The search for new direct bandgap,
earth-abundant semiconductors
for efficient, high-quality optoelectronic devices, as well as photovoltaic
and photocatalytic energy conversion has attracted considerable interest.
One methodology for the search is to study ternary and multiternary
semiconductors with more elements and more flexible properties. Cation
mutation such as binary → ternary → quaternary for ZnS
→ CuGaS2 → Cu2ZnSnS4 and ZnO → LiGaO2 → Li2ZnGeO4 led to a series of new quaternary chalcogenide and oxide
semiconductors with wide applications. Similarly, starting with GaN,
ternary nitrides such as ZnSnN2 and ZnGeN2 have
been designed and synthesized recently. However, quaternary nitride
semiconductors have never been reported either theoretically or experimentally.
Through a combination of the Materials Genome database with the first-principles
calculations, we designed a series of quaternary nitride compounds
I–III–Ge2N4 (I = Cu, Ag, Li, Na,
K; III = Al, Ga, In) following the GaN → ZnGeN2 →
I–III–Ge2N4 mutation. Akin to
Li2ZnGeO4, these quaternary nitrides crystallize
in a wurtzite-derived structure as their ground state. The thermodynamic
stability analysis shows that while most of them are not stable with
respect to phase separation, there are two key exceptions (i.e., LiAlGe2N4 and LiGaGe2N4), which
are stable and can be synthesized without any secondary phases. Interestingly,
they are both lattice-matched to GaN and ZnO, and their band gaps
are direct and larger than that of GaN, 4.36 and 3.74 eV, respectively.
They have valence band edges as low as ZnO and conduction band edges
as high as GaN, thereby combining the best of GaN and ZnO in a single
material. We predict that flexible and efficient band structure engineering
can be achieved through forming GaN/LiAlGe2N4/LiGaGe2N4 heterostructures, which have tremendous
potential for ultraviolet optoelectronics.