The Role of Sintering Temperature and Dual Metal Substitutions (Al3+, Ti4+) in the Development of NASICON-Structured Electrolyte

Abstract

The aim of this study is to synthesize Li1+xAlxTixSn2−2x(PO4) sodium super ion conductor (NASICON) -based ceramic solid electrolyte and to study the effect of dual metal substitution on the electrical and structural properties of the electrolyte. The performance of the electrolyte is analyzed based on the sintering temperature (550 to 950 °C) as well as the composition. The trend of XRD results reveals the presence of impurities in the sample, and from Rietveld Refinement, the purest sample is achieved at a sintering temperature of 950 °C and when x = 0.6. The electrolytes obey Vegard′s Law as the addition of Al3+ and Ti4+ provide linear relation with cell volume, which signifies a random distribution. The different composition has a different optimum sintering temperature at which the highest conductivity is achieved when the sample is sintered at 650 °C and x = 0.4. Field emission scanning electron microscope (FESEM) analysis showed that higher sintering temperature promotes the increment of grain boundaries and size. Based on energy dispersive X-ray spectroscopy (EDX) analysis, x = 0.4 produced the closest atomic percentage ratio to the theoretical value. Electrode polarization is found to be at maximum when x = 0.4, which is determined from dielectric analysis. The electrolytes follow non-Debye behavior as it shows a variety of relaxation times.

Process optimisation for NASICON-type solid electrolyte synthesis using a combination of experiments and bayesian optimisation

“…24,33 Besides compositional control, Li-ion conductivity is also affected by density, morphology, and elemental distribution, 21 and thus, controlling process parameters, such as heating conditions during the solid-state reaction, is effective in improving conduction performance. 34 Optimising the composition and production process is critical for use in the materials industry, but the traditional trial-and-error approach often requires economic cost and time. Previously, machine learning was successful in improving materials, 35 and we successfully demonstrated the efficient compositional optimisation of experimental Li-ion conductivity using Bayesian optimisation (BO) of Li (1+2x+y) Ca x Y y Zr (2ÀxÀy) -(PO 4 ) 3 and (Li y La (1Ày)/3 ) (1Àx) Sr 0.5x NbO 3 .…”

“…24,33 Besides compositional control, Li-ion conductivity is also affected by density, morphology, and elemental distribution, 21 and thus, controlling process parameters, such as heating conditions during the solid-state reaction, is effective in improving conduction performance. 34…”