Abstract
The issue of the allowed sign of the differential capacity C of electrochemical interfaces has a long history dating back to the so‐called ‘‘Cooper–Harrison catastrophe.’’ Previously suggested ‘‘electromechanical models’’ are modified to include entropic contributions; the possibility of C<0 for an isolated electric cell is supported by rigorous solution of the model. We also provide new evidence that for an electric cell in contact with a potentiostat (‘‘extended system’’) the overall C of the cell must be positive; the contribution of an individual double layer may still be negative. The previous statistical mechanical derivation of the upper boundary for C−1 is generalized for a quite general model of an electrolyte in contact with hard charged walls. We also discuss the possibility of electric instabilities and phase transitions in an extended system when an isolated prototype possesses a negative capacity branch.