Comment on “Measurement of Ultrafast Photoinduced Electron Transfer from Chemically Anchored Ru-Dye Molecules into Empty Electronic States in a Colloidal Anatase TiO2 Film”

In a recent Letter published in this journal, Willig et al. reported on the ultrafast electron transfer from photoexcited ruthenium(II) cis-di(isothiocyanato)bis(4,4′-dicarboxy-2,2′-bipyridyl),1 (N3), into titania. This dye has attracted wide attention as a highly efficient charge-transfer sensitizer for mesoporous TiO2 films.2 A key element of Willig et al.’s results was the determination of the absorption spectrum of the oneelectron oxidation product of Ru(II)(dcbpy)2(NCS)2 which was produced by chemical reaction of N3 with oxidants, such as Br2, and was reported to peak at 440 nm with negligible nearinfrared absorption. The apparent stability of this species identified as the [Ru(III)(dcbpy)2(NCS)2] cation by Willig et al. in ethanol is at variance with earlier electrochemical studies of the Ru(II)(dcbpy)2(NCS)2/Ru(III)(dcbpy)2(NCS)2 redox system in organic solvents where a reversible wave in the cyclic voltammetry was obtained only at high scan rates, yielding for the oxidized Ru(III))(dcbpy)2(NCS)2 complex a lifetime of 0.1-1 s, depending on the degree of protonation of the carboxyl substituents.3 The assignment of the 440 nm spectral feature to the Ru(III)(dcbpy)2(NCS)2 cation in ref 1 is also in contradiction with laser photolysis data obtained earlier by Durrant et al. which show this species to exhibit a broad near-IR absorption, probably originating from a LMCT transition with a maximum close to λ ) 800 nm.4 In view of this controversy, we have scrutinized our earlier results and carried out further experimental verifications which we present here. Our data demonstrate unambiguously that the conclusions of Willig et al. are erroneous and that the data reported by these authors have likely been obtained with samples of degraded sensitizer. Instead of abstracting an electron from the Ru(II) center, powerful oxidants such as bromine react with N3 by chemical attack of the thiocyanato ligand. The reason for this is the high (0.85 V/SCE) redox potential for the Ru(II)/Ru(III) couple in N3 which is close to that of the -NCS ligand, favoring attack of the latter by the oxidant. The oxidation of -NCS by bromine has been studied by Werner at the beginning of the century, using thiocyanato complexes of Co(III) which is isoelectronic with Ru(II).5 Cleavage of the CtN bond was observed yielding amino complexes. A similar decomposition of the -NCS ligands in the presence of strong oxidants is also expected to occur in the case of N3. Similar to Willig et al., we obtain a yellow product upon oxidation of Ru(II)(dcbpy)2(NCS)2 in ethanol by a number of oxidants, i.e., Br2, LiOCl, NOBF4, WCl6, or Ce(NH4)2(NO3)6 (Figure 1). The coloration of the solution at this stage is not stable and evolves within seconds (or minutes, depending on the excess of oxidant added) to yield a spectrum which is similar to that of the original N3 dye. The IR spectra of the latter stable compound and of the yellow oxidized product clearly indicate the absence of the characteristic SCN band at 2139 cm-1 (Figure 2), confirming that cleavage of the CtN bond has taken place quantitatively upon chemical oxidation of the dye. Therefore, attribution by Willig et al. of the 440 nm absorption of the product formed by bromine oxidation of N3 to Ru(III)(dcbpy)2(NCS)2 is erroneous. The uncarboxylated analogue of N3, Ru(II)(bpy)2(NCS)2, is a complex which, due to its lower oxidation potential (0.5 V/SCE), is less prone to decomposition of the thiocyanato ligand by chemical oxidants. Except for a 20 nm blue shift of its * Corresponding author. E-mail: [email protected]. § Present address: Elsevier Science S.A., 1003 Lausanne, Switzerland. † Ecole Polytechnique Fédérale. ‡ Imperial College. Figure 1. Absorption spectra of the original Ru(II)(dcbpy)2(NCS)2 dye in ethanolic solution (1) and of the reaction product of the oxidation of the complex by Br2 (2). Absorption spectrum (3) is that of the first oxidation product of the uncarboxylated analogue Ru(II)(bpy)2(NCS)2 by Ce(IV) in acetonitrile.