PNNL

Abstract

Abe et al. illuminated an aqueous suspension of TiO2/Pt with a Hg arc lamp and observed H2 production in the absence of O2 production. An unspecified elemental analysis excluded carbonaceous contaminants, and the source of the electron donor was concluded to be Ti4+ cations. We can suggest at least three more likely (than Ti4?) sources for their mysterious electron donor: 1)an overlooked inorganic species, 2) Ti3?, and/or 3)organic impurities. The authors excluded the latter in their paper, but we view this conclusion as suspect given the lack of information on the limits of detection of the carbon assay, and the fact that the choice of argon(thermal conductivity = 41.33 x 10-6 cal cm-2s-1 (C cm-1)-1)3 as a carrier gas in their thermal conductivity detector4 probably precluded the observation of CO2 (thermal conductivity = 37.61 x 10-6 cal cm-2s-1 (C cm-1)-1)3, which is a common product of organic photochemical oxidations. No mention is made of an analysis for inorganic impurities. It is widely accepted that nanometer scale TiO2 particles can have significant fractions of undercoordinated Ti3? sites which can be easily oxidized to Ti4?, the highest valence of Ti observed in condensed phases. Here we point out that oxidation of Ti3? to Ti4? is much more physically realistic than oxidation of Ti4? to Ti5?. Removal of a fifth electron from titanium requires ~56 eV5.Ti5?, if it were to be generated in a solid oxide, would extract an electron from the valence band composed mainly of oxygen 2s and 2p states rather than remain in the ?5 valence state. This unrealistically high oxidation state seems even more unlikely given the fact that at the room temperatures used in the authors' reactions Ti5? would have been generated and remained at the surface (at room temperature the diffusion of Ti cations in TiO2 is minimal), and the fact that the surface Madelung potential is smaller than that of the bulk6.