-based model. Moreover, temperature-dependent H/D solvent isotope effects imply that tunneling through this barrier plays a role in the mechanism for O2(a(1)Δg) deactivation, even at room temperature. Although we focus on a long-standing problem involving O2(a(1)Δg), our results and interpretation touch fundamental issues of interest to chemists at large.