First Principles Study of Surface Stability and Interfacial Optical Response of InP(001) in Aqueous Environments

Abstract

Shifting our reliance from activities related to greenhouse gas emissions is important to suppress global surface temperature rise. Renewable energy is key to decarbonising industrial sectors; specifically, photoelectrochemical water splitting using III-V tandem cells provides a viable route. However, atomistic-scale insight into interfacial changes--such as phase transitions and surface passivation--is rather limited under electrochemical conditions. Understanding the structural and temporal changes at the semiconductor--electrolyte interface is therefore important. III-V semiconductors, such as indium phosphide and its associated multinary alloys, show high performance in direct solar water splitting. In this work, we use first-principles calculations to investigate the thermodynamic stability of InP(001) surface reconstructions. We further analyse structural changes induced by oxygen exposure to mimic electrochemical conditions, providing a theoretical framework aligned with previous experimental results. Finally, for stable surface reconstructions, ab initio molecular dynamics and reflection anisotropy spectroscopy are used to analyse the evolution of optical anisotropy. These results could provide a correlation with real-time experimental observations at solid--vacuum and solid--liquid interfaces, where the latter is modelled as water in contact with the semiconductor surfaces. This study elucidates the thermodynamic landscape, demonstrating that P--dimer surfaces, specifically the $\beta2(2\times4)$. P-rich (2$\times$2) configurations, dominate upon oxygen exposure. Furthermore, our time-resolved optical anisotropy analysis reveals that the semiconductor--water interface induces significant temporal variations in optical signatures, driven by \ce{H2O} dissociation and subsequent ion migration. The computational insights gained from these InP(001) surfaces provide a foundational reference for investigating more complex III-V ternary alloys such as AlInP or GaInP. Future work could leverage computational hydrogen electrode model to analyse H/OH (co-)adsorption. Additionally transitioning from the IP-RPA toward GW or BSE methods will improve the accuracy of simulated time--resolved optical transitions, thereby capturing the higher-order electronic effects present.