Distinguishing Surface and Bulk Reactivity: Concentration-Dependent Kinetics of Iodide Oxidation by Ozone in Microdroplets

Iodine oxidation reactions play an important role in environmental, biological, and industrial contexts. The multiphase reaction between aqueous iodide and ozone is of particular interest due to its prevalence in the marine atmosphere and unique reactivity at the air-water interface. Here, we explore the concentration dependence of the I- + O3 reaction in levitated microdroplets under both acidic and basic conditions. To interpret the experimental kinetics, molecular simulations are used to benchmark a kinetic model, which enables insight into the reactivity of the interface, the nanometer-scale subsurface region, and the bulk interior of the droplet. For all experiments, a kinetic description of gas- and liquid-phase diffusion is critical to interpreting the results. We find that the surface dominates the iodide oxidation kinetics under concentrated and acidic conditions, with the reactive uptake coefficient approaching an upper limit of 10-2 at pH 3. In contrast, reactions in the subsurface dominate under more dilute and alkaline conditions, with inhibition of the surface reaction at pH 12 and an uptake coefficient that is 10× smaller. The origin of a changing surface mechanism with pH is explored and compared to previous ozone-dependent measurements.

Entropy in catalyst dynamics under confinement

Entropy during the dynamic structural evolution of catalysts has a non-trivial influence on chemical reactions. Confinement significantly affects the catalyst dynamics and thus impacts the reactivity. However, a full understanding has not been clearly established. To investigate catalyst dynamics under confinement, we utilize the active learning scheme to effectively train machine learning potentials for computing free energies of catalytic reactions. The scheme enables us to compute the reaction free energies and entropies of O2 dissociation on Pt clusters with different sizes confined inside a carbon nanotube (CNT) at the timescale of tens of nanoseconds while keeping ab initio accuracy. We observe an entropic effect owing to liquid-to-solid phase transitions of clusters at finite temperatures. More importantly, the confinement effect enhances the structural dynamics of the cluster and leads to a lower melting temperature than those of the bare cluster and cluster outside the CNT, consequently facilitating the reaction to occur at lower temperatures and preventing the catalyst from forming unfavorable oxides. Our work reveals the important influence of confinement on structural dynamics, providing useful insight into entropy in dynamic catalysis.

Enhanced ClNO2 Formation at the Interface of Sea-Salt Aerosol

The reactive uptake of N2O5 on sea-spray aerosol plays a key role in regulating the NOx concentration in the troposphere. Despite numerous field and laboratory studies, a microscopic understanding of its heterogeneous reactivity remains unclear. Here, we use molecular simulation and theory to elucidate the chlorination of N2O5 to form ClNO2, the primary reactive channel within sea-spray aerosol. We find that the formation of ClNO2 is markedly enhanced at the air-water interface due to the stabilization of the charge-delocalized transition state, as evident from the formulation of bimolecular rate theory in heterogeneous environments. We explore the consequences of the enhanced interfacial reactivity in the uptake of N2O5 using numerical solutions of molecular reaction-diffusion equations as well as their analytical approximations. Our results suggest that the current interpretation of aerosol branching ratios needs to be revisited.

Double-Layer Distribution of Hydronium and Hydroxide Ions in the Air-Water Interface

The acid-base nature of the aqueous interface has long been controversial. Most macroscopic experiments suggest that the air-water interface is basic based on the detection of negative charges at the interface that indicates the enrichment of hydroxides (OH-), whereas microscopic studies mostly support the acidic air-water interface with the observation of hydronium (H3O+) accumulation in the top layer of the interface. It is crucial to clarify the interfacial preference of OH- and H3O+ ions for rationalizing the debate. In this work, we perform deep potential molecular dynamics simulations to investigate the preferential distribution of OH- and H3O+ ions at the aqueous interfaces. The neural network potential energy surface is trained based on density functional theory calculations with the SCAN functional, which can accurately describe the diffusion of these two ions both in the interface and in the bulk water. In contrast to the previously reported single ion enrichment, we show that both OH- and H3O+ surprisingly prefer to accumulate in interfaces but at different interfacial depths, rendering a double-layer ionic distribution within ∼1 nm near the Gibbs dividing surface. The H3O+ preferentially resides in the topmost layer of the interface, but the OH-, which is enriched in the deeper interfacial layer, has a higher equilibrium concentration due to the more negative free energy of interfacial stabilization [-0.90 (OH-) vs -0.56 (H3O+) kcal/mol]. The present finding of the ionic double-layer distribution may qualitatively offer a self-consistent explanation for the long-term controversy about the acid-base nature of the air-water interface.

Exploring fracture of H-BN and graphene by neural network force fields

Extreme mechanical processes such as strong lattice distortion and bond breakage during fracture often lead to catastrophic failure of materials and structures. Understanding the nucleation and growth of cracks is challenged by their multiscale characteristics spanning from atomic-level structures at the crack tip to the structural features where the load is applied. Atomistic simulations offer 'first-principles' tools to resolve the progressive microstructural changes at crack fronts and are widely used to explore the underlying processes of mechanical energy dissipation, crack path selection, and dynamic instabilities (e.g. kinking, branching). Empirical force fields developed based on atomic-level structural descriptors based on atomic positions and the bond orders do not yield satisfying predictions of fracture, especially for the nonlinear, anisotropic stress-strain relations and the energy densities of edges. High-fidelity force fields thus should include the tensorial nature of strain and the energetics of bond-breaking and (re)formation events during fracture, which, unfortunately, have not been taken into account in either the state-of-the-art empirical or machine-learning force fields. Based on data generated by density functional theory calculations, we report a neural network-based force field for fracture (NN-F3) constructed by using the end-to-end symmetry preserving framework of deep potential-smooth edition (DeepPot-SE). The workflow combines pre-sampling of the space of strain states and active-learning techniques to explore the transition states at critical bonding distances. The capability of NN-F3is demonstrated by studying the rupture of hexagonal boron nitride (h-BN) and twisted bilayer graphene as model problems. The simulation results elucidate the roughening physics of fracture defined by the lattice asymmetry in h-BN, explaining recent experimental findings, and predict the interaction between cross-layer cracks in twisted graphene bilayers, which leads to a toughening effect.

Prebiotic chemical reactivity in solution with quantum accuracy and microsecond sampling using neural network potentials

While RNA appears as a good candidate for the first autocatalytic systems preceding the emergence of modern life, the synthesis of RNA oligonucleotides without enzymes remains challenging. Because the uncatalyzed reaction is extremely slow, experimental studies bring limited and indirect information on the reaction mechanism, the nature of which remains debated. Here, we develop neural network potentials (NNPs) to study the phosphoester bond formation in water. While NNPs are becoming routinely applied to nonreactive systems or simple reactions, we demonstrate how they can systematically be trained to explore the reaction phase space for complex reactions involving several proton transfers and exchanges of heavy atoms. We then propagate at moderate computational cost hundreds of nanoseconds of a variety of enhanced sampling simulations with quantum accuracy in explicit solvent conditions. The thermodynamically preferred reaction pathway is a concerted, dissociative mechanism, with the transient formation of a metaphosphate transition state and direct participation of water solvent molecules that facilitate the exchange of protons through the nonbridging phosphate oxygens. Associative-dissociative pathways, characterized by a much tighter pentacoordinated phosphate, are higher in free energy. Our simulations also suggest that diprotonated phosphate, whose reactivity is never directly assessed in the experiments, is significantly less reactive than the monoprotonated species, suggesting that it is probably never the reactive species in normal pH conditions. These observations rationalize unexplained experimental results and the temperature dependence of the reaction rate, and they pave the way for the design of more efficient abiotic catalysts and activating groups.

Increase of secondary metabolites in sweet basil (Ocimum basilicum L.) leaves by exposure to N2O5 with plasma technology

Exposure to N2O5 generated by plasma technology activates immunity in Arabidopsis through tryptophan metabolites. However, little is known about the effects of N2O5 exposure on other plant species. Sweet basil synthesizes many valuable secondary metabolites in its leaves. Therefore, metabolomic analyses were performed at three different exposure levels [9.7 (Ex1), 19.4 (Ex2) and 29.1 (Ex3) μmol] to assess the effects of N2O5 on basil leaves. As a result, cinnamaldehyde and phenolic acids increased with increasing doses. Certain flavonoids, columbianetin, and caryophyllene oxide increased with lower Ex1 exposure, cineole and methyl eugenol increased with moderate Ex2 exposure and L-glutathione GSH also increased with higher Ex3 exposure. Furthermore, gene expression analysis by quantitative RT-PCR showed that certain genes involved in the syntheses of secondary metabolites and jasmonic acid were significantly up-regulated early after N2O5 exposure. These results suggest that N2O5 exposure increases several valuable secondary metabolites in sweet basil leaves via plant defense responses in a controllable system.

Molecular Insights into Chemical Reactions at Aqueous Aerosol Interfaces

Atmospheric aerosols facilitate reactions between ambient gases and dissolved species. Here, we review our efforts to interrogate the uptake of these gases and the mechanisms of their reactions both theoretically and experimentally. We highlight the fascinating behavior of N2O5 in solutions ranging from pure water to complex mixtures, chosen because its aerosol-mediated reactions significantly impact global ozone, hydroxyl, and methane concentrations. As a hydrophobic, weakly soluble, and highly reactive species, N2O5 is a sensitive probe of the chemical and physical properties of aerosol interfaces. We employ contemporary theory to disentangle the fate of N2O5 as it approaches pure and salty water, starting with adsorption and ending with hydrolysis to HNO3, chlorination to ClNO2, or evaporation. Flow reactor and gas-liquid scattering experiments probe even greater complexity as added ions, organic molecules, and surfactants alter the interfacial composition and reaction rates. Together, we reveal a new perspective on multiphase chemistry in the atmosphere.

Chemical Kinetics in Microdroplets

Micrometer-sized compartments play significant roles in driving heterogeneous transformations within atmospheric and biochemical systems as well as providing vehicles for drug delivery and novel reaction environments for the synthesis of industrial chemicals. Many reports now indicate that reaction kinetics are accelerated under microconfinement, for example, in sprays, thin films, droplets, aerosols, and emulsions. These observations are dramatic, posing a challenge to our understanding of chemical reaction mechanisms with potentially significant practical consequences for predicting the complex chemistry in natural systems. Here we introduce the idea of kinetic confinement, which is intended to provide a conceptual backdrop for understanding when and why microdroplet reaction kinetics differ from their macroscale analogs.

On the Statistical Mechanics of Mass Accommodation at Liquid-Vapor Interfaces

We propose a framework for describing the dynamics associated with the adsorption of small molecules to liquid-vapor interfaces using an intermediate resolution between traditional continuum theories that are bereft of molecular detail and molecular dynamics simulations that are replete with them. In particular, we develop an effective single particle equation of motion capable of describing the physical processes that determine thermal and mass accommodation probabilities. The effective equation is parametrized with quantities that vary through space away from the liquid-vapor interface. Of particular importance in describing the early time dynamics is the spatially dependent friction, for which we propose a numerical scheme to evaluate from molecular simulation. Taken together with potentials of mean force computable with importance sampling methods, we illustrate how to compute the mass accommodation coefficient and residence time distribution. Throughout, we highlight the case of ozone adsorption in aqueous solutions and its dependence on electrolyte composition.

Determining the mechanical and decomposition properties of high energetic materials (α-RDX, β-HMX, and ε-CL-20) using a neural network potential

Molecular simulations of high energetic materials (HEMs) are limited by efficiency and accuracy. Recently, neural network potential (NNP) models have achieved molecular simulations of millions of atoms while maintaining the accuracy of density functional theory (DFT) levels. Herein, an NNP model covering typical HEMs containing C, H, N, and O elements is developed. The mechanical and decomposition properties of 1,3,5-trinitroperhydro-1,3,5-triazine (RDX), hexahydro-1,3,5-trinitro-1,3,5-triazine (HMX), and 2,4,6,8,10,12-hexanitrohexaazaisowurtzitane (CL-20) are determined by employing the molecular dynamics (MD) simulations based on the NNP model. The calculated results show that the mechanical properties of α-RDX, β-HMX, and ε-CL-20 agree with previous experiments and theoretical results, including cell parameters, equations of state, and elastic constants. In the thermal decomposition simulations, it is also found that the initial decomposition reactions of the three crystals are N-NO2 homolysis, corresponding radical intermediates formation, and NO2-induced reactions. This decomposition trajectory is mainly divided into two stages separating from the peak of NO2: pyrolysis and oxidation. Overall, the NNP model for C/H/N/O elements in this work is an alternative reactive force field for RDX, HMX, and CL-20 HEMs, and it opens up new potential for future kinetic study of nitramine explosives.

Mechanistic insight into the competition between interfacial and bulk reactions in microdroplets through N2O5 ammonolysis and hydrolysis

Reactive uptake of dinitrogen pentaoxide (N2O5) into aqueous aerosols is a major loss channel for NOx in the troposphere; however, a quantitative understanding of the uptake mechanism is lacking. Herein, a computational chemistry strategy is developed employing high-level quantum chemical methods; the method offers detailed molecular insight into the hydrolysis and ammonolysis mechanisms of N2O5 in microdroplets. Specifically, our calculations estimate the bulk and interfacial hydrolysis rates to be (2.3 ± 1.6) × 10-3 and (6.3 ± 4.2) × 10-7 ns-1, respectively, and ammonolysis competes with hydrolysis at NH3 concentrations above 1.9 × 10-4 mol L-1. The slow interfacial hydrolysis rate suggests that interfacial processes have negligible effect on the hydrolysis of N2O5 in liquid water. In contrast, N2O5 ammonolysis in liquid water is dominated by interfacial processes due to the high interfacial ammonolysis rate. Our findings and strategy are applicable to high-chemical complexity microdroplets.

Balance between Physical Interpretability and Energetic Predictability in Widely Used Dispersion-Corrected Density Functionals

We assess the performance of different dispersion models for several popular density functionals across a diverse set of noncovalent systems, ranging from the benzene dimer to molecular crystals. By analyzing the interaction energies and their individual components, we demonstrate that there exists variability across different systems for empirical dispersion models, which are calibrated for reproducing the interaction energies of specific systems. Thus, parameter fitting may undermine the underlying physics, as dispersion models rely on error compensation among the different components of the interaction energy. Energy decomposition analyses reveal that, the accuracy of revPBE-D3 for some aqueous systems originates from significant compensation between dispersion and charge transfer energies. However, revPBE-D3 is less accurate in describing systems where error compensation is incomplete, such as the benzene dimer. Such cases highlight the propensity for unpredictable behavior in various dispersion-corrected density functionals across a wide range of molecular systems, akin to the behavior of force fields. On the other hand, we find that SCAN-rVV10, a targeted-dispersion approach, affords significant reductions in errors associated with the lattice energies of molecular crystals, while it has limited accuracy in reproducing structural properties. Given the ubiquitous nature of noncovalent interactions and the key role of density functional theory in computational sciences, the future development of dispersion models should prioritize the faithful description of the dispersion energy, a shift that promises greater accuracy in capturing the underlying physics across diverse molecular and extended systems.

The role of dynamics in heterogeneous catalysis: Surface diffusivity and N2 decomposition on Fe(111)

Dynamics has long been recognized to play an important role in heterogeneous catalytic processes. However, until recently, it has been impossible to study their dynamical behavior at industry-relevant temperatures. Using a combination of machine learning potentials and advanced simulation techniques, we investigate the cleavage of the N[Formula: see text] triple bond on the Fe(111) surface. We find that at low temperatures our results agree with the well-established picture. However, if we increase the temperature to reach operando conditions, the surface undergoes a global dynamical change and the step structure of the Fe(111) surface is destabilized. The catalytic sites, traditionally associated with this surface, appear and disappear continuously. Our simulations illuminate the danger of extrapolating low-temperature results to operando conditions and indicate that the catalytic activity can only be inferred from calculations that take dynamics fully into account. More than that, they show that it is the transition to this highly fluctuating interfacial environment that drives the catalytic process.

Probing the self-ionization of liquid water with ab initio deep potential molecular dynamics

The chemical equilibrium between self-ionized and molecular water dictates the acid-base chemistry in aqueous solutions, yet understanding the microscopic mechanisms of water self-ionization remains experimentally and computationally challenging. Herein, Density Functional Theory (DFT)-based deep neural network (DNN) potentials are combined with enhanced sampling techniques and a global acid-base collective variable to perform extensive atomistic simulations of water self-ionization for model systems of increasing size. The explicit inclusion of long-range electrostatic interactions in the DNN potential is found to be crucial to accurately reproduce the DFT free energy profile of solvated water ion pairs in small (64 and 128 H2O) cells. The reversible work to separate the hydroxide and hydronium to a distance [Formula: see text] is found to converge for simulation cells containing more than 500 H2O, and a distance of [Formula: see text] 8 Å is the threshold beyond which the work to further separate the two ions becomes approximately zero. The slow convergence of the potential of mean force with system size is related to a restructuring of water and an increase of the local order around the water ions. Calculation of the dissociation equilibrium constant illustrates the key role of long-range electrostatics and entropic effects in the water autoionization process.

Ventilation in a Residential Building Brings Outdoor NOx Indoors with Limited Implications for VOC Oxidation from NO3 Radicals

Energy-efficient residential building standards require the use of mechanical ventilation systems that replace indoor air with outdoor air. Transient outdoor pollution events can be transported indoors via the mechanical ventilation system and other outdoor air entry pathways and impact indoor air chemistry. In the spring of 2022, we observed elevated levels of NOx (NO + NO2) that originated outdoors, entering the National Institute of Standards and Technology (NIST) Net-Zero Energy Residential Test Facility through the mechanical ventilation system. Using measurements of NOx, ozone (O3), and volatile organic compounds (VOCs), we modeled the effect of the outdoor-to-indoor ventilation of NOx pollution on the production of nitrate radical (NO3), a potentially important indoor oxidant. We evaluated how VOC oxidation chemistry was affected by NO3 during NOx pollution events compared to background conditions. We found that nitric oxide (NO) pollution introduced indoors titrated O3 and inhibited the modeled production of NO3. NO ventilated indoors also likely ceased most gas-phase VOC oxidation chemistry during plume events. Only through the artificial introduction of O3 to the ventilation duct during a NOx pollution event (i.e., when O3 and NO2 concentrations were high relative to typical conditions) were we able to measure NO3-initiated VOC oxidation products, indicating that NO3 was impacting VOC oxidation chemistry.

Speciation of La3+-Cl- Complexes in Hydrothermal Fluids from Deep Potential Molecular Dynamics

The stability of rare earth element (REE) complexes plays a crucial role in quantitatively assessing their hydrothermal migration and transformation. However, reliable data are lacking under high-temperature hydrothermal conditions, which hampers our understanding of the association behavior of REE. Here a deep learning potential model for the LaCl3-H2O system in hydrothermal fluids is developed based on the first-principles density functional theory calculations. The model accurately predicts the radial distribution functions compared to ab initio molecular dynamics (AIMD) simulations. Furthermore, species of La-Cl complexes, the dissociation pathway of the La-Cl complexes dissociation process, and the potential of mean forces and corresponding association constants (logK) for LaCln3-n (n = 1-4) are extensively investigated under a wide range of temperatures and pressures. Empirical density models for logK calculation are fitted with these data and can accurately predict logK data from both experimental results and AIMD simulations. The distribution of La-Cl species is also evaluated across a wide range of temperatures, pressures, and initial chloride concentration conditions. The results show that La-Cl complexes are prone to forming in a low-density solution, and the number of bonded Cl- ions increases with rising temperature. In contrast, in a high-density solution, La3+ dominates and becomes the more prevalent species.

Agglomeration Drives the Reversed Fractionation of Aqueous Carbonate and Bicarbonate at the Air-Water Interface

In the course of our investigations of the adsorption of ions to the air-water interface, we previously reported the surprising result that doubly charged carbonate anions exhibit a stronger surface affinity than singly charged bicarbonate anions. In contrast to monovalent, weakly hydrated anions, which generally show enhanced concentrations in the interfacial region, multivalent (and strongly hydrated) anions are expected to show a much weaker surface propensity. In the present work, we use resonantly enhanced deep-UV second-harmonic generation spectroscopy to measure the Gibbs free energy of adsorption of both carbonate (CO32-) and bicarbonate (HCO3-) anions to the air-water interface. Contrasting the predictions of classical electrostatic theory and in support of our previous findings from X-ray photoelectron spectroscopy, we find that carbonate anions do indeed exhibit much stronger surface affinity than do the bicarbonate anions. Extensive computer simulations reveal that strong ion pairing of CO32- with the Na+ countercation in the interfacial region results in the formation of near-neutral agglomerate clusters, consistent with a theory of interfacial ion adsorption based on hydration free energy and capillary waves. Simulated X-ray photoelectron spectra predict a 1 eV shift in the carbonate spectra compared to that of bicarbonate, further confirming our experiments. These findings not only advance our fundamental understanding of ion adsorption chemistry but also impact important practical processes such as ocean acidification, sea-spray aerosol chemistry, and mammalian respiration physiology.

Distinctive Heterogeneous Reaction Mechanism of ClNO2 on the Air-Water Surface Containing Cl

The heterogeneous reaction of nitryl chloride (ClNO2) on the air-water surface plays a significant role in the chloride lifecycle. The air-water surface is ubiquitous on ice surfaces under supercooled conditions, affecting the uptake and heterogeneous reaction processes of trace gases. Previous studies suggest that ClNO2 is formed on Cl-doped ice surfaces following the N2O5 uptake. Herein, a distinctive heterogeneous reaction mechanism of ClNO2 is suggested on an air-water surface containing Cl under supercooled conditions using combined classic molecular dynamics (MD) and Born-Oppenheimer MD simulations. It is found that N2O5 dissociates into a NO2+ and NO3- ionic pair on the top air-water surface. In the top layer of the surface containing barely any Cl-, NO2+ proceeds through hydrolysis and produces H3O+ and HNO3. Thus, surface acidification appears because of H3O+ yields. With NO2+ diffusion to the deep layer of the surface, NO2+ reacts with Cl- and forms ClNO2. Note that ClNO2 formation competes with NO2+ hydrolysis, and the rate of ClNO2 formation is 27.7[Cl-] larger than that of NO2+ hydrolysis. Afterward, the reaction of ClNO2 with Cl- becomes barrierless with the catalysis by H3O+, which is not feasible on a neutral surface. Cl2 is thus generated and escapes into the atmosphere (low solubility of Cl2), contributing to the Cl radical. The proposed mechanism bolsters the current understanding of ClNO2's fate and its role in Cl chemistry in extremely cold environments like the Arctic and other high-latitude regions in wintertime.

DeePMD-kit v2: A software package for deep potential models

DeePMD-kit is a powerful open-source software package that facilitates molecular dynamics simulations using machine learning potentials known as Deep Potential (DP) models. This package, which was released in 2017, has been widely used in the fields of physics, chemistry, biology, and material science for studying atomistic systems. The current version of DeePMD-kit offers numerous advanced features, such as DeepPot-SE, attention-based and hybrid descriptors, the ability to fit tensile properties, type embedding, model deviation, DP-range correction, DP long range, graphics processing unit support for customized operators, model compression, non-von Neumann molecular dynamics, and improved usability, including documentation, compiled binary packages, graphical user interfaces, and application programming interfaces. This article presents an overview of the current major version of the DeePMD-kit package, highlighting its features and technical details. Additionally, this article presents a comprehensive procedure for conducting molecular dynamics as a representative application, benchmarks the accuracy and efficiency of different models, and discusses ongoing developments.

Counterintuitive Oxidation of Alcohols at Air-Water Interfaces

This study shows that the oxidation of alcohols can rapidly occur at air-water interfaces. It was found that methanediols (HOCH₂OH) orient at air-water interfaces with a H atom of the -CH₂- group pointing toward the gaseous phase. Counterintuitively, gaseous hydroxyl radicals do not prefer to attack the exposed -CH₂- group but the -OH group that forms hydrogen bonds with water molecules at the surface via a water-promoted mechanism, leading to the formation of formic acids. Compared with gaseous oxidation, the water-promoted mechanism at the air-water interface significantly lowers free-energy barriers from ∼10.7 to ∼4.3 kcal·mol^-1 and therefore accelerates the formation of formic acids. The study unveils a previously overlooked source of environmental organic acids that are bound up with aerosol formation and water acidity.

Weak Temperature Dependence of the Relative Rates of Chlorination and Hydrolysis of N2O5 in NaCl-Water Solutions

We have measured the temperature dependence of the ClNO₂ product yield in competition with hydrolysis following N₂O₅ uptake to aqueous NaCl solutions. For NaCl-D₂O solutions spanning 0.0054-0.21 M, the ClNO₂ product yield decreases on average by only 4 ± 3% from 5 to 25 °C. Less reproducible measurements at 0.54-2.4 M NaCl also fall within this range. The ratio of the rate constants for chlorination and hydrolysis of N₂O₅ in D₂O is determined on average to be 1150 ± 90 at 25 °C up to 0.21 M NaCl, favoring chlorination. This ratio is observed to decrease significantly at the two highest concentrations. An Arrhenius analysis reveals that the activation energy for hydrolysis is just 3.0 ± 1.5 kJ/mol larger than for chlorination up to 0.21 M, indicating that Cl^- and D₂O attack on N₂O₅ has similar energetic barriers despite the differences in charge and complexity of these reactants. In combination with the measured preexponential ratio favoring chlorination of 300_-200⁺⁴⁰⁰, we conclude that the strong preference of N₂O₅ to undergo chlorination over hydrolysis is driven by dynamic and entropic, rather than enthalpic, factors. Molecular dynamics simulations elucidate the distinct solvation between strongly hydrated Cl^- and the hydrophobically solvated N₂O₅. Combining this molecular picture with the Arrhenius analysis implicates the role of water in mediating interactions between such distinctly solvated species and suggests a role for diffusion limitations on the chlorination reaction.

Oxidation of Catechols at the Air-Water Interface by Nitrate Radicals

Abundant substituted catechols are emitted to, and created in, the atmosphere during wildfires and anthropogenic combustion and agro-industrial processes. While ozone (O3) and hydroxyl radicals (HO•) efficiently react in a 1 μs contact time with catechols at the air-water interface, the nighttime reactivity dominated by nitrate radicals (NO3) remains unexplored. Herein, online electrospray ionization mass spectrometry (OESI-MS) is used to explore the reaction of NO3(g) with a series of representative catechols (catechol, pyrogallol, 3-methylcatechol, 4-methylcatechol, and 3-methoxycatechol) on the surface of aqueous microdroplets. The work detects the ultrafast generation of nitrocatechol (aromatic) compounds, which are major constituents of atmospheric brown carbon. Two mechanisms are proposed to produce nitrocatechols, one (equivalent to H atom abstraction) following fast electron transfer from the catechols (QH2) to NO3, forming NO3- and QH2•+ that quickly deprotonates into a semiquinone radical (QH•). The second mechanism proceeds via cyclohexadienyl radical intermediates from NO3 attack to the ring. Experiments in the pH range from 4 to 8 showed that the production of nitrocatechols was favored under the most acidic conditions. Mechanistically, the results explain the interfacial production of chromophoric nitrocatechols that modify the absorption properties of tropospheric particles, making them more susceptible to photooxidation, and alter the Earth's radiative forcing.

A Kinetic Model for Predicting Trace Gas Uptake and Reaction

A model is developed to describe trace gas uptake and reaction with applications to aerosols and microdroplets. Gas uptake by the liquid is formulated as a coupled equilibria that links gas, surface, and bulk regions of the droplet or solution. Previously, this framework was used in explicit stochastic reaction-diffusion simulations to predict the reactive uptake kinetics of ozone with droplets containing aqueous aconitic acid, maleic acid, and sodium nitrite. With the use of prior data and simulation results, a new equation for the uptake coefficient is derived, which accounts for both surface and bulk reactions. Lambert W functions are used to obtain closed form solutions to the integrated rate laws for the multiphase kinetics; similar to previous expressions that describe Michaelis-Menten enzyme kinetics. Together these equations couple interface and bulk processes over a wide range of conditions and do not require many of the limiting assumptions needed to apply resistor model formulations to explain trace gas uptake and reaction.

Resolving the amine-promoted hydrolysis mechanism of N2O5 under tropospheric conditions

Hydrolysis of N₂O₅ under tropospheric conditions plays a critical role in assessing the fate of O₃, OH, and NO_x in the atmosphere. However, its removal mechanism has not been fully understood, and little is known about the role of entropy. Herein, we propose a removal path of N₂O₅ on the water clusters/droplet with the existence of amine, which entails a low free-energy barrier of 4.46 and 3.76 kcal/mol on a water trimer and droplet, respectively, at room temperature. The free-energy barrier exhibits strong temperature dependence; a barrierless hydrolysis process of N₂O₅ at low temperature (≤150 K) is observed. By coupling constrained ab initio molecular dynamics (constrained AIMD) simulations with thermodynamic integration methods, we quantitively evaluated the entropic contributions to the free energy and compared NH₃-, methylamine (MA)-, and dimethylamine (DMA)-promoted hydrolysis of N₂O₅ on water clusters and droplet. Our results demonstrate that methylation of NH₃ stabilizes the product state and promotes hydrolysis of N₂O₅ by reducing the free-energy barriers. Furthermore, a quantitative analysis of the internal coordinate distribution of the reaction center and the relative position of surrounding species reveals that the significant entropic contribution primarily results from the ensemble effect of configurations observed in the AIMD simulations. Such an ensemble effect becomes more significant with more water molecules included. Lowering the temperature effectively minimizes the entropic contribution, making the hydrolysis more exothermic and barrierless. This study sheds light on the importance of the promoting effect of amines and the entropic effect on gas-phase hydrolysis reactions, which may have far-reaching implications in atmospheric chemistry.

DP Compress: A Model Compression Scheme for Generating Efficient Deep Potential Models

Machine-learning-based interatomic potential energy surface (PES) models are revolutionizing the field of molecular modeling. However, although much faster than electronic structure schemes, these models suffer from costly computations via deep neural networks to predict the energy and atomic forces, resulting in lower running efficiency as compared to the typical empirical force fields. Herein, we report a model compression scheme for boosting the performance of the Deep Potential (DP) model, a deep learning-based PES model. This scheme, we call DP Compress, is an efficient postprocessing step after the training of DP models (DP Train). DP Compress combines several DP-specific compression techniques, which typically speed up DP-based molecular dynamics simulations by an order of magnitude faster and consume an order of magnitude less memory. We demonstrate that DP Compress is sufficiently accurate by testing a variety of physical properties of Cu, H₂O, and Al-Cu-Mg systems. DP Compress applies to both CPU and GPU machines and is publicly available online.

UV-Visible Absorption Spectra of Solvated Molecules by Quantum Chemical Machine Learning

Predicting UV-visible absorption spectra is essential to understand photochemical processes and design energy materials. Quantum chemical methods can deliver accurate calculations of UV-visible absorption spectra, but they are computationally expensive, especially for large systems or when one computes line shapes from thermal averages. Here, we present an approach to predict UV-visible absorption spectra of solvated aromatic molecules by quantum chemistry (QC) and machine learning (ML). We show that a ML model, trained on the high-level QC calculation of the excitation energy of a set of aromatic molecules, can accurately predict the line shape of the lowest-energy UV-visible absorption band of several related molecules with less than 0.1 eV deviation with respect to reference experimental spectra. Applying linear decomposition analysis on the excitation energies, we unveil that our ML models probe vertical excitations of these aromatic molecules primarily by learning the atomic environment of their phenyl rings, which align with the physical origin of the π →π* electronic transition. Our study provides an effective workflow that combines ML with quantum chemical methods to accelerate the calculations of UV-visible absorption spectra for various molecular systems.

Combined QM/MM, Machine Learning Path Integral Approach to Compute Free Energy Profiles and Kinetic Isotope Effects in RNA Cleavage Reactions

We present a fast, accurate, and robust approach for determination of free energy profiles and kinetic isotope effects for RNA 2'-O-transphosphorylation reactions with inclusion of nuclear quantum effects. We apply a deep potential range correction (DPRc) for combined quantum mechanical/molecular mechanical (QM/MM) simulations of reactions in the condensed phase. The method uses the second-order density-functional tight-binding method (DFTB2) as a fast, approximate base QM model. The DPRc model modifies the DFTB2 QM interactions and applies short-range corrections to the QM/MM interactions to reproduce ab initio DFT (PBE0/6-31G*) QM/MM energies and forces. The DPRc thus enables both QM and QM/MM interactions to be tuned to high accuracy, and the QM/MM corrections are designed to smoothly vanish at a specified cutoff boundary (6 Å in the present work). The computational speed-up afforded by the QM/MM+DPRc model enables free energy profiles to be calculated that include rigorous long-range QM/MM interactions under periodic boundary conditions and nuclear quantum effects through a path integral approach using a new interface between the AMBER and i-PI software. The approach is demonstrated through the calculation of free energy profiles of a native RNA cleavage model reaction and reactions involving thio-substitutions, which are important experimental probes of the mechanism. The DFTB2+DPRc QM/MM free energy surfaces agree very closely with the PBE0/6-31G* QM/MM results, and it is vastly superior to the DFTB2 QM/MM surfaces with and without weighted thermodynamic perturbation corrections. 18O and 34S primary kinetic isotope effects are compared, and the influence of nuclear quantum effects on the free energy profiles is examined.

Activation of plant immunity by exposure to dinitrogen pentoxide gas generated from air using plasma technology

Reactive nitrogen species (RNS) play an important role in plant immunity as signaling factors. We previously developed a plasma technology to partially convert air molecules into dinitrogen pentoxide (N₂O₅), an RNS whose physiological action is poorly understood. To reveal the function of N₂O₅ gas in plant immunity, Arabidopsis thaliana was exposed to plasma-generated N₂O₅ gas once (20 s) per day for 3 days, and inoculated with Botrytis cinerea, Pseudomonas syringae pv. tomato DC3000 (Pst), or cucumber mosaic virus strain yellow (CMV(Y)) at 24 h after the final N₂O₅ gas exposure. Lesion size with B. cinerea infection was significantly (P < 0.05) reduced by exposure to N₂O₅ gas. Propagation of CMV(Y) was suppressed in plants exposed to N₂O₅ gas compared with plants exposed to the air control. However, proliferation of Pst in the N₂O₅-gas-exposed plants was almost the same as in the air control plants. These results suggested that N₂O₅ gas exposure could control plant disease depending on the type of pathogen. Furthermore, changes in gene expression at 24 h after the final N₂O₅ gas exposure were analyzed by RNA-Seq. Based on the gene ontology analysis, jasmonic acid and ethylene signaling pathways were activated by exposure of Arabidopsis plants to N₂O₅ gas. A time course experiment with qRT-PCR revealed that the mRNA expression of the transcription factor genes, WRKY25, WRKY26, WRKY33, and genes for tryptophan metabolic enzymes, CYP71A12, CYP71A13, PEN2, and PAD3, was transiently induced by exposure to N₂O₅ gas once for 20 s peaking at 1–3 h post-exposure. However, the expression of PDF1.2 was enhanced beginning from 6 h after exposure and its high expression was maintained until 24–48 h later. Thus, enhanced tryptophan metabolism leading to the synthesis of antimicrobial substances such as camalexin and antimicrobial peptides might have contributed to the N₂O₅-gas-induced disease resistance.

Acids at the Edge: Why Nitric and Formic Acid Dissociations at Air-Water Interfaces Depend on Depth and on Interface Specific Area

Whether the air-water interface decreases or increases the acidity of simple organic and inorganic acids compared to the bulk is critically important in a broad range of environmental and biochemical processes. However, a consensus has not yet been achieved on this key question. Here we use machine learning-based reactive molecular dynamics simulations to study the dissociation of paradigmatic nitric and formic acids at the air-water interface. We show that the local acidity profile across the interface is determined by changes in acid and conjugate base solvation and that the acidity decreases abruptly over a transition region of a few molecular layers. At the interface, both acids are weaker than in the bulk due to desolvation. In contrast, acidities below the interface reach a plateau and are all the stronger compared to those in the bulk as the surface to volume ratio of the aqueous phase is large, due to the growing impact of the stabilization of the released proton at the surface of the water. These results imply that the measured degree of dissociation sensitively depends on the experimental probing length and system size and suggest a molecular explanation for the contrasting experimental results. The aerosol size dependence of acidity has important consequences for atmospheric chemistry.

Smog chamber simulation on heterogeneous reaction of O3 and NO2 on black carbon under various relative humidity conditions

In this study, heterogeneous formation of nitrate from O₃ reaction with NO₂ on black carbon (BC) and KCl-treated BC surface in the presence of NH₃ was simulated under 30-90% RH conditions by using a laboratory smog chamber. We found that O₃ and NO₂ in the chamber quickly reacted into N₂O₅ in the gas phase_, which subsequently hydrolyzed into HNO₃ and further neutralized with NH₃ into NH₄NO₃ on the BC surface, along with a small amount of N₂O₅ decomposed into NO and NO₂ through a reaction with the BC surface active site. Meanwhile, the fractal BC aggregates restructured and condensed to spherical particles during the NH₄NO₃ coating process. Compared to that during the exposure to NO₂ or O₃ alone, the presence of strong signals of CH₂O⁺, CH₂O₂⁺ and CH₄NO⁺ during the simultaneous exposure to both NO₂ and O₃ suggested a synergetic oxidizing effect of NO₂ and O₃, which significantly activated the BC surface by forming carbonyl, carboxylic and nitro groups, promoted the adsorption of water vapor onto the BC surface and enhanced the NH₄NO₃ formation. Under <75 ± 2% RH conditions the coating process of NH₄NO₃ on the BC surface consisted of a diffusion of N₂O₅ onto the surface and a subsequent hydrolysis, due to the limited number of water molecules adsorbed. However, under 90 ± 2% RH conditions N₂O₅ directly hydrolyzed on the aqueous phase of the BC surface due to the multilayer water molecules adsorbed, which caused an instant NH₄NO₃ formation on the surface without any delay. The coating rate of NH₄NO₃ on KCl-treated BC particles was 3-4 times faster than that on the pure BC particles at the initial stage, indicating an increasing formation of NH₄NO₃, mainly due to an enhanced hygroscopicity of BC by KCl salts.

Ab initio neural network MD simulation of thermal decomposition of a high energy material CL-20/TNT

CL-20 (2,4,6,8,10,12-hexanitro-2,4,6,8,10,12-hexaazaisowurtzitane, also known as HNIW) is one of the most powerful energetic materials. However, its high sensitivity to environmental stimuli greatly reduces its safety and severely limits its application. In this work, ab initio based neural network potential (NNP) energy surfaces for both β-CL-20 and CL-20/TNT co-crystals were constructed. To accurately simulate the thermal decomposition processes of these two crystal systems, reactive molecular dynamics simulations based on the NNPs were performed. Many important intermediate species and their associated reaction paths during the decomposition had been identified in the simulations and the direct results on detonation temperatures of both systems were provided. The simulations also showed clearly that 2,4,6-trinitrotoluene (TNT) molecules in the co-crystal act as a buffer to slow down the chain reactions triggered by nitrogen dioxide and this effect is more significant at lower temperatures. Specifically, the addition of TNT molecules in the CL-20/TNT co-crystal introduces intermolecular hydrogen bonds between CL-20 and TNT molecules in the system, thereby increasing the thermal stability of the co-crystal. The current reactive molecular dynamics simulation is performed based on the NNP which helps in accelerating the speed of ab initio molecular dynamics (AIMD) simulation by more than 3 orders of magnitude while preserving the accuracy of density functional theory (DFT) calculations. This enabled us to perform longer-time simulations at more realistic temperatures that traditional AIMD methods cannot achieve. With the advantage of the NNP in its powerful fitting ability and transferability, the NNP-based MD simulation can be widely applied to energetic material systems.

Accessing negative Poisson's ratio of graphene by machine learning interatomic potentials

The negative Poisson's ratio (NPR) is a novel property of materials, which enhances the mechanical feature and creates a wide range of application prospects in lots of fields, such as aerospace, electronics, medicine, etc. Fundamental understanding on the mechanism underlying NPR plays an important role in designing advanced mechanical functional materials. However, with different methods used, the origin of NPR is found different and conflicting with each other, for instance, in the representative graphene. In this study, based on machine learning technique, we constructed a moment tensor potential for molecular dynamics (MD) simulations of graphene. By analyzing the evolution of key geometries, the increase of bond angle is found to be responsible for the NPR of graphene instead of bond length. The results on the origin of NPR are well consistent with the start-of-art first-principles, which amend the results from MD simulations using classic empirical potentials. Our study facilitates the understanding on the origin of NPR of graphene and paves the way to improve the accuracy of MD simulations being comparable to first-principle calculations. Our study would also promote the applications of machine learning interatomic potentials in multiscale simulations of functional materials.

Self-Healing Mechanism of Lithium in Lithium Metal

Li is an ideal anode material for use in state-of-the-art secondary batteries. However, Li-dendrite growth is a safety concern and results in low coulombic efficiency, which significantly restricts the commercial application of Li secondary batteries. Unfortunately, the Li-deposition (growth) mechanism is poorly understood on the atomic scale. Here, machine learning is used to construct a Li potential model with quantum-mechanical computational accuracy. Molecular dynamics simulations in this study with this model reveal two self-healing mechanisms in a large Li-metal system, viz. surface self-healing, and bulk self-healing. It is concluded that self-healing occurs rapidly in nanoscale; thus, minimizing the voids between the Li grains using several comprehensive methods can effectively facilitate the formation of dendrite-free Li.

A deep potential model with long-range electrostatic interactions

Machine learning models for the potential energy of multi-atomic systems, such as the deep potential (DP) model, make molecular simulations with the accuracy of quantum mechanical density functional theory possible at a cost only moderately higher than that of empirical force fields. However, the majority of these models lack explicit long-range interactions and fail to describe properties that derive from the Coulombic tail of the forces. To overcome this limitation, we extend the DP model by approximating the long-range electrostatic interaction between ions (nuclei + core electrons) and valence electrons with that of distributions of spherical Gaussian charges located at ionic and electronic sites. The latter are rigorously defined in terms of the centers of the maximally localized Wannier distributions, whose dependence on the local atomic environment is modeled accurately by a deep neural network. In the DP long-range (DPLR) model, the electrostatic energy of the Gaussian charge system is added to short-range interactions that are represented as in the standard DP model. The resulting potential energy surface is smooth and possesses analytical forces and virial. Missing effects in the standard DP scheme are recovered, improving on accuracy and predictive power. By including long-range electrostatics, DPLR correctly extrapolates to large systems the potential energy surface learned from quantum mechanical calculations on smaller systems. We illustrate the approach with three examples: the potential energy profile of the water dimer, the free energy of interaction of a water molecule with a liquid water slab, and the phonon dispersion curves of the NaCl crystal.

Wildfire smoke destroys stratospheric ozone

Large wildfires inject smoke and biomass-burning products into the mid-latitude stratosphere, where they destroy ozone, which protects us from ultraviolet radiation. The infrared spectrometer on the Atmospheric Chemistry Experiment satellite measured the spectra of smoke particles from the "Black Summer" fires in Australia in late 2019 and early 2020, revealing that they contain oxygenated organic functional groups and water adsorption on the surfaces. These injected smoke particles have produced unexpected and extreme perturbations in stratospheric gases beyond any seen in the previous 15 years of measurements, including increases in formaldehyde, chlorine nitrate, chlorine monoxide, and hypochlorous acid and decreases in ozone, nitrogen dioxide, and hydrochloric acid. These perturbations in stratospheric composition have the potential to affect ozone chemistry in unexpected ways.

Uptake of N2O5 by aqueous aerosol unveiled using chemically accurate many-body potentials

The reactive uptake of N2O5 to aqueous aerosol is a major loss channel for nitrogen oxides in the troposphere. Despite its importance, a quantitative picture of the uptake mechanism is missing. Here we use molecular dynamics simulations with a data-driven many-body model of coupled-cluster accuracy to quantify thermodynamics and kinetics of solvation and adsorption of N2O5 in water. The free energy profile highlights that N2O5 is selectively adsorbed to the liquid-vapor interface and weakly solvated. Accommodation into bulk water occurs slowly, competing with evaporation upon adsorption from gas phase. Leveraging the quantitative accuracy of the model, we parameterize and solve a reaction-diffusion equation to determine hydrolysis rates consistent with experimental observations. We find a short reaction-diffusion length, indicating that the uptake is dominated by interfacial features. The parameters deduced here, including solubility, accommodation coefficient, and hydrolysis rate, afford a foundation for which to consider the reactive loss of N2O5 in more complex solutions.

Understanding High-Temperature Chemical Reactions on Metal Surfaces: A Case Study on Equilibrium Concentration and Diffusivity of C x H y on a Cu(111) Surface

Chemical reactions on metal surfaces are important in various processes such as heterogeneous catalysis and nanostructure growth. At moderate or lower temperatures, these reactions generally follow the minimum energy path, and temperature effects can be reasonably described by a harmonic oscillator model. At a high temperature approaching the melting point of the substrate, general behaviors of surface reactions remain elusive. In this study, by taking hydrocarbon species adsorbed on Cu(111) as a model system and performing extensive molecular dynamics simulations powered by machine learning potentials, we identify several important high-temperature effects, including local chemical environment, surface atom mobility, and substrate thermal expansion. They affect different aspects of a high-temperature surface reaction in different ways. These results deepen our understanding of high-temperature reactions.

Potential Application of Machine-Learning-Based Quantum Chemical Methods in Environmental Chemistry

It is an important topic in environmental sciences to understand the behavior and toxicology of chemical pollutants. Quantum chemical methodologies have served as useful tools for probing behavior and toxicology of chemical pollutants in recent decades. In recent years, machine learning (ML) techniques have brought revolutionary developments to the field of quantum chemistry, which may be beneficial for investigating environmental behavior and toxicology of chemical pollutants. However, the ML-based quantum chemical methods (ML-QCMs) have only scarcely been used in environmental chemical studies so far. To promote applications of the promising methods, this Perspective summarizes recent progress in the ML-QCMs and focuses on their potential applications in environmental chemical studies that could hardly be achieved by the conventional quantum chemical methods. Potential applications and challenges of the ML-QCMs in predicting degradation networks of chemical pollutants, searching global minima for atmospheric nanoclusters, discovering heterogeneous or photochemical transformation pathways of pollutants, as well as predicting environmentally relevant end points with wave functions as descriptors are introduced and discussed.

Peptide Isomerization is Suppressed at the Air-Water Interface

We use molecular dynamics simulations to study the thermodynamics and kinetics of alanine dipeptide isomerization at the air-water interface. Thermodynamically, we find an affinity of the dipeptide to the interface. This affinity arises from stabilizing intramolecular interactions that become unshielded as the dipeptide is desolvated. Kinetically, we consider the rate of transitions between the α_L and β conformations of alanine dipeptide and evaluate it as a continuous function of the distance from the interface using a recent extension of transition path sampling, TPS+U. The rate of isomerization at the Gibbs dividing surface is suppressed relative to the bulk by a factor of 3. Examination of the ensemble of transition states elucidates the role of solvent degrees of freedom in mediating favorable intramolecular interactions along the reaction pathway of isomerization. Near the air-water interface, water is less effective at mediating these intramolecular interactions.

Molecular dynamics simulation of metallic Al-Ce liquids using a neural network machine learning interatomic potential

Al-rich Al-Ce alloys have the possibility of replacing heavier steel and cast irons for use in high-temperature applications. Knowledge about the structures and properties of Al-Ce alloys at the liquid state is vital for optimizing the manufacture process to produce desired alloys. However, reliable molecular dynamics simulation of Al-Ce alloy systems remains a great challenge due to the lack of accurate Al-Ce interatomic potential. Here, an artificial neural network (ANN) deep machine learning (ML) method is used to develop a reliable interatomic potential for Al-Ce alloys. Ab initio molecular dynamics simulation data on the Al-Ce liquid with a small unit cell (∼200 atoms) and on the known Al-Ce crystalline compounds are collected to train the interatomic potential using ANN-ML. The obtained ANN-ML model reproduces well the energies, forces, and atomic structure of the Al₉₀Ce₁₀ liquid and crystalline phases of Al-Ce compounds in comparison with the ab initio results. The developed ANN-ML potential is applied in molecular dynamics simulations to study the structures and properties of the metallic Al₉₀Ce₁₀ liquid, which would provide useful insight into the guiding experimental process to produce desired Al-Ce alloys.

Development of Range-Corrected Deep Learning Potentials for Fast, Accurate Quantum Mechanical/Molecular Mechanical Simulations of Chemical Reactions in Solution

We develop a new deep potential─range correction (DPRc) machine learning potential for combined quantum mechanical/molecular mechanical (QM/MM) simulations of chemical reactions in the condensed phase. The new range correction enables short-ranged QM/MM interactions to be tuned for higher accuracy, and the correction smoothly vanishes within a specified cutoff. We further develop an active learning procedure for robust neural network training. We test the DPRc model and training procedure against a series of six nonenzymatic phosphoryl transfer reactions in solution that are important in mechanistic studies of RNA-cleaving enzymes. Specifically, we apply DPRc corrections to a base QM model and test its ability to reproduce free-energy profiles generated from a target QM model. We perform these comparisons using the MNDO/d and DFTB2 semiempirical models because they differ in the way they treat orbital orthogonalization and electrostatics and produce free-energy profiles which differ significantly from each other, thereby providing us a rigorous stress test for the DPRc model and training procedure. The comparisons show that accurate reproduction of the free-energy profiles requires correction of the QM/MM interactions out to 6 Å. We further find that the model's initial training benefits from generating data from temperature replica exchange simulations and including high-temperature configurations into the fitting procedure, so the resulting models are trained to properly avoid high-energy regions. A single DPRc model was trained to reproduce four different reactions and yielded good agreement with the free-energy profiles made from the target QM/MM simulations. The DPRc model was further demonstrated to be transferable to 2D free-energy surfaces and 1D free-energy profiles that were not explicitly considered in the training. Examination of the computational performance of the DPRc model showed that it was fairly slow when run on CPUs but was sped up almost 100-fold when using NVIDIA V100 GPUs, resulting in almost negligible overhead. The new DPRc model and training procedure provide a potentially powerful new tool for the creation of next-generation QM/MM potentials for a wide spectrum of free-energy applications ranging from drug discovery to enzyme design.

Learning intermolecular forces at liquid-vapor interfaces

By adopting a perspective informed by contemporary liquid-state theory, we consider how to train an artificial neural network potential to describe inhomogeneous, disordered systems. We find that neural network potentials based on local representations of atomic environments are capable of describing some properties of liquid-vapor interfaces but typically fail for properties that depend on unbalanced long-ranged interactions that build up in the presence of broken translation symmetry. These same interactions cancel in the translationally invariant bulk, allowing local neural network potentials to describe bulk properties correctly. By incorporating explicit models of the slowly varying long-ranged interactions and training neural networks only on the short-ranged components, we can arrive at potentials that robustly recover interfacial properties. We find that local neural network models can sometimes approximate a local molecular field potential to correct for the truncated interactions, but this behavior is variable and hard to learn. Generally, we find that models with explicit electrostatics are easier to train and have higher accuracy. We demonstrate this perspective in a simple model of an asymmetric dipolar fluid, where the exact long-ranged interaction is known, and in an ab initio water model, where it is approximated.

The Middle Science: Traversing Scale In Complex Many-Body Systems

A roadmap is developed that integrates simulation methodology and data science methods to target new theories that traverse the multiple length- and time-scale features of many-body phenomena.

Experimental evidence that halogen bonding catalyzes the heterogeneous chlorination of alkenes in submicron liquid droplets

A key challenge in predicting the multiphase chemistry of aerosols and droplets is connecting reaction probabilities, observed in an experiment, with the kinetics of individual elementary steps that control the chemistry that occurs across a gas/liquid interface. Here we report evidence that oxygenated molecules accelerate the heterogeneous reaction rate of chlorine gas with an alkene (squalene, Sqe) in submicron droplets. The effective reaction probability for Sqe is sensitive to both the aerosol composition and gas phase environment. In binary aerosol mixtures with 2-decyl-1-tetradecanol, linoleic acid and oleic acid, Sqe reacts 12-23× more rapidly than in a pure aerosol. In contrast, the reactivity of Sqe is diminished by 3× when mixed with an alkane. Additionally, small oxygenated molecules in the gas phase (water, ethanol, acetone, and acetic acid) accelerate (up to 10×) the heterogeneous chlorination rate of Sqe. The overall reaction mechanism is not altered by the presence of these aerosol and gas phase additives, suggesting instead that they act as catalysts. Since the largest rate acceleration occurs in the presence of oxygenated molecules, we conclude that halogen bonding enhances reactivity by slowing the desorption kinetics of Cl2 at the interface, in a way that is analogous to decreasing temperature. These results highlight the importance of relatively weak interactions in controlling the speed of multiphase reactions important for atmospheric and indoor environments.

Mechanistic Study of the Aqueous Reaction of Organic Peroxides with HSO3 - on the Surface of a Water Droplet

Aqueous reactions between organic peroxides and SO₂ are of intense interest in atmospheric science because of their ubiquitous implications for sulfate formation in secondary aerosols. However, the relative yields of the reaction products (inorganic vs. organic sulfates) remain controversial (i.e., 90 % vs. 40-70 % for inorganic sulfate) due in part to the lack of understanding of the underlying reaction mechanisms. Here, our computational results suggest that the reactions of HSO₃ ^- (dissolved SO₂ ) with organic peroxides are initiated on the surface of water nanodroplets and then proceed under two reaction pathways, in which the S atom of HSO₃ ^- attacks either the O1 or O2 atom of the peroxide group -O(O2)O(O1)H, leading to the formation of inorganic and organic sulfates, respectively. Notably, we find that thse reaction initiated by O1 atom exhibits a relatively low energy barrier and high reaction rate, which favours the formation of inorganic sulfate.