Skip to content

Introduction

Solid-state batteries utilizing solid electrolytes (SEs) promise enhanced safety and energy density compared to conventional liquid electrolyte systems 1 2. The replacement of flammable liquid electrolytes with solid alternatives significantly reduces fire hazards while potentially enabling higher energy densities through the use of lithium metal anodes 5. The development of highly conductive SEs is crucial for realizing practical solid-state batteries, with lithium argyrodites Li\(_6\)MS\(_5\)X (M = P, Ge, Si, Sn; X = Cl, Br, I) emerging as promising candidates due to their high Li-ionic conductivity of more than 1 mS cm\(^{-1}\) 3 4.

Argyrodites are a class of sulfide-based superionic conductors with a complex crystal structure consisting of tetrahedral MS\(_4^{3-}\) units and mobile Li\(^+\) ions. The framework is completed by additional S\(^{2-}\) and halide (X\(^-\)) anions that create an interconnected network of sites for Li\(^+\) diffusion 3.

The performance of SEs fundamentally depends on lithium-ion diffusion through the crystal framework. This ionic transport determines both the power capability and rate performance of the battery. In argyrodites, Li diffusion occurs through sequential shifts among different coordination environments, with the diffusion pathways and barriers being highly sensitive to the local structure 6.

Li\(_6\)PS\(_5\)I is an intriguing system due to its distinct behavior compared to other argyrodites. Unlike Li\(_6\)PS\(_5\)Cl, which shows significant S/Cl site disorder, Li\(_6\)PS\(_5\)I maintains an ordered arrangement of S2\(^-\) and I\(^-\) anions (see the below picture of the crystal structure) 6. This structural difference leads to unexpected variations in Li-ion conductivity, with the ordered Li\(_6\)PS\(_5\)I showing lower conductivity (by three orders of magnitude), despite having a more polarizable iodine sublattice, due to the variation in accessible diffusion channels 6 7. The Li ions in Li\(_6\)PS\(_5\)I can occupy two different crystallographic sites (Wyckoff positions 48\(h\) and 24\(g\)), with a total of six Li ions per formula unit distributed across these positions 9. This partial occupancy (visualized by the partially colored green spheres below) means that at any given time, only some of these possible positions are filled with Li ions, while others remain vacant. To determine the energetically most favorable arrangement of Li ions, different possible configurations of Li distributions across these sites must be sampled computationally.

Crystal structure of Li\(_6\)PS\(_5\)I

As recently demonstrated by Banerjee and Tkatchenko 8, the computational modeling of these systems requires careful consideration of the employed electronic structure methods. The HSE06 hybrid functional combined with the non-local many-body dispersion (MBD-NL) correction was shown to be necessary for the accurate description of both local structure and migration barriers in argyrodites. This approach properly accounts for the interplay between electronic exchange, correlation, and van der Waals interactions that govern the potential energy surface for Li diffusion.

In the next section, we will learn how to model one of the Li diffusion pathways in Li\(_6\)PS\(_5\)I with FHI-aims and aimsChain as described in Ref. 8.

References

  1. Bachman et al., Chem. Rev. 116, 140 (2016). ↩

  2. Zhao et al., Nat. Rev. Mater. 5, 229 (2020). ↩

  3. Deiseroth et al., Angew. Chem. Int. Ed. 47, 755 (2008). ↩↩

  4. Kraft et al., J. Am. Chem. Soc. 139, 10909 (2017). ↩

  5. Janek and Zeier, Nat. Energy 1, 16141 (2016). ↩

  6. Morgan, Chem. Mater. 33, 2004 (2021). ↩↩↩

  7. Yu et al., J. Am. Chem. Soc. 138, 11192 (2016). ↩

  8. Banerjee and Tkatchenko, Accepted in Nat. Comms. ↩↩

  9. Tong Kong et al., C. Chem. - Eur. J. 16 (17), 5138 (2010). ↩