High-dimensional neural network potentials for accurate vibrational frequencies: The formic acid dimer benchmark

Abstract

In recent years, machine learning potentials (MLP) for atomistic simulations have attracted a lot of attention in chemistry and materials science. Many new approaches have been developed with the primary aim to transfer the accuracy of electronic structure calculations to large condensed systems containing thousands of atoms. In spite of these advances, the reliability of modern MLPs in reproducing the subtle details of the multi-dimensional potential-energy surface is still difficult to assess for such systems. On the other hand, moderately sized systems enabling the application of tools for thorough and systematic quality-control are nowadays rarely investigated. In this work we use benchmark-quality harmonic and anharmonic vibrational frequencies as a sensitive probe for the validation of high-dimensional neural network potentials. For the case of the formic acid dimer, a frequently studied model system for which stringent spectroscopic data became recently available, we show that high-quality frequencies can be obtained from state-of-the-art calculations in excellent agreement with coupled cluster theory and experimental data.

References in corpus (21)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• Machine Learning Force Fields
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
• PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges
• Machine learning for molecular simulation
• Less is more: sampling chemical space with active learning
• Active learning of linearly parametrized interatomic potentials
• A Fourth-Generation High-Dimensional Neural Network Potential with Accurate Electrostatics Including Non-local Charge Transfer
• Machine Learning Molecular Dynamics for the Simulation of Infrared Spectra
• Interatomic potentials for ionic systems with density functional accuracy based on charge densities obtained by a neural network
• Raman Spectrum and Polarizability of Liquid Water from Deep Neural Networks
• The Interplay of Structure and Dynamics in the Raman Spectrum of Liquid Water over the Full Frequency and Temperature Range
• High-Dimensional Neural Network Potentials for Magnetic Systems Using Spin-Dependent Atom-Centered Symmetry Functions
• Towards breaking the curse of dimensionality in (ro)vibrational computations of molecular systems with multiple large-amplitude motions
• High-dimensional neural network potentials for accurate vibrational frequencies: The formic acid dimer benchmark
• Exact quantum dynamics background of dispersion interactions: case study for CH${}_4\cdot$Ar in full (12) dimensions
• Full-dimensional (12D) variational vibrational states of CH${}_4\cdot$F${}^{-}$: interplay of anharmonicity and tunneling
• Variational vibrational states of HCOOH
• A Smolyak algorithm adapted to a system-bath separation: application to an encapsulated molecule with large amplitude motions
• Fingerprint region of the formic acid dimer: variational vibrational computations in curvilinear coordinates

Cited by in corpus (6)

• Tutorial: How to Train a Neural Network Potential
• Tensorial properties via the neuroevolution potential framework: Fast simulation of infrared and Raman spectra
• $\Delta$-Machine Learning to Elevate DFT-based Potentials and a Force Field to the CCSD(T) Level Illustrated for Ethanol
• High-dimensional neural network potentials for accurate vibrational frequencies: The formic acid dimer benchmark
• Molecular dynamics-driven global tetra-atomic potential energy surfaces: Application to the AlF dimer
• Neural Canonical Transformation for the Spectra of Fluxional Molecule CH5+