The open-circuit potential (OCP, $V_\mathrm{oc}$) is the potential of the working electrode (WE) relative to the reference electrode (RE) under open-circuit condition (no potential or current is being applied to the cell). It can be established by leaving the potential of the WE uncontrolled, allowing it to achieve equilibrium with the electrolyte. This can be done in both the light and the dark. The open circuit potential is obtained by monitoring the rate of change of the WE potential until it is negligible.
For a PEC water splitting device, the redox couples of interest are the H$^+$/$\mathrm{H_2}$ couple for a p-type semiconductor photocathode, and the $\mathrm{O_2}$/$\mathrm{H_2}$O couple for an n-type semiconductor photoanode. In the case of an n-type semiconductor, the established thermodynamic equilibrium ($E_F = E_\mathrm{O_2/H_2O}$) between the semiconductor and the electrolyte results in an upward band bending of the semiconductor surface. In the dark, the measured $V_\mathrm{oc-dark}$ is thus the $E_F / E_\mathrm{O_2/H_2O}$ relative to the RE. Whereas in the light, the $E_\mathrm{F_n}$ rises because of the higher occupancy of the conduction band and the $E_\mathrm{F_p}$ equilibrates with the $E_\mathrm{redox}$ ($E_\mathrm{F_p} = E_\mathrm{O_2/H_2O}$). The measured $V_\mathrm{oc-light}$ is thus the $E_\mathrm{F_n}$ relative to the RE. Furthermore, under illumination, the photogenerated holes can accumulate at the surface lowering the barrier for electrons with increasing light intensity until the electrons can reach the surface at the same rate as holes, at which condition the measured $V_\mathrm{oc-light}$ approximates the $V_\mathrm{fb}$ versus the RE.
The photovoltage $V_\mathrm{ph}$ generated by the semiconductor is given by the difference between the hole and electron quasi-Fermi levels under illumination, that is, by the free energy difference between the majority carriers and the photoexcited minority carriers. Illuminating the electrode surface will shift the Fermi level of the bulk (measured $V_\mathrm{oc}$) towards more anodic potentials for a p-type material and towards more cathodic potentials for a n-type material. Accordingly,
\[V_\mathrm{ph} = |E_\mathrm{f_n}-E_\mathrm{f_p}|=|V_\mathrm{oc-light}-V_\mathrm{oc-dark}|\]Note, however, that the $V_\mathrm{ph}$ measured at open circuit may not exactly reflect the $V_\mathrm{ph}$ under operating conditions since the latter must also include the kinetic overpotential for the reaction of interest.
The ideal OCP should be equal to a value of $V_\mathrm{bi}$ tha tis determined by the Fermi level difference of the two materials. Neverthless, the actual OCP values display a $V_\mathrm{ph}$ that is determined by quasi-Fermi level splitting and is affected by photovoltaic losses arising because of charge generation, separation, and recombination. In this regard, a dynamic change in
Reference:
-
Photoelectrochemical Water Splitting Standards, Experimental Methods, and Protocols. 2013, Springer.
-
Walter M G, Warren E L, McKone J R, et al. Solar water splitting cells. Chemical Reviews, 2010, 110(11): 6446-6473