Nickel titanium (NiTi) is a protypical shape-memory alloy used in a range of biomedical and engineering devices, but direct molecular dynamics simulations of the martensitic B19' -> B2 phase transition driving its shape-memory behavior are rare and have relied on classical force fields with limited accuracy. Here, we train four machine-learned force fields for equiatomic NiTi based on the LDA, PBE, PBEsol, and SCAN DFT functionals. The models are trained on the fly during NPT molecular dynamics, with DFT calculations and model updates performed automatically whenever the uncertainty of a local energy prediction exceeds a chosen threshold. The models achieve accuracies of 1-2 meV/atom during training and are shown to closely track DFT predictions of B2 and B19' elastic constants and phonon frequencies. Surprisingly, in large-scale molecular dynamics simulations, only the SCAN model predicts a reversible B19' -> B2 phase transition, with the LDA, PBE, and PBEsol models predicting a reversible transition to a previously uncharacterized low-volume phase, which we hypothesize to be a new stable high-pressure phase. We examine the structure of the new phase and estimate its stability on the temperature-pressure phase diagram. This work establishes an automated active learning protocol for studying displacive transformations, reveals important differences between DFT functionals that can only be detected in large-scale simulations, provides an accurate force field for NiTi, and identifies a new phase.

Coarse graining techniques play an essential role in accelerating molecular simulations of systems with large length and time scales. Theoretically grounded bottom-up models are appealing due to their thermodynamic consistency with the underlying all-atom models. In this direction, machine learning approaches hold great promise to fitting complex many-body data. However, training models may require collection of large amounts of expensive data. Moreover, quantifying trained model accuracy is challenging, especially in cases of non-trivial free energy configurations, where training data may be sparse. We demonstrate a path towards uncertainty-aware models of coarse grained free energy surfaces. Specifically, we show that principled Bayesian model uncertainty allows for efficient data collection through an on-the-fly active learning framework and opens the possibility of adaptive transfer of models across different chemical systems. Uncertainties also characterize models’ accuracy of free energy predictions, even when training is performed only on forces. This work helps pave the way towards efficient autonomous training of reliable and uncertainty aware many-body machine learned coarse grain models.

Machine learning interatomic force fields are promising for combining high computational efficiency and accuracy in modeling quantum interactions and simulating atomistic dynamics. Active learning methods have been recently developed to train force fields efficiently and automatically. Among them, Bayesian active learning utilizes principled uncertainty quantification to make data acquisition decisions. In this work, we present a general Bayesian active learning workflow, where the force field is constructed from a sparse Gaussian process regression model based on atomic cluster expansion descriptors. To circumvent the high computational cost of the sparse Gaussian process uncertainty calculation, we formulate a high-performance approximate mapping of the uncertainty and demonstrate a speedup of several orders of magnitude. We demonstrate the autonomous active learning workflow by training a Bayesian force field model for silicon carbide (SiC) polymorphs in only a few days of computer time and show that pressure-induced phase transformations are accurately captured. The resulting model exhibits close agreement with both ab initio calculations and experimental measurements, and outperforms existing empirical models on vibrational and thermal properties. The active learning workflow readily generalizes to a wide range of material systems and accelerates their computational understanding.

The activity of metal catalysts depends sensitively on dynamic structural changes that occur during operating conditions. The mechanistic understanding underlying such transformations in small Pt nanoparticles (NPs) of $\sim1-5$ nm in diameter, commonly used in hydrogenation reactions, is currently far from complete. In this study, we investigate the structural evolution of Pt NPs in the presence of hydrogen using reactive molecular dynamics (MD) simulations and X-ray spectroscopy measurements. To gain atomistic insights into adsorbate-induced structural transformation phenomena, we employ a combination of MD based on first-principles machine-learned force fields with extended X-ray absorption fine structure (EXAFS) measurements. Simulations and experiments provide complementary information, mutual validation, and interpretation. We obtain atomic-level mechanistic insights into `order-disorder' structural transformations exhibited by highly dispersed heterogeneous Pt catalysts upon exposure to hydrogen. We report the emergence of previously unknown candidate structures in the small Pt NP limit, where exposure to hydrogen leads to the appearance of a `quasi-icosahedral' intermediate symmetry, followed by the formation of `rosettes' on the NP surface. Hydrogen adsorption is found to catalyze these shape transitions by lowering their temperatures and increasing the apparent rates, revealing the dual catalytic and dynamic nature of interaction between nanoparticle and adsorbate. Our study also offers a new pathway for deciphering the reversible evolution of catalyst structure resulting from the chemisorption of reactive species, enabling the determination of active sites and improved interpretation of experimental results with atomic resolution.

Computing accurate reaction rates is a central challenge in computational chemistry and biology because of the high cost of free energy estimation with unbiased molecular dynamics. In this work, a data-driven machine learning algorithm is devised to learn collective variables with a multitask neural network, where a common upstream part reduces the high dimensionality of atomic configurations to a low dimensional latent space and separate downstream parts map the latent space to predictions of basin class labels and potential energies. The resulting latent space is shown to be an effective low-dimensional representation, capturing the reaction progress and guiding effective umbrella sampling to obtain accurate free energy landscapes. This approach is successfully applied to model systems including a 5D Müller Brown model, a 5D three-well model, the alanine dipeptide in vacuum, and an Au(110) surface reconstruction unit reaction. It enables automated dimensionality reduction for energy controlled reactions in complex systems, offers a unified and data-efficient framework that can be trained with limited data, and outperforms single-task learning approaches, including autoencoders.

Quantum-mechanically accurate reactive molecular dynamics (MD) at the scale of billions of atoms has been achieved for the heterogeneous catalytic system of H$_2$/Pt(111) using the FLARE Bayesian force field. This achievement provides accelerated time-to-solution from first principles, with Bayesian active learning enabling efficient and autonomous training of the machine learning model. The resulting model is then deployed in LAMMPS on GPUs using the Kokkos performance portability library. The Bayesian force field provides quantitative uncertainty of predictions on every atomic environment, critical for detecting configurations in large reactive simulations that are outside of the training set. Scaling benchmarks were performed using real-application MD of the H$_2$/Pt(111) heterogeneous catalysis on the Summit supercomputer, with simulations reaching 0.5 trillion atoms on 4556 GPU nodes.

Atomistic modeling of chemically reactive systems has so far relied on either expensive ab initio methods or bond-order force fields requiring arduous parametrization. Here, we describe a Bayesian active learning framework for autonomous ``on-the-fly'' training of fast and accurate reactive many-body force fields during molecular dynamics simulations. At each time step, predictive uncertainties of a sparse Gaussian process are evaluated to automatically determine whether additional ab initio training data are needed. We introduce a general method for mapping trained kernel models onto equivalent polynomial models whose prediction cost is much lower and independent of the training set size. As a demonstration, we perform direct two-phase simulations of heterogeneous H 2 turnover on the Pt(111) catalyst surface, at chemical accuracy. The model trains itself in three days and performs at twice the speed of a ReaxFF model, while maintaining much higher fidelity to DFT and excellent agreement with experiment.

Recently, machine learning (ML) has been used to address the computational cost that has been limiting ab initio molecular dynamics (AIMD). Here, we present GNNFF, a graph neural network framework to directly predict atomic forces from automatically extracted features of the local atomic environment that are translationally-invariant, but rotationally-covariant to the coordinate of the atoms. We demonstrate that GNNFF not only achieves high performance in terms of force prediction accuracy and computational speed on various materials systems, but also accurately predicts the forces of a large MD system after being trained on forces obtained from a smaller system. Finally, we use our framework to perform an MD simulation of Li 7 P 3 S 11 , a superionic conductor, and show that resulting Li diffusion coefficient is within 14% of that obtained directly from AIMD. The high performance exhibited by GNNFF can be easily generalized to study atomistic level dynamics of other material systems.

We present a way to dramatically accelerate Gaussian process models for interatomic force fields based on many-body kernels by mapping both forces and uncertainties onto functions of low-dimensional features. This allows for automated active learning of models combining near-quantum accuracy, built-in uncertainty, and constant cost of evaluation that is comparable to classical analytical models, capable of simulating millions of atoms. Using this approach, we perform large-scale molecular dynamics simulations of the stability of the stanene monolayer. We discover an unusual phase transformation mechanism of 2D stanene, where ripples lead to nucleation of bilayer defects, densification into a disordered multilayer structure, followed by formation of bulk liquid at high temperature or nucleation and growth of the 3D bcc crystal at low temperature. The presented method opens possibilities for rapid development of fast accurate uncertainty-aware models for simulating long-time large-scale dynamics of complex materials.

Machine learned force fields typically require manual construction of training sets consisting of thousands of first principles calculations, which can result in low training efficiency and unpredictable errors when applied to structures not represented in the training set of the model. This severely limits the practical application of these models in systems with dynamics governed by important rare events, such as chemical reactions and diffusion. We present an adaptive Bayesian inference method for automating the training of interpretable, low-dimensional, and multi-element interatomic force fields using structures drawn on the fly from molecular dynamics simulations. Within an active learning framework, the internal uncertainty of a Gaussian process regression model is used to decide whether to accept the model prediction or to perform a first principles calculation to augment the training set of the model. The method is applied to a range of single- and multi-element systems and shown to achieve a favorable balance of accuracy and computational efficiency, while requiring a minimal amount of ab initio training data. We provide a fully open-source implementation of our method, as well as a procedure to map trained models to computationally efficient tabulated force fields.

This dataset contains the positions and forces of a 2554-frame AIMD simulation of a 32-atom silver iodide system in the superionic phase. The Quantum Espresso input and output files are included, along with a video of the simulation.

Restructuring of interface plays a crucial role in materials science and heterogeneous catalysis. Bimetallic systems, in particular, often adopt very different composition and morphology at surfaces compared to the bulk. For the first time, we reveal a detailed atomistic picture of the long-timescale restructuring of Pd deposited on Ag, using microscopy, spectroscopy, and novel simulation methods. Encapsulation of Pd by Ag always precedes layer-by-layer dissolution of Pd, resulting in significant Ag migration out of the surface and extensive vacancy pits. These metastable structures are of vital catalytic importance, as Ag-encapsulated Pd remains much more accessible to reactants than bulk-dissolved Pd. The underlying mechanisms are uncovered by performing fast and large-scale machine-learning molecular dynamics, followed by our newly developed method for complete characterization of atomic surface restructuring events. Our approach is broadly applicable to other multimetallic systems of interest and enables the previously impractical mechanistic investigation of restructuring dynamics.

Neural network force field (NNFF) is a method for performing regression on atomic structure–force relationships, bypassing the expensive quantum mechanics calculations that prevent the execution of long ab initio quality molecular dynamics (MD) simulations. However, most NNFF methods for complex multi-element atomic systems indirectly predict atomic force vectors by exploiting just atomic structure rotation-invariant features and network-feature spatial derivatives, which are computationally expensive. Here, we show a staggered NNFF architecture that exploits both rotation-invariant and -covariant features to directly predict atomic force vectors without using spatial derivatives, and we demonstrate 2.2× NNFF–MD acceleration over a state-of-the-art C++ engine using a Python engine. This fast architecture enables us to develop NNFF for complex ternary- and quaternary-element extended systems composed of long polymer chains, amorphous oxide and surface chemical reactions. The rotation-invariant–covariant architecture described here can also directly predict complex covariant vector outputs from local environments, in other domains beyond computational material science. Neural network force fields promise to bypass the computationally expensive quantum mechanical calculations typically required to investigate complex materials, such as lithium-ion batteries. Mailoa et al. accelerate these approaches with an architecture that exploits both rotation-invariant and -covariant features separately.