A Synergistic Approach to Unraveling the Thermodynamic Stability of Binary and Ternary Chevrel Phase Sulfides

Abstract

State-of-the-art high temperature oxide melt solution calorimetry and density functional theory were employed to produce the first systematic study of thermodynamic stability in a series of binary and ternary Chevrel phases. Rapid microwave-assisted solid-state heating methods facilitated the nucleation of pure-phase polycrystalline M y Mo6S8 (M = Fe, Ni, Cu; y = 0, 1, 2) Chevrel phases, and a stability trend was observed wherein intercalation of M y species engenders stability that depends on both the electropositivity and ionic radii of the intercalant species. Ab initio calculations indicate that this stability trend results from competing ionic and covalent contributions, where transition metal intercalation stabilizes the Chevrel structure through increased ionicity but destabilizes the structure through reduced covalency of the Mo6S8 clusters. Our calculations predicted that over intercalation of high-valent M y species leads to slight destabilization of the Mo6 octahedral cores, which we confirm using calorimetry and X-ray absorption spectroscopy. Our combined computational and calorimetric analysis reveals the interplay of the foundational principles of ionic and covalent bonding characteristics that govern the thermodynamic stability of Chevrel and other inorganic phases.

Machine Learning Guided Synthesis of Multinary Chevrel Phase Chalcogenides

“… that is, the difference between M–X ionicity and covalency, indicates the degree of ionicity of M–X along the ionic–covalent bonding spectrum (Figure c). This is consistent with previous work that attributed the stability of intercalated CPs to competing shifts in ionic and covalent bonding that arise from the transfer of up to four electrons from M to the Mo 6 X 8 cluster. , …”

“…Increasing the ratio of cation to chalcogen dipole polarizability pol M /pol X results in larger dipole moments and greater ionic bonding between M and X (i.e., increased ionicity, Figure c). This decreases H δ and increases the stability of the multinary CPs, consistent with our recent results on 3d-TM intercalated sulfide CPs . Conversely, increasing H fus ,tM /CE X increases H δ and decreases stability.…”

“…The inductive effect has important consequences for the stability of multinary systems, where greater covalent bonding between a cation and anion (i.e., M–X) reduces the covalent bonding between the same anion with another element (i.e., Mo–X) . Thus, the degree of ionicity of M–X regulates the inductive effect in CPs and the covalency of the Mo–X bond, which has been shown to drive overall covalent bonding in CPs . A more ionic M–X bond yields a more covalent Mo–X bond and a more stable CP via greater ionicity and covalency, and vice versa .…”

“…Sixty-seven of the CPs in our data set (Table S4) have been synthesized previously with a range of synthesis techniques. ,− ,,,− These techniques can be classified as three general approaches, each used in a different temperature range: (a) direct high-temperature ( T > 1000 C) synthesis from stoichiometric mixtures of the constituent elements and/or non-CP binaries (e.g., MoX 2 ); (b) medium-temperature ( T ≈ 500 C) intercalation (MTI) of elemental M into Mo 6 X 8 ; or (c) room-temperature intercalation of the dissolved cation M into Mo 6 X 8 . The large temperature differences between these different techniques indicate a wide range of stabilities of observed CPs.…”

“…Despite their broad chemical space and favorable properties for multiple energy applications, the CP family is significantly underexplored, with only ∼100 synthesized compositions . This is particularly evident for the CP tellurides, where only five multinary compositions have yet been synthesized. , Unfortunately, synthesizable CP compositions are not easily predicted because cation intercalation gives rise to competing effects on their stability . As a result, determining the synthesizability of CP compositions generally requires computationally demanding ab initio calculations or experimental characterization.…”