MLIP-3: Active learning on atomic environments with moment tensor potentials.

Nowadays, academic research relies not only on sharing with the academic community the scientific results obtained by research groups while studying certain phenomena but also on sharing computer codes developed within the community. In the field of atomistic modeling, these were software packages for classical atomistic modeling, and later for quantum-mechanical modeling; currently, with the fast growth of the field of machine-learning potentials, the packages implement such potentials. In this paper, we present the MLIP-3 package for constructing moment tensor potentials and performing their active training. This package builds on the MLIP-2 package [Novikov et al., "The MLIP package: moment tensor potentials with MPI and active learning," Mach. Learn.: Sci. Technol., 2(2), 025002 (2020)], however, with a number of improvements, including active learning on atomic neighborhoods of a possibly large atomistic simulation.

Nanohardness from First Principles with Active Learning on Atomic Environments.

We propose a methodology for the calculation of nanohardness by atomistic simulations of nanoindentation. The methodology is enabled by machine-learning interatomic potentials fitted on the fly to quantum-mechanical calculations of local fragments of the large nanoindentation simulation. We test our methodology by calculating nanohardness, as a function of load and crystallographic orientation of the surface, of diamond, AlN, SiC, BC2N, and Si and comparing it to the calibrated values of the macro- and microhardness. The observed agreement between the computational and experimental results from the literature provides evidence that our method has sufficient predictive power to open up the possibility of designing materials with exceptional hardness directly from first principles. It will be especially valuable at the nanoscale where the experimental measurements are difficult, while empirical models fitted to macrohardness are, as a rule, inapplicable.

First-Principles Multiscale Modeling of Mechanical Properties in Graphene/Borophene Heterostructures Empowered by Machine-Learning Interatomic Potentials.

Density functional theory calculations are robust tools to explore the mechanical properties of pristine structures at their ground state but become exceedingly expensive for large systems at finite temperatures. Classical molecular dynamics (CMD) simulations offer the possibility to study larger systems at elevated temperatures, but they require accurate interatomic potentials. Herein the authors propose the concept of first-principles multiscale modeling of mechanical properties, where ab initio level of accuracy is hierarchically bridged to explore the mechanical/failure response of macroscopic systems. It is demonstrated that machine-learning interatomic potentials (MLIPs) fitted to ab initio datasets play a pivotal role in achieving this goal. To practically illustrate this novel possibility, the mechanical/failure response of graphene/borophene coplanar heterostructures is examined. It is shown that MLIPs conveniently outperform popular CMD models for graphene and borophene and they can evaluate the mechanical properties of pristine and heterostructure phases at room temperature. Based on the information provided by the MLIP-based CMD, continuum models of heterostructures using the finite element method can be constructed. The study highlights that MLIPs were the missing block for conducting first-principles multiscale modeling, and their employment empowers a straightforward route to bridge ab initio level accuracy and flexibility to explore the mechanical/failure response of nanostructures at continuum scale.

Young's Modulus and Tensile Strength of Ti3C2 MXene Nanosheets As Revealed by In Situ TEM Probing, AFM Nanomechanical Mapping, and Theoretical Calculations.

Two-dimensional transition metal carbides, that is, MXenes and especially Ti3C2, attract attention due to their excellent combination of properties. Ti3C2 nanosheets could be the material of choice for future flexible electronics, energy storage, and electromechanical nanodevices. There has been limited information available on the mechanical properties of Ti3C2, which is essential for their utilization. We have fabricated Ti3C2 nanosheets and studied their mechanical properties using direct in situ tensile tests inside a transmission electron microscope, quantitative nanomechanical mapping, and theoretical calculations employing machine-learning derived potentials. Young's modulus in the direction perpendicular to the Ti3C2 basal plane was found to be 80-100 GPa. The tensile strength of Ti3C2 nanosheets reached up to 670 MPa for ∼40 nm thin nanoflakes, while a strong dependence of tensile strength on nanosheet thickness was demonstrated. Theoretical calculations allowed us to study mechanical characteristics of Ti3C2 as a function of nanosheet geometrical parameters and structural defect concentration.

Machine learning of molecular properties: Locality and active learning.

In recent years, the machine learning techniques have shown great potent1ial in various problems from a multitude of disciplines, including materials design and drug discovery. The high computational speed on the one hand and the accuracy comparable to that of density functional theory on another hand make machine learning algorithms efficient for high-throughput screening through chemical and configurational space. However, the machine learning algorithms available in the literature require large training datasets to reach the chemical accuracy and also show large errors for the so-called outliers-the out-of-sample molecules, not well-represented in the training set. In the present paper, we propose a new machine learning algorithm for predicting molecular properties that addresses these two issues: it is based on a local model of interatomic interactions providing high accuracy when trained on relatively small training sets and an active learning algorithm of optimally choosing the training set that significantly reduces the errors for the outliers. We compare our model to the other state-of-the-art algorithms from the literature on the widely used benchmark tests.