Burrowing of cobalt nanoclusters in copper

https://doi.org/10.1016/S0168-583X(03)00721-3Get rights and content

Abstract

We have examined the burrowing of Co nanoclusters deposited on a Cu surface with classical molecular dynamics simulations. Our simulations show that two different mechanisms of burrowing are dominating depending on the orientation of the deposited nanocluster. The burrowing mechanism for epitaxial nanoclusters was found to be vacancy migration along the Co–Cu interface and for non-aligned nanoclusters the mechanism is disordered motion of atoms along the Co–Cu interface. The re-alignment effect of the Co nanoclusters found in experiments was found to be due to a collective rotational movement of the whole cluster during the burrowing process. We present our results and compare these to previous experimental and simulated results.

Introduction

Experiments performed by Zimmerman et al. show that when a system consisting of Co nanoclusters which have been deposited on a clean Cu surface is heated up to 600 K the nanoclusters burrow into the substrate [1]. Transmission electron microscopy measurements performed on such samples showed that the fully burrowed nanoclusters are aligned with the substrate during the burrowing process. The papers covering this experiment does, however, not clearly establish the atomistic mechanism of the burrowing.

Knowledge about what kind of mechanisms are involved and how they apply on different cluster configurations could play a significant role in the optimization of the manufacturing process of nanostructured films of this nature. How and when the nanocluster aligns with the substrate lattice is certainly also a question of interest.

Burrowing of Co nanoclusters on Cu has previously been examined by means of quasi-ab initio MD simulations [2], [3]. These works examine very small clusters, consisting of tens of Co atoms, which are burrowed into the Cu substrate. The main conclusion of these works is that the magnetic properties of the nanocluster would play a significant role in the burrowing process and that the cluster completely loses its initial shape. As we show in this paper, nanoclusters larger than ∼1 nm in radius will show a different behaviour due to the higher structural integrity.

Section snippets

Computational details

For the simulations we used the classical molecular dynamics method [4], with substrate sizes ranging from 1000 to 140,000 atoms and the clustersizes of 500–6000 atoms. Periodic boundaries were used in all directions except the direction of the surface normal, the three bottom layers were softly quenched to the ambient temperature. The Cu surface normal was always in the 〈001〉 direction.

For atom interaction we used the embedded-atom-method (EAM) by Pasianot and Savino for Co–Co interaction [5],

Results and discussion

During the simulations of the deposition of the nanoclusters we found that the most interesting results from the burrowing point of view is observed for clusters of a diameter larger than 4 nm. Smaller nanoclusters will undergo full contact epitaxy [9] on deposition. We simulated nanoclusters of 2 nm diameter which confirmed these results. Even smaller nanoclusters have been examined in previous work [2], [3] and are discussed in Section 1. The clusters we refer to in this section have a

Conclusion

In this paper we have examined the burrowing of a Co nanocluster on a Cu substrate using classical molecular dynamics simulations. If the cluster has landed on the substrate forming an epitaxial configuration, the burrowing mechanism is found to be vacancy migration along the Co–Cu interface. In the case of the cluster landing on an edge or a corner the burrowing mechanism is disordered motion of atoms along the Co–Cu interface. The burrowing process in this case is considerably swifter than

Acknowledgements

The research was supported by the Academy of Finland under project no. 73722 and by the University of Helsinki under the NAPROMA project. Grants of computer time from the Center for Scientific Computing in Espoo, Finland are gratefully acknowledged. We also want to thank Yinon Ashkenazy and Bob Averback for helpful discussions.

References (12)

  • C. Zimmermann et al.

    Phys. Rev. B

    (2001)
  • J. Izguierdo et al.

    Phys. Rev. B

    (2001)
  • V.S. Stepanyuk et al.

    Phys. Rev. B

    (2001)
  • M.P. Allen et al.

    Computer Simulation of Liquids

    (1989)
  • R. Pasianot et al.

    Phys. Rev. B

    (1992)
  • S.M. Foiles

    Phys. Rev. B

    (1985)
There are more references available in the full text version of this article.
View full text