Na super ionic CONductor (NASICON), Na1+xZr2SixP3-xO12, is an excellent solid-state electrolyte on account of its high sodium-ion conductivity, which can be further enhanced through chemical doping. However, the underlying sodium diffusion mechanism and how it is affected by chemical doping are not fully understood. This, in part, stems from the ambiguity between reported average structure models, where there is variation in the number and location of the sodium crystallographic sites. In this work, we present a neutron scattering study on monoclinic NASICON and Mg2+-doped NASICON, with a focus on the characterization of the sodium sites. Our difference Fourier maps show that the sodium sites have limited long-range order at temperatures as low as 3 K. In addition, our characterization of Mg2+-doped NASICON suggests that although Mg2+ preferentially dopes the secondary Na3PO4 phase rather than the bulk NASICON phase, its NASICON sodium sites are comparatively more diffuse. Furthermore, quasi-elastic neutron scattering data show broader spectra for the Mg-doped NASICON compared to NASICON at 300 K, supporting a more dynamic environment. We propose that the more diffuse character of the sodium sites arising from the introduction of Mg2+ contributes to the improvement of ionic conductivity in Mg-doped NASICON near room temperature.

In parallel with advances in the synthesis of solid-state ionic conductors, there is a need to understand the underlying mechanisms behind their improved ionic conductivities. This can be achieved by obtaining an atomic level picture of the interplay between the structure of materials and the resultant ionic diffusion processes. To this end, the structure and dynamics of Mg²⁺-stabilized rotor phase material γ-Na₃PO₄, characterized by neutron scattering, are detailed in this work. The Mg²⁺-stabilized rotor phase is found to be thermally stable from 4 to 650 K. However, signatures of orientational disorder of the phosphate anions are also evident in the average structure. Long-range Na⁺ self-diffusion was probed by quasi-elastic neutron scattering and subsequently modeled via a jump diffusion matrix with consideration of the phosphate anion rotations. The resultant diffusion model points directly to coupled anion-cation dynamics. Our approach highlights the importance of considering the whole system when developing an atomic level picture of structure and dynamics, which is critical in the rational design and optimization of energy materials.