In this work, we focus on multinuclear solid-state NMR, X-ray, and neutron characterization to explore the relationship between composition, crystal structure, and defect chemistry in a series of complex sodium and lithium niobium oxides. The role of defects on electrochemical transport properties in these new high-rate electrode or solid electrolyte materials will be discussed. Starting with average structure models from X-ray and neutron diffraction, we then turn to a local structure perspective from NMR that is more sensitive to defects and disorder. One- and two-dimensional ^6,7Li and ²³Na NMR spectra provide insights on mobile cation positions and dynamics as well as alkali sublattice vacancies. (Ir)reversible changes upon cycling are identified. DFT calculations and numerical simulations support the spectral assignments in these complex oxides.

A central focus of this work are the compounds NaNb₇O₁₈ and NaNb₁₃O₃₃, which we report as battery materials and study by NMR for the first time. Both structures are derivatives of the Wadsley–Roth structure that can rapidly intercalate lithium. However, the role of sodium in the structure raises interesting questions. Sodium nominally sits in a single crystallographic site with a square planar coordination environment, which is rather unusual for sodium and gives rise to a large and characteristic quadrupolar lineshape in ²³Na NMR. What we find is that even the pristine NaNb₁₃O₃₃ phase, i.e., before lithium is intercalated, is more complicated than the static picture previously reported by X-ray diffraction. Moreover, only half the sodium in NaNb₇O₁₈ has been accounted for in the structure reported from diffraction data. We show that, in both structures, sodium sits not only in the square planar site but occupies perovskite-like sites within tunnels in the Wadsley–Roth structure. This conclusion is supported by DFT and phase boundary mapping. Comparisons are made to related sodium niobium oxide compounds.