Conway [4] proposed model-based diagnostics to identify activation, diffusion, and leakage-resistance controlled mechanisms from voltage vs. time measurements. We show that for the limited voltage range window available, the method is subjective and ambiguous for real measurements. We have proposed and carried out an alternative set of extensive experiments to gain insights into the self-discharge process for an in-house capacitor with sulfuric acid as electrolyte and activated carbon as electrodes. We varied electrode thickness, the distance between the electrodes, and the state of mixing in the electrolyte. We also measured self-discharge of electrochemically isolated charged half-cells (Figures 1a,b). The results bring out interesting and unambiguous findings, independent of the degree of fit with the expected variations for various controlling mechanisms. They rule out leakage resistance as the cause of self-discharge and disprove the shuttle mechanism [2,3] (Figure 1c,d). The analysis of measurements with electrodes of different thicknesses establishes that activation process controlled self-discharge does not dominate either. And the predicted effect of charge-redistribution is in the opposite direction. The role of the complex interplay of diffusion and electrochemical reactions holds promise.
