The photoelectrochemical behavior of TiO2 (rutile) electrodes has been investigated as a function of electrode preparation. Mechanical polishing of the electrodes has been found to introduce a disturbed layer on the order of 1 μm deep. The experimental results indicate that this layer has an increased density of recombination sites. In addition, this layer affects the reduction process by which donors are created in TiO2 . Etching or annealing this layer can remove it or modify its behavior. The behavior of etched and annealed electrodes indicates that the absorption and/or reflection of light by TiO2 can be affected by electrode processing. The stability of TiO2 electrodes is discussed in terms of their usefulness in electrochemical studies.

The capacitance, spectral response, and volt-ampere characteristics of lightly reduced TiO~ photoelectrochemieal electrodes are all significantly altered by immersion in concentrated acids or by operation at sufficiently negative voltages. These changes, which are reversible, are attributed to diffusion of protons along the c-axis channels of rutile. There is an apparent threshold voltage for appreciable accumulation af protons which are presumed to enter the crystal as H atoms.

ACTIVATION OF ELECTRODE SURFACES 83g an essentially linear growth (i.e., constant growth rate) regardless of the relative importance of the oxide formation in the over-all anodic process \ b / L (t) ~ [ (Lo --L1) --L1 log to] Jr Lit [15']Another interesting point would be the variation of the critical time with pertinent process parameters under galvanostatic conditions in view of the welldocumented difference in the growth rate vs. time rerelationship with respect to potentiostatic conditions [e.g., (7)], in the instance of Pt anodes. Since galvanostatic conditions are less inviting than potentiostatic conditions for a rotating wiper-blade electrode, such an investigation would be beyond the scope of our current research. The hereby presented analysis of critical time would most likely prove useful in improved design of wiper-blade electrode systems both for mechanism studies and practical oxidations at the continuously activated electrode surface, by choosing appropriate rotation rates and wiper-blade widths for a given electrode metal and specific electrochemical reactions.

A recent publication by Nakato et al., suggested that thin gold overlayers could be made to protect normalGaP photoelectrodes from corrosion. Repetition of their experiments, including simultaneous measurement of the gold potential, duplicates the volt‐ampere characteristics reported in that paper. Gas evolution was not observed, and the gold potential remains too low to decompose water at its surface. An alternative interpretation of the action of the Au overlayer is suggested.

No abstract