An official website of the United States government
Abstract Machine learning interatomic potentials (IPs) can provide accuracy close to that of first-principles methods, such as density functional theory (DFT), at a fraction of the computational cost. This greatly extends the scope of accurate molecular simulations, providing opportunities for quantitative design of materials and devices on scales hitherto unreachable by DFT methods. However, machine learning IPs have a basic limitation in that they lack a physical model for the phenomena being predicted and therefore have unknown accuracy when extrapolating outside their training set. In this paper, we propose a class of Dropout Uncertainty Neural Network (DUNN) potentials that provide rigorous uncertainty estimates that can be understood from both Bayesian and frequentist statistics perspectives. As an example, we develop a DUNN potential for carbon and show how it can be used to predict uncertainty for static and dynamical properties, including stress and phonon dispersion in graphene. We demonstrate two approaches to propagate uncertainty in the potential energy and atomic forces to predicted properties. In addition, we show that DUNN uncertainty estimates can be used to detect configurations outside the training set, and in some cases, can serve as a predictor for the accuracy of a calculation.
more »
« less
Neural network potentials for reactive chemistry: CASPT2 quality potential energy surfaces for bond breaking
Neural Network potentials are developed which accurately make and break bonds for use in molecular simulations. We report a neural network potential that can describe the potential energy surface for carbon–carbon bond dissociation with less than 1 kcal mol−1 error compared to complete active space second-order perturbation theory (CASPT2), and maintains this accuracy for both the minimum energy path and molecular dynamic calculations up to 2000 K. We utilize a transfer learning algorithm to develop neural network potentials to generate potential energy surfaces; this method aims to use the minimum amount of CASPT2 data on small systems to train neural network potentials while maintaining excellent transferability to larger systems. First, we generate homolytic carbon–carbon bond dissociation data of small size alkanes with density functional theory (DFT) energies to train the potentials to accurately predict bond dissociation at the DFT level. Then, using transfer learning, we retrained the neural network potential to the CASPT2 level of accuracy. We demonstrate that the neural network potential only requires bond dissociation data of a few small alkanes to accurately predict bond dissociation energy in larger alkanes. We then perform additional training on molecular dynamic simulations to refine our neural network potentials to obtain high accuracy for general use in molecular simulation. This training algorithm is generally applicable to any type of bond or any level of theory and will be useful for the generation of new reactive neural network potentials.
more »
« less
- Award ID(s):
- 1945525
- PAR ID:
- 10503668
- Publisher / Repository:
- Digital Discovery
- Date Published:
- Journal Name:
- Digital Discovery
- Volume:
- 2
- Issue:
- 4
- ISSN:
- 2635-098X
- Page Range / eLocation ID:
- 1058 to 1069
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
A new reference state for density functional theory (DFT), termed the independent atom ansatz, is introduced in this work. This ansatz allows for the formally exact representation of electron density in terms of atom-localized orbitals. Self-consistent equations for such states are derived in general and asymptotic forms. The resultant total energy functional is found to closely resemble tight-binding theory. The independent atom ansatz facilitates partial cancellation of inter-atomic electron–electron and nucleus–electron interactions, which allows for the derivation of analytical tight-binding Hamiltonian matrix elements in a weak interaction limit. The formalism provides energy decomposition and charge analyses at no additional cost and links tight-binding, localized orbital, and electronegativity concepts. Numerical accuracy of the total energy functional has been previously reported for hydrogenic systems [Mironenko, J. Phys. Chem. A 127, 7836 (2023)] and is demonstrated here for He2, Li2, Be2, B2, N2, O2, F2, and Ne2. The method accurately reproduces the shapes of potential energy curves, capturing large-basis CCSD(T)-level bond lengths and bond dissociation energies for N2, O2, and F2 using only a minimal basis set. It outperforms both CCSD(T) and some mainstream approximate restricted Kohn–Sham DFT functionals in describing bond dissociation behavior away from equilibrium geometries.more » « less
-
Abstract Vibrational spectroscopy enables critical insight into the structural and dynamic properties of molecules. Presently, the majority of theoretical approaches to spectroscopy employ wavefunction‐basedab initioor density functional methods that rely on the harmonic approximation. This approximation breaks down for large molecules with strongly anharmonic bonds or for molecules with large internuclear separations. An alternative to these methods involves generating molecular anharmonic potential energy surfaces (potentials) and using them to extrapolate the vibrational frequencies. This study examines the efficacy of density functional theory (DFT) and the correlation consistent Composite Approach (ccCA) in generating anharmonic frequencies from potentials of small main group molecules. Vibrational self‐consistent field Theory (VSCF) and post‐VSCF methods were used to calculate the fundamental frequencies of these molecules from their potentials. Functional choice, basis set selection, and mode‐coupling are also examined as factors in influencing accuracy. The absolute deviations for the calculated frequencies using potentials at the ccCA level of theory were lower than the potentials at the DFT level. With DFT resulting in bending modes that are better described than those of ccCA, a multilevel DFT:ccCA approach where DFT potentials are used for single vibrational mode potentials and ccCA is used for vibrational mode‐mode couplings can be utilized for larger polyatomic systems. The frequencies obtained with this multilevel approach using VCIPSI‐PT2 were closer to experimental frequencies than the scaled harmonic frequencies, indicating the success of utilizing post‐VSCF methods to generate more accurate representations of computed infrared spectra.more » « less
-
Neural network potentials (NNPs) trained against density functional theory (DFT) are capable of reproducing the potential energy surface at a fraction of the computational cost. However, most NNP implementations focus on energy and forces. In this work, we modified the NNP model introduced by Behler and Parrinello to predict Fermi energy, band edges, and partial density of states of Cu 2 O. Our NNP can reproduce the DFT potential energy surface and properties at a fraction of the computational cost. We used our NNP to perform molecular dynamics (MD) simulations and validated the predicted properties against DFT calculations. Our model achieved a root mean squared error of 16 meV for the energy prediction. Furthermore, we show that the standard deviation of the energies predicted by the ensemble of training snapshots can be used to estimate the uncertainty in the predictions. This allows us to switch from the NNP to DFT on-the-fly during the MD simulation to evaluate the forces when the uncertainty is high.more » « less
-
Environmental barrier coatings (EBCs) are an enabling technology for silicon carbide (SiC)-based ceramic matrix composites (CMCs) in extreme environments such as gas turbine engines. However, the development of new coating systems is hindered by the large design space and difficulty in predicting the properties for these materials. Density Functional Theory (DFT) has successfully been used to model and predict some thermodynamic and thermo-mechanical properties of high-temperature ceramics for EBCs, although these calculations are challenging due to their high computational costs. In this work, we use machine learning to train a deep neural network potential (DNP) for Y2Si2O7, which is then applied to calculate the thermodynamic and thermo-mechanical properties at near-DFT accuracy much faster and using less computational resources than DFT. We use this DNP to predict the phonon-based thermodynamic properties of Y2Si2O7 with good agreement to DFT and experiments. We also utilize the DNP to calculate the anisotropic, lattice direction-dependent coefficients of thermal expansion (CTEs) for Y2Si2O7. Molecular dynamics trajectories using the DNP correctly demonstrate the accurate prediction of the anisotropy of the CTE in good agreement with the diffraction experiments. In the future, this DNP could be applied to accelerate additional property calculations for Y2Si2O7 compared to DFT or experiments.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1039/D3DD00051F
