Prediction of Novel Li-AgII-F Compounds using Evolutionary Algorithms
Katarzyna Kuder, Wojciech Grochala
TL;DR
This work addresses the lack of characterization for Li–Ag(II)–F ternaries and their potential magnetic properties by combining an XtalOpt-based evolutionary search with DFT+U calculations to identify low-energy Li–Ag–F structures. The study finds several metastable polymorphs relative to LiF and AgF2 on the convex hull, with entropic stabilization unlikely under typical conditions, yet proposes non-equilibrium synthesis routes involving fluorination with $F_2$ or $F^*$ radicals that could realize these phases. A standout result is LiAgF3_2, which exhibits a record-like AFM superexchange of $J = -359$ meV within $[Ag_2F_7]$ dimers, highlighting exceptionally strong magnetic interactions in Ag(II)-fluoride networks. Overall, the work expands the materials space for highly correlated fluorides and provides actionable guidance for experimental attempts to synthesize and study these metastable, magnetically rich systems.
Abstract
This work provides a theoretical exploration of the thermodynamic stability and magnetic behaviour of previously unknown ternary Li AgII F compounds. Convex-hull analysis shows that all predicted structures lie slightly above the LiF plus AgF2 decomposition line, indicating a natural tendency toward phase separation; nevertheless, their negative formation energies relative to AgF, LiF, and F2 or F suggest that alternative synthetic pathways may be feasible for these compounds. All studied structures show preference for antiferromagnetic ground state. Notably, the triclinic LiAgF3 type2 is predicted to exhibit an exceptionally large superexchange constant, J equal to minus 358 meV, within Ag2F7 dimers, placing it above the strongest known magnetic exchange interactions reported to date.
