Depth dependence of vacancy formation energy at (100), (110), and (111) Al surfaces: A first-principles study

S. S. Gupta, M. A. Van Huis, M. H F Sluiter, M. Dijkstra

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Vacancy defects are known to play an important role in the structural and chemical properties of metallic and semiconductor nanoparticles. Here, we investigate the likelihood of vacancy formation at the surface, in the subsurfaces, and in the interior of a model system of Al nanocrystals. The depth dependence of the vacancy formation energy (VFE) in 14-17 layered low-indexed surfaces of aluminium is studied using LDA, PBE, and PBEsol exchange-correlation functionals. Within a depth of two subsurface layers, the functionals make a transition from a similar description of surfaces to the differences in VFEs observed in bulk Al. The VFE converges to the bulk value within 0.01 eV beyond a maximum depth of 3-6 atomic layers, depending on the crystallographic surface plane. We find that the different convergence behaviors are related to the relaxations of atomic planes, normal to the surface, which in turn depend on the packing density of these surfaces. For the (111) subsurfaces, surprisingly, the defect formation energies are found to be higher than that of bulk Al, which is related to the hindered relaxations in its close-packed atomic planes. Although our results predict considerably lower VFE for the topmost layers of all the surfaces, the likelihood of forming a vacancy in the immediate subsurfaces of multifaceted Al nanoparticles is predicted to be lower than in bulk Al, which is in contrast to expectation.

Original languageEnglish
Article number085432
JournalPhysical review. B, Condensed matter and materials physics
Volume93
Issue number8
DOIs
Publication statusPublished - 19 Feb 2016

Fingerprint

Dive into the research topics of 'Depth dependence of vacancy formation energy at (100), (110), and (111) Al surfaces: A first-principles study'. Together they form a unique fingerprint.

Cite this