Developing high-performance all-solid-state batteries is contingent on finding solid electrolyte materials with high ionic conductivity and ductility.Here we report new halide-rich solid solution phases in the argyrodite Li 6 PS 5 Cl family, Li 6Àx PS 5Àx Cl 1+x ,a nd combine electrochemical impedance spectroscopy, neutron diffraction, and 7 Li NMR MAS and PFG spectroscopytoshow that increasing the Cl À /S 2À ratio has as ystematic, and remarkable impact on Li-ion diffusivity in the lattice.T he phase at the limit of the solid solution regime, Li 5.5 PS 4.5 Cl 1.5 ,e xhibits ac old-pressed conductivity of 9.4 AE 0.1 mS cm À1 at 298 K( and 12.0 AE 0.2 mS cm À1 on sintering)almost four-fold greater than Li 6 PS 5 Cl under identical processing conditions and comparable to metastable superionic Li 7 P 3 S 11 .W eakened interactions between the mobile Li-ions and surrounding framework anions incurred by substitution of divalent S 2À for monovalent Cl À play amajor role in enhancing Li + -ion diffusivity,a long with increased site disorder and ahigher lithium vacancy population. Figure 5. a) 7 Li MAS NMR for Li 6Àx PS 5Àx Cl 1+x (x = 0, 0.25, 0.375, 0.5) b) correlation of the activation energies from both techniques with the 7 Li isotropic chemical shift and the Haven ratio for all values of x under study.

Although Na-O2 batteries have a low overpotential and good capacity retention, degradation reactions of glyme-based electrolytes are the primary reason for inefficiency in cell performance. The discharge capacity is accounted for through analysis of the side-products. Although sodium superoxide is the primary product (90 % theoretical), quantitative and qualitative evaluation of the side-products (using (1) H NMR, iodometric titration, and on-line mass spectrometry) shows the presence of sodium acetate (∼3.5 %), and three-fold less sodium formate, methoxy (oxo)acetic anhydride, and sodium carbonate. Our reaction mechanism proposes two paths for their formation. Because the side-products are not fully removed during oxidation, they accumulate on the cathode upon cycling. Resting the cell at open circuit potential during discharge results in consumption of the superoxide through the reaction with diglyme, which greatly increases the fraction of side products, as also confirmed by ex situ reaction studies. These findings have implications in the search for more stable electrolytes.

All-solid-state batteries employing sulfide superionic conductors demand new electrolyte materials with excellent Li-ion transport properties. We report on dual-modified superionic conductors in the Li-argyrodite family. In these materials prepared by a rapid synthesis method, aliovalent doping of the Li+ site with Ca2+ or Al3+ generates vacancies which improve Li+ diffusion and conductivity. This is confirmed by pulsed-field gradient (PFG)-NMR and impedance spectroscopy. The “super Cl-rich” material with overall composition Li5.35Ca0.1PS4.5Cl1.55 exhibits a superionic room temperature conductivity of 10.2 mS·cm–1 in the cold-pressed state and an exceptional diffusivity of 1.21 × 10–11 m2/s. The presence of the aliovalent dopant on the Li sites in the cubic structure is supported by 7Li magic-angle spinning NMR and Rietveld refinement of neutron powder diffraction data. Importantly, we also probed the impact of mechanical modification on grain boundary diffusion in these sulfide electrolytes using PFG-NMR. Detection of a clear difference in activation energies between the powder and pellet-pressed versions indicates that the efficacy of particle-scale, materials engineering modifications of fast-diffusing solid electrolytes can be interrogated with the use of PFG-NMR. Our studies also show that analysis over a wide range of temperatures is often necessary for fitting PFG-NMR Arrhenius plots in order to be able to compare macroscopic measurements with transport coefficients indirectly extrapolated from microscopic measurements, such as NMR relaxometry techniques.

Developing high‐performance all‐solid‐state batteries is contingent on finding solid electrolyte materials with high ionic conductivity and ductility. Here we report new halide‐rich solid solution phases in the argyrodite Li6PS5Cl family, Li6−xPS5−xCl1+x, and combine electrochemical impedance spectroscopy, neutron diffraction, and 7Li NMR MAS and PFG spectroscopy to show that increasing the Cl−/S2− ratio has a systematic, and remarkable impact on Li‐ion diffusivity in the lattice. The phase at the limit of the solid solution regime, Li5.5PS4.5Cl1.5, exhibits a cold‐pressed conductivity of 9.4±0.1 mS cm−1 at 298 K (and 12.0±0.2 mS cm−1 on sintering)—almost four‐fold greater than Li6PS5Cl under identical processing conditions and comparable to metastable superionic Li7P3S11. Weakened interactions between the mobile Li‐ions and surrounding framework anions incurred by substitution of divalent S2− for monovalent Cl− play a major role in enhancing Li+‐ion diffusivity, along with increased site disorder and a higher lithium vacancy population.

No abstract

Solid-state batteries (SSBs) are one of the most promising energy storage technologies due to their low flammability and high energy density compared with currently used liquid-state batteries. The main obstacle to SSB development, however, is the large chemical design space for the solid-state electrolytes (SSEs), as it is significantly time-consuming to screen candidates experimentally or from first-principles simulations. Toward this end, machine learning (ML) offers an efficient strategy. However, current ML models use complex manually created features as inputs based on human intuition, which can introduce human bias, are potentially difficult to obtain for many materials, and can result in a cumbersome feature selection process. This work demonstrates that a neural network-based model utilizing only two simple elemental features (group and period) and one simple structural feature (coordination number) can provide excellent predictive performance comparable to previous manual feature-based studies, while automatically capturing any potential secondary features and reducing the need for human intervention in model training. Such a model is potentially more generalizable than manual feature-based models and can be even applied to other material property predictions, while greatly reducing complexity and training time.