Efficient electrochemical splitting of H 2 O to O 2 and H 2 fuels has become an important goal in the quest for a renewable source of energy.[1] A major source of the inefficiency of this process is the significant overpotential associated with the anodic oxygen evolution reaction (OER). Understanding the mechanism of the OER could help in identifying the elementary processes contributing to OER overpotential and their relationship to the anode composition and morphology. While the OER has been investigated both experimentally and theoretically for over 50 years, its mechanism and the identity of the chemical intermediates involved remain uncertain. [2][3][4][5][6][7] Two principal pathways have been postulated for the OER on metal surfaces, such as gold (Au). The first involves a direct recombination of oxygen atoms to give O 2 , as shown in Equations (1)- (4):The second mechanism consists of a sequence of four consecutive one-electron oxidations, the first two of which are identical to those of the first mechanism, and the next two are as shown in Equations (5) and (6):In this case, the oxygen coupling step produces a hydroperoxy species (MÀOOH), which then dissociates to produce O 2 . Recent theoretical studies indicate that the second mechanism for oxygen coupling should be favored because it has a lower activation barrier. [4,5] Hydroperoxy species have also been suggested as key intermediates in the electrochemical reduction of O 2 (ORR), and initial experimental evidence for their presence has been presented. [8,9] It is notable nonetheless, that while OOH species have been envisioned to be critical for OER, the species have not been observed under electrochemical conditions.Herein we report the first spectroscopic identification of surface-bound OOH as intermediates of oxygen evolution reaction occurring on the surface of a gold catalyst. The presence of OOH species was observed by in situ electrochemical surfaceenhanced Raman spectroscopy (SERS) in both acidic and basic electrolytes. Roughened gold, rather than a more active catalyst such as platinum, was chosen for investigation because it is an excellent SERS substrate.[10] It was also anticipated that the decomposition of hydroperoxy species on Au might be slower than on more active metals, which would result in a higher accumulation of OOH species and facilitate their spectroscopic detection.A confocal Raman microscope coupled with a high numerical aperture water-immersion objective and 633 nm excitation was used to record these spectra. Real-time SER spectra of a Au electrode in 1 m HClO 4 during a linear voltammetry sweep from 1.0 to 1.65 V are shown in Figure 1. At 1.0 V, the Au surface is reduced, as can be seen from its relatively featureless SER spectrum. The peak at 934 cm À1 is assigned to the symmetric stretching mode of ClO 4 À .[11] The elevated spectral background is associated with high SERS activity exhibited by a metal surface, and has been previously assigned to photons emitted during the annihilation of inelastically scattered locali...