On the formation of mixed vacancy-copper clusters in neutron-irradiated Fe-Cu alloys

To investigate the irradiation-induced degradation of the mechanical properties of Cu-containing reactor pressure vessel steels, a set of model alloys was fabricated and neutron-irradiated up to four different doses, 0.026, 0.052, 0.10, and 0.19 dpa, using the same neutron flux (140 x10^-9 dpa/s). A series of complementary experimental techniques have been applied to these samples, including PAS and SANS. We report here the analyses the of small-angle neutron scattering (SANS) experiments for the two binary Fe-Cu model alloys: Fe-0.3%Cu and Fe-0.1%Cu. Size-resolved cluster distributions functions were extracted from the scattering data and characterized by the peak radius of the distribution and the total volume fraction of the detectable precipitate clusters. A significant difference in the results for the Cu-rich and the Cu-poor model alloys could be observed. The aim of the present study is to understand the observed trends on the base of rate theoretical (RT) cluster dynamic simulations.
To this end, the time evolution of the precipitate clusters under the given irradiation conditions were simulated using standard rate theory models for pure Cu clusters. While reasonable model parameters could be found to describe the experimental data of the Fe-0.3%Cu alloy, it was impossible to reproduce the data for the Cu-poor model alloy this way. By linking the rate theory models to classical nucleation theory it could be demonstrated that no RT model could do so – no matter how involved the point defect dynamics – as long as the Cu precipitation process itself is not changed. Different mechanisms can be thought off to result in such modifications, among them heterogeneous nucleation with nanovoids as nucleus for further Cu growth, or the formation of mixed vacancy-copper aggregates throughout.
We therefore augmented our rate theory model to allow for the additional absorption of vacancies by the Cu-rich precipitates and an accompanied change in the interface energy between iron matrix and mixed clusters. Adjusting the model parameters properly the entire set of data point for all four doses and both compositions could be reproduced giving confidence that the basic idea of our RT model are reasonable. In addition, it was possible to provide some estimates on the time evolution of the average composition of the precipitate clusters. They exhibit – without any re-fitting – the same trends as the experimental cluster compositions deduced from the A-ratio of the SANS signals.