Polymer/inorganic composite electrolytes were prepared by mixing of polyether-based macro monomer, NaN(SO2CF3)2 as an alkaline salt, 2, 2-dimethoxy-2-phenylacetophenone as a photoinitiator, acetonitrile as a solvent and Na3Zr2Si2PO12 (NZSP) powder as an inorganic electrolyte of NASICON type electrolyte. The composition percentage of NZSP were changed between 0 to 300wt%. After stirring, vacuum dried viscous liquid were casted to glass substrate and polymerized by UV irradiation to form composite electrolyte films. Thermal, transport and molecular properties of composite electrolytes were investigated by scanning electron microscopy (SEM), differential scanning calorimetry (DSC) and AC impedance method, respectively. Fourier transform infrared spectroscopy (FR-IR) was applied in order to investigate interaction into composite electrolytes. Moreover, interfacial resistances were measured by AC impedance on Na symmetric cells which were maintained 60oC in order to evaluate interfacial stability.
Fig. 1 show prepared polymer electrolyte (0 wt% NZSP) and composite electrolyte containing 30wt% NZSP powder (30wt% NZSP). In comparison with 0wt% sample, 30wt% sample exhibited high flexibility by improving stress distribution due to high ability of dispersion of NZSP particles, following Fig. 1(d). The composition dependence on glass transition temperature (Tg), which is defined as the phase transition point glass state to rubber state, exhibited to decreasing tendency of Tgs with addition NZSP under approximately 100wt% NZSP from Fig. 2. The decreasing Tgs were considered to cause by interaction between polymer and inorganic electrolyte or ions (cation, anion) and inorganic electrolyte particles. Moreover, this phenomena could expect increasing ionic conductivity because of improving segmental motion of polymer and cationic mobility. In the over 100wt% NZSP samples, changing Tg were not confirmed. In the comparison with 0wt% NZSP and 300wt% NZSP sample on Nyquist plots, although 0wt% NZSP sample showed symmetric semicircle spectrum, 300wt% sample showed asymmetric semicircle. In order to investigate detail of this phenomena, 300wt% sample was attempted to separate resistance components. As a result, two resistance components were contained, bulk resistance (Rb) and grain boundary resistance (Rgb) between NZSP particles. Fig. 3 show NZSP composition dependence on ionic conductivity. In the low composition range (<100wt% NZSP), relatively high value were confirmed in comparison with 0wt%. In the contrast, over 100wt% NZSP samples exhibited low value including Rb and Rgb by contribution of presence of high Rgb. The composition dependence on interfacial resistance (Ri) between Na metal and electrolyte is shown Fig. 4. The Ri exhibited lowest value at 200wt% NZSP sample from Fig. 4. The 200wt% NZSP sample was considered to the change point for Na ion transport mechanism. Therefore, Ri were suggested to could decrease by addition NZSP from above results.
The polymer/inorganic electrolytes shifted difference transport mechanism from low to high composition range. Although ionic conductivity showed relatively high value in the low composition range, low value were confirmed over 100 wt% NZSP samples. The increasing ionic conductivity were considered to cause by improving segmental motion because of some interaction of composite electrolyte. The high composition range showed low value due to presence of high Rgb. In the interfacial stability, Ri exhibited lowest value at 200wt% NZSP sample. Therefore, the composite electrolytes of optimal composition are expected to realize high-rate operated all-solid-state Na battery.

