Orbital-free density functional theory study of amorphous Li–Si alloys and introduction of a simple density decomposition formalism

Abstract

We propose a simple density decomposition formalism within orbital-free (OF) density functional theory (DFT) based on the Wang-Govind-Carter-decomposition (WGCD) kinetic energy density functional (KEDF). The resulting simple-WGCD (sWGCD) KEDF provides efficient density optimization, full cell relaxation, reasonable bulk properties for various materials compared to both the original OFDFT-WGCD and the Kohn-Sham (KS) DFT values, and has various numerical benefits including more stable convergence and lower computational cost (twice as fast as the WGCD KEDF).We also study amorphous (a-) Li-Si alloys with KSDFT and OFDFT using the HuangCarter (HC), WGCD, and sWGCD KEDFs. The a-Li-Si alloy samples are prepared with the anneal-and-quench method using NVT molecular dynamics simulations. We report structural properties, equilibrium volumes, bulk moduli, and alloy formation energies for each a-alloy. The HC, WGCD, and sWGCD KEDFs within OFDFT all predict accurate equilibrium volumes compared against KSDFT benchmarks. The HC KEDF bulk moduli 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2 agree with KSDFT benchmarks whereas the WGCD/sWGCD KEDFs generally overestimate the bulk moduli, especially for alloys with low Li concentrations. All three KEDFs show limited ability to predict alloy formation energies, which indicates the lack of transferability of these KEDFs among such systems and motivates future developments in OFDFT and KEDF formalisms.

Implanted neural network potentials: Application to Li-Si alloys

“…The complex nature of the Si-Li interaction in Li x Si, with x = 0 − 4.25, which evolves from pure covalent at x = 0 to pure metallic at x > 2.5, makes this challenge even more daunting. This is confirmed by the recent work of Xia and Carter 6 , who found that methods like kineticenergy-functional approaches show limited ability to predict alloy formation energies for the amorphous Si-Li system, which indicates the limitations in transferability of these functionals for such complex physical systems.…”

Nonlocal kinetic energy functionals by functional integration

“…However, it underestimates the electron density in bonding regions of CD Silicon and III-V semiconductors, it is less accurate than WT for metals, and it is substantially more computationally expensive than any other KEDFs. 10,22 Density decomposition schemes that combine nonlocal with other functionals have found applicability to molecules, materials, and alloys 12,22,23 and thus appear to be an interesting avenue of research.…”

Orbital-Free Density Functional Theory: An Attractive Electronic Structure Method for Large-Scale First-Principles Simulations

“…The switching function must be chosen judiciously to avoid overcomplicated structure that makes implementation itself overly complicated. 169 One issue of complication is whether to make F(r) a function independent of the electron density. In ref 78 that was not the case.…”

“…To our knowledge, the effects of that discrepancy have not been studied. The Carter group applied the density decomposition scheme to transition metals, 403 amorphous Li−Si alloys, 169 and covalently bonded molecules and materials. 78 The results, however, indicated that while the decomposition scheme can improve on the conventional OFDFT methodology (i.e., use of a single KEDF), it does not lead to a substantive, quantitative improvement.…”

“…In Section , we discuss a nonexhaustive selection of illustrative applications. Here we simply note that recent applications with one- and two-point KEDFs include warm dense matter − and molten and zero-temperature metals and semiconductors (bulk, ,− surfaces and other low-dimensional systems, and clusters ,, ).…”