Abstract
My graduate work has been motivated by the booming interest in solid-state batteries (SSB) as the successor technology to the ubiquitous Li-ion for applications of ever-increasing scale (electromobility, grid storage). At the heart of the SSB concept are solid electrolyte materials that need to allow fast ion conduction between electrodes while ensuring the safe operation of the battery. The aim of my research has been the development of deeper understanding of the relevant physicochemical properties of promising ion conductors for SSB applications and in particular Na3PS4, an archetypical Na+ ion conductor in the thiophosphate family particularly interesting for Na SSBs. An RT tetragonal- (α) and closely related cubic (β) phase (T>250 °C) had been described in literature with melting at ~510 °C. Our diffraction experiments at high temperature have brought to evidence the existence of a third crystalline phase above the previously supposed melting point, which we name γ-Na3PS4 and is characterized by impressive mobility of all its constituent atoms: translational for the Na+ atoms and rotational for the PS43- network formers as corroborated by diffraction, molecular dynamics as well as electrical and thermal analysis. Such solid-like macroscopic behavior combined with a liquid-like local disorder are defining features of plastic-crystalline rotor phases. The activation energy for Na+ conduction in γ-Na3PS4 is extremely low (~0.1 eV) leading to a very high ionic conductivity in the order of S/cm at 600 °C and ~50mS/cm extrapolated to RT. In contrast to conventional high-temperature synthesis, stabilization of the β-phase through mechanochemical synthesis have been shown to provide increased RT ionic conductivities of the order of 10-4 S/cm; an effect that has been hitherto not understood. In order to elucidate the effects of mechanochemical synthesis on ion transport, we have explored the hypothesis of mechanical effects being at the origin of the anomalously high conductivity of ball-milled Na3PS4 and related systems. Increased strain is evident from the diffractograms of mechanochemically synthesized samples which, although very rarely considered, can dramatically affect the volume available for ion conduction, specifically around transition states (i.e. “bottlenecks”) for diffusion. Through variable-pressure impedance spectroscopy we have quantified this effect through the activation volume for conduction, which is ~30% higher in mechanochemically synthesized Na3PS4 samples. Finally, we demonstrate that a conventionally-synthesized sample can reach the same high order of magnitude conductivity via the application of external pressure, concluding that pressure and strain are key elements in play when increasing the ionic conductivity of solid electrolytes through mechanochemistry. The significance of these results apply to multiple topical solid-electrolytes (e.g. Li3PS4, LiBH4) and to solid-state batteries in general, where the effects of micromechanics on performance are still poorly understood