On possible existence of pseudobinary mixed valence fluorides of Ag(I)/Ag(II): a DFT study.

The DFT calculations performed within local density approximation disclose conceivable existence of two novel mixed-valence Ag(I)/Ag(II) fluorides, Ag(2)F(3), i.e., Ag(I)Ag(II)F(3) and Ag(3)F(4), i.e., Ag(I)(2)Ag(II)F(4). Ag(2)F(3) is predicted to crystallize in three equally stable NaCuF(3)-, KAgF(3)-, or CuTeO(3)-type structures, while Ag(3)F(4) should be isostructural to Na(2)CuF(4). The calculated vibration-corrected energies of formation at 0 K of Ag(2)F(3) and Ag(3)F(4) (in their most stable polytypes) from binary fluorides are negative but small (respectively, -0.09 eV and -0.21 eV per formula unit). Formation of Ag(3)F(5) (which, in fact, is a mixed valence Ag(I)/Ag(III) salt) from binary fluorides is much less likely, since the energy of formation is quite positive of about a quarter eV. The predicted volumes per formula unit for all forms of Ag(2)F(3) are larger and that for K(2)CuF(4)-type Ag(3)F(4) is smaller than the sum of volumes of the corresponding binary fluorides; Ag(2)F(3) should not form at high pressure conditions due to a decomposition to the binary constituents. Ag(2)F(3) and Ag(3)F(4) should exhibit genuine mixed- and not intermediate-valence with quite different coordination spheres of Ag(I) and Ag(II). Nevertheless, they should not be electric insulators. Ag(2)F(3) is predicted to be a metallic ferrimagnet with a magnetic superexchange coupling constant, J, of -2 meV while Ag(3)F(4) should be a metallic ferromagnet with J of +52 meV. Since Ag(2)F(3) and Ag(3)F(4) are at the verge of thermodynamic stability, a handful of exothermic reactions have been proposed which could yield these as yet unknown compounds.