Synthesis of new oxyfluoride phases, a possible new family of fluoro-LISICON Li+ conductors.

Lithium ion conducting solid electrolytes are of great potential interests for possible all solid state lithium battery applications. Several families of materials are being considered including LISICONs, garnets and argyrodite. This thesis focuses on the synthesis and characterisation of new oxyfluoride materials in the LISICON family following the recent report of high-lithium ion conductivity in Li5SiO4F.

Studies on Li5SiO4F have given four new sets of results. (i) Samples heated for extended times show evidence of Li2O loss in the electrical property data with the appearance of a lithium-deficient grain boundary phase. (ii)The equivalent circuit analysis of high frequency impedance data shows contribution of a dipole reorientation effect to ac conductivity. (iii) Attempts to dope Li5SiO4F with a range of di-, tri-, tetra- and penta-valent cations yielded a large number of new, previously unreported oxyfluoride phases that appear to belong to the LISICON family. (iv) The detailed structure of the two polymorphs of Li5SiO4F is not known, but the X-ray diffraction pattern of the low temperature, α-polymorph has a very strong subcell with hexagonal symmetry that appears to be closely related to the ZnO structure. This is the first example of a tetrahedral, LISICON-based structure with a hexagonal subcell that exhibits complete disorder of both cation and anion positions.

A detailed study on the compositions in the ternary system: Li2O-SiO2-LiF found four new phases in addition to previously reported Li5SiO4F. These are: Phase B, γ-Phase B, Phase T and Phase N. Phase B, with approximate stoichiometry Li2O:SiO2:LiF=47:28:25, also has a ZnO-like hexagonal subcell and a modest ionic conductivity of ~5.6×10-8 Scm-1 at 100 °C. The high temperature polymorph of Phase B, labelled γ-Phase B formed over a wide solid solution range. γ-Phase B has an orthorhombic subcell with doubled volume that is clearly derived from the ZnO-like hexagonal subcell of Phase B. Nevertheless, the supercell structures of both Phase B and γ-Phase B are still unknown. γ-Phase B had an ionic conductivity of ~1.1×10-6 Scm-1 at 100 °C. Phases T and N had similar composition to Li5SiO4F but significantly different XRD patterns and electrical properties, the desired composition for single Phase N had not been found. Phase T showed a poor ionic conductivity similar to that of Li4SiO4.

A preliminary survey of the Li3PO4-Li4SiO4-LiF system found three possible new phases labelled Phase S, γ-Phase S and Phase A. Phase S was also indexed on a ZnO-like hexagonal subcell. γ-Phase S is closely related to Phase S but it is probably thermodynamically metastable and quickly decomposes/transforms to Phase A. Impedance results showed that Phase S and Phase A have only poor ionic conductivity similar to that of bulk Li4SiO4; γ-phase S shows an ionic conductivity which is approximately 2 orders of magnitude lower than that of γ-Li5SiO4F. The preliminary phase diagram study of this system also showed evidence of a significant range of F- doped Li3PO4-Li4SiO4 solid solution. The substitution mechanism is not known for certain, but may involve a novel mechanism with replacement of (PO4)3- by (LiF4)3- tetrahedra.

The various new phases reported in this study, which include Phase S, γ-Phase S, Phase A, Phase B, γ-Phase B, Phase T and Phase N, and Li5SiO4F may all belong to a new oxyfluoride family which have subcell structures derived from hexagonal ZnO, leading to a possible new family of fluoro-LISICON Li+ ion conductors. Phase S, Phase B and α-Li5SiO4F were categorised as belonging to the β-structure family, while γ-Phase S, Phase A, γ-Phase B, Phase T, Phase N and γ -Li5SiO4F belong to the γ-structure family. The β family would have cations only in one set of tetrahedral sites, T+, whereas in γ family, both of the tetrahedral sites, T- and T+, are partially occupied. All of these structure have a cation:anion ratio greater than the 1:1 ratio in ZnO. This mains that the structures also contain extra Li+, probably in interstitial sites, and this may be the reason for the observed high level of Li+ ion conductivity. We do not know if the O2-, F- distribution is random. However, the stoichiometry of Li5SiO4F suggests that the O, F arrangement is ordered and in particular, this phase may be regarded as containing (SiO4F)5- pentagonal bipyramids instead of the more common SiO4 tetrahedra.