The impedance of a growing anodic oxide layer is inductive in a rcertain frequency range. This effect is related to the occurrence of transients after changes in current density or applied electric field. Measurements on aluminum samples in the frequency range 5 mHz-10 kHz and the current density range 3 X 10-6-5 • ]~O -3 A/cm 2 and in various electrolytes are presented and compared with a theory in which the relaxation effect is ascribed to a buildup of surface charge at the metal-oxide interface.It has long been recognized that the study of nonsteady-state situations during anodization is helpful to an understanding of the mechanism of anodization. A very early measurement was made by Baumann in 1939 (1) who studied the frequency dependence of the impedance during porous oxide formation on aluminum. The techniques used most commonly are: a sudden change of the applied field, which is kept constant afterward (potentiostatic transient); a sudden change of the current density which is kept constant afterward (galvanostatic transient); and the superposition of a small a-c signal on the d-c signal.Potentiostatic transients have been measured on tantalum by Vermilyea (2), Young (3), and Taylor and Dignam (4), and on aluminum by Vermilyea (2) and Dignam and Ryan (5). Galvanostatic transients have been measured on tantalum by Dewald (6), Vermilyea (2), and Young (3) and on aluminum by Goad and Dignam (7). A-C measurements have been performed on aluminum by Baumann (1), Winkel, Pistorius, and van Geel (8), and Goad and Dignam (9), and on tantalum by Taylor and Dignam (4). In all these experiments it was shown that the system after a change does not immediately assume the new steady-state values for current or overpotential, but that it relaxes to the new steady state in a certain time. This relaxation time is inversely proportional to the d-c ionic current density.Several theories have been put forward to explain this relaxation. Bean, Fisher, and Vermilyea (10) ascribe it to a sluggishness in the response of the number of charge carriers, assumed to be metal interstitials, to the variation of the applied field. Their approach has come in for criticism, on the one hand, from Young (3), who showed that it does not satisfactorily describe the acceleration of the ionic current which is found when a potential step is applied, and, on the other hand, from Dignam ( 11), who challenged it on the grounds of free path considerations. Dignam and co-workers (4, 7, 11) consider a relaxation of the dielectric polarization of the oxide to be responsible for the effects. Their approach leads to a set of equations which satisfactorily accounts for the experiments and which is used as a starting point for the theoretical considerations of the present paper. The mechanism, however, by which these polarization changes occur is unclear.The present work presents a continuation and extension of the cited work. It gives the results of impedance measurements during anodization of aluminum in much wider frequency and current ranges than used hi...
