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

Bull-Vulpe, Ethan F.; Riera, Marc; Goetz, Andreas; Paesani, Francesco

doi:10.33774/chemrxiv-2021-pjr3l-v2

“…Briefly, the MB-nrg PEFs use a combination of short-range permutationally invariant polynomials (PIPs) 58 trained on electronic structure data and physicsbased functions to represent the 1B, 2B, and 3B terms of the MBE in Eq. 1, 57,59 while all other nB terms with n > 3 are represented by an implicit many-body term derived from the Thole model of classical polarization. 60 Within the MB-nrg theoretical/computational framework, Eq.…”

Section: Mb-nrg Pefsmentioning

confidence: 99%

Accurate Modeling of Bromide and Iodide Hydration with Data-Driven Many-Body Potentials

Caruso

¹

,

Zhu

²

,

Fulton

³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Ion–water interactions play a central role in determining the properties of aqueous systems in a wide range of environments. However, a quantitative understanding of how the hydration properties of ions evolve from small aqueous clusters to bulk solutions and interfaces remains elusive. Here, we introduce the second generation of data-driven many-body energy (MB- nrg) potential energy functions (PEFs) representing bromide–water and iodide–water interactions. The MB-nrg PEFs use permutationally invariant polynomials to reproduce two-body and three-body energies calculated at the coupled cluster level of theory, and implicitly represent all higher-body energies using classical many-body polarization. A systematic analysis of the hydration structure of small Br−(H2O)n and I−(H2O)n clusters demonstrates that the MB-nrg PEFs predict interaction energies in quantitative agreement with “gold standard” coupled cluster reference values. Importantly, when used in molecular dynamics simulations carried out in the isothermal-isobaric ensemble for single bromide and iodide ions in liquid water, the MB-nrg PEFs predict extended X-ray absorption fine structure (EXAFS) spectra that accurately reproduce the experimental spectra, which thus allows for characterizing the hydration structure of the two ions with high level of confidence.

show abstract

“…The accuracy and predictive power of the MB-nrg potentials have been demonstrated in simulations of ion hydration, [30][31][32][33][34][35][36][37][38] and various fluid mixtures, including CO 2 -water, 25 ammonia-water, 39 , and methane-water 26 mixtures. The MB-nrg potentials are implemented in the open-access MBX software 27 which is interfaced with LAMMPS 40 and i-PI.…”

Section: A Molecular Dynamics Simulationsmentioning

confidence: 99%

The Behavior of Methane-Water Mixtures Under Elevated Pressures Using Many-Body Potentials

Robinson

¹

,

Ghosh

²

,

Egan

³

et al. 2022

Preprint

Self Cite

0

View full text Add to dashboard Cite

Non-polarizable empirical potentials have been shown not to be able to capture the mixing of methane–water mixtures at elevated pressures. Although DFT-based ab initio simulations may circumvent this discrepancy, they are limited in terms of the relevant time and length scales associated with mixing phenomena. Here we show that the many-body MB-nrg potential, designed to reproduce methane–water interactions with coupled cluster accuracy, successfully captures this phenomenon up to 3 GPa and 500 K with varying methane concentration. Two-phase simulations and long time scales that are required to fully capture the mixing, affordable due to the speed and accuracy of the MBX software, are assessed. Constructing the methane–water equation of state across the phase diagram shows that the stable mixtures are denser than the sum of their parts at a given pressure and temperature. We find that many-body polarization plays a central role, enhancing the induced dipole moments of methane by 0.20 D during mixing under pressure. Overall, the mixed system adopts a denser state, which involves a significant enthalpic driving force as elucidated by a systematic many-body energy decomposition analysis.

show abstract

“…Since the MB-nrg theoretical/computational framework has already been described in the literature, [47][48][49][50]57 we will only describe here the salient features and provide details specific to the bromide-water and iodide-water PEFs. Briefly, the MB-nrg PEFs use a combination of short-range permutationally invariant polynomials (PIPs) 58 trained on electronic structure data and physicsbased functions to represent the 1B, 2B, and 3B terms of the MBE in Eq.…”

Section: Mb-nrg Pefsmentioning

confidence: 99%

“…Briefly, the MB-nrg PEFs use a combination of short-range permutationally invariant polynomials (PIPs) 58 trained on electronic structure data and physicsbased functions to represent the 1B, 2B, and 3B terms of the MBE in Eq. 1, 57,59 while all other nB terms with n > 3 are represented by an implicit many-body term derived from the Thole model of classical polarization. 60 Within the MB-nrg theoretical/computational framework, Eq.…”

Section: Mb-nrg Pefsmentioning

confidence: 99%

See 1 more Smart Citation

Accurate Modeling of Bromide and Iodide Hydration with Data-Driven Many-Body Potentials

Caruso

¹

,

Zhu

²

,

Paesani

³

2022

Preprint

0

View full text Add to dashboard Cite

Ion–water interactions play a central role in determining the properties of aqueous systems in a wide range of environments. However, a quantitative understanding of how the hydration properties of ions evolve from small aqueous clusters to bulk solutions and interfaces remains elusive. Here, we introduce the second generation of data-driven many-body energy (MB- nrg) potential energy functions (PEFs) representing bromide–water and iodide–water interactions. The MB-nrg PEFs use permutationally invariant polynomials to reproduce two-body and three-body energies calculated at the coupled cluster level of theory, and implicitly represent all higher-body energies using classical many-body polarization. A systematic analysis of the hydration structure of small Br−(H2O)n and I−(H2O)n clusters demonstrates that the MB-nrg PEFs predict interaction energies in quantitative agreement with “gold standard” coupled cluster reference values. Importantly, when used in molecular dynamics simulations carried out in the isothermal-isobaric ensemble for single bromide and iodide ions in liquid water, the MB-nrg PEFs predict extended X-ray absorption fine structure (EXAFS) spectra that accurately reproduce the experimental spectra, which thus allows for characterizing the hydration structure of the two ions with high level of confidence.

show abstract

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

Cited by 4 publications

References 122 publications

Accurate Modeling of Bromide and Iodide Hydration with Data-Driven Many-Body Potentials

Accurate Modeling of Bromide and Iodide Hydration with Data-Driven Many-Body Potentials

The Behavior of Methane-Water Mixtures Under Elevated Pressures Using Many-Body Potentials

Accurate Modeling of Bromide and Iodide Hydration with Data-Driven Many-Body Potentials

Contact Info

Product

Resources

About