Potentiodynamic polarization curves of AA7075 at high scan rates interpreted using the high field model

The study addresses how high potential scan rates distort potentiodynamic polarization (PDP) curves for aluminum alloy AA7075, impeding accurate extraction of corrosion kinetics via conventional Tafel analysis. Automotive lightweighting increases reliance on AA, but micro-galvanic coupling drives localized corrosion, making reliable electrochemical characterization crucial. Standard PDP recommendations favor low scan rates (≈0.167 mV s⁻¹) and separate anodic/cathodic branches from Ecorr; however, such conditions suffer from irreversible surface changes and accumulation of corrosion products, and macroscale PDP can misrepresent localized corrosion phenomena. Microscale PDP mapping requires high scan rates (>10 mV s⁻¹) in a single direction, placing the system far from steady-state and biasing Tafel parameters. Prior work has largely attributed high-scan-rate distortions to double-layer capacitance and mass transport and explored only a narrow scan-rate range. This work hypothesizes that oxide growth/passivation kinetics, described by the high field model, dominate PDP behavior at high scan rates. The goal is to measure PDP up to 100 mV s⁻¹ in two electrolytes (0.62 M NaH₂PO₄ and 3.5 wt% NaCl) and interpret trends via simulations including capacitive contributions and high field oxide growth.

Previous studies of scan-rate effects on PDP have mostly been limited to ~0.1–10 mV s⁻¹, attributing distortions primarily to double-layer capacitance and, to a lesser extent, mass transport (e.g., Zhang et al. 2009; ASTM G102 guidance; Birbilis et al. 2005; Fischer et al. 2019). Far fewer works have considered the scan-rate dependence inherent to oxide growth/passivation kinetics. Seminal contributions include the high field model applications to aluminum passivation by White and Isaacs, as well as the point defect model by Macdonald, all indicating that ionic migration in the oxide film and associated field effects yield scan-rate-dependent anodic responses. Extensive work on anodic oxide growth kinetics on valve metals supports the notion that oxide growth can dominate transient currents during dynamic polarization, yet this has not been fully integrated into interpreting high-scan-rate PDP of AA alloys.

Experimental: AA7075 sheets (8×5 cm, 2 mm thick) were lightly abraded, ethanol/water rinsed, air dried. Electrolytes: 0.62 M NaH₂PO₄ (acidic, pH ~3.6) and 3.5 wt% NaCl (near-neutral, pH ~7). PDP protocol (ASTM G5-94 and G61-86): In aerated solutions, hold at OCP until stable (<10 mV/min; ~20 min to −0.82 V vs SCE in NaH₂PO₄; ~5 min to −0.78 V in NaCl), precondition at −0.25 V vs OCP for 20 s, then scan from −0.25 V vs OCP to +1 V vs SCE at 0.167, 5, 25, 100 mV s⁻¹ (cutoff 1 mA cm⁻²). Three-electrode cell (1 cm² area) with SCE reference and Pt mesh counter; Biologic VSP-300 potentiostat. Five replicates per scan rate using fresh 1 cm² areas; currents normalized to geometric area. EIS: Potentiostatic EIS (10 mV amplitude; 10⁵–1 Hz) at multiple potentials (−1 V to OCP) and before/after PDP in both electrolytes. Equivalent circuit: Rs in series with parallel (Rp || CPE). Capacitance extracted via Brug equation to estimate effective C(t). Measured C(t) differed between electrolytes but varied little with potential or conditioning. Surface/solution analyses: XPS (Al Kα) to measure native oxide thickness d0 on AA7075; ICP-OES after 24 h immersion (and nitric acid rinse) to quantify Al dissolution and estimate Al₂O₃ dissolution rates in each electrolyte. Modeling/simulation: PDP current density modeled as jpdp = jhf + jc + jcap. Capacitive current jcap = C(t)·(dV/dt) with C(t) from EIS (assumed constant over the scan ranges: ~16 µF cm⁻² in NaH₂PO₄, ~21 µF cm⁻² in NaCl). Anodic high field current jhf follows jhf = A·exp(βa·E), with E proportional to overpotential divided by oxide thickness; oxide thickness evolves from Faraday’s law of growth with passivation efficiency εp and chemical dissolution rate Rdiss. Cathodic current jc modeled by a single Tafel-like expression (combining HER and ORR) with parameters j0c, Ec (taken as OCP), βc. Simulations implemented in COMSOL Multiphysics v6.2 (DAE/ODE) using literature values for A, βa, ρ, n=6, EAl2O3 from Pourbaix, and fitted variables (d0, Rdiss, εp, j0c, βc). For NaCl, pitting onset was represented by step changes in Rdiss and εp beyond a threshold potential to reproduce rapid current rise. A buffered phosphate experiment (0.564 M NaH₂PO₄/0.056 M Na₂HPO₄, pH 5.5) was also used to assess pH effects on steady-state current.

• Experimental PDP: In 0.62 M NaH₂PO₄ at 0.167 mV s⁻¹, Ecorr ≈ −0.77 V vs SCE and jcorr ≈ 9.8 µA cm⁻²; the anodic branch exhibits a passivation plateau. In 3.5 wt% NaCl at 0.167 mV s⁻¹, Ecorr ≈ −0.78 V vs SCE and jcorr ≈ 1.60 µA cm⁻²; pitting occurs with Epit close to Ecorr and no clear passive region.
• High-scan-rate behavior: Increasing scan rate (5, 25, 100 mV s⁻¹) causes apparent Ecorr shifts (e.g., in phosphate to ~−0.848 V vs SCE) and higher anodic currents; in NaCl, Ecorr shifts are larger and the cathodic branch slope increases with scan rate, with Epit moving positive.
• Capacitive current insufficiency: Simulations adding only jcap (using measured C(t)) reproduce some Ecorr shifts but fail to match overall PDP shapes or anodic trends across scan rates and electrolytes; explaining all distortions would require unrealistically large, scan-rate- and potential-dependent C(t), inconsistent with EIS.
• High field model success: Incorporating jhf (oxide growth kinetics) with jc and measured C(t) accurately captures key PDP features across scan rates in both electrolytes, including Ecorr,app movement and anodic current increases. In NaCl, the model reproduces pre-pitting behavior but not the pitting regime (requiring ad hoc steps in Rdiss and εp).
• Parameterization and validation: Simultaneous fitting across scan rates yields physically reasonable parameters: d0 ≈ 2.44 nm (sim) vs 2.38 ± 0.51 nm (XPS); Rdiss on the order of 10⁻¹¹–10⁻¹² mol cm⁻² s⁻¹ (phosphate higher than chloride), consistent within an order of magnitude with ICP-OES; εp < 1 in acidic phosphate, indicating incomplete passivation; higher jc and Rdiss in acidic conditions align with cathodic dissolution literature.
• pH/buffering effects: A buffered phosphate solution (pH 5.5) stabilizes steady-state current jss relative to unbuffered acidic phosphate, indicating surface pH significantly affects passivation efficiency and oxide growth kinetics.
• ln(jss) vs v/jss: Although absolute kinetic parameters A and βa derived from ln(jss) vs v/jss are unreliable under non-ideal assumptions, the linearity confirms high-field-controlled oxide growth kinetics and enables extrapolation of jss at low scan rates from high-scan-rate data.
• Variability at high scan rate: Replicate variance increases markedly with scan rate, broadening Ecorr,app and Epit distributions (notably in NaCl). Simulations show small variations in native oxide thickness (e.g., d0 ~2.60–2.75 nm) can produce large differences in high-scan-rate PDP, explaining reduced reproducibility.
• Quantitative examples of model inputs: For simulations, C(t) ≈ 16 µF cm⁻² (NaH₂PO₄) and 21 µF cm⁻² (NaCl); EAl2O3 ≈ −1.9 V (phosphate) and −2.0 V (chloride) vs SCE; fitted jc parameters in phosphate j0c ~30 µA cm⁻², βc ~−250 mV/dec; in chloride j0c ~3.7 µA cm⁻², βc ~−600 mV/dec. In NaCl, emulating pitting required stepping Rdiss to ~1×10⁻⁹ mol cm⁻² s⁻¹ and εp to 0 above −0.773 V vs SCE.

The findings demonstrate that oxide growth kinetics, not charging currents, dominate the transient currents that reshape PDP at high scan rates for AA7075. The high field model inherently links scan rate, oxide thickness evolution, and electric field within the growing oxide, producing larger anodic currents and apparent Ecorr shifts with faster scans, consistent with experiments in both acidic phosphate and neutral chloride media. Capacitive contributions alone are too small (with measured C values) and lack the complex potential dependence needed to reproduce observed shapes. The model clarifies why Ecorr,app and cathodic branch overlap across multiple high scan rates in phosphate: jc intersects jhf before divergence in jhf becomes pronounced. It also explains scan-rate-dependent increases in anodic current via delayed attainment of steady-state jss as oxide growth lags the changing potential window. Surface pH strongly modulates passivation efficiency and dissolution, altering steady-state behavior; buffering stabilizes jss, affirming the sensitivity of oxide growth to local chemistry. While the model is effective pre-pitting in NaCl, it cannot capture localized breakdown, indicating separate mechanisms govern pit initiation/propagation. The increased variance among high-scan-rate replicates is consistent with the model’s sensitivity to small differences in initial oxide thickness and local properties, which are averaged out at lower scan rates. Overall, incorporating high field oxide kinetics provides a more physically grounded interpretation of high-scan-rate PDP than Tafel extrapolation and supports quantitative use of microscale PDP when properly modeled.

High field oxide growth kinetics govern the distortions observed in high-scan-rate PDP of AA7075, whereas capacitive currents alone cannot explain the measured trends. A single parameter set within a high field framework reproduces the main scan-rate-dependent features across two electrolytes, tying kinetics to oxide properties with physically interpretable parameters (e.g., native oxide thickness, dissolution rate, passivation efficiency). The approach advances the prospect of making microscale, high-scan-rate PDP quantitative and enables extrapolation of steady-state currents from fast scans. Future work should: (1) improve independent characterization of key high field parameters and their dependence on electrolyte composition and pH; (2) incorporate localized corrosion phenomena (micro-galvanic effects, chloride-induced oxide breakdown) to better model pitting and spatial variability.

• The high field model, as implemented, does not capture localized pitting corrosion; pitting behavior in NaCl was emulated via step changes in dissolution rate and passivation efficiency rather than mechanistic modeling.
• Several parameters (Rdiss, εp, jc Tafel parameters) were fitted and assumed constant across conditions; real systems may exhibit time-, potential-, and environment-dependent properties (e.g., ion incorporation, evolving pH) not explicitly modeled.
• Kinetic constants A and βa derived from ln(jss) vs v/jss under ideal assumptions are inconsistent with fitted values when dissolution and incomplete passivation occur, limiting direct parameter extraction from steady-state analyses.
• Capacitive properties were treated as constant within each electrolyte based on high-frequency EIS; any frequency or structure-dependent capacitance changes during polarization were not resolved.
• Increased variance at high scan rates suggests sensitivity to local surface heterogeneity; macroscale measurements average these effects, and microscale validation is needed to generalize findings.