Optical Gaps of Ionic Materials from GW/BSE-in-DFT and CC2-in-DFT

This work presents a density functional theory (DFT)-based
embedding
technique for the calculation of optical gaps in ionic solids. The
approach partitions the supercell of the ionic solid and embeds a
small molecule-like cluster in a periodic environment using a cluster-in-periodic
embedding method. The environment is treated with DFT, and its influence
on the cluster is captured by a DFT-based embedding potential. The
optical gap is estimated as the lowest singlet excitation energy of
the embedded cluster, obtained using a wave function theory method:
second-order approximate coupled-cluster singles and doubles (CC2),
and a many-body perturbation theory method: GW approximation combined
with the Bethe–Salpeter equation (GW/BSE). The calculated excitation
energies are benchmarked against the periodic GW/BSE values, equation-of-motion
coupled-cluster singles and doubles (EOM-CCSD) results, and experiments.
Both CC2-in-DFT and GW/BSE-in-DFT deliver excitation energies that
are in good agreement with experimental values for several ionic solids
(MgO, CaO, LiF, NaF, KF, and LiCl) while incurring negligible computational
costs. Notably, GW/BSE-in-DFT exhibits remarkable accuracy with a
mean absolute error (MAE) of just 0.38 eV with respect to experiments,
demonstrating the effectiveness of the embedding strategy. In addition,
the versatility of the method is highlighted by investigating the
optical gap of a 2D MgCl2 system and the excitation energy
of an oxygen vacancy in MgO, with results in good agreement with reported
values.