Computational design of polaronic conductive Li-NASICON mixed ionic–electronic conductors

Abstract

Li-NASICONs featuring three-dimensional corner-sharing frameworks have been experimentally and theoretically demonstrated to be fast Li-ion conductors, with high room-temperature ionic conductivities. In this study, we conduct first-principles calculations to investigate the polaronic conduction in Li-NASICONs comprising different transition metal (TM) elements (Ti, V, Cr, Mn, Fe, Mo) and evaluate their potential as mixed electron-ion conductors. We find that the electron polaron conductivity of octahedral–octahedral (oct–oct) hopping in Li-NASICONs is strongly correlated with the Jahn–Teller activity on the TM site. While LiMn₂(PO₄)₃ and Li₃Cr₂(PO₄)₃ have a low predicted electron polaron conductivity owing to the J–T distortion resulting from the polaron formation, LiFe₂(PO₄)₃ is predicted to have the most facile polaron transport, which stems from the elimination of its intrinsic J–T distortion upon polaron formation. We also find that polaron hopping between octahedral and tetrahedral sites contributes limited electronic conductivity because of the large difference in the polaron formation energies on these sites. For polaron transport along purely tetrahedral sites, we find limited transport through the V⁵⁺/V⁴⁺ pair in LiGe₂(VO₄)₃ and LiTi₂(VO₄)₃ but better conductivity through the Mo⁶⁺/Mo⁵⁺ pair in LiTi₂(MoO₄)₃. Our research suggests that Li-NASICONs with TM elements involving no J–T distortion upon polaron formation on octahedral sites and Mo on tetrahedral sites are promising candidates for mixed ionic electronic conductors (MIECs).