Microstructure of a Cu film grown on bcc Ta (100) by large-scale molecular-dynamics simulations

Journal Article (2010)
Copyright
© 2010 The Author(s); American Physical Society
To reference this document use:
https://doi.org/doi:10.1103/PhysRevB.81.045410
More Info
expand_more
Publication Year
2010
Copyright
© 2010 The Author(s); American Physical Society
Related content
DOI:
https://doi.org/doi:10.1103/PhysRevB.81.045410
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Molecular-dynamics simulations using embedded atom method potentials were carried out to study the growth and subsequent annealing of a 6.3 monolayer thick Cu film on a bcc-Ta (100) substrate of a very large area (100×100?nm2). The purpose is to obtain, for this typical example of a crystallographically incompatible system, atomic-level insight into the microstructural evolution of the film/substrate interface and the film itself, with the limiting influence of in-plane boundary conditions kept to a minimum. It is found that the first Cu plane grows heteroepitaxially on the bcc-Ta (100) surface. The second Cu plane is the most interesting plane; it grows in the form of a pattern of 26-atom misfit supercells which after prolonged film deposition relax energetically, by breaking up in groups and forming 30?Å wide in-plane island strips separated by fcc (1/6)(211) vectors. The distorted monolayer represents a way of enabling epitaxy between very different crystal structures without introducing misfit dislocations. The third and higher Cu planes are fcc (111) planes. The Cu film is polycrystalline, starting from the second plane up, with two different in-plane crystal orientations. The influence of a one-monolayer Ta terrace on the substrate is minimal. Upon annealing the film at elevated temperatures, a tendency toward island coarsening and agglomeration is observed, but a full three-dimensional island formation such as found experimentally is not observed. Compared to an earlier simulation on a smaller substrate area (18×18?nm2), the current simulations have produced a much richer microstructure. The present results therefore warn against the use of too small simulation domains when complex strain fields can be expected.

Files

Lazic_2010.pdf
(pdf | 2.92 Mb)
License info not available