Deciphering Fast Ion Transport in Glasses: A Case Study of Sodium and Silver Vitreous Sulfides

Abstract

High-capacity solid-state batteries are promising future products for large-scale energy storage and conversion. Sodium fast ion conductors including glasses and glass ceramics are unparalleled materials for these applications. Rational design and tuning of advanced sodium sulfide electrolytes need a deep insight into the atomic structure and dynamics in relation with ion-transport properties. Using pulsed neutron diffraction and Raman spectroscopy supported by first-principles simulations, we show that preferential diffusion pathways in vitreous sodium and silver sulfides are related to isolated sulfur S iso , that is, the sulfur species surrounded exclusively by mobile cations with a typical stoichiometry of M/S iso ≈ 2. The S iso /S tot fraction appears to be a reliable descriptor of fast ion transport in glassy sulfide systems over a wide range of ionic conductivities and cation diffusivities. The S iso fraction increases with mobile cation content x, tetrahedral coordination of the network former and, in case of thiogermanate systems, with germanium disulfide metastability and partial disproportionation, GeS 2 → GeS + S, leading to the formation of additional sulfur, transforming into S iso . A research strategy enabling to achieve extended and interconnected pathways based on isolated sulfur would lead to glassy electrolytes with superior ionic diffusion.

Glassy and liquid Sb2S3: insight into the structure and dynamics of a promising functional material

“…8. As reported earlier, [56][57][58][59] the FPMD with hybrid functional PBE0 yields good agreement with experimental data for chalcogenide systems. The FPMD replicas in Q-space reproduce well both the positions and amplitudes of oscillations for Q[S X (Q) À 1], Fig.…”

“…53 The generalized gradient approximation (GGA) and the PBE0 hybrid 54,55 exchange-correlation functional combining the exact Hartree-Fock and DFT approaches were used, providing better agreement with experiment. [56][57][58][59] The Grimme dispersion corrections D3BJ 60 were also employed, improving first-principles molecular dynamics (FPMD) results for chalcogenide systems. 61,62 The applied FPMD technique was similar to previous reports.…”

Evidence for Complex Dynamics in Glassy Fast Ion Conductors: The Case of Sodium Thiosilicates

“…A recent measurement 38 from 22 Na tracer diffusion on the isochemical Na 2 S−GeS 2 glass led to a value of D Na = 6 × 10 −12 cm 2 s −1 for 10 3 /T = 2.23, i.e., somewhat smaller than our present extrapolated data. However, this can be ascribed to the facts that (i) the glass transition temperatures are different (10 3 /T g = 2.01 and 1.83 for Ge 38 and Si-based 12 glasses, respectively) and (ii) sodium thiosilicate conductivities are usually higher than those of corresponding thiogermanates, 44 which underscores an increased Na mobility/diffusivity in the former. For the 56Na 2 S−44SiS 2 system, 43 the Na diffusivity is 5.83 × 10 −8 cm 2 s −1 at 10 3 /T = 2.11, and as such, in somewhat good agreement with extrapolations of our obtained diffusivities.…”

“…All values appear to be of the same order as those calculated from the present MSDs. A recent measurement from 22 Na tracer diffusion on the isochemical Na 2 S–GeS 2 glass led to a value of D Na = 6 × 10 –12 cm 2 s –1 for 10 3 /T = 2.23, i.e., somewhat smaller than our present extrapolated data. However, this can be ascribed to the facts that (i) the glass transition temperatures are different (10 3 / T g = 2.01 and 1.83 for Ge and Si-based glasses, respectively) and (ii) sodium thiosilicate conductivities are usually higher than those of corresponding thiogermanates, which underscores an increased Na mobility/diffusivity in the former.…”

“…At low temperatures, because of the breakdown of ergodicity which results from a limited exploration of the phase space, deep supercooled liquids (and eventually glasses) will be out of equilibrium so that the invariance fails and aging onsets . As a result, an estimation , of diffusivities or conductivities in the glassy state from computer simulations is unwise, if not simply misleading. Separately, simulations of more frequent dumping were performed at 600 K for 1 ns to allow for accessing structural changes in a continuous nonlogarithmic scale.…”