LAUSR

Despite the efforts to implement the idea of a deep level impurity intermediate band (IB) into bulk solar cell materials, a breakthrough in efficiency increase has not yet been achieved. Taking Sn-doped ${\mathrm{CuGaS}}_{2}$ as an example, we investigate the problem here from the perspective of defect physics, considering all possible charge states of the dopant and its interaction with native defects. Using an optimized hybrid functional, we find that ${\mathrm{Sn}}_{\mathrm{Ga}}$ has not only a donor-type (+/0), but also an acceptor-type ($0/\ensuremath{-}$) charge transition level. We estimate the probability of the optical transition of an electron from/to the neutral defect to/from the conduction-band edge to be about equal, therefore, the lifetimes of the excited carriers are probably quite short, limiting the enhancement of the photocurrent. In addition, we find that doping with ${\mathrm{Sn}}_{\mathrm{Ga}}$ leads to the spontaneous formation of the intrinsic acceptor ${\mathrm{Cu}}_{\mathrm{Ga}}$ defects which passivate the donor ${\mathrm{Sn}}_{\mathrm{Ga}}$ and pin the Fermi level to a position (1.4 eV above the valence-band edge) where both defects are ionized. As a result, the possibility of absorption in the middle of the visible range gets lost. These two recombination and passivation mechanisms appear to be quite likely the case for other donors and other similar host materials as well, explaining some of the experimental bottlenecks with IB solar cells based on deep level impurities.