Semiconductor junctions are believed to form a depletion region at the band edge of the semiconductor as the chemical potentials for electrons (work functions) are aligned to the same level. Here, we demonstrated that ultrathin Ni film (less than 4 nm thick)/Si junction-based photoelectrochemical (PEC) devices have no depletion region due to three distinct phenomena: (i) the electrostatic or electrochemical potential extrinsically charged to the electrolytic-capacitive Ni surface dominates rather than the chemical potential of electrons (work function) of the bulk Ni, (ii) the charged potential is dynamically variable depending on the reaction and is rapidly volatile so as not to be constant; therefore, (iii) the charged potential is misaligned with the chemical potential of Si under equivalent circuit conditions. Such junction PEC devices were shown to follow a novel operating principle in which the output voltage (open circuit potential) is generated by the electrochemical potential charged at the Ni surface, and not by the light-induced potential (photovoltage) in Si. In addition, due to the bipolar charging nature, the ultrathin Ni film was effective in achieving a high open circuit potential in both p-Si photocathodes (0.57 V) and n-Si photoanodes (0.45 V). These anomalous results were not explained by the classical Schottky diode model based on the equilibrium of diffusion-drift current but by establishing a new model based on the equilibrium of the diffusion-charging current without accounting for the depletion region. Our findings provide an explanation for the unexpected results of the nanostructured PEC devices and insight into a new design that can overcome conventional limitations.