Solvent dependent photosensitized singlet oxygen production from an Ir(III) complex: pointing to problems in studies...

A cationic cyclometallated Ir(III) complex with 1,10-phenanthroline and 2-phenylpyridine ligands photosensitizes the production of singlet oxygen, O2(a(1)Δ(g)), with yields that depend appreciably on the solvent. In water, the quantum yield of photosensitized O2(a(1)Δ(g)) production is small (ϕ(Δ) = 0.036 ± 0.008), whereas in less polar solvents, the quantum yield is much larger (ϕ(Δ) = 0.54 ± 0.05 in octan-1-ol). A solvent effect on ϕ(Δ) of this magnitude is rarely observed and, in this case, is attributed to charge-transfer-mediated processes of non-radiative excited state deactivation that are more pronounced in polar solvents and that kinetically compete with energy transfer to produce O2(a(1)Δ(g)). A key component of this non-radiative deactivation process, electronic-to-vibrational energy transfer, is also manifested in pronounced H2O/D2O isotope effects that indicate appreciable coupling between the Ir(III) complex and water. This Ir(III) complex is readily incorporated into HeLa cells and, upon irradiation, is cytotoxic as a consequence of the O2(a(1)Δ(g)) thus produced. The data reported herein point to a pervasive problem in mechanistic studies of photosensitized O2(a(1)Δ(g))-mediated cell death: care must be exercised when interpreting the effective cytotoxicity of O2(a(1)Δ(g)) photosensitizers whose photophysical properties depend strongly on the local environment. Specifically, the photophysics of the sensitizer in bulk solutions may not accurately reflect its intracellular behavior, and the control and quantification of the O2(a(1)Δ(g)) "dose" can be difficult in vivo.