Dispersion corrected RPBE studies of liquid water

Toward Explicit Solvation for Simulations of Electrocatalytic Reactions: AIMD for pKa and Redox Potentials of Transition Metal Compounds and Catalyst Models

In this work, we study the possibility to extend electronic structure simulations for electrocatalysis by explicit solvation models. In previous work, we proposed a simulation scheme that explicitly includes the effects of pH and electrochemical potential in density functional theory (DFT) simulations with implicit solvation. Based on calculations of protonation and oxidation reactions, the pH and electrochemical potential can be included given appropriate reference values. In this work, we compute the pKa values and oxidation potentials for a series of transition metal aquo complexes and compare the results including implicit, explicit static and explicit dynamic (AIMD) models for the aqueous solvent and compare vs experimental pKa and redox potential data. This allows the construction of a pKa/redox potential scale that can in principle be extrapolated to the simulation of other transition metal-based materials. An explicit dynamic solvent model is then proposed and applied to a model system for iridium oxide-based catalysts for the oxygen evolution reaction. We outline the advantages and disadvantages of the different approaches and demonstrate that, at the expense of a larger computational effort, the microsolvation environment of a given model can be described in a robust way using a limited amount of solvent molecules and AIMD. Especially for reactions in which water is solvent and reactant like the oxygen evolution reaction (OER) or oxygen reduction reaction (ORR), this model provides a more detailed and complete description that can be exploited in mechanistic studies.

Random Sampling Versus Active Learning Algorithms for Machine Learning Potentials of Quantum Liquid Water

Training accurate machine learning potentials requires electronic structure data comprehensively covering the configurational space of the system of interest. As the construction of this data is computationally demanding, many schemes for identifying the most important structures have been proposed. Here, we compare the performance of high-dimensional neural network potentials (HDNNPs) for quantum liquid water at ambient conditions trained to data sets constructed using random sampling as well as various flavors of active learning based on query by committee. Contrary to the common understanding of active learning, we find that for a given data set size, random sampling leads to smaller test errors for structures not included in the training process. In our analysis, we show that this can be related to small energy offsets caused by a bias in structures added in active learning, which can be overcome by using instead energy correlations as an error measure that is invariant to such shifts. Still, all HDNNPs yield very similar and accurate structural properties of quantum liquid water, which demonstrates the robustness of the training procedure with respect to the training set construction algorithm even when trained to as few as 200 structures. However, we find that for active learning based on preliminary potentials, a reasonable initial data set is important to avoid an unnecessary extension of the covered configuration space to less relevant regions.

Machine learning the electric field response of condensed phase systems using perturbed neural network potentials

The interaction of condensed phase systems with external electric fields is of major importance in a myriad of processes in nature and technology, ranging from the field-directed motion of cells (galvanotaxis), to geochemistry and the formation of ice phases on planets, to field-directed chemical catalysis and energy storage and conversion systems including supercapacitors, batteries and solar cells. Molecular simulation in the presence of electric fields would give important atomistic insight into these processes but applications of the most accurate methods such as ab-initio molecular dynamics (AIMD) are limited in scope by their computational expense. Here we introduce Perturbed Neural Network Potential Molecular Dynamics (PNNP MD) to push back the accessible time and length scales of such simulations. We demonstrate that important dielectric properties of liquid water including the field-induced relaxation dynamics, the dielectric constant and the field-dependent IR spectrum can be machine learned up to surprisingly high field strengths of about 0.2 V Å-1 without loss in accuracy when compared to ab-initio molecular dynamics. This is remarkable because, in contrast to most previous approaches, the two neural networks on which PNNP MD is based are exclusively trained on molecular configurations sampled from zero-field MD simulations, demonstrating that the networks not only interpolate but also reliably extrapolate the field response. PNNP MD is based on rigorous theory yet it is simple, general, modular, and systematically improvable allowing us to obtain atomistic insight into the interaction of a wide range of condensed phase systems with external electric fields.

Hybrid Quantum Mechanical, Molecular Mechanical, and Machine Learning Potential for Computing Aqueous-Phase Adsorption Free Energies on Metal Surfaces

Performing reliable computer simulations of elementary processes occurring at metal-water interfaces is pivotal for novel catalyst design in sustainable energy applications. Computational catalyst design hinges on the ability to reliably and efficiently compute the potential energy surface (PES) of the system. Due to the large system sizes needed for studying processes at liquid water-metal interfaces, these systems can currently not be described using density functional theory (DFT). In this work, we used a hybrid quantum mechanical, molecular mechanical, and machine learning potential for studying the adsorption behavior of phenol, atomic hydrogen, 2-butanol, and 2-butanone on the (0001) facet of Ru under reducing conditions when Ru is not oxidized. Specifically, we describe the adsorbate and the surrounding metal atoms at the DFT level of theory. Here, we also considered the electrostatic field effect of the water molecules on adsorbate-metal interactions. Next, for the water-water and water-adsorbate interactions, we used established classical force fields. Finally, for the water-Ru surface interaction, for which no reliable force fields have been published, we used Behler-Parrinello high-dimensional neural network potentials (HDNNPs). Employing this setup, we used our explicit solvation for metal surface (eSMS) approach to compute the aqueous-phase effect on the low-coverage adsorption of selected molecules and atoms on the (0001) facet of Ru. In agreement with previous experimental and computational studies of oxygenated molecules over transition metal facets, we found that liquid water destabilizes the tested adsorbates on Ru(0001). Interestingly, our findings indicate that adsorbates on Ru are less affected by the presence of an aqueous phase than on other transition metals (e.g., Pt), highlighting the necessity of experimental investigations of Ru-based catalytic systems in liquid water.

Metal-water interface formation: Thermodynamics from ab initio molecular dynamics simulations

Metal-water interfaces are central to many electrochemical, (electro)catalytic, and materials science processes and systems. However, our current understanding of their thermodynamic properties is limited by the scarcity of accurate experimental and computational data and procedures. In this work, thermodynamic quantities for metal-water interface formation are computed for a range of FCC(111) surfaces (Pd, Pt, Au, Ag, Rh, and PdAu) through extensive density functional theory based molecular dynamics and the two-phase entropy model. We show that metal-water interface formation is thermodynamically favorable and that most metal surfaces studied in this work are completely wettable, i.e., have contact angles of zero. Interfacial water has higher entropy than bulk water due to the increased population of low-frequency translational modes. The entropic contributions also correlate with the orientational water density, and the highest solvation entropies are observed for interfaces with a moderately ordered first water layer; the entropic contributions account for up to ∼25% of the formation free energy. Water adsorption energy correlates with the water orientation and structure and is found to be a good descriptor of the internal energy part of the interface formation free energy, but it alone cannot satisfactorily explain the interfacial thermodynamics; the interface formation is driven by the competition between energetic and entropic contributions. The obtained results and insight can be used to develop, parameterize, and benchmark theoretical and computational methods for studying metal-water interfaces. Overall, our study yields benchmark-quality data and fundamental insight into the thermodynamic forces driving metal-water interface formation.

Benchmarking water adsorption on metal surfaces with ab initio molecular dynamics

Solid-water interfaces are ubiquitous in nature and technology. In particular, technologies evolving in the green transition, such as electrocatalysis, heavily rely on the junction of an electrolyte and an electrode as a central part of the device. For the understanding of atomic-scale processes taking place at the electrolyte-electrode interface, density functional theory (DFT) has become the de facto standard. The validation of DFT's ability to simulate the interfacial solid/water interaction is crucial, and ideal simulation setups need to be identified in order to prevent avoidable systematic errors. Here, we develop a rigorous sampling protocol for benchmarking the adsorption/desorption energetics of water on metallic surfaces against experimental temperature programmed desorption, single crystal adsorption calorimetry, and thermal energy atom scattering. We screened DFT's quality on a series of transition metal surfaces, applying three of the most common exchange-correlation approximations: PBE-D3, RPBE-D3, and BEEF-vdW. We find that all three xc-functionals reflect the pseudo-zeroth order desorption of water rooted in the combination of attractive adsorbate-adsorbate interactions and their saturation at low and intermediate coverages, respectively. However, both RPBE-D3 and BEEF-vdW lead to more accurate water adsorption strengths, while PBE-D3 clearly overbinds near-surface water. We relate the variations in binding strength to specific variations in water-metal and water-water interactions, highlighting the structural consequences inherent in an uninformed choice of simulation parameters. Our study gives atomistic insight into water's complex adsorption equilibrium. Furthermore, it represents a guideline for future DFT-based simulations of solvated solid interfaces by providing an assessment of systematic errors in specific setups.

Cavity formation at metal-water interfaces

The free energy cost of forming a cavity in a solvent is a fundamental concept in rationalizing the solvation of molecules and ions. A detailed understanding of the factors governing cavity formation in bulk solutions has inter alia enabled the formulation of models that account for this contribution in coarse-grained implicit solvation methods. Here, we employ classical molecular dynamics simulations and multistate Bennett acceptance ratio free energy sampling to systematically study cavity formation at a wide range of metal-water interfaces. We demonstrate that the obtained size- and position-dependence of cavitation energies can be fully rationalized by a geometric Gibbs model, which considers that the creation of the metal-cavity interface necessarily involves the removal of interfacial solvent. This so-called competitive adsorption effect introduces a substrate dependence to the interfacial cavity formation energy that is missed in existing bulk cavitation models. Using expressions from scaled particle theory, this substrate dependence is quantitatively reproduced by the Gibbs model through simple linear relations with the adsorption energy of a single water molecule. Besides providing a better general understanding of interfacial solvation, this paves the way for the derivation and efficient parametrization of more accurate interface-aware implicit solvation models needed for reliable high-throughput calculations toward improved electrocatalysts.

Accelerating explicit solvent models of heterogeneous catalysts with machine learning interatomic potentials

Realistically modelling how solvents affect catalytic reactions is a longstanding challenge due to its prohibitive computational cost. Typically, an explicit atomistic treatment of the solvent molecules is needed together with molecular dynamics (MD) simulations and enhanced sampling methods. Here, we demonstrate the utility of machine learning interatomic potentials (MLIPs), coupled with active learning, to enable fast and accurate explicit solvent modelling of adsorption and reactions on heterogeneous catalysts. MLIPs trained on-the-fly were able to accelerate ab initio MD simulations by up to 4 orders of magnitude while reproducing with high fidelity the geometrical features of water in the bulk and at metal-water interfaces. Using these ML-accelerated simulations, we accurately predicted key catalytic quantities such as the adsorption energies of CO*, OH*, COH*, HCO*, and OCCHO* on Cu surfaces and the free energy barriers of C-H scission of ethylene glycol over Cu and Pd surfaces, as validated with ab initio calculations. We envision that such simulations will pave the way towards detailed and realistic studies of solvated catalysts at large time- and length-scales.

Accurate diffusion coefficients of the excess proton and hydroxide in water via extensive ab initio simulations with different schemes

Despite its simple molecular formula, obtaining an accurate in silico description of water is far from straightforward. Many of its very peculiar properties are quite elusive, and in particular, obtaining good estimations of the diffusion coefficients of the solvated proton and hydroxide at a reasonable computational cost has been an unsolved challenge until now. Here, I present extensive results of several unusually long ab initio molecular dynamics (MD) simulations employing different combinations of the Born-Oppenheimer and second-generation Car-Parrinello MD propagation methods with different ensembles (NVE and NVT) and thermostats, which show that these methods together with the RPBE-D3 functional provide a very accurate estimation of the diffusion coefficients of the solvated H₃O⁺ and OH^- ions, together with an extremely accurate description of several properties of neutral water (such as the structure of the liquid and its diffusion and shear viscosity coefficients). In addition, I show that the estimations of D_H3O⁺ and D_OH^- depend dramatically on the simulation length, being necessary to reach timescales in the order of hundreds of picoseconds to obtain reliable results.

Simulating the solvation structure of low- and high-spin [Fe(bpy)3]2+: long-range dispersion and many-body effects

When characterizing transition metal complexes and their functionalities, the importance of including the solvent as an active participant is becoming more and more apparent. Whereas many studies have evaluated long-range dispersion effects inside organic molecules and organometallics, less is known about their role in solvation. Here, we have analysed the components within solute-solvent and solvent-solvent interactions of one of the most studied iron-based photoswitch model systems, in two spin states. We find that long-range dispersion effects modulate the coordination significantly, and that this is accurately captured by density functional theory models including dispersion corrections. We furthermore correlate gas-phase relaxed complex-water clusters to thermally averaged molecular densities. This shows how the gas-phase interactions translate to solution structure, quantified through 3D molecular densities, angular distributions, and radial distribution functions. We show that finite-size simulation cells can cause the radial distribution functions to have artificially enlarged amplitudes. Finally, we quantify the effects of many-body interactions within the solvent shells, and find that almost a fifth of the total interaction energy of the solute-shell system in the high-spin state comes from many-body contributions, which cannot be captured by by pair-wise additive force field methods.

Ab Initio Simulations of Water/Metal Interfaces

Structures and processes at water/metal interfaces play an important technological role in electrochemical energy conversion and storage, photoconversion, sensors, and corrosion, just to name a few. However, they are also of fundamental significance as a model system for the study of solid-liquid interfaces, which requires combining concepts from the chemistry and physics of crystalline materials and liquids. Particularly interesting is the fact that the water-water and water-metal interactions are of similar strength so that the structures at water/metal interfaces result from a competition between these comparable interactions. Because water is a polar molecule and water and metal surfaces are both polarizable, explicit consideration of the electronic degrees of freedom at water/metal interfaces is mandatory. In principle, ab initio molecular dynamics simulations are thus the method of choice to model water/metal interfaces, but they are computationally still rather demanding. Here, ab initio simulations of water/metal interfaces will be reviewed, starting from static systems such as the adsorption of single water molecules, water clusters, and icelike layers, followed by the properties of liquid water layers at metal surfaces. Technical issues such as the appropriate first-principles description of the water-water and water-metal interactions will be discussed, and electrochemical aspects will be addressed. Finally, more approximate but numerically less demanding approaches to treat water at metal surfaces from first-principles will be briefly discussed.

A first principle study of water adsorbed on flat and stepped silver surfaces

The structural, electronic and vibrational properties of a water layer on Ag(100) and Ag(511) have been studied by first principles calculations and ab initio molecular dynamics simulations. The most stable...

Suitable acid groups and density in electrolytes to facilitate proton conduction

Proton conducting materials suffer from low proton conductivity under low-relative humidity (RH) conditions. Previously, it was reported that acid-acid interactions, where acids interact with each other at close distances, can facilitate proton conduction without water movement and are promising for overcoming this drawback [T. Ogawa, H. Ohashi, T. Tamaki and T. Yamaguchi, Chem. Phys. Lett., 2019, 731, 136627]. However, acid groups have not been compared to find a suitable acid group and density for the interaction, which is important to experimentally synthesize the material. Here, we performed ab initio calculations to identify acid groups and acid densities as a polymer design that effectively causes acid-acid interactions. The evaluation method employed parameters based on several different optimized coordination interactions of acids and water molecules. The results show that the order of the abilities of polymer electrolytes to readily induce acid-acid interactions is hydrocarbon-based phosphonated polymers > phosphonated aromatic hydrocarbon polymers > perfluorosulfonic acid polymers ≈ perfluorophosphonic acid polymers > sulfonated aromatic hydrocarbon polymers. The acid-acid interaction becomes stronger as the distance between acids decreases. The preferable distance between phosphonate moieties is within 13 Å.

Machine learning potentials for complex aqueous systems made simple

Understanding complex materials, in particular those with solid–liquid interfaces, such as water on surfaces or under confinement, is a key challenge for technological and scientific progress. Although established simulation approaches have been able to provide important atomistic insight, ab initio techniques struggle with the required time and length scales, while force field methods can often be limited in terms of their accuracy. Here we show how these limitations can be overcome in a simple and automated machine learning procedure to provide accurate models of interactions at the ab initio level, as illustrated for a variety of complex aqueous systems. These developments open up the prospect of the straightforward exploration of many technologically relevant systems by molecular simulations.

Simulation techniques based on accurate and efficient representations of potential energy surfaces are urgently needed for the understanding of complex systems such as solid–liquid interfaces. Here we present a machine learning framework that enables the efficient development and validation of models for complex aqueous systems. Instead of trying to deliver a globally optimal machine learning potential, we propose to develop models applicable to specific thermodynamic state points in a simple and user-friendly process. After an initial ab initio simulation, a machine learning potential is constructed with minimum human effort through a data-driven active learning protocol. Such models can afterward be applied in exhaustive simulations to provide reliable answers for the scientific question at hand or to systematically explore the thermal performance of ab initio methods. We showcase this methodology on a diverse set of aqueous systems comprising bulk water with different ions in solution, water on a titanium dioxide surface, and water confined in nanotubes and between molybdenum disulfide sheets. Highlighting the accuracy of our approach with respect to the underlying ab initio reference, the resulting models are evaluated in detail with an automated validation protocol that includes structural and dynamical properties and the precision of the force prediction of the models. Finally, we demonstrate the capabilities of our approach for the description of water on the rutile titanium dioxide (110) surface to analyze the structure and mobility of water on this surface. Such machine learning models provide a straightforward and uncomplicated but accurate extension of simulation time and length scales for complex systems.

Solvent Effects in the Ultraviolet and X-ray Absorption Spectra of Pyridazine in Aqueous Solution

Electrostatic interaction of the solvent with the solute and fluctuations of the solvent configurations may make excitation energies of the solute different from those in the gas phase. These effects may dominate photoinduced or chemical reaction dynamics in solution systems and can be observed as shifts or broadening of peaks in absorption spectra. In this work, the nitrogen K-edge X-ray absorption spectra were measured for pyridazine in the gas phase and in aqueous solution. The ultraviolet and X-ray absorption spectra of pyridazine in aqueous solution, as well as those in the gas phase, were then calculated with models based on the algebraic-diagrammatic construction through second order [ADC(2)] with the resolution-of-identity (RI) approximation and compared with the spectra obtained in experiments. For aqueous solution, explicit local solvation structures were extracted from an ab initio molecular dynamics (AIMD) trajectory of pyridazine in bulk water, and RI-ADC(2) was combined with the conductor-like screening model (COSMO). The experimental absorption spectra of pyridazine in aqueous solution were reproduced with good accuracy by theoretical treatment of an ensemble containing the explicit local solvation structures of pyridazine with relevant water molecules combined with the COSMO solvation model of water for long-range solvation.

Aqueous TMAO solution under high hydrostatic pressure

Trimethylamine N-oxide (TMAO) is a well known osmolyte in nature, which is used by deep sea fish to stabilize proteins against High Hydrostatic Pressure (HHP). We present a combined ab initio molecular dynamics, force field molecular dynamics, and THz absorption study of TMAO in water up to 12 kbar to decipher its solvation properties upon extreme compression. On the hydrophilic oxygen side of TMAO, AIMD simulations at 1 bar and 10 kbar predict a change of the coordination number from a dominating TMAO·(H₂O)₃ complex at ambient conditions towards an increased population of a TMAO·(H₂O)₄ complex at HHP conditions. This increase of the TMAO-oxygen coordination number goes in line with a weakening of the local hydrogen bond network, spectroscopic shifts and intensity changes of the corresponding intermolecular THz bands. Using a pressure-dependent HHP force field, FFMD simulations predict a significant increase of hydrophobic hydration from 1 bar up to 4-5 kbar, which levels off at higher pressures up to 10 kbar. THz spectroscopic data reveal two important pressure regimes with spectroscopic inflection points of the dominant intermolecular modes: The first regime (1.5-2 kbar) is barely recognizable in the simulation data. However, it relates well with the observation that the apparent molar volume of solvated TMAO is nearly constant in the biologically relevant pressure range up to 1 kbar as found in the deepest habitats on Earth in the ocean. The second inflection point around 4-5 kbar is related to the amount of hydrophobic hydration as predicted by the FFMD simulations. In particular, the blueshift of the intramolecular CNC bending mode of TMAO at about 390 cm^-1 is the spectroscopic signature of increasingly pronounced pressure-induced changes in the solvation shell of TMAO. Thus, the CNC bend can serve as local pressure sensor in the multi-kbar pressure regime.

Quantum Simulations of Hydrogen Bonding Effects in Glycerol Carbonate Electrolyte Solutions

The need for environmentally friendly nonaqueous solvents in electrochemistry and other fields has motivated recent research into the molecular-level solvation structure, thermodynamics, and dynamics of candidate organic liquids. In this paper, we present the results of quantum density functional theory simulations of glycerol carbonate (GC), a molecule that has been proposed as a solvent for green industrial chemistry, nonaqueous alternatives for biocatalytic reactions, and liquid media in energy storage devices. We investigate the structure and dynamics of both the pure GC liquid and electrolyte solutions containing KF and KCl ion pairs. These simulations reveal the importance of hydrogen bonding that controls the structural and dynamic behavior of the pure liquid and ion association in the electrolyte solutions. The results illustrate the difficulties associated with classical modeling of complex organic solvents. The simulations lead to a better understanding of the underlying mechanisms behind the previously observed peculiar ion-specific behavior in GC electrolyte solutions.

Model Studies on the Formation of the Solid Electrolyte Interphase: Reaction of Li with Ultrathin Adsorbed Ionic-Liquid Films and Co3 O4 (111) Thin Films

In this work we aim towards the molecular understanding of the solid electrolyte interphase (SEI) formation at the electrode electrolyte interface (EEI). Herein, we investigated the interaction between the battery‐relevant ionic liquid (IL) 1‐butyl‐1‐methylpyrrolidinium bis(trifluoromethylsulfonyl)imide (BMP‐TFSI), Li and a Co₃O₄(111) thin film model anode grown on Ir(100) as a model study of the SEI formation in Li‐ion batteries (LIBs). We employed mostly X‐ray photoelectron spectroscopy (XPS) in combination with dispersion‐corrected density functional theory calculations (DFT‐D3). If the surface is pre‐covered by BMP‐TFSI species (model electrolyte), post‐deposition of Li (Li⁺ ion shuttle) reveals thermodynamically favorable TFSI decomposition products such as LiCN, Li₂NSO₂CF₃, LiF, Li₂S, Li₂O₂, Li₂O, but also kinetic products like Li₂NCH₃C₄H₉ or LiNCH₃C₄H₉ of BMP. Simultaneously, Li adsorption and/or lithiation of Co₃O₄(111) to Li_nCo₃O₄ takes place due to insertion via step edges or defects; a partial transformation to CoO cannot be excluded. Formation of Co⁰ could not be observed in the experiment indicating that surface reaction products and inserted/adsorbed Li at the step edges may inhibit or slow down further Li diffusion into the bulk. This study provides detailed insights of the SEI formation at the EEI, which might be crucial for the improvement of future batteries.

Mechanisms of Al3+ Dimerization in Alkaline Solutions

The molecular speciation of aluminum (Al³⁺) in alkaline solutions is fundamental to its precipitation chemistry within a number of industrial applications that include ore refinement and industrial processing of Al wastes. Under these conditions, Al³⁺ is predominantly Al(OH)₄^-, while at high [Al³⁺] dimeric species are also known to form. To date, the mechanism of dimer formation remains unclear and is likely influenced by complex ion···ion interactions. In the present work, we investigate a suite of potential dimerization pathways and the role of ion pairing on energetics using static DFT calculations and DFT and density functional tight binding molecular dynamics. Specific cation effects imparted by the background electrolyte cations Na⁺, Li⁺, and K⁺ have been examined. Our simulations predict that, when the Al species are ion-paired with either cation, the formation of the oxo-bridged Al₂O(OH)₆^2- is favored with respect to the dihydroxo-bridged Al₂(OH)₈^2-, in agreement with previous spectroscopic work. The formation of both dimers first proceeds by bridging of two monomeric units via one hydroxo ligand, leading to a labile Al₂(OH)₈^2- isomer. The effect of contact ion pairing of Li⁺ and K⁺ on the dimerization energetics is distinctly more favorable than that of Na⁺, which may have an effect on further oligomerization.

Sulfate, Bisulfate, and Hydrogen Co-adsorption on Pt(111) and Au(111) in an Electrochemical Environment

The co-adsorption of sulfate, bisulfate and hydrogen on Pt(111) and Au(111) electrodes was studied based on periodic density functional calculations with the aqueous electrolyte represented by both explicit and implicit solvent models. The influence of the electrochemical control parameters such as the electrode potential and pH was taken into account in a grand-canonical approach. Thus, phase diagrams of the stable coadsorption phases as a function of the electrochemical potential and Pourbaix diagrams have been derived which well reproduce experimental findings. We demonstrate that it is necessary to include explicit water molecules in order to determine the stable adsorbate phases as the (bi)sulfate adsorbates rows become significantly stabilized by bridging water molecules.

Revisiting the Volmer-Heyrovský mechanism of hydrogen evolution on a nitrogen doped carbon nanotube: constrained molecular dynamics versus the nudged elastic band method

Density functional theory (DFT) based computational electrochemistry has the potential to serve as a tool with predictive power in the rational development and screening of electrocatalysts for renewable energy technologies. It is, however, of paramount importance that simulations are conducted rigorously at a level of theory that is sufficiently accurate in order to obtain physicochemically sensible results. Herein, we present a comparative study of the performance of the static climbing image nudged elastic band method (CI-NEB) vs. DFT based constrained molecular dynamics simulations with thermodynamic integration in estimating activation and reaction (free) energies of the Volmer-Heyrovský mechanism on a nitrogen doped carbon nanotube. Due to cancellation of errors within the CI-NEB calculations, static and dynamic activation barriers are observed to be surprisingly similar, while a substantial decrease in reaction energies is seen upon incorporation of solvent dynamics. This finding is attributed to two competing effects; (1) solvent reorganization that stabilizes the transition and, in particular, the product states with respect to the reactant state and (2) destabilizing entropic contributions due to solvent fluctuations. Our results highlight the importance of explicitly sampling the interfacial solvent dynamics when studying hydrogen evolution at solid-liquid interfaces.

Assessing the properties of supercritical water in terms of structural dynamics and electronic polarization effects

Evolution of water's structural dynamics from ambient liquid to supercritical dense liquid-like and dilute gas-like conditions.

Water structures on a Pt(111) electrode from ab initio molecular dynamic simulations for a variety of electrochemical conditions

A structural analysis of solvating water layers on a Pt(111) electrode has been performed based on extensive ab initio molecular dynamics simulations. We have emulated different electrochemical conditions by varying the concentration of hydrogen ions in the water layers, which effectively corresponds to a variation in the electrode potential. We present a detailed analysis of the arrangement and orientation of the water molecules and also address their mobility in the solvation layer.

Ab Initio Structure and Dynamics of CO2 at Supercritical Conditions

Green technologies rely on green solvents and fluids. Among them, supercritical CO₂ already finds many important applications. The molecular-level understanding of the dynamics and structure of this supercritical fluid is a prerequisite for rational design of future green technologies. Unfortunately, the commonly employed Kohn-Sham density functional theory (DFT) is too computationally demanding to produce meaningfully converged dynamics within a reasonable time and with a reasonable computational effort. Thanks to subsystem DFT, we analyze finite-size effects by considering simulation cells of varying sizes (up to 256 independent molecules in the cell) and finite-time effects by running 100 ps trajectories. We find that the simulations are in reasonable and semiquantitative agreement with the available neutron diffraction experiments and that, as opposed to the gas phase, the CO₂ molecules in the fluid are bent with an average OCO angle of 175.8°. Our simulations also confirm that the dimer T-shape is the most prevalent configuration. Our results further strengthen the experiment-simulation agreement for this fluid when comparing radial distribution functions and diffusion coefficient, confirming subsystem DFT as a viable tool for modeling structure and dynamics of condensed-phase systems.

Interaction of H2O with the Platinum Pt (001), (011), and (111) Surfaces: A Density Functional Theory Study with Long-Range Dispersion Corrections

Platinum is a noble metal that is widely used for the electrocatalytic production of hydrogen, but the surface reactivity of platinum toward water is not yet fully understood, even though the effect of water adsorption on the surface free energy of Pt is important in the interpretation of the morphology and catalytic properties of this metal. In this study, we have carried out density functional theory calculations with long-range dispersion corrections [DFT-D3-(BJ)] to investigate the interaction of H₂O with the Pt (001), (011), and (111) surfaces. During the adsorption of a single H₂O molecule on various Pt surfaces, it was found that the lowest adsorption energy (E _ads) was obtained for the dissociative adsorption of H₂O on the (001) surface, followed by the (011) and (111) surfaces. When the surface coverage was increased up to a monolayer, we noted an increase in E _ads/H₂O with increasing coverage for the (001) surface, while for the (011) and (111) surfaces, E _ads/H₂O decreased. Considering experimental conditions, we observed that the highest coverage was obtained on the (011) surface, followed by the (111) and (001) surfaces. However, with an increase in temperature, the surface coverage decreased on all the surfaces. Total desorption occurred at temperatures higher than 400 K for the (011) and (111) surfaces, but above 850 K for the (001) surface. From the morphology analysis of the Pt nanoparticle, we noted that, when the temperature increased, only the electrocatalytically active (111) surface remained.

Structure and thermodynamics of aqueous urea solutions from ambient to kilobar pressures: From thermodynamic modeling, experiments, and first principles simulations to an accurate force field description

Molecular simulations based on classical force fields are a powerful method for shedding light on the complex behavior of biomolecules in solution. When cosolutes are present in addition to water and biomolecules, subtle balances of weak intermolecular forces have to be accounted for. This imposes high demands on the quality of the underlying force fields, and therefore force field development for small cosolutes is still an active field. Here, we present the development of a new urea force field from studies of urea solutions at ambient and elevated hydrostatic pressures based on a combination of experimental and theoretical approaches. Experimental densities and solvation shell properties from ab initio molecular dynamics simulations at ambient conditions served as the target properties for the force field optimization. Since urea is present in many marine life forms, elevated hydrostatic pressure was rigorously addressed: densities at high pressure were measured by vibrating tube densitometry up to 500 bar and by X-ray absorption up to 5 kbar. Densities were determined by the perturbed-chain statistical associating fluid theory equation of state. Solvation properties were determined by embedded cluster integral equation theory and ab initio molecular dynamics. Our new force field is able to capture the properties of urea solutions at high pressures without further high-pressure adaption, unlike trimethylamine-N-oxide, for which a high-pressure adaption is necessary.

Structure and Dynamics of Water at the Water-Air Interface Using First-Principles Molecular Dynamics Simulations. II. NonLocal vs Empirical van der Waals Corrections

van der Waals (vdW) correction schemes have been recognized to be essential for an accurate description of liquid water in first-principles molecular dynamics simulation. The description of the structure and dynamics of water is governed by the type of the vdW corrections. So far, two vdW correction schemes have been often used: empirical vdW corrections and nonlocal vdW corrections. In this paper, we assess the influence of the empirical vs nonlocal vdW correction schemes on the structure and dynamics of water at the water–air interface. Since the structure of water at the water–air interface is established by a delicate balance of hydrogen bond formation and breaking, the simulation at the water–air interface provides a unique platform to testify as to the heterogeneous interaction of water. We used the metrics [Ohto et al. J. Chem. Theory Comput., 2019, 15, 595−60230468702] which are directly connected with the sum-frequency generation spectroscopic measurement. We find that the overall performance of nonlocal vdW methods is either similar or worse compared to the empirical vdW methods. We also investigated the performance of the optB88-DRSLL functional, which showed slightly less accuracy than the revPBE-D3 method. We conclude that the revPBE-D3 method shows the best performance for describing the interfacial water.

Effect of Functional and Electron Correlation on the Structure and Spectroscopy of the Al2O3(001)-H2O Interface

Oxide-water interfaces are ubiquitous, with many applications in industry and the environment, yet there is a great deal of controversy over their properties and microscopic structure. This controversy stems, in part, from the unique H-bond networks formed at different surface terminations and mineral compositions. Density functional theory simulations of these interfaces require an accurate description of both the oxide mineral and water in diverse H-bond environments. Thus, herein we simulate the Al₂O₃(001)-H₂O interface using the PBE, PBE-TS, RPBE, SCAN, and HSE06-TS functionals to determine how calculated interfacial properties depend on the choice of functional. We find that the structure of the first few layers of water at the surface is determined by electron correlation in a way that cannot be approximated using semiemipirical van der Waals corrections. Of the functionals investigated, we find that SCAN yields the most accurate interfacial structure, dynamics, and sum frequency generation spectrum. Furthermore, SCAN leads to a reduction in the order of the 2D H-bond network of water at the alumina surface predicted by GGA functionals, leading to a significant decrease in the anisotropy of the diffusion coefficient at the surface. We emphasize the importance of using a functional which accurately describes electron correlation for more complex oxides, such as transition-metal oxides, where electron correlation may play an even greater role in determining the structure and dynamics of the oxide-water interface.

Nanoconfined Water within Graphene Slit Pores Adopts Distinct Confinement-Dependent Regimes

In view of the increasing importance of nanoconfined aqueous solutions for various technological applications, it has become necessary to understand how strong confinement affects the properties of water at the level of molecular and even electronic structure. By performing extensive ab initio simulations of two-dimensionally nanoconfined water lamellae between graphene sheets subject to different interlayer spacings, we find new regimes at interlayer distances of 10 Å and less where water can be described neither to behave like interfacial water nor to be bulklike at the level of its H-bonding characteristics and electronic structure properties. It is expected that this finding will offer new opportunities to tune both diffusive and reactive processes taking place in aqueous environments that are strongly confined by chemically inert hard walls.

Ab Initio Molecular Dynamics Simulations of the Influence of Lithium Bromide on the Structure of the Aqueous Solution-Air Interface

We present the results of ab initio molecular dynamics simulations of the solution–air interface of aqueous lithium bromide (LiBr). We find that, in agreement with the experimental data and previous simulation results with empirical polarizable force field models, Br^– anions prefer to accumulate just below the first molecular water layer near the interface, whereas Li⁺ cations remain deeply buried several molecular layers from the interface, even at very high concentration. The separation of ions has a profound effect on the average orientation of water molecules in the vicinity of the interface. We also find that the hydration number of Li⁺ cations in the center of the slab N_{c,Li⁺–H2O} ≈ 4.7 ± 0.3, regardless of the salt concentration. This estimate is consistent with the recent experimental neutron scattering data, confirming that results from nonpolarizable empirical models, which consistently predict tetrahedral coordination of Li⁺ to four solvent molecules, are incorrect. Consequently, disruption of the hydrogen bond network caused by Li⁺ may be overestimated in nonpolarizable empirical models. Overall, our results suggest that empirical models, in particular nonpolarizable models, may not capture all of the properties of the solution–air interface necessary to fully understand the interfacial chemistry.

Quantum nature of the hydrogen bond from ambient conditions down to ultra-low temperatures

Quantum simulations reveal strong temperature effects for weak hydrogen bonds and differences in quantum delocalization between various hydrogen-bonded systems.

Structure formation and surface chemistry of ionic liquids on model electrode surfaces-Model studies for the electrode | electrolyte interface in Li-ion batteries

Ionic liquids (ILs) are considered as attractive electrolyte solvents in modern battery concepts such as Li-ion batteries. Here we present a comprehensive review of the results of previous model studies on the interaction of the battery relevant IL 1-butyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)imide ([BMP]⁺[TFSI]^-) with a series of structurally and chemically well-defined model electrode surfaces, which are increasingly complex and relevant for battery applications [Ag(111), Au(111), Cu(111), pristine and lithiated highly oriented pyrolytic graphite (HOPG), and rutile TiO₂(110)]. Combining surface science techniques such as high resolution scanning tunneling microscopy and X-ray photoelectron spectroscopy for characterizing surface structure and chemical composition in deposited (sub-)monolayer adlayers with dispersion corrected density functional theory based calculations, this work aims at a molecular scale understanding of the fundamental processes at the electrode | electrolyte interface, which are crucial for the development of the so-called solid electrolyte interphase (SEI) layer in batteries. Performed under idealized conditions, in an ultrahigh vacuum environment, these model studies provide detailed insights on the structure formation in the adlayer, the substrate-adsorbate and adsorbate-adsorbate interactions responsible for this, and the tendency for chemically induced decomposition of the IL. To mimic the situation in an electrolyte, we also investigated the interaction of adsorbed IL (sub-)monolayers with coadsorbed lithium. Even at 80 K, postdeposited Li is found to react with the IL, leading to decomposition products such as LiF, Li₃N, Li₂S, Li_xSO_y, and Li₂O. In the absence of a [BMP]⁺[TFSI]^- adlayer, it tends to adsorb, dissolve, or intercalate into the substrate (metals, HOPG) or to react with the substrate (TiO₂) above a critical temperature, forming LiO_x and Ti³⁺ species in the latter case. Finally, the formation of stable decomposition products was found to sensitively change the equilibrium between surface Li and Li⁺ intercalated in the bulk, leading to a deintercalation from lithiated HOPG in the presence of an adsorbed IL adlayer at >230 K. Overall, these results provide detailed insights into the surface chemistry at the solid | electrolyte interface and the initial stages of SEI formation at electrode surfaces in the absence of an applied potential, which is essential for the further improvement of future Li-ion batteries.

OH formation and H2 adsorption at the liquid water-Pt(111) interface

The liquid water-Pt(111) interface is studied with constant temperature ab initio molecular dynamics to explore the importance of liquid water dynamics of catalytic reactions such as the oxygen reduction reaction in PEM fuel cells. The structure and energetics of hydroxyls formed at the liquid water-Pt(111) interface are found to be significantly different from those of the hydroxyl formed on a bare Pt(111) surface and the hydroxyl formed on a Pt(111) surface with a static water layer. We identify 1/12 ML *OH, 5/12 ML *OH and 2/3 ML *OH as particularly stable hydroxyl coverages in highly dynamic liquid water environments, which - contrary to static water-hydroxyl models - contain adjacent uncovered Pt sites. Atomic surface oxygen is found to be unstable in the presence of liquid water, in contrast to static atomic level simulations. These results give an improved understanding of hydroxide and surface oxide formation from Pt(111) cyclic voltammetry and allow us to draw detailed connections between the electrostatic potential and the interface structure. The study of hydrogen adsorption at the liquid water-Pt(111) interface finds competitive adsorption between the adsorbed hydrogen atoms and water molecules. This does not adhere with experimental observations, and this indicates that the Pt(111) surface has to be negatively charged for a correct description of the liquid water-Pt(111) interface at potentials where hydrogen adsorption occurs.

Ab Initio Molecular Dynamics Reveal Spectroscopic Siblings and Ion Pairing as New Challenges for Elucidating Prenucleation Aluminum Speciation

The characterization of prenucleation species is essential to understand crystallization mechanisms across many chemical systems and often involves the use of vibrational spectroscopy. Nowhere is this more evident than in the development of "green" aluminum processing technologies, where detailed understanding of the speciation of aluminum and its polynuclear analogues in highly alkaline, low water solutions is elusive. The aluminate anion Al(OH)₄^- predominates in alkaline conditions, yet equilibrium with dimeric species, either μ-oxo Al₂O(OH)₆^2- or di-μ-hydroxo Al₂(OH)₈^2-, can be assumed. Using ab initio molecular dynamics with full solvation and the presence of counterions, this work reconciles previous contradictory studies that had concluded only a single species under relevant solution conditions. We reveal that the two dimers are energetically separated by 2 kcal/mol in pure water but that the stability of each can be reversed by ion pairing expected in saturated salt solutions. Simulated Raman and IR spectra for each species (accounting for anharmonicity and the fluctuating solvating environment) provide the first proof that the considered species are "spectroscopic siblings", whose multiple overlapping bands prevent definitive assertions in terms of speciation when compared to the experimental spectra. These observations are likely to hold in higher order aluminate oligomers and as such present a massive challenge toward understanding the crystallization mechanisms relevant to aluminum processing.

Structure of aqueous NaOH solutions: insights from neural-network-based molecular dynamics simulations

Sodium hydroxide, NaOH, is one of the most widely-used chemical reagents, but the structural properties of its aqueous solutions have only sparingly been characterized. Here, we automatically classify the cation coordination polyhedra obtained from molecular dynamics simulations. We find that, for example, with increasing concentration, octahedral coordination geometries become less favored, while the opposite is true for the trigonal prism. At high concentrations, the coordination polyhedra frequently deviate considerably from "ideal" polyhedra, because of an increased extent of interligand hydrogen-bonding, in which hydrogen bonds between two ligands, either OH₂ or OH^-, around the same Na⁺ are formed. In saturated solutions, with concentrations of about 19 mol L^-1, ligands are frequently shared between multiple Na⁺ ions as a result of the deficiency of solvent molecules. This results in more complex structural patterns involving certain "characteristic" polyhedron connectivities, such as octahedra sharing ligands with capped trigonal prisms, and tetrahedra sharing ligands with trigonal bipyramids. The simulations were performed using a density-functional-theory-based reactive high-dimensional neural network potential, that was extensively validated against available neutron and X-ray diffraction data from the literature.