Li and group-III impurity doping in ZnSnN2 : Potential and limitations

II--IV nitrides and their alloys represent an earth-abundant and potentially cost-efficient alternative to the well-developed AlN-GaN-InN system. A major drawback with the II--IV nitrides is that $\mathrm{Zn}\mathrm{Sn}{\mathrm{N}}_{2}$, the lowest band gap material, exhibits unfavorably high carrier concentrations for as-grown, stoichiometric material, limiting the material systems potential use in applications such as solar cells and light-emitting diodes. Lithium (Li) has been suggested as a shallow acceptor defect in $\mathrm{Zn}\mathrm{Sn}{\mathrm{N}}_{2}$ if substituting for Zn, and hence doping with Li has been identified as a possible way to improve the electronic properties. Herein, theoretical calculations by hybrid functional density functional theory have been employed and extended to include defect complexes as well, which to this point remained unexplored. The calculations reveal that even though Li on the Zn site (the ${\mathrm{Li}}_{\text{Zn}}$) is an acceptor, the defect may easily complex with the ${\mathrm{Li}}_{i}$ donor, rendering the complex neutral. Our theoretical findings are supported by a Li-doping series of $\mathrm{Zn}\mathrm{Sn}{\mathrm{N}}_{2}$, where a doping concentration ranging from $2.10\ifmmode\times\else\texttimes\fi{}{10}^{19}\phantom{\rule{4pt}{0ex}}{\mathrm{cm}}^{\ensuremath{-}3}$ to $1.85\ifmmode\times\else\texttimes\fi{}{10}^{20}\phantom{\rule{4pt}{0ex}}{\mathrm{cm}}^{\ensuremath{-}3}$ was obtained. The $\mathit{n}$-type carrier concentration was found to be unaffected by the doping concentration, and no systematic change in the absorption onset, probably affected by a Burstein-Moss shift, was observed. Possible group-III dopants, as have been found to yield interesting results for $\mathrm{Zn}\mathrm{Ge}{\mathrm{N}}_{2}$, such as In, Ga, Al, and B, have also been investigated as an alternative dopant in $\mathrm{Zn}\mathrm{Sn}{\mathrm{N}}_{2}$.