Structural and Dynamical Studies of Lithium-Based Solid-State Electrolytes

Abstract

Current energy storage research is increasingly directed towards all-solid-state batteries (ASSBs), which can potentially offer improved safety, higher energy densities, longer cycle life, and faster charging when compared to conventional lithium-ion (Li-ion) batteries that employ liquid electrolytes. However, advancing ASSBs depends on overcoming several key challenges, particularly, discovering materials that exhibit fast ion conduction in the solid state.

Ion conductivity in crystalline solids is intrinsically linked to local structural features, and a detailed atomic-level characterisation is essential both for understanding ion transport mechanisms and identifying strategies to enhance performance. This thesis investigates the relationship between the crystal structure and ion dynamics in two candidate solid-state electrolyte materials: the anti-perovskite Li2OHCl and the complex hydride Li2(NH2)(BH4).

Reviewing the literature on Li2OHCl reveals recurring inconsistencies in its structural description. The first research chapter seeks to address this by providing a more definitive structural characterisation. Given the high mobility of both the Li-ions and protons in this material, approaches beyond conventional diffraction-based experiments were necessary. A nuclear magnetic resonance (NMR) crystallography methodology was developed and implemented, combining 35Cl solid-state NMR spectroscopy with ab initio molecular dynamics (AIMD) simulations, machine-learned interatomic potentials (MLIPs) and density functional theory (DFT) calculations. Using this approach, NMR lineshape simulations were obtained, providing validation for an orthorhombic (space group Pban) model for Li2OHCl at room temperature.

Fluorine substitution in Li2OHCl has been proposed to enhance Li-ion conductivity, but its mechanism of action and the fluorine site preference remain unclear. The second research chapter explores the structural and dynamical effects of fluorine incorporation in Li2OHCl. The results highlight that only low fluorine substitution levels stabilise the conductive cubic phase of Li2OHCl at room temperature, while higher fluorine contents leads to disorder and impurities. A combined approach, using powder diffraction (X-ray and neutron), NMR spectroscopy, and AIMD simulations indicates that fluorine substitution suppresses Li-ion dynamics in Li2OHCl, in contrast to what has previously been reported, and only selective chlorine-site substitution is predicted to enhance Li-ion mobility.

The third research chapter focuses on Li2(NH2)(BH4), an underexplored complex hydride that has emerged as a promising solid electrolyte for next-generation lithium batteries. While its sodium analogue, Na2(NH2)(BH4) has been shown to exhibit fast sodium ion conduction via a “double paddle-wheel”-assisted mechanism, the atomistic origins of Li-ion transport in Li2(NH2)(BH4) remain elusive. Li2(NH2)(BH4) was prepared and analysed via powder X-ray diffraction, multinuclear NMR spectroscopy, and AIMD simulations to probe the interplay between crystal structure, anion reorientation and Li-ion transport. The results confirm that both borohydride and amide anions are highly mobile even in the crystalline solid, with the borohydride units in particular exhibiting liquid-like dynamics across the investigated temperature range. However, no consistent spatial or temporal correlation was observed between anion motion and Li-ion hopping, and the measured activation energy for Li-ion diffusion indicates a thermally activated, defect-mediated process rather than a concerted “paddle-wheel” mechanism. Instead, the high-frequency anion dynamics appear largely decoupled from any Li-ion hopping events, and the ionic conductivity is more plausibly attributed to structural disorder and lattice softening as the material approaches its low melting point.

7Li spin–lattice relaxation measurements are routinely employed in the literature to obtain activation energies for Li-ion mobility in solid-state electrolytes. A widespread oversimplification persists that 7Li (I = 3/2) undergoes exponential spin–lattice relaxation, governed by a single time constant (T1). Owing to a combination of dipolar and non-thermal (octupolar) spin populations unique to I > 1/2 systems, spin–lattice relaxation is known to be described biexponentially. This has previously been observed in for I = 3/2 systems in liquids, where rapid, isotropic motion averages out orientation-dependant NMR interactions. However, in solids, where anisotropic NMR interactions remain, no experiments have been performed to assess whether biexponential spin–lattice relaxation occurs. The fourth research chapter explores 7Li spin–lattice relaxation in Li2OHCl, and, using a triple-quantum-filtered NMR pulse sequence previously utilised in liquid-state NMR, show the unambiguous observation of biexponential I = 3/2 spin–lattice relaxation in a solid.