Photoelectrochemical (PEC) water splitting offers a scalable route to solar hydrogen production; however, plasma treatments of oxide photoanodes often produce inconsistent performance gains because the dominant loss mechanisms vary strongly with film thickness. Here, we establish a thickness-guided design rule for remote plasma engineering of WO3 photoanodes with thicknesses ranging from 300 to 900 nm, using Ar (physical control), O2 (donor-suppressing), and N2 (donor-compensating) plasmas. X-ray diffraction and UV–Vis spectroscopy confirm phase-pure monoclinic WO3 with essentially unchanged light absorption and optical bandgap following plasma treatment, indicating that performance variations are not governed by bulk crystallinity or bandgap modification. In contrast, X-ray photoelectron spectroscopy reveals plasma-dependent changes in surface chemical states, consistent with modulation of reduced tungsten oxidation states and oxygen vacancy defects.
PEC performance evaluated by linear sweep voltammetry and chopped-light chronoamperometry, together with electrochemical impedance spectroscopy and Mott–Schottky analysis, reveals a clear thickness-dependent regime switch. In thin films (300 nm), plasma treatment leads to reduced photocurrent, consistent with enhanced surface recombination and plasma-induced surface damage. At intermediate thickness (600 nm), N2 plasma achieves the optimal balance between carrier compensation and recombination, delivering the highest photocurrent density (~0.6–0.8 mA cm-2 at 1.23 V vs RHE). For thicker films (900 nm), where charge transport and interfacial charge transfer increasingly limit performance, N2 plasma produces the strongest enhancement, consistent with improved interfacial kinetics. Overall, these results demonstrate that identical plasma chemistries can lead to opposite PEC outcomes depending on WO3 thickness, providing a practical and scalable strategy to optimise WO3 photoanodes by matching plasma treatment to the thickness-dependent dominant loss mechanism.