Highly Accurate Many-Body Potentials for Simulations of N₂O₅ in Water: Benchmarks, Development, and Validation

Abstract

Dinitrogen pentoxide (N2O5) is an important intermediate in the atmospheric chemistry of nitrogen oxides.Although there has been much research, the processes that govern the physical interactions between N2O5 and water are still not fully understood at a molecular level. Gaining quantitative insight from computer simulations requires going beyond the accuracy of classical force fields, while accessing length scales and time scales that are out of reach for high-level quantum chemical approaches. To this end we present the development of MB-nrg many-body potential energy functions for simulations of N2O5 in water. This MB-nrg model is based on electronic structure calculations at the coupled cluster level of theory and is compatible with the successful MB-pol model for water. It provides a physically correct description of long-range many-body interactions in combination with an explicit representation of up to three-body short-range interactions in terms of multidimensional permutationally invariant polynomials. In order to further investigate the importance of the underlying interactions in the model, a TTM-nrg model was also devised. TTM-nrg is a more simplistic representation that contains only two-body short-range interactions represented through Born-Mayer functions. In this work an active learning approach was employed to efficiently build representative training sets of monomer, dimer and trimer structures, and benchmarks are presented to determine the accuracy of our new models in comparison to a range of density functional theory methods. By assessing binding curves, distortion energies of N2O5, and interaction energies in clusters of N2O5 and water, we evaluate the importance of two-body and three-body short-range potentials. The results demonstrate that our MB-nrg model has high accuracy with respect to the coupled cluster reference, outperforms current density functional theory models, and thus enables highly accurate simulations of N2O5 in aqueous environments. File list (2)download file view on ChemRxiv paper_n2o5.pdf (1.37 MiB) download file view on ChemRxiv paper_n2o5_si.pdf (731.31 KiB)

MB-Fit: Software Infrastructure for Data-Driven Many-Body Potential Energy Functions

“…Specifically, building upon the MB-pol PEF for water, TTM-nrg and MB-nrg PEFs were introduced for halide 91,92 and alkali-metal ions 94,95 in water, water-carbon dioxide 97 and water-methane 98 mixtures, and small molecules in water. 99 When used in computer simulations, the TTM-nrg and MB-nrg PEFs consistently provide remarkable agreement with experimental data, which is effectively quantitative in the case of the MB-nrg PEFs. 97,98,[100][101][102][103][104][105]107 In the most general form, the TTM-nrg and MB-nrg PEFs approximate the MBE of Eq.…”

“…[86][87][88][89][90] The hybrid data-driven/physics-based scheme originally developed for MB-pol was later extended to generic molecules through the introduction of two families of MB-PEFs, the TTM-nrg (for "Thole-type model energy") and MB-nrg (for "many-body energy") PEFs, for halide [91][92][93] and alkali-metal [94][95][96] ions in water, molecular fluids, 97,98 and small molecules in water. 99 When employed in computer simulations, the MB-nrg PEFs have been shown to consistently provide remarkable agreement with experimental data for both gas-phase and condensed-phase systems. [100][101][102][103][104][105][106][107] Here, we present MB-Fit, a complete software infrastructure for the automated development of TTM-nrg and MB-nrg PEFs for generic molecules.…”

“…As shown in previous studies, 92,95,[97][98][99] higher accuracy in modeling 2B energies is achieved by the MB-nrg PEFs that adopt short-range PIPs to effectively represent 2B quantum-mechanical in- teractions that arise from the overlap of the monomers' electron densities (e.g., Pauli repulsion, and charge transfer and penetration). 58,59,61,92,95,97,98,103,105,139 The 2B training (7261 configurations) and test (1449 configurations) sets for the MB-nrg PEFs were generated by including configurations extracted from three different sources.…”

MB-Fit: Software Infrastructure for Data-Driven Many-Body Potential Energy Functions

“…[86][87][88][89][90] The hybrid data-driven/physics-based scheme originally developed for MB-pol was later extended to generic molecules through the introduction of two families of MB-PEFs, the TTM-nrg (for "Thole-type model energy") and MB-nrg (for "many-body energy") PEFs, for halide [91][92][93] and alkali-metal [94][95][96] ions in water, molecular fluids, 97,98 and small molecules in water. 99 When employed in computer simulations, the MB-nrg PEFs have been shown to consistently provide remarkable agreement with experimental data for both gas-phase and condensed-phase systems. [100][101][102][103][104][105][106][107] Here, we present MB-Fit, a complete software infrastructure for the automated development of TTM-nrg and MB-nrg PEFs for generic molecules.…”

MB-Fit: Software Infrastructure for Data-Driven Many-Body Potential Energy Functions

“…As shown in previous studies, 92,95,[97][98][99] higher accuracy in modeling 2B energies is achieved by the MB-nrg PEFs that adopt short-range PIPs to effectively represent 2B quantum-mechanical in- teractions that arise from the overlap of the monomers' electron densities (e.g., Pauli repulsion, and charge transfer and penetration). 58,59,61,92,95,97,98,103,105,139 The 2B training (7261 configurations) and test (1449 configurations) sets for the MB-nrg PEFs were generated by including configurations extracted from three different sources.…”