posted on 2017-12-06, 00:00authored byJaren Harrell, Muhammed Acikgoz, Hela Lieber Sasson, Iris Visoly-Fisher, Alessandro Genova, Michele Pavanello
Indium
oxide (IO) and indium tin oxide (ITO) are important metal
oxide materials with a wide array of applications. Particularly, ITO
is employed as a transparent conductive electrode in photovoltaic
systems. While bulk metal oxides are typically well characterized,
their surfaces, especially in real-life applications, can be hydroxylated
and intrinsically disordered to a level that a structure–function
prediction becomes a daunting task. We tackle this problem by carrying
out simulations based on Density Functional Theory. We propose IO
and ITO hydroxylated surfaces derived from the bcc and rombohedral
IO polymorphs (100%, 66%, 33%, and 0% hydroxylation coverages were
considered). By correlating computed quantities such as surface partial
density of states, work functions, and surface dipole strength, a
clear picture of the structure–function relationships in these
model systems emerges. In line with conclusions drawn from experiments,
we find that the density of states of 100% hydroxylated surfaces and
bulk models are unaltered by Sn doping, with the only difference being
the position of the Fermi level. The partially hydroxylated surfaces,
instead show a rich array of behaviors, including appearance of surface
states in the gap and appearance of interesting morphologies, such
as chemisorbed molecular oxygen. We also find that the hydroxylation
level affects surface dipoles in a systematic way, that is, the higher
the hydroxylation level, the higher the surface dipole (screening/reducing
the work function). Furthermore, models with In-atom vacancies show
a relatively small decrease in surface dipole with hydroxyl coverage
due to surface distortions.