Publication date: 22nd July 2026
Artificial leaves based on III-V tandem photoabsorbers currently deliver the highest solar-to-hydrogen (STH) efficiencies for direct solar water splitting [1]. These devices typically employ a thin, wide-bandgap AlInP window layer to electronically passivate the photoabsorber surface, suppress surface recombination, and promote selective electron transport, together with a TiO2 overlayer [2] that protects the III-V photoabsorber from corrosion while enabling electron extraction to the catalyst. However, their unassisted PEC performance is critically limited by the uncontrolled oxidation of the AlInP surface. During storage and upon subsequent wet-chemical removal of the protective GaAs cap layer, the AlInP surface oxidizes, forming a chemically complex mixture of Al-, In-, and P-containing oxides and hydroxides that becomes buried at the AlInP/TiO2 interface. This interfacial oxide introduces an electronic barrier that hinders selective charge carrier transport, leading to increased interfacial losses and reduced PEC performance.
Here, we present a systematic interface-engineering strategy to optimize the AlInP surface oxides prior to atomic layer deposition of the TiO2 protection layer. Two complementary surface modification pathways are investigated: selective wet-chemical etching using HCl and NH4OH, and in situ H2 plasma treatment immediately before TiO2 deposition. X-ray photoelectron spectroscopy reveals that wet-chemical treatment selectively removes the electronically unfavorable In(OH)3 species while largely preserving phosphate/phosphite surface species. The H2 plasma treatment eliminates residual carbon contamination and creates a chemically activated surface that, as confirmed by in situ spectroscopic ellipsometry, promotes uniform TiO2 nucleation, and thereby minimizes interfacial charge carrier transfer losses. The resulting interface chemistry, electronic structure, film growth, and PEC performance are correlated using X-ray and ultraviolet photoelectron spectroscopy, atomic force microscopy, and PEC characterization.
While wet-chemical treatment alone provides only a modest improvement in fill factor, H2 plasma treatment completely restores the fill factor by minimizing interfacial charge carrier transfer losses. Owing to the limited penetration depth of neutral hydrogen radicals, the highest performance is achieved by combining wet-chemical etching with H2 plasma treatment, thereby simultaneously reducing oxide thickness and optimizing the interfacial electronic structure. Transient photocurrent measurements further reveal efficient selective charge carrier transport with strongly suppressed interfacial charge carrier trapping. The optimized interface enables photocurrent densities exceeding 14 mA/cm2, STH efficiencies approaching 18%, and stable unassisted water splitting for more than 5 h in pH 0 electrolyte. To the best of our knowledge, this represents one of the highest reported stability–efficiency combinations for 1 cm2 III-V artificial leaves operating under highly acidic conditions. These results demonstrate that precise control of the buried semiconductor/oxide heterointerface is a key design principle for realizing durable, high-efficiency artificial leaves for scalable solar hydrogen production.
We gratefully acknowledge the Helmholtz Association of German Research Centers (HGF) and the Federal Ministry of Research, Technology and Space (BMFTR), Germany for supporting the development of demonstrators for direct solar water splitting within the framework of the H2Demo project (No. 03SF0619A-K).
