Lead-halide perovskites have emerged as a promising class of semiconductors;
however they suffer from issues related to lead-toxicity and instability. We report results of a firstprinciples-based design of heavy-metal-based oxynitrides as alternatives to lead-halide
perovskites. We have used density-functional-theory calculations to search a vast composition
space of ABO2N and ABON2 compounds, where B is a p-block cation, and A is an alkaline, alkaliearth, rare-earth or transition metal cation, and identify 10 new ABO2N oxynitride semiconductors
that we expect to be formable. Specifically, we discover a new family of ferroelectric
semiconductors with A3+SnO2N stoichiometry (A = Y, Eu, La, In, and Sc) in the LuMnO3-type
structure, which combine the strong bonding of metal oxides with the low carrier effective mass
and small, tunable band gaps of the lead-halide perovskites. These tin oxynitrides have predicted
direct band gaps ranging from 1.6 – 3.3 eV, and a sizeable electric polarization up to 17 μC/cm2
,
which is predicted to be switchable by an external electric field through a non-polar phase. With
their unique combination of polarization, low carrier effective mass and band gaps spanning the
entire visible spectrum, we expect ASnO2N ferroelectric semiconductors will find useful
applications as photovoltaics, photocatalysts, and for optoelectronics.