Molecular clustering and associated dynamic processes of lithium polysulfide species were unraveled using classical molecular dynamics and ab initio metadynamics calculations. The spectroscopic signatures of polysulfide clusters were analyzed using a multimodal analysis including experimental and computational nuclear magnetic resonance (NMR) and X-ray absorption spectroscopies. Lithium polysulfide solutes (Li2S4, Li2S6, and Li2S8) and their mixtures in a 1,3-dioxolane and 1,2-dimethoxyethane (DOL/DME) solvent undergo aggregation driven by intramolecular lithium-sulfur (Li-S) interactions, leading to distributions of cluster sizes, which could critically influence the functioning of lithium-sulfur batteries. Representative polysulfide clusters with systematic increases in molecular size were extracted from the classical molecular dynamics trajectories for subsequent structural and spectroscopic property calculations using density functional theory analysis. Structural analysis of these clusters reveals progressively decreasing solvent involvement in Li+ coordination varying from Li2S4 to Li2S8, with more pronounced variation and changes in DME compared with those in DOL. These observations are reflected in the analysis of the experimental and theoretical 7Li and 17O NMR chemical shifts and pulsed field gradient-NMR diffusion measurements. A comparison of experimental and theoretical S K-edge X-ray absorption near edge spectra shows that relatively large lithium sulfide chain clusters are likely to occur in the DOL/DME-solvated lithium sulfide systems. Ab initio metadynamics simulations and NMR analysis indicate that Li+ solvated by only the solvent can occur through Li+ dissociation from sulfide chains. However, the occurrence of "sulfide-free" Li+ is a minor mechanism compared with the dynamic aggregation and shuttling processes of polysulfide solvates in DOL/DME-based electrolytes of Li-S batteries. Overall, atomistic insights gained about clustering and lithium exchange dynamics will be critical for the predictive understanding of the polysulfide shuttling and nucleation process that dictates the Li-S battery performance.

Multivalent batteries represent an important beyond Li-ion energy storage concept. The prospect of calcium batteries, in particular, has emerged recently due to novel electrolyte demonstrations, especially that of a ground-breaking combination of the borohydride salt Ca(BH₄)₂ dissolved in tetrahydrofuran. Recent analysis of magnesium and calcium versions of this electrolyte led to the identification of divergent speciation pathways for Mg²⁺ and Ca²⁺ despite identical anions and solvents, owing to differences in cation size and attendant flexibility of coordination. To test these proposed speciation equilibria and develop a more quantitative understanding thereof, we have applied pulsed-field-gradient nuclear magnetic resonance and dielectric relaxation spectroscopy to study these electrolytes. Concentration-dependent variation in anion diffusivities and solution dipole relaxations, interpreted with the aid of molecular dynamics simulations, confirms these divergent Mg²⁺ and Ca²⁺ speciation pathways. These results provide a more quantitative description of the electroactive species populations. We find that these species are present in relatively small quantities, even in the highly active Ca(BH₄)₂/tetrahydrofuran electrolyte. This finding helps interpret previous characterizations of metal deposition efficiency and morphology control and thus provides important fundamental insight into the dynamic properties of multivalent electrolytes for next-generation batteries.

The solvation free energy of organic molecules is a critical parameter in determining emergent properties such as solubility, liquid-phase equilibrium constants, and pKa and redox potentials in an organic redox flow battery. In this work, we present a machine learning (ML) model that can learn and predict the aqueous solvation free energy of an organic molecule using Gaussian process regression method based on a new molecular graph kernel. To investigate the performance of the ML model on electrostatic interaction, the nonpolar interaction contribution of solvent and the conformational entropy of solute in solvation free energy, three data sets with implicit or explicit water solvent models, and contribution of conformational entropy of solute are tested. We demonstrate that our ML model can predict the solvation free energy of molecules at chemical accuracy with a mean absolute error of less than 1 kcal/mol for subsets of the QM9 dataset and the Freesolv database. To solve the general data scarcity problem for a graph-based ML model, we propose a dimension reduction algorithm based on the distance between molecular graphs, which can be used to examine the diversity of the molecular data set. It provides a promising way to build a minimum training set to improve prediction for certain test sets where the space of molecular structures is predetermined.

Efforts to expand the technological capability of batteries have generated increased interest in divalent cationic systems. Electrolytes used for these electrochemical applications often incorporate cyclic ethers as electrolyte solvents; however, the detailed solvation environments within such systems are not well-understood. To foster insights into the solvation structures of such electrolytes, Ca(TFSI)₂ and Zn(TFSI)₂ dissolved in tetrahydrofuran (THF) and 2-methyl-tetrahydrofuran were investigated through multi-nuclear magnetic resonance spectroscopy (¹⁷O, ⁴³Ca, and ⁶⁷Zn NMR) combined with quantum chemistry modeling of NMR chemical shifts. NMR provides spectroscopic fingerprints that readily couple with quantum chemistry to identify a set of most probable solvation structures based on the best agreement between the theoretically predicted and experimentally measured values of chemical shifts. The multi-nuclear approach significantly enhances confidence that the correct solvation structures are identified due to the required simultaneous agreement between theory and experiment for multiple nuclear spins. Furthermore, quantum chemistry modeling provides a comparison of the solvation cluster formation energetics, allowing further refinement of the preferred solvation structures. It is shown that a range of solvation structures coexist in most of these electrolytes, with significant molecular motion and dynamic exchange among the structures. This level of solvation diversity correlates with the solubility of the electrolyte, with Zn(TFSI)₂/THF exhibiting the lowest degree of each. Comparisons of analogous Ca²⁺ and Zn²⁺ solvation structures reveal a significant cation size effect that is manifested in significantly reduced cation-solvent bond lengths and thus stronger solvent bonding for Zn²⁺ relative to Ca²⁺. The strength of this bonding is further reduced by methylation of the cyclic ether ring. Solvation shells containing anions are energetically preferred in all the studied electrolytes, leading to significant quantities of contact ion pairs and consequently neutrally charged clusters. It is likely that the transport and interfacial de-solvation/re-solvation properties of these electrolytes are directed by these anion interactions. These insights into the detailed solvation structures, cation size, and solvent effects, including the molecular dynamics, are fundamentally important for the rational design of electrolytes in multivalent battery electrolyte systems.

Conventional electrolytes made by mixing simple Mg2+ salts and aprotic solvents, analogous to those in Li-ion batteries, are incompatible with Mg anodes because Mg metal readily reacts with such electrolytes, producing a passivation layer that blocks Mg2+ transport. Here, we report that, through tuning a conventional electrolyte - Mg(TFSI)2 (TFSI- is N(SO2CF3)2-) - with an Mg(BH4)2 cosalt, highly reversible Mg plating/stripping with a high Coulombic efficiency is achieved by neutralizing the first solvation shell of Mg cationic clusters between Mg2+ and TFSI- and enhanced reductive stability of free TFSI-. A critical adsorption step between Mg0 atoms and active Mg cation clusters involving BH4- anions is identified to be the key enabler for reversible Mg plating/stripping through analysis of the distribution of relaxation times (DRT) from operando electrochemical impedance spectroscopy (EIS), operando electrochemical X-ray absorption spectroscopy (XAS), nuclear magnetic resonance (NMR), and density functional theory (DFT) calculations.