Publications

Caccin, M., Li, Z., Kermode, J. R. & De Vita, A. A framework for machine-learning-augmented multiscale atomistic simulations on parallel supercomputers. Int. J. Quantum Chem. (2015). doi:10.1002/qua.24952

Recent advances in quantum mechanical (QM)-based molecular dynamics (MD) simulations have used machine-learning (ML) to predict, rather than recalculate, QM-accurate forces in atomic configurations sufficiently similar to previously encountered ones. Here, we discuss how ML approaches can be deployed within large-scale QM/MM materials simulations on massively parallel supercomputers, making QM zones of ≳1000 atoms routinely attainable. We argue that the ML approach allows computational effort to be concentrated on the most chemically active subregions of the QM zone, significantly improving the overall efficiency of the simulation. We thus propose a novel method to partition large QM regions into multiple subregions, which can be computed in parallel to achieve optimal scaling. Then we review a recently proposed QM/ML MD scheme (Z. Li, J.R. Kermode, A. De Vita Phys. Rev. Lett., 2015, 114, 096405), discussing how this could be efficiently combined with QM-zone partitioning.

Zhenwei Li, James R. Kermode and Alessandro De Vita, Phys. Rev. Lett. 114, 096405 (2015)
doi:10.1103/PhysRevLett.114.096405

We present a molecular dynamics scheme which combines first-principles and machine-learning (ML) techniques in a single information-efficient approach. Forces on atoms are either predicted by Bayesian inference or, if necessary, computed by on-the-fly quantum-mechanical (QM) calculations and added to a growing ML database, whose completeness is, thus, never required. As a result, the scheme is accurate and general, while progressively fewer QM calls are needed when a new chemical process is encountered for the second and subsequent times, as demonstrated by tests on crystalline and molten silicon.

N. Mousseau, L.K. Béland, P. Brommer, F. El-Mellouhi, J.-F. Joly, G.K. N'Tsouaglo, O. Restrepo, M. Trochet, Comp. Mater. Sci. 100 B, 111–123 (2015), doi:10.1016/j.commatsci.2014.11.047

The properties of materials, even at the atomic level, evolve on macroscopic time scales. Following this evolution through simulation has been a challenge for many years. For lattice-based activated diffusion, kinetic Monte Carlo has turned out to be an almost perfect solution. Various accelerated molecular dynamical schemes, for their part, have allowed the study on long time scale of relatively simple systems. There is still a need, however, for methods able to handle complex materials such as alloys and disordered systems. Here, we review the kinetic Activation–Relaxation Technique (k-ART), one of a handful of off-lattice kinetic Monte Carlo methods, with on-the-fly cataloging, that have been proposed in the last few years.

E. Bitzek, J. R. Kermode and P. Gumbsch, Atomistic aspects of fracture, Int. J. Fract. (2015), doi: 10.1007/s10704-015-9988-2

Any fracture process ultimately involves the rupture of atomic bonds. Processes at the atomic scale therefore critically influence the toughness and overall fracture behavior of materials. Atomistic simulation methods including large-scale molecular dynamics simulations with classical potentials, density functional theory calculations and advanced concurrent multiscale methods have led to new insights e.g. on the role of bond trapping, dynamic effects, crack- microstructure interactions and chemical aspects on the fracture toughness and crack propagation patterns in metals and ceramics. This review focuses on atomistic aspects of fracture in crystalline materials where significant advances have been achieved over the last ten years and provides an outlook on future perspectives for atomistic modelling of fracture.

M. Aldegunde and K. Kalna, "Energy conserving, self-force free Monte Carlo simulations of semiconductor devices on unstructured meshes", Comput. Phys. Commun. 189, pp. 31-36 (2015), doi:10.1016/j.cpc.2014.11.020

Unphysical self-forces resulting from the particle–mesh coupling occur when ensemble Monte Carlo simulations of semiconductor devices use an unstructured mesh to describe device geometry. We report on the development of a correction to the driving electric field on arbitrary meshes and show that self-forces can be virtually eliminated on a finite element mesh at a small additional computational cost. The developed methodology is included into a self-consistent 3D finite element Monte Carlo device simulator. We show the efficiency of the method simulating an isolated particle and obtaining kinetic energy conservation down to a magnitude of 10⁻¹⁰ $\mathrm{}{\mathrm{}}^{\mathrm{}}$ meV. The methodology is later applied to a FinFET simulation to show what impact can be expected from the self-forces using traditional electric field interpolation strategies. We find that for a large enough ensemble of particles, the impact of self-forces on the final I_D–V_G is almost negligible.

Anke Peguiron, Lucio Colombi Ciacchi, Alessandro De Vita, James R. Kermode and Gianpietro Moras, Accuracy of buffered-force QM/MM simulations of silica, J. Chem. Phys., 142, 064116 (2015), doi: 10.1063/1.4907786

We report comparisons between energy-based quantum mechanics/molecular mechanics (QM/MM) and buffered force-based QM/MM simulations in silica. Local quantities—such as density of states, charges, forces, and geometries—calculated with both QM/MM approaches are compared to the results of full QM simulations. We find the length scale over which forces computed using a finite QM region converge to reference values obtained in full quantum-mechanical calculations is ∼10 Å rather than the ∼5 Å previously reported for covalent materialssuch as silicon. Electrostatic embedding of the QM region in the surrounding classical point charges gives only a minor contribution to the force convergence. While the energy-based approach provides accurate results in geometry optimizations of point defects, we find that the removal of large force errors at the QM/MM boundary provided by the buffered force-based scheme is necessary for accurate constrained geometry optimizations where Si–O bonds are elongated and for finite-temperature molecular dynamics simulations of crack propagation. Moreover, the buffered approach allows for more flexibility, since special-purpose QM/MM coupling terms that link QM and MM atoms are not required and the region that is treated at the QM level can be adaptively redefined during the course of a dynamical simulation.

Peter Brommer and David Quigley: Automated effective band structures for defective and mismatched supercells. J. Phys.: Condens. Matter 26 485501 (2014). doi: 10.1088/0953-8984/26/48/485501

In plane-wave density functional theory codes, defects and incommensurate structures are usually represented in supercells. However, interpretation of E versus k band structures is most effective within the primitive cell, where comparison to ideal structures and spectroscopy experiments are most natural. Popescu and Zunger recently described a method to derive effective band structures (EBS) from supercell calculations in the context of random alloys. In this paper, we present bs_sc2pc, an implementation of this method in the CASTEP code, which generates an EBS using the structural data of the supercell and the underlying primitive cell with symmetry considerations handled automatically. We demonstrate the functionality of our implementation in three test cases illustrating the efficacy of this scheme for capturing the effect of vacancies, substitutions and lattice mismatch on effective primitive cell band structures.

Peter Brommer, Laurent Karim Béland, Jean-François Joly, and Normand Mousseau: Understanding long-time vacancy aggregation in iron: A kinetic activation-relaxation technique study. Physical Review B 90, 134109 (2014). doi:10.1103/PhysRevB.90.134109