The hydrogen overpotential of pure and impure aluminum in 0.1 N HzSOr a t 25" has been studied. The effect of the autocorrosion of the metal has been interpreted and the role of impurities in determining the overpotential and the rest potential has been shown. As well, the build-up and decay of the overpotential has been studied; the behavior is much more complicated than that of noble metals.Studies of hydrogen overpotential on aluminum have been few (2,3,5,6,7,8) and in most cases details of procedure and the degree of purity of the metal are lacking or there is indication that proper experimental precautions (I) have not been taken. In 110 case was the rate of corrosion reported, although it may have been substantial. The investigation reported here was an exploration a t low current densities, for which there are no serious doubts about the homogeneity of current density distribution.
E X P E R I M E N T A LT h e electrolyte used for all experiments was 0.1 N H2S01, a concentration selected as a compromise between the desires for high conductivity and low rate of corrosion. The water was of "conductivity" grade, k = 8X10-7 ohm-' cm.-I, the acid redistilled reagent quality. T h e electrolyte was purged with purified hydrogen and subjected t o lengthy pre-electrolysis, according t o the method recommended by Bocltris and co-workers (1).T h e aluminum cathode and platinum anode of the cell were in separate compartments conilected through closed ungreased stopcocks wet with solution. Hydrogel1 was bubbled continuously about both electrodes. An auxiliary cathode of platinum was available for pre-electrolysis. A capillary probe near the cathode surface led t o a reference electrode, 0.1 N H2S04/Hg2S0,, Hg. Four types of cathode material were selected for study: (I) rolled sheet, 99.995% pure, (11) cast polycrystalline, 99.995y0 pure, (111) single crystal, 99.995% pure, and (IV) "commercially pure" ingot, 99.5% pure. T h e cathodes were electropolished in a perchloric acid bath, then mounted in a Teflon box which exposed only one flat face. Unless stated otherwise, current densities mentioned refer to the apparent area of this face.T h e currents used were electronically regulated. T h e potential diit'erence between the probe and cathode was passed to a direct-coupled amplifier of very high impedance (about 1013 ohms) and thence to a torsion string galvanometer (Model TSB-A29, Kipp en Zonen, Delft). The deflection of the galvanometer could be recorded on film in a drum camera. Alternatively, a potentiometer was used for steady state readings. Rates of corrosion were determined by measurement of the weight loss of the sample; a resistance-capacitance bridge 'iManuscript