Freshly formed Ru/Ti oxide anodes, containing between 5 and 40 atom % Ru, have been examined for their Tafel behavior during chlorine evolution, as well as their cyclic voltammetric ͑CV͒ and ac impedance response at the open-circuit potential, in chlorinefree NaCl solutions. Also, 30 atom % Ru electrodes have been electrochemically deactivated, as seen by an increase in the anode potential and the Tafel slope for the chlorine evolution reaction during long-term electrolysis. A comparison of the data for the fresh and the deactivated anodes suggests that the deactivated anodes have similar electrochemical characteristics as freshly formed, low-Ru-content ͑ca. 5 atom %͒ oxide films. To understand this better, the experimentally obtained ac impedance data were compared to the calculated impedance, based on a porous film model in which a one-electron surface redox reaction occurs. While the fit is good at medium-to-high frequencies, the inclusion of a diffusion-controlled process for the low atom percent Ru films is required to achieve a good fit also at low frequencies. Taken together, these results support the hypothesis that the deactivation of originally high atom percent Ru anodes is due to the depletion of Ru from the oxide film, causing its electrochemical behavior to become more similar to that of freshly formed low atom percent Ru oxide films. Ru/Ti oxide anodes, formed by thermal decomposition techniques on titanium substrates and used in the chlor-alkali industry, typically contain Ͼ30 atom % RuO 2 , the remaining being TiO 2 .
1,2The Ru metal loading in these oxide mixtures is generally in the range of 4-10 g/m 2 , depending on whether their use is in diaphragm or in membrane ͑or chlorate͒ cells. Ru/Ti oxides of this composition typically exhibit chlorine evolution overpotentials of 40-60 mV at 200-250 mA/cm 2 in 5 M NaCl solutions at 80-90°C. However, with extended usage in cells in practice, the chlorine overpotential sometimes increases to 300-400 mV, and the anode is considered to have become ''deactivated.'' This is clearly an undesirable situation, and hence, the focus in this paper is to understand the origin of anode deactivation more clearly.The possible mechanisms relevant to the deactivation of RuO 2 /TiO 2 -based anodes during the course of the chlorine ͑and oxygen͒ evolution reactions have been the subject of several recent reviews 1-3 and various publications. One possible cause of anode deactivation is a loss of active sites for chlorine evolution by blockage of insoluble surface layers ͑e.g., MnO 2 , BaSO 4 , Fe oxides, etc.͒ or impurities ͑e.g., organics͒ which arise from impurities in the feed brine to the cells.Another possibility is the formation of an insulating TiO 2 layer at the Ti/coating interface with time of anodic polarization, [8][9][10]24 although there is no direct supporting evidence for this hypothesis. A further option is that a loss of Ru occurs from the coating, either throughout the film or only at its outer surface. 9,12,20,23,25 This may take place by erosion and/o...