While photoelectrochemical (PEC) solar-to-hydrogen efficiencies have greatly improved over the past few decades, advances in PEC durability have lagged behind. Corrosion of semiconductor photoabsorbers in the aqueous conditions needed for water splitting is a major challenge that limits device stability. In addition, a precious-metal catalyst is often required to efficiently promote water splitting. Herein, we demonstrate unassisted water splitting using a non-precious metal molybdenum disulfide nanomaterial catalytic protection layer paired with a GaInAsP/GaAs tandem device. This device was able to achieve stable unassisted water splitting for nearly 12 hours, while a sibling sample with a PtRu catalyst was only stable for 2 hours, highlighting the advantage of the non-precious metal catalyst. In situ optical imaging illustrates the progression of macroscopic degradation that causes device failure. In addition, this work compares unassisted water splitting devices across the field in terms of the efficiency and stability, illustrating the need for improved stability.
TOC GRAPHICThe primary strategy that has emerged to mitigate semiconductor surface corrosion is depositing thin films, such as titanium dioxide, that can act as protective barriers to prevent the electrolyte from coming into contact with the semiconductor surface. 7,8 These films have to be stable, thin enough to prevent significant light blocking, conformal, and conductive to create a stable and functional device. 9-11 Furthermore, if these films do not demonstrate intrinsic catalytic activity, an additional hydrogen and/or oxygen evolution catalyst is needed to promote efficient water splitting. Molybdenum disulfide nanomaterials have been shown to stabilize a variety of singlejunction Si and III-V PEC systems, functioning as a hydrogen evolution reaction (HER) catalyst and protection layer. [9][10][11][12][13][14][15] Because of the promising performance in single-junction photocathodes, it is of interest to use MoS2 with tandem semiconductor systems to improve the stability during unassisted solar water splitting.While most III-V-based unassisted water splitting devices to date have incorporated a Ga0.51In0.49P (hereafter GaInP2) (1.8 eV) top cell, device lifetimes have been limited to <100 h. 7,16 GaxIn1-xAsyP1-y (1.7 eV) has shown promise as a PV material and has been paired with a GazIn1-zAs bottom cell (1.1 eV) for efficient tandem PV systems, motivating efforts to incorporate GaxIn1-xAsyP1-y into PEC systems and investigate the stability of this quaternary top cell. [16][17][18][19] The composition of GaxIn1-xAsyP1-y (hereafter GaInAsP), nominally x ~ 0.68 and y ~ 0.34, gives the desired bandgap of 1.7 eV and a lattice constant matching that of GaAs. 16,18 A GaInAsP/GaAs (1.7/1.4 eV) pairing has a predicted maximum STH efficiency of ~12%, 16,19 far from the ideal combination of absorbers to achieve the highest of efficiencies, however sufficiently high to perform durability studies on active, unassisted water splitting systems, guiding t...