Directly monitoring the shift in corrosion mechanisms of a model FeCrNi alloy driven by electric potential

Austenitic stainless steels exhibit excellent corrosion resistance due to a thin protective chromium oxide layer but can suffer severe corrosion in acidic chloride-containing environments, with or without electrochemical bias. While corrosion mechanisms of stainless steels have long been studied, nanoscale, spatially resolved, quantitative understanding has only recently emerged. In situ EC-AFM enables tracking topographical changes under controlled electrochemical conditions, offering direct observation of pit initiation and growth. The research question addressed here is how applied electric potential influences early-stage corrosion mechanisms—specifically pit nucleation and growth kinetics—in a model austenitic Fe-18Cr-14Ni alloy. The goal is to quantify vertical and lateral pit growth kinetics with and without bias and to correlate these with crystallographic orientation and subsurface chemistry, thereby informing predictive corrosion models.

In situ AFM has been widely applied to study corrosion at solid–liquid interfaces due to high spatial resolution and controllable solution parameters. Prior EC-AFM studies have monitored surface changes during polishing and corrosion under various electrochemical biases and stresses, including early-stage pit initiation and the role of grain boundaries and surface inhomogeneities in austenitic steels. High-speed AFM has captured rapid intergranular pit formation in sensitized stainless steels. Crystallographic orientation significantly affects pitting susceptibility: planes with higher atomic density (e.g., (111), (100)) generally show higher resistance under certain conditions, though potential-dependent behavior has been observed (e.g., {100} dissolving faster at noble potentials, {111} at less noble). Grain-specific corrosion rates can vary with deviation from [111]. Dissolution kinetics have been modeled as diffusion-limited in some systems, and classical views consider chloride adsorption and passive film breakdown via intermediate complexes as precursors to pitting.

Materials: A high-purity model Fe-18Cr-14Ni (wt%) austenitic alloy was arc melted, cast, homogenized by remelting five times, cold-rolled to 3 mm (50% reduction), and recrystallized at 900 °C for 4 h. Microstructure was characterized by SEM, EBSD, and synchrotron XRD (beamline 11-ID-C, APS; λ = 0.1173 Å), confirming equiaxed grains with annealing twins and fully stabilized FCC austenite. Regions of interest (ROIs) for AFM were marked using PFIB fiducials and mapped by high-resolution EBSD. In situ EC-AFM: Experiments used a Bruker Nanoscope 8 (J scanner) in contact mode at ~25 °C with hybrid silicon-on-SiN cantilevers (AppNano HYDRA-ALL, k = 0.405 N/m, tip radius <10 nm). The three-electrode setup comprised the Fe-18Cr-14Ni substrate (working), Pt wire (counter), and leakless Ag/AgCl (reference, ET072, eDAQ). Electrolyte was 0.5 M deuterium chloride (DCI) in D2O. Imaging conditions: scan rate 1 line/s, 128 lines/image, aspect ratio 4, horizontal size 20 or 40 μm. Zero time was defined at acid injection. Two samples were monitored: (1) Nonbiased: 5 μm × 20 μm continuously scanned for ~1.5 h. (2) Biased: initially corroded ~30 min at open-circuit voltage (OCV ~ −0.9 V vs Ag/AgCl), then an additional −0.5 V bias was applied and held for 1 h; area 10 μm × 40 μm to include multiple grains. Electrochemical control used a CH Instruments 600E workstation. For biased tests, three grains near [012], [124], and [323] orientations were analyzed. Post-corrosion TEM/STEM: After EC-AFM, SEM documented surface changes; site-specific FIB lamellae were prepared. STEM/TEM (JEOL ARM200CF, 200 keV, Cs-corrected) with Gatan Orius CCD and Centurio EDS provided imaging and elemental mapping. STEM-EDS line scans across oxide layers characterized compositional stratification.

Microstructure: The alloy exhibited equiaxed, random-textured FCC austenite with annealing twins (SEM/EBSD/XRD). Nonbiased EC-AFM: Pits appeared between 2.2–4.4 min; pit number density increased rapidly initially, with no new pits after ~20 min. Both pit depth and width of isolated pits increased with time over 30 min. Power-law fits D = βp t^n1 and W = βv t^n2 showed vertical dissolution exponents n1 = 0.624–0.900 and lateral exponents n2 ≈ 0.391–0.580 (average ≈ 0.5). Representative rates: Pit #1 (near a grain boundary) had the highest vertical rate βp = 1.68 ± 0.09 nm min^−n1; lateral rates βv ~ 4.16–4.75 nm min^−n2. Interpretation: vertical growth is surface-kinetics/diffusion hybrid-controlled (0.5 < n1 < 1), while lateral growth is diffusion-controlled (n2 ≈ 0.5). Grain boundaries acted as preferential initiation sites and exhibited faster dissolution. Biased EC-AFM (−0.5 V vs Ag/AgCl after 30 min): Pit depth growth remained approximately linear with time and showed no obvious change in vertical dissolution rate upon applying bias. Pit width often saturated before bias, and after bias, width changes were small, with growth constrained by increased pit number density. The applied bias substantially increased pit nucleation: pit number density rose sharply after bias, particularly in the [323]-oriented grain (closest to [111]). Total pitting area increased linearly with time, with slopes (βA) more than doubling upon bias. Quantitatively (μm^2/min): [012] grain 11.68 ± 1.94 → 31.05 ± 1.38; [124] 17.89 ± 3.40 → 43.51 ± 1.25; [323] 25.49 ± 5.67 → 55.89 ± 1.51 (before → after bias). The [323] grain exhibited the highest pit density and area growth rate both before and after bias. TEM/STEM-EDS: After corrosion without bias, surface showed pits with planar islands between them; with bias, adjacent pits merged and islands were not observed. Both conditions produced a dual-layer oxide: outer Fe-rich oxide and inner Cr-rich oxide, with Ni segregation near the metal–oxide interface. Nanoscale Cr2O3 grains were identified in the inner oxide layer. Mechanistic insight: Chloride interactions with the passive film are consistent with a multistep mechanism involving adsorbed intermediates and a rate-controlling electron-transfer/complex step. Under negative bias, decreased surface Cl− and increased H+ likely inhibit lateral growth (steps 1–3) while promoting pit nucleation and sustaining vertical dissolution (step 4), shifting the dominant corrosion mechanism toward nucleation-driven area growth.

The study quantitatively dissects early-stage pit growth kinetics and the role of electric potential in a model Fe-18Cr-14Ni alloy. Without bias, vertical pit growth follows a surface kinetics/diffusion hybrid regime, while lateral growth is diffusion-limited, consistent with the need to break more bonds for vertical detachment and the dominance of species transport for lateral expansion. Grain boundaries provide defective bonding environments that facilitate interfacial reactions and faster vertical dissolution. Applying a negative potential does not accelerate vertical growth of existing pits but markedly increases nucleation of new pits, thereby controlling the total pitting area. This mechanistic shift is attributed to redistribution of charged species at the interface: negative bias reduces Cl− coverage and increases H+, suppressing lateral propagation steps associated with chloride-complex mediation while enhancing nucleation and maintaining vertical dissolution rate. Crystallographic orientation modulates susceptibility, with grains near [111] ([323]) showing higher pit densities and area growth rates, potentially due to lower adsorption of passivating species and reduced oxidation activation barriers. These findings address how potential orchestrates the balance between nucleation and growth, providing parameters and mechanistic context for predictive corrosion models.

In situ EC-AFM combined with TEM demonstrates that in chloride-containing acidic media, a model Fe-18Cr-14Ni alloy exhibits anisotropic pit growth kinetics: vertical dissolution is surface kinetics/diffusion hybrid-controlled, and lateral dissolution is diffusion-controlled. Grain boundaries initiate pits and accelerate vertical growth. Under an applied −0.5 V bias, corrosion accelerates primarily via enhanced nucleation of new pits rather than increased lateral growth; vertical growth rates of existing pits remain largely unchanged. Grain orientation strongly influences corrosion, with a [323] (near-[111]) grain exhibiting the highest pit density and pitting area growth. Post-corrosion, a dual-layer oxide forms (outer Fe-rich, inner Cr-rich with Ni segregation; inner Cr2O3 nanograins). These quantitative relationships (depth/width vs time; pitting area vs time) under unbiased and biased conditions provide critical experimental inputs for mechanistic and predictive models of stainless steel corrosion. Future work will investigate the effects of different ionic species, mechanical deformation, and a broader range of potentials on corrosion mechanisms and orientation dependence.

The study focuses on a high-purity model Fe-18Cr-14Ni alloy, which simplifies chemistry and defects relative to commercial stainless steels and may limit direct generalization. Experiments probe early-stage corrosion in a specific aggressive electrolyte (0.5 M DCl in D2O) and at a single applied negative potential (−0.5 V vs Ag/AgCl), with limited grains/areas monitored. Contact-mode AFM and the finite scan window may bias observed lateral growth due to pit coalescence and density effects. Orientation-dependent mechanisms are inferred for a subset of orientations; broader orientation mapping and varied electrochemical conditions are needed for comprehensive generalization.