Radiation-induced segregation on dislocation loops in austenitic Fe-Cr-Ni alloys

Radiation-induced segregation of alloying elements to crystallographic defects is commonly observed in irradiated austenitic stainless steels. The interaction between solutes and radiation-induced defects changes the physical distribution of solutes and thus affects the formation and growth of defects. The change of the microstructure consequently affects the mechanical properties of the material. A qualitative and quantitative understanding of the interaction between solutes and defects is desirable to better predict the service lifetime of nuclear materials. We used atom probe tomography to measure the distribution of solutes at dislocation loops in 304L stainless steel, irradiated with 2 MeV protons up to 1.5 displacements per atom at 373 and 633 K. No segregation at dislocation loops was found in samples irradiated at 373 K, whereas Ni and Si enrichment and Cr depletion were detected at dislocation loops irradiated at 633 K. The experimentally observed perfect and faulted dislocation loops in vacancy and interstitial types were reproduced by molecular dynamics (MD). A hybrid MD/Monte Carlo method was used to predict the redistribution of alloying atoms at all possible types of dislocation loops in face-centered cubic Fe-Cr-Ni alloys at the same irradiated temperatures (373 and 633 K). The simulations show that, at both temperatures, Cr clusters were formed and distributed randomly, and Ni atoms enriched or depleted at interstitial or vacancy dislocation loops, respectively. The change of solute concentration reaches the highest at the edge of the loop. Ni profiles exhibit characteristic behavior in terms of the stress field of the loops: tension inside of vacancy loops showing depletion of Ni atoms compared with compression inside of interstitial loops showing enrichment of Ni atoms. In addition, the stress field is reduced after solute redistribution. The absence of alloying segregation observed in experiments at a lower temperature (373 K) is explained by a rate theory model: Low-temperature irradiation requires significantly longer irradiation time to see the same amount of segregation as at high temperatures because of the extremely low diffusion of vacancies at low temperatures.

Figure 1

Atom probe tomography (APT) analysis of proton-irradiated 304L stainless steel with 2 MeV protons ($\sim 1.5\mathrm{dpa}$) at (a) 373 K and (b) 633 K with isosurfaces for 4.5 at. % Si reveal segregation regions, and a marked region indicates a single dislocation loop that is isolated in (c). (d) 3 at. % Si isosurface to position the region of interest (ROI). (e) Cylindrical ROI with 4 nm thickness. (f) Elemental concentration from the ROI.

Figure 2

Measured at 373 K. (1, 2, 3, 4, 5) are displaced atoms colored by common neighbor analysis (CNA), strain energy map, and concentration map of Fe/Ni/Cr for (a) vacancy-type perfect loop, (b) interstitial-type perfect loop, (c) vacancy-type Frank loop, and (d) interstitial-type Frank loop.

Figure 3

Measured at 633 K. (1, 2, 3, 4, 5) are displaced atoms colored by common neighbor analysis (CNA), strain energy map, and concentration map of Fe/Ni/Cr for (a) vacancy-type perfect loop, (b) interstitial-type perfect loop, (c) vacancy-type Frank loop, and (d) interstitial-type Frank loop.

Figure 4

The concentration profile of each element and the strain energy as a function of the distance from the loop center for (a) and (e) vacancy-type perfect loops, (b) and (f) interstitial-type perfect loops, (c) and (g) vacancy-type Frank loops, and (d) and (h) interstitial-type Frank loops. (a)–(d) are measured at 373 K, and (e)–(h) are measured at 633 K. The loop edge is highlighted by a black dashed line at 6 nm.

Figure 5

After annealing at 373 K, under the condition of (a)–(d) no alloying segregation and (e)–(h) with alloying segregation: the radial stress component of stress ($S_{rr}$) for (a) and (e) a vacancy-type perfect loop, (b) and (f) an interstitial-type perfect loop, (c) and (g) a vacancy-type Frank loop, and (d) and (h) an interstitial Frank loop.

Figure 6

After annealing at 633 K, under the condition of (a)–(d) no alloying segregation and (e)–(h) with alloying segregation: the radial stress component of stress ($S_{rr}$) for (a) and (e) a vacancy-type perfect loop, (b) and (f) an interstitial-type perfect loop, (c) and (g) a vacancy-type Frank loop, and (d) and (h) an interstitial Frank loop.

Figure 7

The averaged radial stress component ($S_{rr}$) after annealing at 373 K (red data) and 633 K (blue data), under the condition of no alloying segregation (open circles) and with alloying segregation (closed circles): (a) vacancy-type perfect loops, (b) interstitial-type perfect loops, (c) vacancy-type Frank loops, and (d) interstitial-type Frank loops. The loop edge is at 6 nm.

Figure 8

After alloying redistribution, Ni concentration (red) and the averaged radial stress (black) as a function of the distance from the loop center for (a) and (e) a vacancy-type perfect loop, (b) and (f) an interstitial-type perfect loop, (c) and (g) a vacancy-type Frank loop, and (d) and (h) an interstitial-type Frank loop. (a)–(d) are measured at 373 K, and (e)–(h) are measured at 633 K. The loop edge is at 6 nm.

Figure 9

The representative concentration profiles of (a) Cr, (b) Ni, and (c) Si across an ideal sink at 373 K (upper) and 633 K (bottom) based on the prediction of the model (1.5 dpa at an irradiation rate of ${10}^{-5}\mathrm{dpa}{\mathrm{s}}^{-1}$). Purple square and green cross markers represent atom probe tomography (APT) measurements for grain boundaries [45, 72] and the dislocation loop, respectively.

Figure 10

Total segregation of the representative concentration as a function of irradiation damage at (a)–(c) 373 K and (d)–(f) 633 K at three irradiation rates: ${10}^{-5}$, ${10}^{-4}$, and ${10}^{-3}\mathrm{dpa}{\mathrm{s}}^{-1}$.