Atomic-level computer simulations of copper-vacancy clusters in alpha-Fe

Reactor pressure vessel (RPV) steels consist of polycrystalline alpha-Fe with different alloying elements, e.g. nickel, and different impurities, e.g. copper. During the operation of a nuclear fission reactor at a temperature of about 300 °C this material is continuously irradiated by neutrons and both vacancies and self-interstitials are formed. The presence of point defects enhances the diffusion of impurities and leads to precipitation if their solid solubility is small. The precipitates may not only contain the impurity species but also point defects since vacancies and/or self-interstitials take part in the process of clustering. Furthermore, pure vacancy and self-interstitial clusters may be formed. Copper-rich precipitates (CRP) are assumed to be the main cause of hardening and embrittlement of Cu-bearing RPV steels since these defects act as obstacles to dislocation motion within the grains of polycrystalline alpha-Fe. There is clear evidence that these nanoclusters remain small and have the bcc structure of the surrounding matrix. CRP and nanovoids have been observed by different experimental methods such as small angle neutron scattering, tomographic atom probe, positron annihilation spectroscopy, and high-resolution transmission electron microscopy. On the other hand, multiscale modeling contributes to a better understanding of the various physical processes that occur during the formation of clusters and precipitates. Rate theory is a useful and efficient tool to simulate defect evolution on realistic time and length scales. However, the values of many parameters used in rate theory, such as diffusion coefficients of mobile species and free binding energies of clusters, are not very well known from experimental investigations. Atomic-level computer simulations can provide these data.

In the present work a combination of Metropolis Monte Carlo simulations on a rigid bcc lattice and molecular dynamics simulations [1,2] is applied in order to determine the most stable configuration of numerous CunVm clusters. In all calculations the most recent Fe-Cu interatomic potential by Pasianot and Malerba [3] is used. Present investigations do not only yield formation energies of the most stable clusters but also the corresponding binding energies. The results are compared with literature data [1,2] obtained by the potentials of Ackland-Bacon [4] and Ludwig-Farkas [5]. The configuration of some clusters containing both copper and vacancies are visualized and their morphology is compared with the interpretation of recent experimental investigations.

References
[1] A. Takahashi, N. Soneda, S. Ishino, and G. Yagawa, Phys. Rev. B 67, 024104 (2003).
[2] D. Kulikov, L. Malerba, and M. Hou, Philos. Mag. 86, 141 (2006).
[3] R. C. Pasianot and L. Malerba, J. Nucl. Mater. 360, 118 (2007).
[4] G. J. Ackland, D. J. Bacon, A .F. Calder, and T. Harry, Philos. Mag. A 75, 713 (1997).
[5] M. Ludwig, D. Farkas, D. Pedraza, and S. Schmauder, Modelling Simul. Mater. Sci. Eng.
6, 19 (1998).