Fig. 1. A cross-sectional SEM image of the SMRT IF steel sample.

Fig. 2. (a) Bright-field and (b) dark-filed cross-sectional TEM images of the top surface layer of the SMRT sample. The insert in (a) shows the corresponding SAED pattern. (c) Statistic distribution of lamellar thickness derived from dark-field TEM images. (d) In-depth distribution of lamellar thickness (or grain size) in the surface layer of the SMRT sample. At least 200 lamellae or grains were counted to determine the mean size value at each depth.

Fig. 3. (a) Variation of surface microhardness of the SMRT sample with annealing temperature. (b) In-depth distributions of microhardness in the as-SMRT sample and the SMRT sample annealed at 500 °C for 120?min. Measurements were performed on the sample surface from a planar view in (a) and on the surface layer from a cross-sectional view in (b).

Fig. 4. A cross-sectional SEM image of the annealed SMRT sample.

Fig. 5. (a) Measured Cr diffusion profiles in the GNL samples listed in Table 1, in comparison with those in the CG samples without annealing (zero profile) and annealed at 500 °C for 720?min (CG500). (b) Diffusion profiles of the GNL samples are replotted by choosing the origin at the position where the last section begins (i.e. for the diffusion in the bcc phase). The solid lines in (b) show error-function solutions according to Eq. (2) to fit the experimental data.

Fig. 6. Comparison of the temperature dependence of the effective diffusivity of Cr in the GNL IF steel with those in the Fe lattice [23], in the SMAT Fe [24], and along GBs of the relaxed CG Fe [25].