Reactive Monte Carlo sampling with an ab initio potential

Facilitating ab initio configurational sampling of multicomponent solids using an on-lattice neural network model and active learning

We propose a scheme for ab initio configurational sampling in multicomponent crystalline solids using Behler-Parinello type neural network potentials (NNPs) in an unconventional way: the NNPs are trained to predict the energies of relaxed structures from the perfect lattice with configurational disorder instead of the usual way of training to predict energies as functions of continuous atom coordinates. An active learning scheme is employed to obtain a training set containing configurations of thermodynamic relevance. This enables bypassing of the structural relaxation procedure which is necessary when applying conventional NNP approaches to the lattice configuration problem. The idea is demonstrated on the calculation of the temperature dependence of the degree of A/B site inversion in three spinel oxides, MgAl2O4, ZnAl2O4, and MgGa2O4. The present scheme may serve as an alternative to cluster expansion for `difficult' systems, e.g., complex bulk or interface systems with many components and sublattices that are relevant to many technological applications today.

Generation of Quantum Configurational Ensembles Using Approximate Potentials

Conformational analysis is of paramount importance in drug design: it is crucial to determine pharmacological properties, understand molecular recognition processes, and characterize the conformations of ligands when unbound. Molecular Mechanics (MM) simulation methods, such as Monte Carlo (MC) and molecular dynamics (MD), are usually employed to generate ensembles of structures due to their ability to extensively sample the conformational space of molecules. The accuracy of these MM-based schemes strongly depends on the functional form of the force field (FF) and its parametrization, components that often hinder their performance. High-level methods, such as ab initio MD, provide reliable structural information but are still too computationally expensive to allow for extensive sampling. Therefore, to overcome these limitations, we present a multilevel MC method that is capable of generating quantum configurational ensembles while keeping the computational cost at a minimum. We show that FF reparametrization is an efficient route to generate FFs that reproduce QM results more closely, which, in turn, can be used as low-cost models to achieve the gold standard QM accuracy. We demonstrate that the MC acceptance rate is strongly correlated with various phase space overlap measurements and that it constitutes a robust metric to evaluate the similarity between the MM and QM levels of theory. As a more advanced application, we present a self-parametrizing version of the algorithm, which combines sampling and FF parametrization in one scheme, and apply the methodology to generate the QM/MM distribution of a ligand in aqueous solution.

First principles reactive simulation for equation of state prediction

The high cost of density functional theory (DFT) has hitherto limited the ab initio prediction of the equation of state (EOS). In this article, we employ a combination of large scale computing, advanced simulation techniques, and smart data science strategies to provide an unprecedented ab initio performance analysis of the high explosive pentaerythritol tetranitrate (PETN). Comparison to both experiment and thermochemical predictions reveals important quantitative limitations of DFT for EOS prediction and thus the assessment of high explosives. In particular, we find that DFT predicts the energy of PETN detonation products to be systematically too high relative to the unreacted neat crystalline material, resulting in an underprediction of the detonation velocity, pressure, and temperature at the Chapman-Jouguet state. The energetic bias can be partially accounted for by high-level electronic structure calculations of the product molecules. We also demonstrate a modeling strategy for mapping chemical composition across a wide parameter space with limited numerical data, the results of which suggest additional molecular species to consider in thermochemical modeling.

Complete equations of state for PETN and its products from atomistic simulations

The complete caloric and thermal equations of state for pentaerythritol tetranitrate (PETN) and its decomposition products are developed. The equation for the crystalline state is obtained with quasiharmonic approximation for the vibrational energy, with the force constants being calculated using density functional theory. The equation of state for the products is derived from equilibrium ReaxFF molecular dynamics simulations. Two equations are coupled through the heat of thermal decomposition calculated using ReaxFF at high temperature. Our hydrodynamic code utilizing the developed EOSs reproduces well the detonation velocity and Chapman-Jouguet pressure obtained in the molecular dynamics simulations.

Direct coupling of first-principles calculations with replica exchange Monte Carlo sampling of ion disorder in solids

We demonstrate the feasibility of performing sufficient configurational sampling of disordered oxides directly from first-principles without resorting to the use of fitted models such as cluster expansion. This is achieved by harnessing the power of modern-day cluster supercomputers using the replica exchange Monte Carlo method coupled directly with structural relaxation and energy calculation performed by density functional codes. The idea is applied successfully to the calculation of the temperature-dependence of the degree of inversion in the cation sublattice of MgAl₂O₄ spinel oxide. The possibility of bypassing fitting models will lead to investigation of disordered systems where cluster expansion is known to perform badly, for example, systems with large lattice deformation due to defects, or systems where long-range interactions dominate such as electrochemical interfaces.

Towards operando computational modeling in heterogeneous catalysis

An increased synergy between experimental and theoretical investigations in heterogeneous catalysis has become apparent during the last decade. Experimental work has extended from ultra-high vacuum and low temperature towards operando conditions. These developments have motivated the computational community to move from standard descriptive computational models, based on inspection of the potential energy surface at 0 K and low reactant concentrations (0 K/UHV model), to more realistic conditions. The transition from 0 K/UHV to operando models has been backed by significant developments in computer hardware and software over the past few decades. New methodological developments, designed to overcome part of the gap between 0 K/UHV and operando conditions, include (i) global optimization techniques, (ii) ab initio constrained thermodynamics, (iii) biased molecular dynamics, (iv) microkinetic models of reaction networks and (v) machine learning approaches. The importance of the transition is highlighted by discussing how the molecular level picture of catalytic sites and the associated reaction mechanisms changes when the chemical environment, pressure and temperature effects are correctly accounted for in molecular simulations. It is the purpose of this review to discuss each method on an equal footing, and to draw connections between methods, particularly where they may be applied in combination.

Molecular Simulation of Chemical Reaction Equilibrium by Computationally Efficient Free Energy Minimization

The molecular simulation of chemical reaction equilibrium (CRE) is a challenging and important problem of broad applicability in chemistry and chemical engineering. The primary molecular-based approach for solving this problem has been the reaction ensemble Monte Carlo (REMC) algorithm [Turner et al. Molec. Simulation2008, 34, (2), 119-146], based on classical force-field methodology. In spite of the vast improvements in computer hardware and software since its original development almost 25 years ago, its more widespread application is impeded by its computational inefficiency. A fundamental problem is that its MC basis inhibits the implementation of significant parallelization, and its successful implementation often requires system-specific tailoring and the incorporation of special MC approaches such as replica exchange, expanded ensemble, umbrella sampling, configurational bias, and continuous fractional component methodologies. We describe herein a novel CRE algorithm (reaction ensemble molecular dynamics, ReMD) that exploits modern computer hardware and software capabilities, and which can be straightforwardly implemented for systems of arbitrary size and complexity by exploiting the parallel computing methodology incorporated within many MD software packages (herein, we use GROMACS for illustrative purposes). The ReMD algorithm utilizes these features in the context of a macroscopically inspired and generally applicable free energy minimization approach based on the iterative approximation of the system Gibbs free energy function by a mathematically simple convex ideal solution model using the composition at each iteration as a reference state. Finally, we additionally describe a simple and computationally efficient a posteriori method to estimate the equilibrium concentrations of species present in very small amounts relative to others in the primary calculation. To demonstrate the algorithm, we show its application to two classic example systems considered previously in the literature: the N₂-O₂-NO system and the ammonia synthesis system.

Reaction Ensemble Monte Carlo Simulations of CO2 Absorption in the Reactive Ionic Liquid Triethyl(octyl)phosphonium 2-Cyanopyrrolide

The absorption of CO₂ into an aprotic heterocyclic anion ionic liquid (IL) is modeled using reaction ensemble Monte Carlo (RxMC) with the semigrand reaction move. RxMC has previously been unable to sample chemical equilibrium involving molecular ions in nanostructured liquids due to the high free-energy requirements to open and close cavities and restructure the surrounding environment. Our results are validated by experiments in the modeled IL, triethyl(octyl)phosphonium 2-cyanopyrrolide ([P₂₂₂₈][cnp]), and in a close analog with longer alkyl chains on the cation. Heats of absorption and reaction from both experiment and simulation are exothermic and of comparable magnitude. Replacing experimental Henry's constants with their simulated counterparts improves the accuracy of a Langmuir-type model at moderate pressures. Nonidealities that affect chemical equilibrium are identified and calculated with high precision.

Water-Lean Solvents for Post-Combustion CO2 Capture: Fundamentals, Uncertainties, Opportunities, and Outlook

This review is designed to foster the discussion regarding the viability of postcombustion CO₂ capture by water-lean solvents, by separating fact from fiction for both skeptics and advocates. We highlight the unique physical and thermodynamic properties of notable water-lean solvents, with a discussion of how such properties could translate to efficiency gains compared to aqueous amines. The scope of this review ranges from the purely fundamental molecular-level processes that govern solvent behavior to bench-scale testing, through process engineering and projections of process performance and cost. Key discussions of higher than expected CO₂ mass transfer, water tolerance, and compatibility with current infrastructure are presented along with current limitations and suggested areas where further solvent development is needed. We conclude with an outlook of the status of the field and assess the viability of water-lean solvents for postcombustion CO₂ capture.