Molecular dynamics simulations have become an indispensable tool for investigating the structural and dynamic properties of highly concentrated aqueous electrolytes, commonly referred to as water-in-salt electrolytes (WiSE). These systems exhibit unique solvation behaviors that deviate significantly from dilute aqueous solutions, enabling electrochemical stability windows far beyond conventional aqueous electrolytes. By tracking the trajectories of individual atoms over time, MD simulations provide atomic-level insights into the solvation structures, ion clustering phenomena, and hydrogen bonding networks that govern the exceptional properties of WiSE.
In highly concentrated electrolytes, the scarcity of free water molecules leads to the formation of distinct solvation structures. MD simulations reveal that cations become coordinated not only by water molecules but also by anions, creating a hybrid solvation shell. For lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) solutions at concentrations exceeding 20 mol/kg, simulations show that each lithium ion coordinates with approximately 1.5 water molecules and 2.5 TFSI anions on average. This contrasts sharply with dilute solutions where lithium ions are fully hydrated by 4-6 water molecules. The partial dehydration of cations significantly alters the electrochemical stability of the solution by shifting the redox potentials of water decomposition.
Ion clustering represents another critical feature revealed by MD simulations. At high salt concentrations, anions and cations form extended networks with percolating structures rather than existing as isolated ion pairs. Radial distribution functions calculated from MD trajectories demonstrate strong correlations between lithium ions and TFSI anions at distances below 5 Å, with coordination numbers increasing linearly with salt concentration. These clusters create a rigid matrix that restricts water mobility and reorganizes the hydrogen bonding network. The simulations show water molecules participating in fewer hydrogen bonds compared to pure water, with an average of 2.1 hydrogen bonds per water molecule in 21 mol/kg LiTFSI versus 3.4 in pure water.
The hydrogen bonding network undergoes profound changes in WiSE, as quantified through various analysis techniques applied to MD trajectories. Angular distribution functions reveal that water molecules in WiSE exhibit broader O-H···O angles compared to bulk water, indicating distorted hydrogen bonds. Lifetime correlation functions show that hydrogen bonds between water molecules persist for shorter durations in WiSE, typically 1-2 ps compared to 5-7 ps in pure water. This dynamic behavior arises from the competition between water-water and water-anion interactions, where the strongly coordinating TFSI anions disrupt the tetrahedral network of bulk water.
MD simulations provide crucial insights into the mechanisms behind the expanded electrochemical stability window observed experimentally in WiSE. The simulations demonstrate that the free water molecules remaining in WiSE exist in a chemically distinct state with elevated oxidation potential. This occurs because the electron density on oxygen atoms decreases when water molecules coordinate to multiple anions, making them more difficult to oxidize. Similarly, the reduction potential of water increases because the limited availability of free water molecules suppresses hydrogen evolution. Simulation-based calculations of the highest occupied molecular orbital and lowest unoccupied molecular orbital energies confirm these shifts in redox potentials.
The suppression of water decomposition reactions correlates strongly with the solvation structure observed in MD simulations. Water molecules incorporated into the primary solvation shell of anions exhibit higher activation barriers for oxidation compared to bulk-like water. Potential of mean force calculations reveal that the energy required to remove a proton from a water molecule coordinated to TFSI is approximately 0.5 eV higher than from a water molecule in the bulk phase. This explains the experimental observation that oxygen evolution in WiSE begins at potentials above 3 V versus Li/Li+, compared to 1.23 V thermodynamically expected for pure water.
Validation of MD simulation results comes from multiple experimental techniques. Raman spectroscopy measurements show excellent agreement with simulated vibrational spectra, particularly in the regions corresponding to anion-water interactions. The splitting of the SO2 symmetric stretching mode of TFSI observed experimentally matches precisely with the coordination environments predicted by MD. Similarly, X-ray scattering structure factors calculated from MD trajectories reproduce all major features observed experimentally, including the pre-peak around 0.5 Å^-1 that signifies the formation of nanoscale ionic aggregates.
The transport properties derived from MD simulations also align with experimental measurements. Calculated ionic conductivities from mean squared displacement analysis typically fall within 15% of experimental values for concentrations ranging from 5 to 21 mol/kg. The simulations correctly predict the maximum in conductivity observed at intermediate concentrations and the subsequent decrease at higher salt loadings. This non-monotonic behavior arises from competing effects: increasing charge carrier concentration versus decreasing mobility due to enhanced ion clustering.
Several key implications for battery applications emerge from MD studies of WiSE. The simulations explain how the unique solvation environment enables the use of high-voltage cathode materials such as LiMn2O4 and LiCoO2 in aqueous systems. By quantifying the fraction of free water molecules as a function of concentration, MD provides guidelines for designing electrolytes with optimal stability windows. The simulations also suggest strategies for further improving stability, such as introducing anions with stronger water-binding capability or cations with higher charge density to further reduce water activity.
Recent advances in MD methodology have enhanced the accuracy of WiSE simulations. Polarizable force fields now account for electronic polarization effects that become significant at high ion concentrations. Machine learning potentials trained on quantum mechanical calculations offer improved descriptions of ion-water interactions while maintaining computational efficiency for nanosecond-scale simulations. These developments allow more reliable prediction of thermodynamic and kinetic properties across wide concentration and temperature ranges.
The mechanistic understanding derived from MD simulations guides the development of next-generation aqueous batteries. Simulations have identified specific concentration thresholds where percolating ion networks form, suggesting optimal electrolyte compositions for balancing stability and conductivity. The detailed view of interfacial structures near electrodes helps interpret unusual phenomena such as the formation of anion-derived interphases that further suppress water decomposition. As simulation techniques continue advancing alongside experimental validation, molecular dynamics remains a powerful tool for unlocking the full potential of water-in-salt electrolytes for safe, high-energy battery systems.