The Potential of Neural Network Potentials.

In the next half-century, physical chemistry will likely undergo a profound transformation, driven predominantly by the combination of recent advances in quantum chemistry and machine learning (ML). Specifically, equivariant neural network potentials (NNPs) are a breakthrough new tool that are already enabling us to simulate systems at the molecular scale with unprecedented accuracy and speed, relying on nothing but fundamental physical laws. The continued development of this approach will realize Paul Dirac's 80-year-old vision of using quantum mechanics to unify physics with chemistry and providing invaluable tools for understanding materials science, biology, earth sciences, and beyond. The era of highly accurate and efficient first-principles molecular simulations will provide a wealth of training data that can be used to build automated computational methodologies, using tools such as diffusion models, for the design and optimization of systems at the molecular scale. Large language models (LLMs) will also evolve into increasingly indispensable tools for literature review, coding, idea generation, and scientific writing.

Understanding the Electrochemical Extraction of Lithium from Ultradilute Solutions.

The electrochemical extraction of lithium (Li) from aqueous sources using electrochemical means is a promising direct Li extraction technology. However, to this date, most electrochemical Li extraction studies are confined to Li-rich brine, neglecting the practical and existing Li-lean resources, with their overall extraction behaviors currently not fully understood. More still, the effect of elevated sodium (Na) concentrations typically found in most Li-lean water sources on Li extraction is unclear. Hence, in this work, we first understand the electrochemical Li extraction behaviors from ultradilute solutions using spinel lithium manganese oxide as the model electrode. We discovered that Li extraction depends highly on the Li concentration and cell operation current density. Then, we switched our focus on low Li to Na ratio solutions, revealing that Na can dominate the electrostatic screening layer, reducing Li ion concentration. Based on these understandings, we rationally employed pulsed electrochemical operation to restructure the electrode surface and distribute the surface-adsorbed species, which efficiently achieves a high Li selectivity even in extremely low initial Li/Na concentrations of up to 1:20,000.

High-Throughput Aqueous Electrolyte Structure Prediction Using IonSolvR and Equivariant Graph Neural Network Potentials.

Neural network potentials have recently emerged as an efficient and accurate tool for accelerating ab initio molecular dynamics (AIMD) in order to simulate complex condensed phases such as electrolyte solutions. Their principal limitation, however, is their requirement for sufficiently large and accurate training sets, which are often composed of Kohn-Sham density functional theory (DFT) calculations. Here we examine the feasibility of using existing density functional tight-binding (DFTB) molecular dynamics trajectory data available in the IonSolvR database in order to accelerate the training of E(3)-equivariant graph neural network potentials. We show that the solvation structure of Na+ and Cl- in aqueous NaCl solutions can be accurately reproduced with remarkably small amounts of data (i.e., 100 MD frames). We further show that these predictions can be systematically improved further via an embarrassingly parallel resampling approach.

Effect of fluoro and hydroxy analogies of diglyme on sodium-ion storage in graphite: a computational study.

Diglyme co-intercalation with sodium ion (Na+) into graphite can enable the use of graphite as a potential anode for sodium-ion batteries (NIBs). However, the presence of diglyme molecules in Na+ intercalated graphite limits Na+ storage capacity and increases volume changes. In this work, the effect of functionalising diglyme molecules with fluoro and hydroxy groups on Na+ storage properties in graphite were computationally studied. It was found that the functionalisation can significantly alter the binding between sodium and the solvent ligand as well as between the sodium-solvent complex and the graphite. The hydroxy-functionalised diglyme exhibits the strongest binding to the graphite of the other functionalised diglyme compounds considered. The calculations also reveal that the graphene layer affects the electron distribution on the diglyme molecule and Na, so the diglyme complexed Na binds more strongly to the graphene layer than the Na alone. We also propose a mechanism for the early stages of the intercalation mechanism that involves a reorientation of the sodium-diglyme complex and suggest how the solvent can be designed to optimise the co-intercalation process.

Salting-Up of Surfactants at the Surface of Saline Water as Detected by Tensiometry and SFG and Supported by Molecular Dynamics Simulation.

Surfactant adsorption at the air-water interface is critical to many industrial processes but its dependence on salt ions is still poorly understood. Here, we investigate the adsorption of sodium dodecanoate onto the air-water interface using model saline waters of Li+ or Cs+ at pH values 8 and 11. Both cations enhance the surfactant adsorption, as expected, but their largest effects on the adsorption also depend on pH. Specifically, surface tension measurements, sum-frequency generation spectroscopy, and microelectrophoresis show that small (hard) Li+ enhances the surfactant adsorption more than large (soft) Cs+ at pH 11. This effect is fully reversed at pH 8. We argue that this salting-up (increasing adsorption) reversal is attributable to the conversion of the neutralized carboxylic (-COOH) headgroup at pH 8 into the charged carboxylate (-COO-) headgroup at pH 11, which, respectively, interact with Cs+ and Li+ favorably. Molecular dynamics simulation shows that the affinity of Cs+ to the interface is decreased and eventually overtaken by Li+ as the carboxylic groups are deprotonated. This study highlights the importance of the charge and size of salt ions in selecting surfactants and electrolytes for industrial applications.

Lithium-Ion Transport Behavior in Thin-Film Graphite Electrodes with SEI Layers Formed at Different Current Densities.

There has been rapidly growing interest in developing fast-charging batteries for electric vehicles. The solid electrolyte interphase (SEI) layer formed at the graphite/electrolyte interface plays an important role in determining the lithiation rate of lithium-ion batteries (LIBs). In this work, we investigated lithium-ion transport behavior in thin-film graphite electrodes with different graphite particle sizes and morphologies for understanding the role of the SEI layer in fast charging LIBs. We varied the properties of the SEI by changing the current rate during the SEI formation. We observed that forming the SEI layer at a much higher current density than is traditionally used leads to a substantial reduction in electrode impedance and a corresponding increase in ion diffusivity. This enables thin-film graphite electrodes to be charged at current rates as high as 12 C (i.e., about 5 min charging time), demonstrating that graphite is not necessarily prevented from fast charging. By comparing the SEI layers formed at different current densities, we observed that lithium-ion diffusivity across the SEI layer formed on a 23 µm commercial graphite at a current density currently used in the industry (e.g., 0.1 C) is approximately 8.9 × 10-10 cm2/s.

Toward a First-Principles Framework for Predicting Collective Properties of Electrolytes.

Given the universal importance of electrolyte solutions, it is natural to expect that we have a nearly complete understanding of the fundamental properties of these solutions (e.g., the chemical potential) and that we can therefore explain, predict, and control the phenomena occurring in them. In fact, reality falls short of these expectations. But, recent advances in the simulation and modeling of electrolyte solutions indicate that it should soon be possible to make progress toward these goals. In this Account, we will discuss the use of first-principles interaction potentials based in quantum mechanics (QM) to enhance our understanding of electrolyte solutions. Specifically, we will focus on the use of quantum density functional theory (DFT) combined with molecular dynamics simulation (DFT-MD) as the foundation for our approach. The overarching concept is to understand and accurately reproduce the balance between local or short-ranged (SR) structural details and long-range (LR) correlations, allowing the prediction of the thermodynamics of both single ions in solution as well as the collective interactions characterized by activity/osmotic coefficients. In doing so, relevant collective motions and driving forces characterized by chemical potentials can be determined.In this Account, we will make the case that understanding electrolyte solutions requires a faithful QM representation of the SR nature of the ion-ion, ion-water, and water-water interactions. However, the number of molecules that is required for collective behavior makes the direct application of high-level QM methods that contain the best SR physics untenable, making methods that balance accuracy and efficiency a practical goal. Alternatives such as continuum solvent models (CSMs) and empirically based classical molecular dynamics have been extensively employed to resolve this problem but without yet overcoming the fundamental issue of SR accuracy. We will demonstrate that accurately describing the SR interaction is imperative for predicting both intrinsic properties, namely, at infinite dilution, and collective properties of electrolyte solutions.DFT has played an important role in our understanding of condensed phase systems, e.g., bulk liquid water, the air-water interface, ions in bulk, and at the air-water interface. This approach holds huge promise to provide benchmark calculations of electrolyte solution properties that will allow for the development and improvement of more efficient methods, as well as an enhanced understanding of fundamental phenomena. However, the standard protocol using the generalized gradient approximation with van der Waals (vdW) correction requires improvement in order to achieve a high level of quantitative accuracy. Simply simulating with higher level DFT functionals may not be the best route considering the significant computational cost. Alternative methods of incorporating information from higher levels of QM should be explored; e.g., using force matching techniques on small clusters, where high level benchmark calculations are possible, to develop ideal correction terms to the DFT functional is a promising possibility. We argue that DFT with statistical mechanics is becoming an increasingly useful framework enabling the prediction of collective electrolyte properties.

The surface potential explains ion specific bubble coalescence inhibition.

HYPOTHESIS: Some ions can prevent bubbles from coalescing in water. The Gibbs-Marangoni pressure has been proposed as an explanation of this phenomenon. This repulsive pressure occurs during thin film drainage whenever surface enhanced or surface depleted solutes are present. However, bubble coalescence inhibition is known to depend on which particular combination of ions are present in a peculiar and unexplained way. This dependence can be explained by the electrostatic surface potential created by the distribution of ions at the interface, which will alter the natural surface propensity of the ions and hence the Gibbs-Marangoni pressure. CALCULATIONS: A generalised form of the Gibbs-Marangoni pressure is derived for a mixture of solutes and the modified Poisson-Boltzmann equation is used to calculate this pressure for five different electrolyte solutions made up of four different ions. FINDINGS: Combining ions with differing surface propensities, i.e., one enhanced and one depleted, creates a significant electrostatic surface potential which dampens the natural surface propensity of these ions, resulting in a reduced Gibbs-Marangoni pressure, which allows bubble coalescence. This mechanism explains why the ability of electrolytes to inhibit bubble coalescence is correlated with surface tension for pure electrolytes but not for mixed electrolytes.

From Surface Tension to Molecular Distribution: Modeling Surfactant Adsorption at the Air-Water Interface.

Surfactants are centrally important in many scientific and engineering fields and are used for many purposes such as foaming agents and detergents. However, many challenges remain in providing a comprehensive understanding of their behavior. Here, we provide a brief historical overview of the study of surfactant adsorption at the air-water interface, followed by a discussion of some recent advances in this area from our group. The main focus is on incorporating an accurate description of the adsorption layer thickness of surfactant at the air-water interface. Surfactants have a wide distribution at the air-water interface, which can have a significant effect on important properties such as the surface excess, surface tension, and surface potential. We have developed a modified Poisson-Boltzmann (MPB) model to describe this effect, which we outline here. We also address the remaining challenges and future research directions in this area. We believe that experimental techniques, modeling, and simulation should be combined to form a holistic picture of surfactant adsorption at the air-water interface.

Exploring the effect of interlayer distance of expanded graphite for sodium ion storage using first principles calculations.

Expanded graphite (EG) has been shown to be able to store a significant amount of sodium ions. Understanding the alkali metal ion storage in EG is of importance for improving EG electrode performance. In this work, the effect of interlayer distance of pure EG on sodium ion storage was investigated using the density functional theory calculation method. EG structure models with interlayer distances ranging from 3.4 Å to 10.0 Å were simulated. It was found that EG can store a fairly large amount of sodium ions through an intercalation mechanism without any contributions from the co-intercalation mechanism or adsorption mechanism if the interlayer distance is larger than 4.4 Å and smaller than 6.0 Å. It was also found that an interlayer distance of 6.0 Å gives strong binding energy of sodium ions with EG forming thermodynamically stable sodium-graphite intercalation compound (Na-GIC). However, when the interlayer distance becomes larger than 6.0 Å, the binding energy between sodium ions and EG becomes weaker. Computational results have also shown that the enthalpy of formation of the Na-GIC of EG is energetically more favourable when the interlayer distance is increased. An optimal d-spacing of EG for sodium ion storage was identified in this work. These findings provide atomistic insights into sodium ion storage in EG, providing guidelines for the design of graphite-based anode materials for sodium-ion batteries.

The Born model can accurately describe electrostatic ion solvation.

Accurate models of the free energies of ions in solution are crucially important. They can be used to predict and understand the properties of electrolyte solutions in the huge number of important applications where these solutions play a central role such as electrochemical energy storage. The Born model, developed to describe ion solvation free energies, is widely considered to be critically flawed as it predicts a linear response of water to ionic charge, which fails to match water's supposed intrinsic preference to solvate anions over cations. Here, we demonstrate that the asymmetric response observed in simulation is the result of an arbitrary choice of the oxygen atom to be the centre of a water molecule. We show that an alternative and reasonable choice, which places the centre 0.5 Å towards the hydrogen atoms, results in a linear and charge symmetric response of water to ionic charge for a classical water model consistent with the Born model. Therefore, this asymmetry should be regarded as a property of the specific short-range repulsive interaction not an intrinsic electrostatic property of water and so the fact that the Born model does not reproduce it is not a limitation of this approach. We also show that this new water centre results in a more reasonable surface potential contribution to the solvation free energies.

Quantifying the Counterion-Specific Effect on Surfactant Adsorption Using Modeling, Simulation, and Experiments.

Ionic surfactants behave differently in the presence of various counterions, which plays an important role in many scientific and engineering processes. Previous work has shown that the counterion-specific surface tension can be reproduced with classical adsorption models, but the underlying origin of this effect has not been explained. In this paper, we extend our previously developed adsorption model to account for the specific counterion adsorption. This model can accurately predict the surface tension of surfactant solutions like sodium dodecyl sulfate (SDS) in the presence of the monovalent salts LiCl, NaCl, KCl, and CsCl. The predicted surface excess and surface potential are validated by corresponding sum-frequency generation (SFG) spectroscopy experiments. We also used molecular dynamic (MD) simulation to explain the origin of the counterion-specific effect for surfactant behavior. Our study shows that for SDS, binding of the counterion to both the headgroup and a few CH2 fragments close to the surfactant head contributes to the counterion-specific effect. In general, SDS behaves like a large ion, and it prefers to bind with large counterions such as Cs+, which is consistent with Collins's law of matching water affinity. Therefore, large counterions enhance the surface adsorption and lower the surface tension the most.

Method for Accurately Predicting Solvation Structure.

Accurately predicting the molecular structure of solutions is a fundamental scientific challenge. Using quantum mechanical density functional theory (DFT) to make these predictions is hindered by significant variation depending on which DFT functional is used. Here, we present a simple metric that can determine the reliability of a DFT functional for predicting solvation structure. We then show that including a simple interaction term to correct this metric leads to quantitative agreement with experimental measurements of liquid structure. We demonstrate the utility of this method by using it to accurately describe the hydration structure around the Na+ and K+ ions as well as the structural properties of pure water with a computationally cheap functional.

Significant Effect of Surfactant Adsorption Layer Thickness in Equilibrium Foam Films.

Foam films formed at the air-water interface do not have fixed adsorption sites where adsorbed surfactants can arrange themselves, resulting in the formation of thick adsorption layers. Current theories of equilibrium foam films fail to account for this feature and significantly underestimate the adsorption layer thickness. Here we show that this thickness has a significant effect on the disjoining pressure in foam films. If ignored, the theory predicts unphysical electrostatic potential profiles, which underestimate the disjoining pressure. We apply a previously developed adsorption model that incorporates a realistic thickness for the adsorption layer. This new model reproduces experimental measurements of the disjoining pressure of foam films very well over a wide surfactant concentration range without fitting parameters. Our work shows that a thick adsorption layer is less effectively screened by counterions, resulting in a higher electrostatic potential inside the film and therefore a higher disjoining pressure.

Surface Potential Explained: A Surfactant Adsorption Model Incorporating Realistic Layer Thickness.

Soluble surfactants form thick adsorption layers at the air-liquid interface, but classical adsorption models fail to account for it as they treat the adsorption layer as a mathematical plane (of zero thickness). This simplification has produced several inconsistencies between theoretical predictions and experimental results, especially for the surface potential. Here, we develop a new adsorption model for ionic surfactants at the air-water interface that incorporates the effect of the adsorption layer thickness using a modified Poisson-Boltzmann equation that integrates information from molecular dynamics simulation. We show that the surface potential depends sensitively on both the thickness of the adsorption layer and the interfacial depth at which the surface potential is probed. This model, therefore, provides a much more accurate picture of the surface potential than classical models.

Quantifying the hydration structure of sodium and potassium ions: taking additional steps on Jacob's Ladder.

The ability to reproduce the experimental structure of water around the sodium and potassium ions is a key test of the quality of interaction potentials due to the central importance of these ions in a wide range of important phenomena. Here, we simulate the Na+ and K+ ions in bulk water using three density functional theory functionals: (1) the generalized gradient approximation (GGA) based dispersion corrected revised Perdew, Burke, and Ernzerhof functional (revPBE-D3) (2) the recently developed strongly constrained and appropriately normed (SCAN) functional (3) the random phase approximation (RPA) functional for potassium. We compare with experimental X-ray diffraction (XRD) and X-ray absorption fine structure (EXAFS) measurements to demonstrate that SCAN accurately reproduces key structural details of the hydration structure around the sodium and potassium cations, whereas revPBE-D3 fails to do so. However, we show that SCAN provides a worse description of pure water in comparison with revPBE-D3. RPA also shows an improvement for K+, but slow convergence prevents rigorous comparison. Finally, we analyse cluster energetics to show SCAN and RPA have smaller fluctuations of the mean error of ion-water cluster binding energies compared with revPBE-D3.

Improvement of Hard Carbon Electrode Performance by Manipulating SEI Formation at High Charging Rates.

There is a growing demand for high-rate rechargeable batteries for powering electric vehicles and portable electronics. Here, we demonstrate a strategy for improving electrode performance by controlling the formation of solid electrolyte interphase (SEI). A composite electrode consisting of hard carbon (HC) and carbon nanotubes (CNTs) was used to study the formation of the SEI at different charging rates in an electrolyte consisting of 1 M NaClO4 in a mixed solvent with ethylene carbonate (EC) and propylene carbonate (PC), as well as fluoroethylene carbonate (FEC) additive. The half-cell method was used to form the SEI at different charging rates (e.g., 1, 10, and 100 A/g). Symmetric capacitor cells were employed to study ion transport properties through the SEI. It was found that the SEI is a primary factor responsible for limiting the capacity of the composite anode material in conventional ester-based electrolytes. The electrode with the SEI formed at 100 A/g exhibited the lowest impedance and delivered nearly twice the capacity of the electrode with the SEI formed at 1 A/g. This significant difference is due to a thin SEI formed at the fast charging rate, as has been observed with ether-based electrolytes. An identical decay rate (0.11 mA h/g per cycle) was observed on the electrodes with SEIs formed at different charging rates in an ester electrolyte. No chemical difference among the three SEI layers was found. However, morphological differences of the SEI layers were observed. This difference is believed to account for the different electrochemical behaviors of the electrodes. This work shows that high charging rates can result in the formation of an optimal SEI layer, contradicting the widely accepted practice of using low charging rates during the SEI formation in alkali-ion batteries.

Detecting the undetectable: The role of trace surfactant in the Jones-Ray effect.

The surface tension of dilute salt water is a fundamental property that is crucial to understanding the complexity of many aqueous phase processes. Small ions are known to be repelled from the air-water surface leading to an increase in the surface tension in accordance with the Gibbs adsorption isotherm. The Jones-Ray effect refers to the observation that at extremely low salt concentration, the surface tension decreases. Determining the mechanism that is responsible for this Jones-Ray effect is important for theoretically predicting the distribution of ions near surfaces. Here we use both experimental surface tension measurements and numerical solution of the Poisson-Boltzmann equation to demonstrate that very low concentrations of surfactant in water create a Jones-Ray effect. We also demonstrate that the low concentrations of the surfactant necessary to create the Jones-Ray effect are too small to be detectable by surface sensitive spectroscopic measurements. The effect of surface curvature on this behavior is also examined, and the implications for unexplained bubble phenomena are discussed. This work suggests that the purity standards for water may be inadequate and that the interactions between ions with background impurities are important to incorporate into our understanding of the driving forces that give rise to the speciation of ions at interfaces.

Cavitation energies can outperform dispersion interactions.

The accurate dissection of binding energies into their microscopic components is challenging, especially in solution. Here we study the binding of noble gases (He-Xe) with the macrocyclic receptor cucurbit[5]uril in water by displacement of methane and ethane as 1H NMR probes. We dissect the hydration free energies of the noble gases into an attractive dispersive component and a repulsive one for formation of a cavity in water. This allows us to identify the contributions to host-guest binding and to conclude that the binding process is driven by differential cavitation energies rather than dispersion interactions. The free energy required to create a cavity to accept the noble gas inside the cucurbit[5]uril is much lower than that to create a similarly sized cavity in bulk water. The recovery of the latter cavitation energy drives the overall process, which has implications for the refinement of gas-storage materials and the understanding of biological receptors.

Understanding the scale of the single ion free energy: A critical test of the tetra-phenyl arsonium and tetra-phenyl borate assumption.

The tetra-phenyl arsonium and tetra-phenyl borate (TATB) assumption is a commonly used extra-thermodynamic assumption that allows single ion free energies to be split into cationic and anionic contributions. The assumption is that the values for the TATB salt can be divided equally. This is justified by arguing that these large hydrophobic ions will cause a symmetric response in water. Experimental and classical simulation work has raised potential flaws with this assumption, indicating that hydrogen bonding with the phenyl ring may favor the solvation of the TB- anion. Here, we perform ab initio molecular dynamics simulations of these ions in bulk water demonstrating that there are significant structural differences. We quantify our findings by reproducing the experimentally observed vibrational shift for the TB- anion and confirm that this is associated with hydrogen bonding with the phenyl rings. Finally, we demonstrate that this results in a substantial energetic preference of the water to solvate the anion. Our results suggest that the validity of the TATB assumption, which is still widely used today, should be reconsidered experimentally in order to properly reference single ion solvation free energy, enthalpy, and entropy.