Effect of rolling
Prior deformation samples after recrystallization were characterized by EBSD to calculate grain boundary character distribution, grain size and analysis of macroscopic features. Bulk texture analysis of the ferrite phase was done on Brukers™ D-8. Figure 1(a) shows the IPF (inverse pole Fig.) obtained after the EBSD scan and IPF texture from bulk crystallography showing a strong texture of {111}; its fraction increases with the increase of deformation on the sample, but there is a reduction in grain size, respectively. Samjadar et al. [29] have attributed the texture {111} formation to favorable Taylor factor and stored energy. Texture influences mechanical properties and has been reported in the literature [29, 30]. But, the relationship of corrosion properties with texture has limited exposure [31–33]. Our study has revealed a significant correlation between the increase in the {111} texture fraction and heightened susceptibility to corrosion, as shown in Fig. 2(a) & 2(b). Analysis of grains, grain boundary nature after EBSD shows the distribution of different grain boundaries of low (< 15º) angle, high (> 15º) angle and CSL (coincidence site lattice).
Figure 2 depicts an increase in corrosion current and non-passivation with increasing prior deformation recrystallized samples. The observed increase in corrosion current cannot be solely attributed to the texture or the presence of impurities at the grain boundary. Since the alloying elements are only present in trace amounts, we investigated the grain boundary's nature and associated energy role. To facilitate this analysis, we derived a specialized formula [34, 35]. Utilizing this formula, we calculated the Effective Grain Boundary Energy (EGBE) for each prior deformation, as depicted in Fig. 2(c).
Grain boundary energy role and corrosion properties:
Figure 2(a) and 2(b) show the potentiodynamic results exhibiting a lower corrosion current for samples with lower prior deformation (35%). The depth of attack, denoted as τav, was measured using a non-contact optical profilometer, specifically the Zeta™ instrument. Both bulk measurements (covering the entire specimen surface) and local measurements (limited to individual CSL boundaries) were considered. A higher corrosion current and greater depth of attack were observed in samples with higher prior deformation, indicating an increased susceptibility to corrosion. Notably, a direct correlation was observed between these factors and the EGBE. During the calculation of EGBE, the presence of CSL boundaries and their deviation from the Brandon criteria (θ ≤ 15°) were taken into consideration, along with grain size [36, 37]. It has been reported that grain size influences the nature of corrosion [32, 36, 38, 39], and in our calculation, EGBE was influenced by gain size. To assess the influence of grain size on these properties, we conducted an investigation using a higher prior deformation (80%), resulting in a smaller grain size. We performed grain boundary characterization and calculated the EGBE for the prior deformation (80%) with a coarser grain size. The results demonstrated a similar nature to the prior deformation (80%) with a smaller grain size, as depicted in Fig. 2(c). This comparison confirms the validity of our findings and suggests that grain size impacts the observed properties. The anodic polarization test of coarser prior deformation shows a lower corrosion current than finer prior deformation (80%).
Additionally, the depth of attack on the grain boundary after the potentiodynamic test was also lower for these samples, as shown in Fig. 3. Furthermore, the depth of attack (bulk) exhibits a consistent trend, as observed previously, even in the absence of grain growth. This finding proves that grain size does not play a substantial role in our case. However, it should be noted that grain size exerts some influence, as previously reported in studies on corrosion behavior [32]. The slight variation in corrosion current observed for the same deformation (80%) but different grain sizes can be attributed to the difference in grain boundary area. With a smaller grain size, larger grain boundaries are present as compared to the larger grain size, resulting in a higher grain boundary area per unit area of observation. A relationship between EGBE and the depth of attack was plotted for all prior deformations at three different concentrations of HCl (0.1, 0.5M, and 1M) as shown in Fig. 3, 4(a) & 4(b) for EIS curve and charge transfer resistance, performed using different 0.1M HCl solution shows the solution concentration affect to a great extent the corrosion of the materials due to higher ion species presence.
Role of nature of grain boundary and corrosion properties:
A comprehensive investigation was conducted to explore the grain boundary characteristics, their distribution and their influence on the corrosion behavior of IF steel. The study focused on examining the depth of attack of specific grain boundaries and their role in the corrosion susceptibility of the material. The investigation included an analysis of various Σ (sigma) coincidence site lattice (CSL) boundaries, specifically Σ1, Σ3, Σ9, Σ13b, and Σ21b. These specific boundaries were selected due to their higher occurrence than other Σ CSL boundaries. A square region measuring 500mm × 500mm was demarcated to facilitate easy identification using a light indention and EBSD scanning was performed within the region. This technique allowed for the characterization of grain boundaries and their crystallographic orientations. Following the EBSD analysis, a potentiodynamic test was conducted to evaluate the corrosion behavior of the IF steel. This electrochemical testing method provided insights into the material's susceptibility to corrosion and allowed for a correlation between the specific grain boundaries and their impact on the corrosion nature of the steel. The study revealed that the lower CSL boundary, specifically Σ1 CSL, which was selected for the grains with misorientation angles ranging from 10–15º [40], exhibited a lower depth of attack compared to other CSL boundaries and random boundaries that experienced higher levels of attack as shown in the SEM image in Fig. 5. The SEM images in Fig. 5 depict the nature of corrosion in interstitial-free steel samples subjected to different levels of prior deformation (35% and 80%) and immersed in varying concentrations of HCl solution. In samples with lower prior deformation, intergranular corrosion was evident, while samples with higher prior deformation exhibited both intergranular and transgranular (surface attack) corrosion. This suggests that a higher corrosion current and greater bulk depth of attack (material removal) may be attributed to the combined effect of higher EGBE, a higher number of random boundaries, and the occurrence of Σ3 CSL boundaries with greater deviations. Furthermore, the EGBE (effective grain boundary energy) and depth of attack (bulk) calculated from the zeta profilometer along with grove angles of the attacked special boundaries were calculated, and the corresponding plot is presented in Fig. 6(a) and 6(b), respectively. It was observed that the boundaries with a higher degree of attack exhibited higher grove angles. Interestingly, our findings indicated that the Σ3 CSL boundary, which falls between all the CSL boundaries, displayed a greater depth of attack compared to other CSL boundaries but still lower than that of random boundaries (boundaries with misorientation angles ≥ 15º).
The literature has reported that lower CSL boundaries are more coherent, so they get attacked less, but that was not the case here. This could be attributed to the presence of such CSL boundaries, which shows how much they deviate from exact CSL boundaries based on Brandon criteria = 15/Σ1/2. The analysis revealed a higher number of boundaries exhibiting greater deviations, indicating an increased depth of attack. The relationship between Σ3 CSL boundaries and their deviations is illustrated in Fig. 6(c). A minimum of 50 CSL boundaries were considered for this calculation. It was observed that the population of Σ3 CSL boundaries increased with higher levels of prior deformation. However, random boundaries still dominated in all cases of prior deformation, and their numbers also increased with increasing prior deformation.
Following cold rolling and subsequent recrystallization processes, a significant grain refinement and the dominance of gamma fiber texture were observed in our case. This phenomenon has been widely reported in the literature [6, 41]. It can be attributed to stress localization during rolling and the stored energy associated with the gamma texture compared to other textures[29, 42]. Previous studies have demonstrated that low-angle boundaries and specific Σ3 and Σ9 CSL boundaries exhibit improved corrosion resistance compared to random boundaries, which tend to be more susceptible to attack. CSL boundaries, owing to their low energy state, are often considered favorable parameters in grain boundary engineering for enhancing material properties [32, 43]. The selection of CSL boundaries is typically based on different criteria, such as the Brandon and Palumbo-Aust misorientation angles when analyzing the desired properties [44]. It has been reported in various studies that a higher fraction of special boundaries, particularly low Σ CSL boundaries, offers improved corrosion resistance in aluminum. These findings further support the significance of grain boundary engineering in enhancing material performance against corrosion [45, 46].
Another significant parameter for estimating the corrosive nature of materials is the boundary attack, which can be described as the depth of attack experienced by grain boundaries as a result of exposure to aggressive solutions. Grain boundaries serve as potential sites for corrosion initiation due to their higher energy and entropy, leading to a positive Gibbs free energy compared to the inner topography of the grains [47]. A higher depth of attack at the grain boundary indicates a more corrosive nature, as it signifies the dissolution of a larger surface area. Another influential factor impacting the behavior of special boundaries is their deviation from the exact CSL, as determined by the Brandon criteria.