Alignment of Redox Levels at Semiconductor–Water Interfaces

Abstract

We determine the band alignment between various semiconductors and liquid water by combining molecular dynamics (MD) simulations of atomistic interface models, electronic-structure calculations at the hybrid-functional and GW level, and a computational standard hydrogen electrode. Our study comprises GaAs, GaP, GaN, CdS, ZnO, SnO 2 , rutile TiO 2 , and anatase TiO 2 . For each semiconductor, we generate atomistic interface models with liquid water at the pH corresponding to the point of zero charge. The molecular dynamics are started from two kinds of initial configurations, in which the water molecules are either molecularly (m) or dissociatively (d) adsorbed on the semiconductor surface. The calculated band offsets are found to be strongly influenced by the adsorption mode at the semiconductor−water interface, leading to differences larger than 1 eV between m and d models of the same semiconductor. We first assess the accuracy of various ab initio electronic-structure schemes. The use of a standard hybrid functional leads to large errors for the conduction band edge but nevertheless accounts accurately for the position of the valence band edge. One-shot GW calculations with a starting point at the semilocal density functional level do not yield any improvement. It is necessary to turn to one-shot GW calculations based on a hybrid-functional starting point to achieve a noticeable improvement in the determination of the band edges, with mean average errors ranging between 0.23 and 0.27 eV. The use of state-of-the-art quasiparticle self-consistent GW schemes does not lead to any further improvement for the set of semiconductors under investigation. Further improvement with mean average errors of 0.20 eV is obtained when turning to hybrid-functional and GW methods, in which the experimental band gap of the semiconductor is enforced by construction. The present work sets a benchmark for the accuracy by which band edges at semiconductor−water interfaces can be obtained with current advanced electronic-structure methods. In particular, the importance of providing an atomistic description of the semiconductor−water interface is emphasized.

Nonempirical dielectric-dependent hybrid functional with range separation for semiconductors and insulators

“…delocalization error) prevalent in approximate xc functionals [9], which further contributes to improved band gaps as well as to more accurate lattice parameters and atomization energies [10]. Hence, hybrid functionals have been widely applied to the problems of point defects in solids [11][12][13] and of band alignments at interfaces [14][15][16][17][18].…”

Absolute band alignment at semiconductor-water interfaces using explicit and implicit descriptions for liquid water

“…Step e1/i1: Bulk levels PBE band gaps, and band alignment with respect to the average electrostatic potential were calculated for the bulk materials using the lattice constants of Ref. 28…”

Ab initio simulations of liquid electrolytes for energy conversion and storage

“…For example, a large number of first-principles simulations have been reported for solid/liquid interfaces in ion batteries [53,138,[183][184][185][186][187] and photoelectrochemical cells for solarwater-splitting. [31,32,59,77,[188][189][190][191][192][193][194][195][196] Despite the impressive development in the FPMD method, making the connection between theoretical observables and experimental measurements remains a challenging task, which motivates further developments in the community. A primary example is the electrode/electrolyte interface, a key component in several energy conversion and storage devices, for which additional physical factors should be considered to provide a realistic description of the systems.…”