Evaluation of Machine Learning Interatomic Potentials for the Properties of Gold Nanoparticles

Theoretical Prediction of Thermal Expansion Anisotropy for Y2Si2O7 Environmental Barrier Coatings Using a Deep Neural Network Potential and Comparison to Experiment

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.

Structures, Electronic, and Magnetic Properties of CoKn (n = 2-12) Clusters: A Particle Swarm Optimization Prediction Jointed with First-Principles Investigation

Transition-metal-doped clusters have long been attracting great attention due to their unique geometries and interesting physical and/or chemical properties. In this paper, the geometries of the lowest- and lower-energy CoKn (n = 2-12) clusters have been screened out using particle swarm optimization and first principles relaxation. The results show that except for CoK2 the other CoKn (n = 3-12) clusters are all three-dimensional structures, and CoK7 is the transition structure from which the lowest energy structures are cobalt atom-centered cage-like structures. The stability, the electronic structures, and the magnetic properties of CoKn clusters (n = 2-12) clusters are further investigated using the first principles method. The results show that the medium-sized clusters whose geometries are cage-like structures are more stable than smaller-sized clusters. The electronic configuration of CoKn clusters could be described as 1S1P1D according to the spherical jellium model. The main components of petal-shaped D molecular orbitals are Co-d and K-s states or Co-d and Co-s states, and the main components of sphere-like S molecular orbitals or spindle-like P molecular orbitals are K-s states or Co-s states. Co atoms give the main contribution to the total magnetic moments, and K atoms can either enhance or attenuate the total magnetic moments. CoKn (n = 5-8) clusters have relatively large magnetic moments, which has a relation to the strong Co-K bond and the large amount of charge transfer. CoK4 could be a magnetic superatom with a large magnetic moment of 5 μB.

Evaluation of Machine Learning Interatomic Potentials for Gold Nanoparticles-Transferability towards Bulk

We analyse the efficacy of machine learning (ML) interatomic potentials (IP) in modelling gold (Au) nanoparticles. We have explored the transferability of these ML models to larger systems and established simulation times and size thresholds necessary for accurate interatomic potentials. To achieve this, we compared the energies and geometries of large Au nanoclusters using VASP and LAMMPS and gained better understanding of the number of VASP simulation timesteps required to generate ML-IPs that can reproduce the structural properties. We also investigated the minimum atomic size of the training set necessary to construct ML-IPs that accurately replicate the structural properties of large Au nanoclusters, using the LAMMPS-specific heat of the Au147 icosahedral as reference. Our findings suggest that minor adjustments to a potential developed for one system can render it suitable for other systems. These results provide further insight into the development of accurate interatomic potentials for modelling Au nanoparticles through machine learning techniques.