Effect of Soluble Chromate Inhibitor on the Macro- and Meso-Scale Damage Morphology and Electrochemical Corrosion Behavior of AA7050-T7451 Coupled to Type 316 Stainless Steel

It is well documented that chromate conversion coatings (CCC) and anodization of Al precipitation age hardened alloys, such as AA2024 (UNS A92024⁽¹⁾) and AA7075 (UNS A97075), improve corrosion resistance.^1-9  While chromate is a known human carcinogen, many legacy structural components still rely on chromate for corrosion inhibition.⁹  A typical chromate coating system contains many layers including a surface treatment, the CCC, a primer, and a topcoat. A thin layer (10 nm to 60 nm) CCC provides corrosion protection and improved adhesion between the next layers is typical. The second layer, the primer, is comprised of a pigmented resin matrix. The primer is the reservoir that houses soluble chromate.^10-11  Finally, a topcoat is applied, which serves as a main barrier against environmental influences.¹²  The topcoat is typically a UV resistant polyurethane that provides resistance to photonic degradation.¹¹  Barrier coatings are used to suppress the cathodic reaction by limiting diffusion of the electrolyte, oxygen, and water to the substrate.^2,13  Ionic inhibitors released from coatings and pretreatments are a key approach to inhibit corrosion.¹⁰  Therefore, a key stage in protection is after the barrier degrades enough to allow moisture and corrodent ingress when soluble chromate release is a primary factor in protection of fastener-plate galvanic corrosion in an occluded site.

In this work, the NaCl-based solutions were tested with selected additions of the inhibitor species as an oxyanion circumventing the storage/release stages to probe the influence of chromate inhibitor on corrosion damage evolution on bare AA7050-T7451 galvanically coupled to stainless steel (SS). The bare fastener-plate arrangement isolates the coating breakdown stage and requirement for moisture ingress and salt deliquescence from the local corrosion stage.

The inhibition of corrosion of aluminum alloys by chromate and CCCs has been extensively reported.^1,10,12-21  The properties of Cr allow for very different behaviors for oxy- and hydroxy compounds of the Cr⁶⁺ and the Cr³⁺.²²  The tetrahedral d⁰ hexavalent oxyanion compounds of chromium, including chromate, dichromate, bichromate, and chromic acid, dissolve as stable complexes in water, transport easily, and adsorb on oxide surfaces.¹⁸   inhibitor decreases the metastable pit nucleation rate at a given potential and Cl⁻ concentration, which minimizes the probability for pit stabilization.²³  On pure Al, in mixed chloride-chromate solution ranging from 10⁻⁶ M to 10⁻² M, the pitting potential began to be ennobled when the chloride to chromate ratio was between 10∶1 to 1∶1.²⁴   lowered pit growth rates and reduced the metastable pit nucleation rate.²³  However, pre-existing deep pits were not inhibited.²³  Passivating inhibitors such as chromate operate under site specific conditions.^{1,4,12-14,16,18-19,25-27}  Chromate dissolves or is solubilized in a +6 valence state oxyanion in aqueous solution. One hypothesized method on the inhibition of corrosion is that chromate is reduced to its more stable Cr³⁺ oxide or hydroxide at active anodic sites which creates a soluble barrier that protects the underlying substrate from local attack. As stated above, hexavalent chromate is reduced to the trivalent state forming a chromium(III) hydroxide following the overall or full cell reaction given in Equation (1):^{12,14,18,25-26}  

In general, the mechanism for chromate inhibition of Al alloy dissolution is that chromate is a very soluble, higher valent, oxidizing ion with a lower valent form that is insoluble and creates an extremely protective film.^14,17  Bulk Cr³⁺ hydroxides form by sol-gel polymerization, giving rise to well-hydrated amorphous materials.¹³  CCCs form on aluminum through reduction of Cr⁶⁺ in solution. CCC consists of a mixture of regions of Cr³⁺ hydroxide at sites where chromate was reduced along with regions of Al₂O₃. Cr⁶⁺ and Cr³⁺ may coexist within the CCC in several specific forms.²⁶  Cr⁶⁺ reduced on the bare aluminum surface can lead to a Cr³⁺-rich barrier oxide. Anodic inhibition of Al alloys by soluble chromate has been shown to occur as well.¹² 

Chromates are also excellent inhibitors of oxygen reduction near neutral and alkaline solutions by suppression of the cathodic reaction.^18,25-26  The oxygen reduction reaction (ORR) rate was reduced by about an order of magnitude during cathodic polarization scans of AA2024-T3 with the addition of 10⁻⁵ M sodium dichromate.^13,28  This typically means that chromate inhibits cathodic reaction on reactive sites such as Cu- and Fe-rich constituent particles in aluminum alloys. The speculated mechanisms are through the blocking films discussed above or soluble chromate competition with oxygen for adsorption. However, a debated issue in the understanding the role of chromate on cathodic reactions is whether prior exposure in chromate at open circuit results in a reduction in corrosion of constituent particles and subsequent reduction in copper replating, which accounted for reduced ORR.

The stages of damage evolution on bare AA2024-T3 around Cd-plated SS fasteners during exposure in marine atmosphere have also been studied. Crack initiation and propagation occurred in the rivet holes.²⁹  The damage evolution from the Al and fastener corrosion did not occur by a single continuous corrosion mode. The damage evolution was found to evolve through several stages and transition among corrosion modes. In addition to classic galvanic corrosion, galvanic action had a significant effect on crevice corrosion, intergranular cracking (IGC), and stress corrosion cracking (SCC).³⁰  Previous work on bare AA7050 shows that localized corrosion sites form elongated fissures in the longitudinal (L) direction, hemispherical pits, and can transition to IGC.^31-32 

Galvanic attack has been studied on coated AA2024-T3 in laboratory and field exposures.³³  AA2024-T3 panels with different scribed coatings with and without galvanic connections to Type 316 SS (UNS S31600) fasteners were exposed in the field and to ASTM B117.³³  The panels were pretreated with a CCC and then painted with a chromate epoxy primer. During field exposures, it was found that current transients associated with galvanic corrosion depended strongly on the electrolyte responding to changes in relative humidity (RH), water drop evaporations, condensation, and salt precipitation. Field exposures had much lower galvanic charge than found in the case of ASTM B117 exposures.³³  The CCC panels without a SS fastener did not exhibit detectable corrosion attack after 460 d of field exposure. However, CCC panels with SS fasteners displayed corrosion damage even in short field exposure periods despite the presence of chromate. In other words, the chromate coating system was not effective when corrosion attack was mainly driven by galvanic coupling to bare SS fasteners. The SS fasteners were physically distant from any released chromate and did not benefit from inhibition of their cathodic reactions.³³  However, these studies contained coated or otherwise covered rivet holes. A flaw in the study was that corrosion in the rivet hole was not analyzed. To date, no work has studied the effect of chromate on the damage evolution inside a fastener hole.

The effect of pretreatment on galvanic attack of coated Al alloy panels was also studied.³⁴  The substrates for the galvanic samples were AA7075-T6 panels containing four through holes with five different pretreatments. All of the pretreated panels were coated with the same chromate-containing epoxy primer and polyurethane topcoat. The holes were created before surface pretreatment and were not intentionally treated. However, the inside surfaces were exposed during processing and eventually were unevenly coated with topcoat. The Al panels exhibited isolated narrow but deep corrosion attack with the fastest pit growth compared to the other pretreatments which produced wide general corrosion attack with a trivalent chromium pretreatment and CCC pretreatment. These findings point out that chromate might produce one dominant pit that might create a more severe site for fatigue or SCC than without chromate.

The objective of this work is to investigate the macro- and mesoscale effects of chromate as an inhibitor on the damage morphology and corrosion electrochemistry when AA7050-T7451 (UNS A97050) is coupled to Type 316 SS and, furthermore, to investigate the effects of chromate on replated Cu on AA7050. The use of corrosion inhibitors such as chromate is hypothesized to significantly slow both the development of corrosion damage and fatigue crack growth rate in aerospace aluminum alloys.³⁵  However, no work has studied these effects inside a fastener hole. This work will quantitatively study the corrosion damage initiation and morphology with the use of soluble chromate in solution as an inhibitor in NaCl environments and elucidate the controlling electrochemical factors altered by the anionic inhibitor and the remaining damage morphology.

This work seeks to understand the spatially distributed damage morphology occurring in a fastener plate arrangement when soluble chromate chemical species is added as an aqueous phase inhibitor. The study herein accounts for sources of cathodic reactions that can support local corrosion with and without chromate.

The Al alloy used in this study was machined from Al-Zn-Mg-Cu alloy plate of 50 mm thickness provided by Alcoa. The grains were elongated in both the longitudinal (L) and long transverse (T) directions and thin in the short transverse (S) directions, forming a pancake-like microstructure. The grains in the L direction ranged from 22 μm to 1,250 μm in length, while the grains in the T and S directions ranged from 15 μm to 265 μm and 12 μm to 112 μm, respectively. The Al-Zn-Mg-Zr was cast at the Kroehling Advanced Materials Foundry, Virginia Tech Foundry Institute for Research and Education. The alloy was overaged at 121°C for 120 h. This composition and processing resulted in an S phase (Al₂CuMg) free microstructure. Copper was replated on AA7050-T7451 by potentiostatically holding for 600 s at −1 V_SCE in 0.1 M CuSO₄ + H₂SO₄ at pH 3. This produced a 100 nm layer of Cu on the aluminum alloy surface. The Al₂CuMg precipitate phase (S phase) was also studied on this work. Coupons were freshly ground with successively finer grit SiC abrasive paper (LECO wet/dry 600, 800, 1200 grit) using water lubrication. Residual polishing compounds were removed using a clean polishing pad saturated in ethanol prior to ultrasonication in ethanol for 30 s, and blown dry with clean, compressed air. Table 1 shows the compositions for AA7050 and Type 316 SS.

Sodium chromate (Na₂CrO₄) was added to 0.5 M NaCl and 4.0 M NaCl solutions. Three concentrations of Na₂CrO₄ were studied in this work: 0.1 mM (pH 7.8), 1 mM (pH 8.3), and 10 mM (pH 8.8). The pHs of the solutions were not adjusted, only measured. Selected tests were conducted in 0.5 M NaCl + Na₂CrO₄ adjusted to pH 5.5 with H₂SO₄. The concentrations of Na₂CrO₄ were chosen based on previous work conducted on leaching of chromate into solution.^12,36  Full immersion exposures were conducted in 0.5 M NaCl to match the equilibrium NaCl concentration of 4 M NaCl, from salt drops equilibrated at 98% RH.³⁷ 

The x-ray fastener geometry was a 1 mm diameter cylindrical AA7050-T7451 pin with a 250 μm rivet hole embedded with a SS wire. The rivet hole was 400 μm to 500 μm deep. The gaps between the SS wire and the AA7050-T7451 varied from 1 μm to 50 μm, depending on the 360° location and the tilt of the pin with respect to the long axis of the hole. The area ratio inside the hole of AA7050-T7451 to SS was 1∶1. The entire pin was encapsulated with a plastic tube that contained filter paper with saturated potassium sulfate (K₂SO₄) controlling the RH to approximately 98%. The plastic was sealed with an Al stopper at the top. A schematic of the simulated fastener is shown in Figure 1. Three pin samples were exposed under a 0.4 μL droplet of 4 M NaCl with 0.1 mM, 1 mM, and 10 mM Na₂CrO₄ for 62 h. The electrolyte concentrations supplied by the droplet on the surface of the pin equilibrated to the droplet’s equilibrium concentration of 0.5 M NaCl at 98% RH.³⁷  The steel rivet was removed and ex situ x-ray tomography was performed with an XRadia MicroXCT-200^† at 10× magnification and 80 kV voltage. The voxel size was 4.2 × 4.2 × 4.2 with 3 μm resolution. Ex situ x-ray computed tomography (XCT) conducted in this work was compared to a previous study by the authors under the exact corrosion conditions under a 4 M NaCl droplet with no additions of chromate. Details of the inhibitor-free exposure can be found elsewhere.^32,38-39 

From the XCT results the anodic charge associated with cumulative damage, Q_A, can be found using Equation (2): where V_XCT is the volume lost as a result of corrosion found by the XCT, F is Faraday’s constant, ρ is the density of AA7050 (2.83 g/cm²), and E.W. is the equivalent weight of AA7050 (9.77 g/equivalents assuming congruent dissolution).

Chromate-containing exposures were compared to previous work on geometrically identical simulated fasteners under a droplet of inhibitor-free 4 M NaCl exposure and assessed with operando XCT at the diamond light source in Oxfordshire, United Kingdom.^31-32,39 

Planar electrode testing augmented fastener plate studies. All electrochemical testing, unless noted otherwise, was performed using a three-electrode flat cell with a platinum niobium mesh counter electrode and a saturated calomel reference electrode (SCE). Samples were cleaned in dilute nitric acid according to ASTM Standard G1 to remove corrosion products before determining mass loss. All experimental procedures were conducted after 1 h at open circuit, unless noted otherwise. Potentiostatic polarization was conducted on AA7050-T7451 coupons at the galvanic couple potential (−0.73 V_SCE) for 144 h. The environments that were tested were inhibitor-free 0.5 M NaCl (pH 5.5), 0.5 M NaCl + NaAlO₂ (pH 8), and 0.5 M NaCl + NaOH (pH 8), and chromate-containing environments 0.5 M NaCl with additions of 10 mM, 1 mM, and 10 mM Na₂CrO₄. Constant potential holds were conducted in duplicates for three to ensure replicability.

Cu was replated on AA7050-T7451 and exposed in 0.5 M NaCl + 1 mM Na₂CrO₄. Cyclic voltammetry (CV) was conducted on select AA7050-T7451 coupons after exposure to Cu replating as evidenced by Cu₂O and CuO oxidation and reduction peaks. CVs were conducted in borate buffer pH 8.4. Borate buffer pH 8.4 was deaerated for 2 h prior to testing. CVs were conducted at 1 mV/s.

Potentiodynamic polarization was conducted on AA7050-T7451, Type 316 SS, pure Cu, Al-Zn-Mg-Zr, and Cu replated on AA7050-T7451 in chromate-containing environments to understand the effect of chromate in solution on the galvanic couple potential and galvanic couple current contributions. Anodic and cathodic potentiodynamic polarization scans were conducted with a scan rate of 1 mV/s.

The grain structure of one AA7050-T7451 pin contained small recrystallized deformation-free grains and the other two contained large unrecrystallized deformed grains. The dissimilar grain structure could be a function of position on the AA7050-T7451 plate (Figure 2).

AA7050-T7451 simulated SS fasteners were exposed to droplets and assessed with XCT. Chromate-containing droplets of 4 M NaCl + 0.1 mM NaCrO₄ at pH 7.7, 4 M NaCl + 1 mM NaCrO₄ at pH 8.3, and 4 M NaCl + 10 mM NaCrO₄ at pH 8.7 were placed on the surface of the simulated fasteners. 2D horizontal slices through an XCT for the simulated fasteners exposed to 4 M NaCl with additions of 0.1 mM, 1 mM, and 10 mM Na₂CrO₄ can be observed in Figure 3, relative to the inhibitor-free 4 M NaCl exposure at the top right. This shows a horizontal section in the LT plane at different depths along the long axis of the fastener from the top of the cylindrical AA7050 test piece. The large gray disk is the AA7050 with light spots indicating constituent particles. The large dark gray circle is the rivet hole, while the SS wire simulated a rivet. The dark corrosion fissures (shown in color) were observed in AA7050 around the rivet hole and extended from 1 μm to ∼500 μm in depth. The corrosion damage was identified by phase contrast and verified using scanning electron microscopy (SEM).

Four separate fissures grew throughout the 500 μm rivet hole in inhibitor-free 4 M NaCl after 62 h.^31-32,39  One fissure formed at the mouth and the other three fissures formed near the bottom of the rivet. The largest corrosion fissure was observed to be at the mouth of the crevice. During the 4 M NaCl exposure, the corrosion fissures did not follow the bands of intermetallic clusters. The path that fissures corroded was intragranular. A 3D reconstruction of the corrosion fissure damage after the 4 M NaCl inhibitor-free exposure can be seen in Figure 4.^31-32,39 

In 4 M NaCl + 0.1 mM Na₂CrO₄, approximately nine small fissures formed in the top half of the rivet hole. These pits were all formed on one side of the rivet and were at most 100 μm in depth. The 3D reconstruction can be observed in Figure 5. The SS rivet is shown in blue and the 3D-shaped volume is representing the corrosion damage in red. In 4 M NaCl + 1 mM Na₂CrO₄, one small fissure developed in the middle of the 400 μm rivet hole. The corrosion fissure was approximately 55 μm deep. This exposure produced the least amount of corrosion damage. The 3D reconstruction of the 4 M NaCl + 1 mM Na₂CrO₄ exposure can be observed in Figure 6. In 4 M NaCl + 10 mM Na₂CrO₄, the density of fissures was reduced and the most significant damage was in the form of one large corrosion fissure located at the bottom of rivet hole (Figure 7). The corrosion fissure was approximately 150 μm deep.

The fissure area was tracked over time and depth for each exposure. The area was quantified using Image J^† contrast analysis. The image for every slice was individually adjusted via threshold image processing.⁽²⁾ The damage morphology of the first fissure in the 4 M NaCl inhibitor-free exposure and each of three chromate exposures can be observed in Figure 8. The damage morphology seen in this figure shows one 2D slice after threshold and binary image processing. The 2D area of damage in each of the slices analyzed were plotted as a function of rivet depth in Figure 9. The bars in red show the corrosion damage in chromate-containing exposures and the bars with the white stripes behind the chromate bars indicate the in operando 4 M NaCl inhibitor-free corrosion area bars. In 4 M NaCl + 10 mM Na₂CrO₄, only one corrosion fissure formed. However, the corrosion volume was almost as large as the sum total of four fissures observed in the inhibitor-free 4 M NaCl exposure.

Additions of 1 mM and 0.1 mM Na₂CrO₄ reduced the corrosion volume by two orders of magnitude after a 62 h exposure. The corrosion volume from the ex situ XCT exposure was calculated after the 62 h exposures. The corrosion volume enabled determination of the required anodic charge to produce the damage observed. The anodic charge density after 62 h exposure can be seen in Figure 10. The anodic charge density associated with the inhibitor-free 4 M NaCl exposure is indicated by the dark blue bar compared with the red bars for each of the different chromate-containing environments. The levels of chromate tested reduced the anodic charge density associated with corrosion by 10-fold. Here, none of the fissures with chromate were deeper than seen during the inhibitor-free exposure. The deepest pit in each exposure correlated with the anodic charge density (Figure 11). The low chromate concentration produced a lower anodic charge than inhibitor-free exposure while producing narrow deep pits.

To further support the effects of chromate inhibitor on damage morphology in fastener-rivet coupons observed with XCT, exposures on planar electrodes were conducted. Potentiostatic polarization was conducted at the galvanic couple potential (−0.73 V_SCE) on planar electrodes of AA7050-T7451 (SL surface exposed), coupled to Type 316 SS in chromate-containing environments and inhibitor-free environments. The current density and charge density plots over the 144 h exposures are shown in Figure 12. The material environment combinations that produced the highest charge densities were uninhibited 0.5 M NaCl, 0.5 M NaCl + 10 mM Na₂CrO₄ at pH 8.8, and Cu replated on AA7050-T7451 in 0.5 M NaCl + 1 mM Na₂CrO₄ at pH 8.8.⁽³⁾ In 0.1 mM Na₂CrO₄, corrosion kinetics were not inhibited relative to 0.5 M NaCl at pH 5 and pH 8. The corrosion kinetics were reduced by approximately one order of magnitude in 1 mM Na₂CrO₄ relative to these inhibitor-free exposures. The charge densities of the chromate-containing exposures and the inhibitor-free exposures are summarized in Table 2. The damage morphologies of these samples were analyzed with SEM on the surface and in cross section with optical microscopy. Figure 13 shows optical images of the SL surface of AA7050-T7451. Figure 14 shows the secondary electron micrographs of the SL surface of AA7050-T7451. In inhibitor-free environments, hundreds of small pits were observed. In the environment containing the least amount of chromate, 0.1 mM Na₂CrO₄, approximately 20 pits were visible. In 1 mM Na₂CrO₄, no pitting was observed. In the most concentrated Cr environment (i.e., 10 mM Na₂CrO₄), approximately 42 deep trench-like pits were seen. On Cu replated on AA7050-T7451 in 1 mM Na₂CrO₄, 18 deep trench-like pits were observed. Figure 15 shows the cross-sectioned optical micrographs of AA7050-T7451 LS. The highest density local corrosion sites occurred on replated Cu on AA7050-T7451 and on AA7050-T7451 in inhibitor-free NaCl solutions at pH 8. In the chromate-containing environments 0.1 mM, 1 mM, and 10 mM Na₂CrO₄, pit density was reduced but the deepest pit depths were 127 μm, 0 μm, and 627 μm, respectively. In chromate-free environments, the deepest pits observed in 0.5 M NaCl + NaAlO₂ (pH 8) were 247 μm. (Table 2). Potentiostatic polarization agreed with the XCT that in chromate-containing solutions, the damage morphology produces deep and narrow pitting. However, potentiostatic polarization at the galvanic couple potential proved to be more aggressive than the damage observed in galvanic couples.

To substantiate the damage morphologies observed on AA7050, the effect of chromate on the ORR kinetics of Type 316 SS, pure Cu, Cu replated on AA7050-T7451, and S phase was explored (Figure 16). Chromate substantially reduced ORR kinetics on Type 316 SS. On Type 316 SS, ORR kinetics were reduced by one order of magnitude with additions of chromate. Chromate also reduced the ORR kinetics on the constituent particle phase Al₂CuMg. The effects of chromate on Cu were less significant. On pure Cu, ORR kinetics were slightly reduced in 1 mM and 10 mM Na₂CrO₄. No reduction on ORR cathodic kinetics was observed in 0.1 mM Na₂CrO₄ (Figure 16). Cu replated on AA7050-T7451 did not experience a reduction on ORR kinetics with additions of chromate. In this study, in fact, a slight increase in ORR kinetics was observed in 10 mM and 1 mM Na₂CrO₄. In all three concentrations of Na₂CrO₄, chromate significantly reduced cathodic kinetics by one order of magnitude on Al₂CuMg relative to inhibitor-free environment. In general, the addition of chromate substantially reduced ORR kinetics on Type 316 SS and S phase but only slightly reduced ORR kinetics on pure Cu. No reduction in cathodic ORR kinetics was observed on Cu replated on AA7050-T7451; in fact, ORR cathodic kinetics increased slightly. The finding that chromate does not inhibit corrosion on replated Cu correlated with potentiostatic polarization (Figure 15).

In 0.5 M NaCl, anodic kinetics of AA7050-T7451, Al-Zn-Mg-Zr, and S phase were reduced with additions of Na₂CrO₄ as shown in Figure 17. Na₂CrO₄ changes the pH of the solution to 7.8 to 8.8, depending on the concentration. Cathodic and anodic potentiodynamic polarization were conducted in 0.5 M NaCl + 1 mM Na₂CrO₄ with the addition of H₂SO₄ to adjust the pH to 5.5 to attain comparable pH (Figure 18). The pH change was observed to negligibly effect both the anodic and cathodic electrochemical kinetics.

On all three materials, the pitting potential was significantly increased with additions of Na₂CrO₄. The pitting potentials of AA7050-T7451 and Al-Zn-Mg-Zr were highest in the 1 mM Na₂CrO₄ environments, followed by 0.1 mM, and lastly 10 mM Na₂CrO₄. On S phase, the most significant improvement on the critical potential was observed in 10 mM Na₂CrO₄.⁽⁴⁾ In 0.5 M NaCl, the pitting potentials of AA7050-T7451 were −0.71 V_SCE, −0.63 V_SCE, −0.58 V_SCE, and −0.65 V_SCE with additions of 0 mM, 0.1 mM, 1 mM, and 10 mM Na₂CrO₄, respectively. The pitting potential of Cu-free Al-Zn-Mg-Zr in 0.5 M NaCl with additions of 0 mM, 0.1 mM, 1 mM, and 10 mM were −0.95 V_SCE, −0.88 V_SCE, −0.85 V_SCE, and −0.92 V_SCE, respectively. In general, on AA7050-T7451 and Al-Zn-Mg-Zr, the best corrosion inhibition was observed in environments of 0.5 M NaCl + 1 mM Na₂CrO₄ and the worst corrosion inhibition was observed in the inhibitor-free environment and that with the most concentrated amount of Na₂CrO₄, 0.5 M NaCl + 10 mM Na₂CrO₄. On S phase, the best corrosion inhibition was observed in environments of 0.5 M NaCl + 10 mM Na₂CrO₄. On all three materials, all three concentrations of chromate were observed to reduce anodic kinetics relative to inhibitor-free 0.5 M NaCl, suggesting that chromate inhibited Al and Mg dealloying. Anodic kinetics were most significantly improved by inhibitor additions in AA7050-T7451 and S phase.

Cyclic anodic polarization in 0.5 M NaCl + 1 mM Na₂CrO₄ (pH 8.3) on AA7050-T7451 after 24 h at open-circuit potential (OCP) showed reduced ORR cathodic kinetics by two orders of magnitude relative to the anodic polarization behavior examined without an OCP pre-exposure (Figure 19). Moreover, the OCP was ennobled by 200 mV. While no change was observed in the width of the passive region, I_pass was reduced with the longer OCP period. The passive current density was 2.9 × 10⁻⁶ A/cm² in 0.5 M NaCl chromate-free environment with 1 h OCP. The passive current densities in chromate-containing environments after no OCP, 1 h, and 24 h OCP were 1.1 × 10⁻⁶ A/cm², 3.5 × 10⁻⁷ A/cm², and 5.7 × 10⁻⁸ A/cm², respectively. The pitting potential on AA7050-T7451 increased by 150 mV after 24 h at OCP. After 24 h OCP, cathodic polarization showed a decrease in the limiting current density by one order of magnitude relative to that of 1 h at OCP.

XCT on AA7050-T7451 fasteners under NaCl chromate-containing droplets showed that chromate, in general, reduced the total metal volume loss and furthermore reduced the number density of corrosion sites when compared to the inhibitor-free exposure (Figures 4 through 7). The XCT showed that while the anodic charge was reduced, the depths of the remaining corrosion fissures were not necessarily reduced (Figure 11). In the previous work, inhibitor-free exposures showed that multiple corrosion fissures were growing simultaneously over time, indicating that corrosion fissure growth was not severely cathodically limited to a single fissure.^32,38,40  This was attributed the presence of strong cathodes such as the SS fastener and replated Cu on AA7050-T7451, which both support ORR. In this work, it was evident that true anodic charge density associated with the metal volume loss was less with chromate in solution than in the inhibitor-free exposure (Figure 12). This can be attributed to both soluble chromate and a lower propensity for replated Cu (as evident from chromate-inhibited anodic electrochemistry of AA7050-T7451, Al-Zn-Mg, and S phase which limits Cu release) and, therefore, a reduced ORR to support the growth of new fissures. Figure 16 shows that the ORR cathodic reaction rate was reduced by over an order of magnitude in the presence of chromate. Lower anodic current on Al₂CuMg might suppress dealloying in chromate-containing environments and thus, little Cu will be leached into solution to replate on the surface of AA7050-T7451. Constant potential holds at the galvanic couple potential showed that after exposure in chromate-containing environments, Cu replating was not visible (Figure 13). Cyclic voltammetry was conducted on postexposure AA7050 samples to assess the surface coverage of Cu on AA7050-T7451. Figure 20 shows that after exposure in inhibitor-free 0.5 M NaCl at pH 8, a Cu oxidation peak was observed. An important note is that this peak was not observed after exposure in 0.5 M NaCl + 1 mM Na₂CrO₄ at pH 8. The lower anodic charge associated with metal volume loss in chromate-containing environments can in part be speculated to be a result of the lack of the Cu replating process.

Kelly, et al., studied the effect of soluble chromate additions vs. CCCs in small volume cells. They determined that under conditions where a thin water layer exists on the surface in high humidity, the accumulation rate was 1 mM/h. Therefore, after 10 h, a chromate concentration of 10 mM would be expected.²⁶  In this work, concentrations of 0.1 mM, 1 mM, and 10 mM were chosen to represent the amount of chromate present in solution when a CCC is under thin film corrosion conditions for 0.1 h, 1 h, and 10 h of exposure.

Corrosion electrochemistry in this work agrees with the findings of other authors that low concentrations of soluble chromate in solution suppresses corrosion kinetics (Figures 16 and 17).^{12,14,16,18,34}  Figure 21 showed that the pitting potentials of AA7050-T7451, Al-Zn-Mg-Zr, and Al₂CuMg were all raised in 0.1 mM and 1 mM Na₂CrO₄. S phase experienced the most significant ennoblement of critical dealloying potentials in all three chromate concentrations. This suggests that Cu replating is likely inhibited indirectly through inhibition of Al₂CuMg corrosion. However, ennoblement of the pitting potential indicates fewer corrosion sites but does not necessarily indicate a lower anodic charge density once pits form.

The ORR current density for Type 316 SS, Cu, Cu replated on AA7050, and Al₂CuMg (S phase) are plotted as a function of Na₂CrO₄ concentration in Figures 16 and 22. The limiting current density on Type 316 SS and Al₂CuMg was significantly reduced with increasing chromate. Pure Cu experienced slightly less ORR kinetic inhibition. Interestingly, for Cu replated on AA7050-T7451, the ORR current density increased slightly with the additions of chromate, which is not understood. Inhibition of Cu release through passivation of Al₂CuMg would lower cathodic reaction rates. However, any Cu replated on the AA7050-T7451 surface during corrosion in Cl⁻ or alkaline treatments would raise cathodic reactions rates when Cu is negatively polarized to the OCP of AA7050-T7451.

XCT (Figures 5 through 7) and constant potential holds (Figure 15) showed that even though chromate reduced the ORR kinetics and ennobled the pitting potential of AA7050-T7451 (Figure 22), the effect on damage morphology is nuanced. Table 2 summarizes the constant potential hold exposures. Fissure density was significantly reduced with the addition of soluble chromate. However, it was observed that aggressive environments, such as high anodic potentials or galvanic coupling, a deep fissure could survive in concentrated chromate environments (Figure 15). XCT showed that one deep fissure formed in 10 mM Na₂CrO₄ at the bottom of the rivet hole (Figure 7), while one small pit formed and survived inside the rivet hole in 1 mM Na₂CrO₄ (Figure 6). The XCT also showed that the large corrosion pit that formed in 10 mM Na₂CrO₄ was nearly as large as the sum of all of the corrosion sites formed in the inhibitor-free exposure (Figure 9). In the case of an exceptionally deep corrosion pit, all of the cathodic reactions occurring at all sites near the rivet hole could support the anodic current for the single fissure. It is possible that in the case of chromate, pit density is lower but residual cathodic current can support one large fissure whose corrosion metal volume loss was equal or larger than the sum of all of the corrosion pits in exposure with no inhibitor.

In 10 mM Na₂CrO₄, the few corrosion pits observed were over seven times deeper than the typical result in the case of the inhibitor-free exposure. AA7050-T7451 coupons that deliberately had Cu replated and were exposed in 1 mM Na₂CrO₄ formed deep corrosion fissures that were six times larger than observed in the equivalent exposure with no Cu replated on AA7050. Such deep pits might develop a low pH due to metal ion hydrolysis. Cr⁶⁺ is hydrolyzed in aqueous solutions and exists as an oxyanion in all but the most acidic conditions.¹³  Figure 23 shows the predominance diagram for chromate as a function of pH based on standard free enthalpies given by Pourbaix.⁴¹  The predominant species at pH levels between 1 and 6 is and above pH 6, is stable. The dichromate ion (⁠⁠) forms when the concentration of chromium exceeds 1 g/L. Even if pH is maintained constant for a particular pit growth experiment, there will be a distribution of Cr⁶⁺ between the mononuclear (⁠⁠) and the dichromate ion (⁠⁠), which varies with concentration and pH.^13,42  While corrosion inhibition is detected in chromate solutions with widely varying pH, a possible exception to this is the fully protonated H₂CrO₄, which exists in strongly acidic solutions.¹³  Evidence suggests H₂CrO₄ is present in active pits, but does not inhibit corrosion because it is not charged and cannot compete with Cl⁻ and adsorb or interact with the corroding surface.⁴³  Therefore, it can be speculated that in aggressive environments, when halide is present and pitting does occurs, the pH local to a deep fissure becomes acidic.³⁵  Under this condition, HCrO₄ and H₂CrO₄ may be the preferred species and in this condition, Cr⁶⁺ would not contribute strongly to the corrosion inhibition as pointed out above. In this case, the deep fissure might develop sufficient acidity such that chromate is not effective. From a structural integrity viewpoint, the large flaw associated with single fissure could be an initiation site for fatigue despite the presence of chromate.

• •

  Aggressive conditions created by occluded sites in fastener-plate systems experiencing galvanic coupling of stainless steel to AA7050-T7451 can induce many pit-like fissure sites and formation of multiple deep fissures. Similar corrosion morphology can be duplicated with application of high anodic potentials on a planar electrode instead of a fastener-plate configuration.

• •

  Damage morphology studies showed that the fissure density was reduced with the addition of soluble chromate. The addition of 1 mM Na₂CrO₄ inhibited pitting-fissure corrosion on AA7050-T7451 almost entirely and defined an optimal chromate level in fasteners. In one case, remaining fissures were over seven times deeper in more concentrated chromate environments than chromate-free environments.

• •

  The growth of deep corrosion fissures in chromate-containing environments can be attributed to several factors, for instance, the concentration of cathodic reaction from all available sites toward the growth of a single deep fissure. It is also speculated that the fully protonated H₂CrO₄ may exist in strongly acidic solutions associated with deep pit fissures. H₂CrO₄ present in active pits would not inhibit corrosion because it is not charged and cannot compete with Cl⁻ to adsorb or interact with the corroding surface.

• •

  Additions of sodium chromate reduced cathodic kinetics by one order of magnitude on Type 316 stainless steel and by less than one order of magnitude on pure Cu interpreted through effects of chromate in suppressing the oxygen reduction reaction. However, additions of sodium chromate slightly increased the ORR cathodic kinetics on replated Cu on AA7050-T77451 for reasons that remain unclear.

• •

  In NaCl environments, the pitting potential was ennobled with the addition of Na₂CrO₄. On AA7050-T7451 and Al-Zn-Mg-Zr, the best corrosion inhibition was observed in environments of 0.5 M NaCl + 1 mM Na₂CrO₄ and the least significant corrosion inhibition was observed in the environment with the most concentrated amount of Na₂CrO₄, 0.5 M NaCl + 10 mM Na₂CrO_4.

We gratefully acknowledge support from the Office of Naval Research under the contract ONR: N00014-14-1-0012 with William Nickerson as contract manager.