Accelerating crystal structure prediction by machine-learning interatomic potentials with active learning
Authors:
Evgeny V. Podryabinkin,
Evgeny V. Tikhonov,
Alexander V. Shapeev,
Artem R. Oganov
Abstract:
In this letter we propose a new methodology for crystal structure prediction, which is based on the evolutionary algorithm USPEX and the machine-learning interatomic potentials actively learning on-the-fly. Our methodology allows for an automated construction of an interatomic interaction model from scratch replacing the expensive DFT with a speedup of several orders of magnitude. Predicted low-en…
▽ More
In this letter we propose a new methodology for crystal structure prediction, which is based on the evolutionary algorithm USPEX and the machine-learning interatomic potentials actively learning on-the-fly. Our methodology allows for an automated construction of an interatomic interaction model from scratch replacing the expensive DFT with a speedup of several orders of magnitude. Predicted low-energy structures are then tested on DFT, ensuring that our machine-learning model does not introduce any prediction error. We tested our methodology on a problem of prediction of carbon allotropes, dense sodium structures and boron allotropes including those which have more than 100 atoms in the primitive cell. All the the main allotropes have been reproduced and a new 54-atom structure of boron have been found at very modest computational efforts.
△ Less
Submitted 26 February, 2019; v1 submitted 21 February, 2018;
originally announced February 2018.
Electron correlations and silicon nanocluster energetics
Authors:
N. L. Matsko,
Yu. A. Uspenskii,
E. V. Tikhonov,
V. S. Baturin,
S. V. Lepeshkin
Abstract:
The first-principle prediction of nanocluster stable structure is often hampered by the existence of many isomer configurations with energies close to the ground state. This fact attaches additional importance to many-electron effects going beyond density functional theory (DFT), because their contributions may change a subtle energy order of competitive structures. To analyze this problem, we con…
▽ More
The first-principle prediction of nanocluster stable structure is often hampered by the existence of many isomer configurations with energies close to the ground state. This fact attaches additional importance to many-electron effects going beyond density functional theory (DFT), because their contributions may change a subtle energy order of competitive structures. To analyze this problem, we consider, as an example, the energetics of silicon nanoclusters passivated by hydrogen Si$_{10}$H$_{2n}$ with $0\le n\le 11$, the structure of which varies with passivation from compact to loose-packed, similar to branching polymers. Our calculations performed by the DFT, hybrid functionals and Hartree-Fock (H-F) methods, as well as by the GW approximation (GWA), confirm a considerable sensitivity of structure prediction and isomer energy ordering to many-electron effects and show some results which may be obtained with the methods less computationally demanding than the GWA.
△ Less
Submitted 5 March, 2016;
originally announced March 2016.
