Multistate Reactive Molecular Dynamics Simulations of Proton Diffusion in Water Clusters and in the Bulk

Grotthuss Molecular Dynamics Simulations for Modeling Proton Hopping in Electrosprayed Water Droplets

Excess protons in water exhibit unique transport properties because they can rapidly hop along H-bonded water wires. Considerable progress has been made in unraveling this Grotthuss diffusion mechanism using quantum mechanical-based computational techniques. Unfortunately, high computational cost tends to restrict those techniques to small systems and short times. Molecular dynamics (MD) simulations can be applied to much larger systems and longer time windows. However, standard MD methods do not permit the dissociation/formation of covalent bonds, such that Grotthuss diffusion cannot be captured. Here, we bridge this gap by combining atomistic MD simulations (using Gromacs and TIP4P/2005 water) with proton hopping. Excess protons are modeled as hydronium ions that undergo H₃O⁺ + H₂O → H₂O + H₃O⁺ transitions. In accordance with ab initio MD data, these Grotthuss hopping events are executed in "bursts" with quasi-instantaneous hopping across one or more waters. The bursts are separated by regular MD periods during which H₃O⁺ ions undergo Brownian diffusion. The resulting proton diffusion coefficient agrees with the literature value. We apply this Grotthuss MD technique to highly charged water droplets that are in a size regime encountered during electrospray ionization (5 nm radius, ∼17,000 H₂O). The droplets undergo rapid solvent evaporation and occasional H₃O⁺ ejection, keeping them at ca. 81% of the Rayleigh limit. The simulated behavior is consistent with phase Doppler anemometry data. The Grotthuss MD technique developed here should be useful for modeling the behavior of various proton-containing systems that are too large for high-level computational approaches. In particular, we envision future applications related to electrospray processes, where earlier simulations used metal cations while in reality excess protons dominate.

Machine Learning for Chemical Reactions

Machine learning (ML) techniques applied to chemical reactions have a long history. The present contribution discusses applications ranging from small molecule reaction dynamics to computational platforms for reaction planning. ML-based techniques can be particularly relevant for problems involving both computation and experiments. For one, Bayesian inference is a powerful approach to develop models consistent with knowledge from experiments. Second, ML-based methods can also be used to handle problems that are formally intractable using conventional approaches, such as exhaustive characterization of state-to-state information in reactive collisions. Finally, the explicit simulation of reactive networks as they occur in combustion has become possible using machine-learned neural network potentials. This review provides an overview of the questions that can and have been addressed using machine learning techniques, and an outlook discusses challenges in this diverse and stimulating field. It is concluded that ML applied to chemistry problems as practiced and conceived today has the potential to transform the way with which the field approaches problems involving chemical reactions, in both research and academic teaching.

Insights into protonation for cyclohexanol/water mixtures at the zeolitic Brønsted acid site

Proton transfer from Brønsted acid sites (BASs) to alcohol molecules ignites the acid-catalyzed alcohol dehydration reactions. For aqueous phase dehydration reactions in zeolites, the coexisting water molecules around BASs in the zeolite pores significantly affect the alcohol dehydration activity. In the present work, proton transfer processes among the BASs of H-BEA zeolites, the adsorbed cyclohexanol and surrounding water clusters with different sizes up to 8 water molecules were investigated using ab initio molecular dynamics (AIMD) simulations combined with the multiple-walker well-tempered metadynamics algorithm. The plausible proton locations and proton transfer processes were characterized using two/three-dimensional free energy landscapes. The strong proton affinity makes the protonated cyclohexanol stable species until a water trimer is formed. The proton either is shared between protonated cyclohexanol and the water trimer or remains with the water trimer (H7O3+). With a further increase in water concentrations, the proton prefers to remain with the water clusters.

Protonation Dynamics in the K-Channel of Cytochrome c Oxidase Estimated from Molecular Dynamics Simulations

Proton transfer reactions are one of the most fundamental processes in biochemistry. We present a simplistic approach for estimating proton transfer probabilities in a membrane protein, cytochrome c oxidase. We combine short molecular dynamics simulations at discrete protonation states with a Monte Carlo approach to exchange between those states. Requesting for a proton transfer the existence of a hydrogen-bonded connection between the two source and target residues of the exchange, restricts the acceptance of transfers to only those in which a proton-relay is possible. Together with an analysis of the hydrogen-bonded connectivity in one of the proton-conducting channels of cytochrome c oxidase, this approach gives insight into the protonation dynamics of the hydrogen-bonded networks. The connectivity and directionality of the networks are coupled to the conformation of an important protein residue in the channel, K362, rendering proton transfer in the entire channel feasible in only one of the two major conformations. Proton transport in the channel can thus be regulated by K362 not only through its possible role as a proton carrier itself, but also by allowing or preventing proton transport via water residues.

Synergistic Computational Modeling Approaches as Team Players in the Game of Solubility Predictions

Several approaches to predict and model drug solubility have been used in the drug discovery and development processes during the last decades. Each of these approaches have their own benefits and place, and are typically used as standalone approaches rather than in concert. The synergistic effects of these are often overlooked, partly due to the need of computational experts to perform the modeling and simulations as well as analyzing the data obtained. Here we provide our views on how these different approaches can be used to retrieve more information on drug solubility, ranging from multivariate data analysis over thermodynamic cycle modeling to molecular dynamics simulations. We are discussing aqueous solubility as well as solubility in more complex mixed solvents and media with colloidal structures present. We conclude that the field of computational pharmaceutics is in its early days but with a bright future ahead. However, education of computational formulators with broad knowledge of modeling and simulation approaches is imperative if computational pharmaceutics is to reach its full potential.

A Reactive Force Field with Coarse-Grained Electrons for Liquid Water

Nonreactive force fields are defined by perturbations of electron density that are relatively small, whereas chemical reactivity involves wholesale electronic rearrangements that make and break bonds. Thus, reactive force fields are incredibly difficult to develop compared to nonreactive force fields, yet at the same time, they fill a critical need when ab initio molecular dynamics methods are not affordable. We introduce a new reactive force field model for water that combines modified nonbonded terms of the ReaxFF model and its embedding in the electrostatic interactions described by our recently introduced coarse-grained electron model (C-GeM). The ReaxFF/C-GeM force field is characterized for many energetic and dissociative water properties for water clusters, structure, and dynamical properties under ambient conditions in the condensed phase, as well as the temperature dependence of density and water diffusion, with very good agreement with experiment. The ReaxFF/C-GeM force field should be more transferable and more broadly applicable to a range of reactive systems involving both proton and electron transfer in the condensed phase.

Decoding the 2D IR spectrum of the aqueous proton with high-level VSCF/VCI calculations

The aqueous proton is a common and long-studied species in chemistry, yet there is currently intense interest devoted to understanding its hydration structure and transport dynamics. Typically described in terms of two limiting structures observed in gas-phase clusters, the Zundel H₅O₂ ⁺ and Eigen H₉O₄ ⁺ ions, the aqueous structure is less clear due to the heterogeneity of hydrogen bonding environments and room-temperature structural fluctuations in water. The linear infrared (IR) spectrum, which reports on structural configurations, is challenging to interpret because it appears as a continuum of absorption, and the underlying vibrational modes are strongly anharmonically coupled to each other. Recent two-dimensional IR (2D IR) experiments presented strong evidence for asymmetric Zundel-like motifs in solution, but true structure-spectrum correlations are missing and complicated by the anharmonicity of the system. In this study, we employ high-level vibrational self-consistent field/virtual state configuration interaction calculations to demonstrate that the 2D IR spectrum reports on a broad distribution of geometric configurations of the aqueous proton. We find that the diagonal 2D IR spectrum around 1200 cm^-1 is dominated by the proton stretch vibrations of Zundel-like and intermediate geometries, broadened by the heterogeneity of aqueous configurations. There is a wide distribution of multidimensional potential shapes for the proton stretching vibration with varying degrees of potential asymmetry and confinement. Finally, we find specific cross peak patterns due to aqueous Zundel-like species. These studies provide clarity on highly debated spectral assignments and stringent spectroscopic benchmarks for future simulations.