Performance of a Magnesium-Rich Primer on Pretreated AA2024-T351 in Selected Laboratory and Field Environments: Conversion Coating Pretreatments

Aerospace structures using precipitation age hardened aluminum alloys rely on the use of multilayered coatings to provide corrosion resistance, improved adhesion, and other specialized functions.¹  The most commonly used active protection system includes corrosion inhibition by hexavalent chromium.^1-2  Typical formulations consist of chromated pigments enveloped in epoxy resin.^1-2  Carcinogenicity, high handling costs, and lack of environmental safety required an accelerated phase-out of hexavalent chromium, concurrent with a push to find effective alternatives.^1-2  Over the past few years, a commercial organic coating system containing a Mg-rich primer (MgRP) has been developed for the active corrosion protection of aluminum alloys.^3-16  The commercial MgRP coating consists of a surface pretreatment, an epoxy resin with Mg pigment, and a polyurethane-based topcoat (TC). The primary function of the pretreatment is to provide good adhesion between substrate and primer.¹  Pretreatments also provide additional functions such as improved corrosion resistance by acting as a barrier layer and/or enabling inhibitor ion release.¹  The metal pigment system is designed to galvanically couple the relatively more active, metallic Mg pigment in the primer to the AA2024-T351 (UNS A92024⁽¹⁾) substrate, thereby providing sacrificial anode based cathodic protection.¹¹  This approach has been well established previously in the design of zinc-rich primers for use on various steel.^17-21  The organic resin in the epoxy-based primer mediates the global galvanic protection potential and also provides barrier protection to the substrate.¹¹  A TC functions as main barrier for environmental influences such as extreme climate and ultraviolet (UV) rays, and can also serve additional purposes.¹  A systematic evaluation of each of these components and their effect on overall performance of coating systems in different lab-accelerated life testing (LALT) environments and different environmental exposures needs to be understood.

Previous LALT and field exposures conducted to determine optimal primer formulation concluded that approximately 45% pigment volume concentration Mg provides a balance between moderated sacrificial anode based cathodic protection, long-term barrier protection, and the beneficial characteristics of preserved, isolated clusters of Mg pigment available for the future protection of defects.¹¹  Two possible modes of galvanic protection were proposed: long-range protection of remote defects by global galvanic corrosion protection afforded to the substrate, and local or short-range Mg pigment-based protection of local, as well as buried, defects in close proximity to a buried pigment particle.¹¹  Both modes of protection are mediated by the high ionic and electrical resistances of the coating systems as a function of Mg pigment volume concentration (PVC), surface pretreatments, primer polymer, and TC properties. The regulation of cathodic protection abilities is an important aspect in optimizing the MgRP performance.¹¹ 

Environmental degradation of MgRP on NFF pretreated AA2024-T351 in selected field and laboratory environment in the presence and absence of TC has been extensively studied.^6-8  Full immersion in ambiently aerated 5% NaCl solution, ASTM B117 in 5% NaCl, and ASTM B117 in ASTM substitute ocean water all resulted in significant depletion of metallic Mg pigment in the MgRP (without TC) far from the scribe after 1,000 h.⁸  Field exposures in Charlottesville, Virginia and Kennedy Space Center, Florida also resulted in similar levels of Mg pigment depletion far from the scribe after 2,000 h and 4,000 h of exposure.⁸  This implies Mg loss via self-corrosion. The global galvanic protection potential of the coating system, with respect to remote scratches, became more positive with exposure time in each environment, from values approximately equal to that of bare Mg (−1.6 V vs. saturated calomel electrode [SCE]) to those approximately equal to that of bare AA2024-T351 (−0.7 V_SCE). It was found that this rise took approximately 300 h in full immersion in ambiently aerated 5% NaCl solution, ASTM B117 in 5% NaCl, and in ASTM B117 with ASTM substitute ocean water and approximately 1,000 h in the field at Birdwood Golf Course (CHO) in Charlottesville, Virginia and Kennedy Space Center (KSC) in Florida. These times do not represent actual service times on account of periods of drying and wetting. Residual barrier properties of the MgRP with an initial Mg PVC of 45% coating system also degrade with time in each environment.⁸  However, corrosion was not observed under the residual coating polymer after Mg pigment depletion.⁸  Therefore, the primer provides some residual barrier protection.⁸  The Aerodur 5000^† (high-performance advanced coating) TC significantly regulated the depletion of Mg pigment from the MgRP in all exposure environments as compared to identical environmental exposures of non-topcoated samples as measured by x-ray diffraction (XRD).⁷  Full immersion in ambiently aerated 5% NaCl solution, ASTM B117 using 5% NaCl, and ASTM B117 modified with ASTM substitute ocean water and UV all resulted in very limited depletion of metallic Mg pigment in the MgRP far from the scribe after 1,000 h.⁷  Field exposures in CHO and KSC also resulted in partial Mg pigment depletion far from the scribe after 1 y of exposure.⁷  The global galvanic protection potential of the coating system, with respect to remote scratches, increased slightly with exposure time in each environment, from initial values of approximately −1.0 V_SCE to −0.7 V_SCE after extensive environmental exposure. These values fall between the open-circuit potentials (OCP) of bare AA2024-T351 (−0.6 V_SCE) and bare Mg (−1.6 V_SCE), and are predicted by mixed potential theory.^7-8  This suggests that Mg pigment that is both electrically and ionically connected to the AA2024-T351 can provide sacrificial galvanic protection of the AA2024-T351 substrate in extended time-of-wetness events.^7-8  Barrier properties of the MgRP primer coating, as assessed by electrochemical impedance, also only slightly degrade with time in each environment but, overall, remain very high (≥10⁹ Ω·cm² at 0.01 Hz) throughout exposure in each environment, indicating significant barrier protection remains after all environmental exposures studied.^7-8 

However, all of these studies investigated MgRP over a bare or NFF pretreated AA2024-T351. In practical applications, several pretreatments are of interest to improve adhesion between the substrate and the polymer and also to impart additional corrosion protection.^22-30  Chromate conversion coatings (CCC) offer strong corrosion resistance properties and are noted for their ability to self-heal.^29-35  This phenomenon is attributed to the release of hexavalent chromium from the coating into the corrosive solution in contact with the surface.^29-30  Because of environmental hazards posed by hexavalent chromium, non-chromium pretreatment (NCP)^23,25  and trivalent chromium pretreatment (TCP)²⁴  coatings have been explored as alternatives for CCC. NCP conversion coatings are based on titanium/zirconium oxides,^23,25  whereas TCP conversion coatings are trivalent chromium enriched zirconium oxide coatings.²⁴ 

A systematic evaluation of each of these coating system components, including the conversion coatings previously mentioned, and their effect on the overall performance of coating systems containing MgRP in different LALT environments and different environmental exposures is of interest. The objective of this study was to investigate MgRP system on AA2024-T351 with various pretreatments. The thickness of the pretreatment, its chemistry, and electrical properties imparted by the pretreatments were examined to understand the role of pretreatment in global galvanic protection potential mediation.^5-6  Degradation of pretreatment, MgRP, and TC as a function of time for different pretreatments were studied using diagnostic tests including the accelerated electrochemical cycle test with ex situ Mg depletion studies using XRD.^5-6  Environmental degradation of different pretreated AA2024-T351 with MgRP and with and without TC in relevant lab and field environments needs to be understood to expand the knowledge of the role of pretreatments in corrosion protection function. Comparison of environmental degradation of conversion coatings to non-film forming (NFF) pretreatment based coating systems are discussed in this work. The sacrificial anode protection function, barrier properties, and alternative protection modes afforded by AA2024-T351/Pretreatment/MgRP and AA2024-T351/Pretreatment/MgRP/TC were characterized by utilizing electrochemical as well as non-electrochemical post-mortem analysis techniques including XRD, Raman spectroscopy, optical profilometry, and scanning electron microscopy (SEM)/energy dispersive spectroscopy (EDS) after pertinent lab and field exposures. The behavior in field is particularly pertinent based on the lack of applicability of standard ASTM B117²⁸  (5 wt% NaCl) to assessments of field performance as a result of discrepancies between lab and field exposures.^7-8  Effect of pretreatment on Mg depletion, self-corrosion of Mg, and scratch protection are reported herein.

AA2024-T351 sheet (1.6 mm thickness) was pretreated with four different pretreatments for comparison including: (i) NFF surface pretreatment, (ii) CCC, (iii) TCP, and (iv) NCP. Prekote^† was the NFF chromium-free surface pretreatment used, containing approximately 95% water and less than 3% each of diethylene glycol monobutyl ether and N-methyl-2-pyrrolidone.³⁶  The other pretreatments used were Alodine 1200s^†,37  (CCC), Alodine 5200^†,38  (NCP), and Surtec 650^†,39  (TCP).

A 40 μm primer layer of MgRP and a 50 μm thick TC of high-performance advanced coating, both produced by Akzo Nobel Coatings, were applied. The MgRP consists of one part epoxy matrix with 20 μm diameter Mg metal flake pigment with PVC of 45% (3rd generation 2100P003, Lot: 493-190). The high-performance advanced coating (gloss white finish product: ECM-G7875) is a two component polyurethane TC developed for military applications in a variety of exposure environments.⁴⁰ 

MgRP coated AA2024-T351 panels were exposed to salt spray in a QFog Cyclic Corrosion Tester (QFog model CCT 1100^†) according to ASTM B117⁴¹  “with neutral” 5 wt% NaCl solution (pH = 6.9±0.4) as the electrolyte at a temperature of 35°C for at least 1,000 h. During a second exposure, the standard ASTM B117 salt fog was altered such that the standard 5 wt% NaCl solution electrolyte was replaced with acidified ASTM substitute ocean water⁴²  (SOW; pH = 3.2±0.2) and UV radiation. The salt fog exposure cabinet was modified to include four hanging ultraviolet A (UVA) fluorescent lights. The UVA lamps (Q-Lab Corporation model UVA-340^†) were chosen to simulate sunlight in the critical short-wave UV region from 365 nm down to the solar cutoff of 295 nm. The ASTM substitute ocean water was produced according to ASTM D-1141⁴²  and acidified by the addition of 10 mL of glacial acetic acid per 1 L of salt solution following ASTM G85 A3.⁴³  In all salt fog exposures reported herein, ambient air was supplied to the chamber and to the atomizer for fog production. Ambient concentrations of CO₂ were measured in situ and were found to be approximately 425 ppm. Other ambient gas concentrations were not measured. Natural weathering exposures of MgRP-coated AA2024-T351 panels were conducted in two different environments: at a coastal marine site 30 m from the high tide line at KSC Corrosion Technology Lab in Titusville, FL (28.6°N, 80.6°W, elevation = 0 m) and a rural inland site at CHO (38.0402°N, 78.54.27°W, elevation = 172 m). The sample test racks at KSC face the ocean and the one in CHO face south. During exposure, panels were mounted on unsheltered atmospheric test racks with full exposure to natural elements according to ASTM G-4.⁴⁴  Pertinent environmental parameters such as mean temperature, mean relative humidity, mean dew point, mean precipitation rate, precipitation pH, and chloride deposition rate were measured in all LALT/field environments. These parameters are summarized in Table 1.

All full-immersion studies, as well as post-mortem analysis after salt fog and field exposures, reported herein were conducted in quiescent 5 wt% NaCl (pH: 6.9±0.4) open to laboratory air. Electrochemical experiments were performed using a Gamry potentiostat (Ref 600/ PCI4^†) with computer interface software. SCE and Pt mesh were used as the reference and counter electrode, respectively. The intact coating area tested was far away (≥2 cm away) from the scribe. A typical electrochemical impedance spectroscopy (EIS) scan was acquired in sine sweep mode from 100 kHz to 0.01 Hz with six points per decade. MgRP and MgRP/TC coated panels were scanned with an AC amplitude of 80 mV to 100 mV to reduce noise. The tests were conducted in quiescent 5 wt% NaCl, as discussed above, after 1 h exposure at open circuit for MgRP coated panels and 12 h exposure for MgRP/TC coated panels.

XRD was conducted to characterize global Mg depletion as a function of exposure time in different lab and field environments as described elsewhere.^7-8,11  A powder diffractometer utilizing a Cu-Kα source was utilized for all measurements. All samples were scanned continuously from 30° to 50° at 5° per minute. XRD measurements of pristine and environmentally exposed samples were made on panels far away (≥2 cm away) from any edge or scribe. These measurements were confirmed to be representative of global coating degradation. XRD obtained from MgRP panels were normalized against the face-centered cubic (fcc) Al <200> 2θ = 44.74° peak from the underlying substrate. Peak normalization and integration was performed with Origin Lab 7.5^† software. The lower detection limit for crystalline phases was approximately 3% to 5% of the sample by volume.

Corrosion products formed after environmental exposure were characterized using Raman spectroscopy. Raman spectroscopy was conducted using a Renishaw InVia Raman Microscope^†. Measurements were conducted using a 514 nm laser at 1% to 50% power under the 20× objective with a 3,000 L/mm (vis) grating. Scans with 15 s exposure time were taken with two accumulations under standard confocality. If Raman spectra showed heavy fluorescence, then a pre-measurement sample bleaching was conducted, where the sample was subjected to laser exposure under the aforementioned conditions for 450 s to 600 s prior to taking the spectra. For all measurements, prior calibration of Raman spectroscope was performed using a silicon standard.

SEM and EDS were used for post-mortem analysis of corrosion products in the scribe. A field emission Quanta SEM^† was used to conduct these investigations. For EDS measurement, a working distance of 15 mm and an accelerating voltage of at least three times the energy of the maximum characteristic peak of interest were used (~15 kV). At an accelerating voltage of 15 kV, EDS has a penetration depth of roughly 2 μm to 5 μm into the materials investigated in this study. Elemental maps and line profiles were collected and the analysis was performed using the Aztec^† software.

Optical profilometry of the scribe was conducted using a zygo optical profilometer (Newview 7200/7300^† model). The environmentally exposed samples were first exposed to concentrated nitric acid for 20 min to remove corrosion products present in the scribe as per the ASTM Standard G1.⁴⁵  Image refinement and corrosion volume loss calculations were performed using MountainsMaps^†,46  imaging topography software.

The low-frequency EIS and break point frequency data for the AA2024-T351/Pretreatment/MgRP system as a function of exposure time in different LALT and field environments are summarized in Figures 1 and 2, respectively. The approximate barrier low-frequency EIS at 0.01 Hz did indicate significant coating degradation of AA2024-T351/NFF/MgRP in full immersion. Similar full immersion exposure led to lowering of low-frequency impedance by one to two orders of magnitude for all chosen conversion coatings based systems, indicating degradation of conversion coatings. However, there is substantial degradation in barrier properties of AA2024-T351/Pretreatment/MgRP after exposure in the standard and modified ASTM B117 environment. After 1,000 h exposure in standard/modified ASTM B117, the low-frequency EIS lowered by two to three orders of magnitude, indicating rapid coating degradation. Barrier properties slightly improved at longer exposure times, indicating sealing of pores by the conversion of Mg pigment to magnesium hydroxide. Similar exposure conducted in field (KSC/CHO) showed moderate coating degradation in initial and intermediate exposure times (2 weeks to 24 weeks) and significant coating degradation after 52 weeks. There is significant degradation of primer polymer by UV exposure, eventually exposing the underlying substrate after 52 weeks. This can be correlated to the significant degradation in barrier properties of AA2024-T351/Pretreatment/MgRP after 52 weeks. The breakpoint frequency increased with time in chosen LALT/field exposures, suggesting porosity development. This could result from the combination of metallic Mg pigment corrosion and the degradation of primer polymer. However, in selected LALT environments corrosion of metallic Mg pigments led to formation of Mg corrosion products, which sealed the pores and improved the barrier properties. Breakpoint frequency analysis indicated that severity of coating degradation as a function of exposure environment is in the following order: full immersion (5% NaCl) < KSC ~  CHO  < ASTM B117 < modified ASTM B117.

X-ray spectra of AA2024-T351/NFF/MgRP after exposure in selected LALT/field environments are shown in Figure 3. Elemental Mg HCP <001>, <002>, and <101> peaks, and fcc Al <111> and <200> peaks were observed in initial spectra. The integral intensity of <200> peaks was normalized so as to compare the elemental Mg depletion as a function of exposure time in all chosen environments. Partial depletion (~80%) of Mg occurred after 550 h exposure in full immersion. Mg in MgRP was completely depleted after 96 h, 1,000 h, and 8,736 h in modified ASTM B117, ASTM B117, and field exposure (KSC/CHO), respectively. This indicates that self-corrosion of Mg occurs in all chosen environments, but only in absence of a TC. Mg depletion was not complemented by detection of any crystalline Mg corrosion products, as indicated in Figure 3. Similar Mg depletion trends were observed in other chosen AA2024-T351/Pretreatment/MgRP systems that were conversion coatings based. The results are summarized in Figure 4. Irrespective of the nature of the pretreatments, the Mg depletion trends were similar, indicating self-corrosion is the predominant factor in Mg depletion in chosen environments, as these studies involved intact coatings. Mg depletion trends indicate that Mg depletion as a function of exposure environment is in the following order: full immersion (5% NaCl) < KSC ~ CHO < ASTM B117 < modified ASTM B117.

The global galvanic protection potential of the AA2024-T351/NFF/MgRP system as a function of amount of Mg available is shown in Figure 5(a). The average OCP of AA2024-T351 and Mg in ambiently aerated 5 wt% NaCl solution are −0.71 V and −1.48 V, respectively. In NFF pretreatment, magnesium is galvanically coupled to AA2024-T351 immediately, as evident from the more negative potential, and Mg is available for cathodic protection from the beginning of exposure, as indicated in Figure 5(a). In field exposures, the global galvanic protection potential rises to the OCP of AA2024-T351 after 24 weeks. This could be correlated to visual inspection of samples wherein the coating has completely degraded by UV exposure and the bare substrate is exposed. The global galvanic protection potential of the AA2024-T351/TCP/MgRP system as a function of amount of Mg available in three different environments including full immersion, ASTM B117, and field exposure (at CHO) is shown in Figure 5(b). In TCP pretreatment, the global galvanic protection potential is initially more positive compared to OCP of bare AA2024-T351. This indicates that initially no galvanic coupling occurs between AA2024-T351 and MgRP because of the high resistance between the anode and cathode given the resistive nature of conversion coating. However, prolonged exposure up to 85 h brought about lowering in the resistance, enabling galvanic coupling. The OCP shifted to more negative potentials during this process, signaling improved galvanic coupling over time. In field exposures, the global galvanic protection potential reaches the OCP of AA2024-T351 after 52 weeks. This could be correlated to visual inspection of samples wherein coating has completed degraded by UV exposure and the bare substrate was exposed. A similar trend of delayed galvanic coupling was observed in CCC and NCP-based conversion coatings. In modified ASTM B117 (acidified SOW and UV), Mg is completely depleted within 96 h of exposure and there is no galvanic coupling, as could be inferred from the much higher potentials. Similar behavior of rapid Mg degradation was seen in all chosen AA2024-T351/Pretreatment/MgRP systems. The acidic pH (3.2±0.2) resulted in immediate depletion of Mg by self-corrosion.

The global breakpoint frequency of the AA2024-T351/NFF/MgRP system as a function of amount of Mg available is shown in Figure 6. The coating degradation as a function of exposure time is similar in full immersion and field exposures. The ASTM B117 environments resulted in rapid degradation of MgRP. Acidified SOW/UV modified ASTM B117 environment had even more rapid depletion of Mg in addition to the coating degradation resulting from highly acidic environments. As the Mg was depleted, extensive porosity development occurred, as indicated by the rapid rise of the high breakpoint frequency (F_bpt). Similar behavior was seen in all chosen pretreatments.

Raman spectroscopy was conducted for pristine, as well as field exposed, AA2024-T351/NFF/MgRP samples to study the nature of different corrosion products formed in pertinent lab and field environments. For pristine samples, the Raman studies indicated that Mg present on surface of MgRP is immediately converted to magnesium oxide. After LALT exposure in standard ASTM B117 for 400 h, the peaks corresponding to Mg-O bonds disappeared and a more prominent peak corresponding to MgCO₃ appeared at 1,090 cm⁻¹, as shown in Figure 7(a). Similar trends were observed in field exposures at KSC and CHO after 24 weeks. The Raman spectra of samples exposed at CHO for 24 weeks are shown in Figure 7(b). This indicates that Mg in the MgRP is converted to MgCO₃ at the surface as a result of ambient concentration of carbon dioxide present in the environments. The reference peaks for magnesite (MgCO₃) and brucite (Mg(OH)₂) are summarized in Table 2. Numerous additional peaks also appears in the range of 1,200 cm⁻¹ to 2,200 cm⁻¹ after pertinent lab and field exposures. This may be a result of degradation of polymer, and further work on change in polymer chemistry is out of the scope of this current work.

In the ASTM B117 salt fog exposures (LALT), the electrolyte layer is subject to continuous wetting, enabling a fairly accurate determination of throwing power across the scribe. This means that Mg depleted from the primer can be transported over the scratch and precipitated in the scratch. Figure 8 shows EDS elemental Mg maps of NFF/TCP pretreated AA2024-T351/MgRP after different ASTM B117 exposure times. Mg EDS intensity in the scribe increased with exposure time and corresponding intensity of Mg in the primer near to the scribe edge decreased, as indicated in EDS line profiles (Figure 9[a]). Integral intensity of Mg EDS line profiles across the scribe and primer as a function of exposure time in ASTM B117, KSC, and modified ASTM B117 are summarized in Figures 9(b) through (d). Preliminary results indicate that throwing power for Mg redeposition is not greatly inhibited by the presence of the additional resistance resulting from pretreatments. The presence of Mg(OH)₂ in the scribe indicates production of hydroxide, resulting in change of equilibrium pH from 6.9 to 10.45 via the oxygen reduction reaction, as well as Mg²⁺/Mg(OH)₂/H₂O equilibrium and also transport and deposition of Mg to the scribe.⁴⁷  The alkaline pH suppresses the chemical dissolution of species used to identify zones of cathodic protection. The increase in Mg EDS intensity as a function of exposure time in ASTM B117 and KSC environment could be exploited to investigate the transfer of Mg and its chemical precipitation. The modified ASTM B117 environment (acid) indicated initial deposition of Mg(OH)₂ after 48 h, followed by dissolution of Mg corrosion products, as well as leaching of Mg from the AA2024-T351 substrate, on prolonged exposure, as shown in Figure 9[d]. This is a result of dissolution of Mg corrosion products in the low pH solution. Field exposures studies conducted at CHO indicated no presence of Mg corrosion products. The acidic nature of CHO environment resulted in complete dissolution of Mg corrosion products. The Mg corrosion products, in addition to being chemical markers, could also provide a secondary form of AA2024-T351 protection. This mode of protection will be examined in future work.

Optical profilometry maps of scribe exposing AA2024-T351 in NFF and TCP pretreated AA2024-T351/MgRP after exposure in ASTM B117 for 1,000 h is indicated in Figure 10. The optical profilometry map of AA2024-T351/NFF/MgRP after exposure for 1,000 h in ASTM B117 is indicative of the sacrificial corrosion protection or other corrosion mitigation processes provided by Mg across the scribe, as indicated by the presence of fewer and shallower pits. Similar behavior was exhibited by AA2024-T351/TCP/MgRP after exposure for 1,000 h in ASTM B117, indicating that any determent resulting from delayed galvanic protection of AA2024-T351 scribe was minimal. The sacrificial protection was enabled immediately after breakdown of pretreatments and the localized corrosion damage during the initial delay in galvanic coupling was minimal. Because of low pH (3.2±0.2), the acidified ASTM SOW/UV environment exposed scribe suffered a very large amount of localized corrosion, exhibiting pits as deep as 60 μm to 70 μm. Irrespective of the nature of pretreatment, this behavior was prominent in acidified ASTM B117 environment. Figures 11(a) and (b) summarize the corrosion volume loss for AA2024-T351/NFF/MgRP and AA2024-T351/TCP/MgRP without TC in the various environments indicated. Corrosion volume loss increased as a function of exposure time in all environments. The rate of loss of corrosion volume was higher in acidified ASTM B117 modified with SOW/UV environment and the least in field environments with moderate protection in ASTM B117. Non-topcoated conditions with both pretreatments perform better than control experiments conducted on bare AA2024-T351 without any coating in ASTM B117 environment, as shown in Figures 11(a) and (b), indicating that MgRP provides scribe protection in all environments except for the modified ASTM B117 environment. This indicates that these results can be correlated to the observations during post-exposure scribe chemical analysis using EDS and, hence, could be effective to determine scribe protection in environments such as CHO. The chemical markers approach is not feasible in CHO as a result of dissolution of corrosion products in scribe.

The low-frequency EIS and break point frequency of AA2024-T351/Pretreatment/MgRP/TC as a function of exposure time in different LALT/field environments are summarized in Figures 12 and 13, respectively. Irrespective of the nature of the pretreatment, the barrier low-frequency EIS at 0.01 Hz did not indicate significant coating degradation of AA2024-T351/Pretreatment/MgRP/TC in full immersion and standard ASTM B117. After 5,000 h exposure in field, the low-frequency EIS lowered by one to two orders of magnitude, indicative of moderate coating degradation. Substantial degradation in barrier properties of the coating was observed in modified ASTM B117 environment. Breakpoint frequency increased with time in chosen LALT/field exposures, suggesting porosity development. This could be primarily a result of degradation of TC polymer. Breakpoint frequency analysis indicates that the severity of coating degradation as a function of exposure environment is in the following order: full immersion (5% NaCl) < ASTM B117 < KSC ~ CHO < modified ASTM B117.

X-ray spectra of AA2024-T351/NFF/MgRP/TC after exposure in selected LALT/field environments are shown in Figure 14. Elemental Mg HCP <001>, <002>, and <101> peaks, rutile TiO₂ HCP <101>, <200>, <111>, and <210> peaks, and fcc Al <111> and <200> peaks were observed in initial spectra. The integral intensity of <200> peaks was normalized so as to compare the elemental Mg depletion as a function of exposure time in all chosen environments. There is no significant depletion of Mg in full immersion, ASTM B117, or field exposures. However, there is a partial depletion of Mg in the modified ASTM B117 environment. This indicates that self-corrosion of Mg is minimized in the presence of TC. Mg depletion was not complemented by detection of any crystalline corrosion products, as indicated in Figure 14. Similar Mg depletion trends were observed in other chosen AA2024-T351/Pretreatment/MgRP/TC systems, which were conversion coatings based. The results are summarized in Figure 15. Irrespective of the nature of the pretreatments, the Mg depletion trends were similar, indicating that the degradation of TC polymer is a predominant factor for subsequent Mg depletion in chosen environments.

The average OCP of AA2024-T351 and Mg in naturally aerated 5 wt% NaCl solution are −0.71 V and −1.48 V, respectively. In TCP pretreatment, the global galvanic protection potential is heavily mediated by the presence of resistive TC layers. Mg pigment remains in the MgRP coating and is not coupled to AA2024-T351 substrate and, hence, is not used up; therefore, global galvanic couple potential was above the OCP of AA2024-T351. Similar behavior of limited galvanic coupling and minimal depletion of Mg was observed all through the lifetime of coating for all chosen AA2024-T351/Pretreatment/MgRP/TC systems.

The global breakpoint frequency of the AA2024-T351/NFF/MgRP/TC system as a function of amount of Mg available is shown in Figure 16. There is neither appreciable coating degradation nor Mg depletion in full immersion, standard ASTM B117, and field environment. Modified ASTM B117 showed partial depletion of Mg and substantial degradation of coating barrier properties.

EDS elemental Mg maps of NFF/TCP pretreated AA2024-T351/MgRP/TC after different ASTM B117 exposure times showed no significant Mg presence in the scribe. The limited scratch protection is a result of resistive layers of TC which render galvanic coupling and Mg transport to the scribe more difficult. Similar observations were made for AA2024-T351/Pretreatment/MgRP/TC exposed to other LALT and field exposures. As no appreciable amount of Mg was observed in the scribe or coating adjacent to the scribe, further line profiles to quantify the amount of Mg in and near the scribe region were not conducted. It is to be noted that, from previous global Mg depletion trends by XRD (Figure 15), elemental Mg is still preserved beneath the TC polymers and would be available for sacrificial protection and Mg transport to the scribe once further degradation of the TC occurs.

Optical profilometry maps of the scribe exposing AA2024-T351 in NFF and TCP pretreated AA2024-T351/MgRP/TC after exposure in ASTM B117 for 1,000 h are indicated in Figure 17. The optical profilometry map of AA2024-T351/NFF/MgRP/TC after exposure for 1,000 h in ASTM B117, indicative of the moderate sacrificial corrosion protection provided by Mg across the scribe, as indicated by the presence of fewer and shallower pits in comparison to control experiments run for bare AA2024-T351. Similar behavior was exhibited by AA2024-T351/TCP/MgRP/TC after exposure for 1,000 h in ASTM B117, indicating that any determent in galvanic protection of the AA2024-T351 scribe is minimal. As a result of low pH (3.2±0.2), the modified ASTM environment exposed scribe suffered a very large amount of localized corrosion, exhibiting pits as deep as 60 μm to 70 μm. Irrespective of nature of pretreatment, this behavior was prominent in the modified ASTM B117 environment. Figures 18(a) and (b) summarize the corrosion volume loss for AA2024-T351/NFF/MgRP/TC and AA2024-T351/TCP/MgRP/TC with TC in the various environments indicated. Corrosion volume loss increased as a function of exposure time in all environments. The rate of increase of corrosion volume loss was the highest in acidified ASTM B117 modified with SOW/UV environment and the least in field environments with intermediate protection in ASTM B117. However, TC conditions with both pretreatments performed better than control experiments conducted for bare AA2024-T351 without any coating in ASTM B117 environment, as shown in Figures 18(a) and (b), indicating that MgRP provides scribe protection in all environments except for the modified ASTM B117 environment. In spite of no substantial corrosion product redeposition in the scribe as evident from EDS, in TC condition there was moderate sacrificial protection indicative of galvanic protection at the scribe after TC polymer degradation. Higher corrosion volume loss densities for the MgRP coated system with TC, in comparison to non-topcoated condition, is an indication of the TC limiting the sacrificial protection as a result of the additional resistance. There is a need for optimized TC barrier properties in a way that both scribe protection by cathodic protection and global barrier protection are balanced.

Tables 3 and 4 summarize observations concerning degradation times for global Mg and barrier properties for AA2024-T351 panels coated with MgRP and MgRP/polyurethane TC, respectively. The threshold value for |Z|_0.01 Hz and breakpoint frequency changes were approximated to be 1/100th of low-frequency impedance at 0 h and 10³ Hz, respectively, to assess the performance of the coating in selected LALT and field exposures. These values correspond to significant coating degradation and were utilized to compare the performance of the coating. There was no significant macroscopic blistering and coating delamination resulting from underpaint corrosion in the coating systems exposed to standard ASTM B117 and field exposures. However, the modified ASTM B117 with acidified SOW and UV showed significant blister formation. Global Mg depletion from XRD measurements (Figure 3) can be correlated to self-corrosion of Mg. Mg self-corrosion is minimized in systems with a TC (Figure 14). The coating degradation trends (Figures 1 and 2) for systems with MgRP and without a TC indicate that there is rapid degradation in barrier properties of coatings after exposure in ASTM B117 and modified ASTM B117 environments. The barrier properties’ degradation in field environments is similar, moderately severe initially and substantially high at the end of 1 y resulting from UV degradation of polymer (Figures 1 and 2). The coating degradation rate is relatively low in systems with a TC, indicative of the excellent barrier properties of polyurethane TC (Figures 12 and 13).

The post-exposure characterization of AA2024-T351/Pretreatment/MgRP and AA2024-T351/Pretreatment/MgRP/TC after exposure in different LALT/field exposures indicated that initially there was no sacrificial protection mechanism because of high electrical resistance imparted by pretreatments. There was a delayed sacrificial protection mechanism in the bare scribe once the resistance dropped (Figure 5[b]). The corrosion volume loss in the bare scribe was very minimal during this initial time and Mg corrosion products were seen in the scribe even though global galvanic protection potential indicated no initial sacrificial anode protection (Figure 5[b]). The protection could be a result of self-corrosion of Mg and redeposition of Mg²⁺ ion in the scribe as Mg(OH)₂ when local pH of the electrolyte in the scribe increased during Mg dissolution (Figure 8). The role of Mg corrosion products in residual protection of AA2024-T351 must be further investigated by electrochemically depositing these corrosion products on AA2024-T351 and studying the anodic and cathodic kinetics of AA2024-T351 in presence of Mg(OH)₂ and MgCO₃ rich films. This will be reported in future work.

The NFF pretreated AA2024-T351/MgRP exhibited galvanic protection from the beginning of exposure in both LALT and field, as indicated by global galvanic protection potential trends (Figure 5[a]). The scribe protection could be the result of a combination of sacrificial anode cathodic protection by coupling with MgRP and protection of the scratch by Mg(OH)₂ redeposition (Figure 8[b]). The conversion coatings pretreated AA2024-T351/MgRP had no galvanic protection at the beginning of exposure but exhibited delayed sacrificial protection in both LALT as well as field environments, as indicated by global galvanic protection potential trends (Figure 5[b]). The scribe protection could be a result of the combination of delayed sacrificial anode cathodic protection by coupling with MgRP and protection of the scratch by Mg(OH)₂ redeposition (Figure 8[c]). Two of the three conversion coatings examined are Cr-based and could provide some additional corrosion protection by leaching and supplying of Cr⁶⁺ species. However, EDS/Raman detected no presence of chromium-based oxide redeposition on the scribe. Moreover, further environmental exposure studies of pretreated AA2024-T351 without any coating would be of interest to understand any residual additional corrosion protection provided by the pretreatments. In the TC-based systems, both chemical analysis of the scribe as well as the global protection potential trends indicated limited scratch protection. However, corrosion volume loss analysis by optical profilometry indicates that they still have better corrosion protection than non-topcoated samples, indicating that in spite of limited ionic and Mg transport, TC-based systems do provide moderate scratch protection (Figure 18). The protection trends of these systems are summarized in Figures 19(a) and (b).

Based on the threshold |Z|_0.01 Hz and threshold breakpoint frequency in Table 2, for non-topcoated systems, the time required for coating degradation in standard and modified ASTM B117 environments is less than 48 h. In contrast, field exposure results indicated that substantial coating degradation occurs only after 1 y. Break point frequency trends of field exposures indicate significant pore development around 1,344 h. LALT are generally not recommended, especially in acidified B117, for a fairly accurate comparison to field exposures. Nevertheless, the acceleration factor with respect to time for equivalent degradation in LALT environment in comparison to field is roughly two orders of magnitude. The Mg depletion trends showed different acceleration factors in comparison to coating degradation trends. Standard ASTM B117 resulted in complete depletion of metallic Mg pigment after 1,000 h. Similar exposure studies in modified ASTM B117 environments and field indicated that exposure time required for complete depletion of Mg are 48 h and 8,736 h, respectively. This translates to an acceleration factor of 10 and 100 for standard ASTM B117 and modified ASTM B117, respectively. In the TC system, the coating degradation trends indicated that threshold breakpoint frequency was achieved only in modified ASTM B117, so a quantitative definition of the acceleration factor in TC systems cannot be accurately determined. The differences in environmental factors (Table 1) such as mean chloride ion concentration, pH, mean precipitation, and relative humidity could explain the difference in acceleration factors. Extent of scribe protection in standard ASTM B117 can be correlated to field environment with high chloride ion concentration, such as KSC. Modified ASTM B117 resulting from highly acidic environment cannot be directly correlated to the field exposures. To simulate an actual field environment that has episodes of drying and wetting, a modified ASTM B117 exposure with SOW and accelerated UV testing would need to be conducted with cycling for better correlation with field environments.

The “throwing power” pertains to the distance over which the MgRP coating system can protect bare AA2024-T351 by sacrificial anode based cathodic protection or other means of corrosion mitigation. Optical profilometry analysis of corrosion volume loss after exposure in different environments for MgRP-based systems in coated and uncoated conditions could be utilized to understand its efficacy in terms of scribe protection. Optical microscopy studies show very low corrosion volume loss in all environments studied except for acidic environments (Figures 11 and 18). Comparison of corrosion volume loss densities to control experiments in bare AA2024-T351 indicates that the corrosion volume loss densities are lowered by two to three orders of magnitude in non-topcoated conditions (Figure 11) and an order of magnitude for TC conditions (Figure 18), indicating they both provide corrosion protection to the whole width of the scribe. It should be noted that during a given episodic drying or wetting event, throwing power may be temporarily increased or diminished, making a definitive determination of throwing power difficult. At the end of the exposure, the definitive throwing power and inverse throwing power is complicated by corrosion during drying or isolated drop formation, which leads to attack of those areas, which were previously protected. Additional factors that complicate the determination of a throwing power via post-mortem characterization of environmentally exposed panels result from the chemical dissolution of species stable at high pH used as markers indicative of zones of cathodic protection (such as Mg(OH)₂ and CaCO₃), as well as difficulties in distinguishing between definitive regions of protection and substrate corrosion in defect areas. Chemical dissolution of the precipitates that are common to zones of cathodic protection is more likely in acidic, high time-of-wetness environments, like that of CHO, which is subject to regular acidic precipitation. EDS maps obtained throughout the width of the scribe after exposure at CHO showed very little indication of Mg or calcareous deposits common to regions of cathodic protection, making observation for throwing power in this environment difficult. Moreover, it is likely that the throwing power of the MgRP could be detected in EDS maps obtained throughout the width of the scribe after exposure at KSC, presumably resulting from the more alkaline exposure conditions (Figure 8[c]). Not only is the rain precipitation at KSC slightly alkaline as compared to CHO, but the proximity of the test racks to the ocean make the samples susceptible to spray from the ocean surf, which has a pH of roughly 8.2. This alkaline pH suppresses the chemical dissolution of species (such as Mg(OH)₂ and CaCO₃) used to identify zones of cathodic protection. Additionally, in most exposure environments and moderate pH ranges, aluminum is well known to form a barrier oxide film that reforms rapidly when damaged, leaving the primary form of attack in the scribe observed after exposure in most service environments to be non-uniform pitting corrosion.

• Full immersion in ambiently aerated 5% NaCl solution resulted in partial depletion of metallic Mg pigment in the AA2024-T351/Pretreatment/MgRP after 543 h. Exposure in ASTM B117 in 5% NaCl and modified ASTM B117 in ASTM acidified substitute ocean water with UV all resulted in complete depletion of metallic Mg pigment in the AA2024-T351/Pretreatment/MgRP without a topcoat far from the scribe after 1,000 h and 96 h, respectively. These harsh LALT environments are not necessarily recommended to assess field performance.

• Field exposures in CHO and KSC also resulted in complete depletion of metallic Mg pigment in the AA2024-T351/Pretreatment/MgRP without a topcoat far from the scribe after 1 y of exposure. It also resulted in complete degradation of polymer by UV in field environments. Testing without a topcoat is also not recommended unless the specific application does not involve a topcoat.

• The global galvanic protection potential of the AA2024-T351/NFF/MgRP system, with respect to remote scratches, increased slightly with exposure time in each environment, from initial values of approximately −1.0 V_SCE to −0.7 V_SCE after extensive environmental exposure. These potential values fall between the open-circuit potentials of bare AA2024-T351 (−0.7 V_SCE) and bare Mg (−1.6 V_SCE) and could be predicted by mixed potential theory. This suggests that Mg pigment is both electrically and ionically connected to the AA2024-T351 and can provide immediate sacrificial galvanic protection to the AA2024-T351 substrate.

• The global galvanic protection potential of the AA2024-T351/conversion coatings/MgRP system, with respect to remote scratches, was initially more positive, decreasing after initial exposure times and then shifting back to more positive values after longer exposure times. This suggests that Mg pigment is not initially electrically connected to the AA2024-T351. A resistive pretreatment can provide delayed sacrificial galvanic protection to the AA2024-T351 substrate as the pretreatment degrades over time.

• The low-frequency EIS measurement and breakpoint frequency analysis of AA2024-T351/Pretreatment/MgRP indicate rapid degradation of coating barrier properties in ASTM B117 and modified ASTM B117 in ASTM acidified substitute ocean water with UV environments. The barrier properties degradation are relatively moderate in full immersion in ambiently aerated 5% NaCl solution and field exposures in CHO and KSC.

• Raman spectroscopy of non-topcoated conditions indicate that Mg in the MgRP is converted to an outer layer of MgCO₃ both in LALT and field exposures and possibly an inner layer of Mg(OH)₂ because of the presence of ambient concentration of CO₂ in the environment.

• The throwing power measurements using SEM/EDS suggest that for all chosen pretreatments, AA2024-T351/Pretreatment/MgRP system could protect the entire half width of the scribe (~350 μm) in standard ASTM B117 and field exposures at KSC, as indicated by magnesium hydroxide detection across the scribe. Modified ASTM B117 in ASTM acidified substitute ocean water with UV environment indicated initial deposition of Mg(OH)₂ after 48 h followed by dissolution of Mg corrosion products, as well as leaching of Mg from AA2024-T351 substrate. Field exposures studies conducted at CHO indicated no presence of Mg corrosion products. The acidic nature of CHO environment resulted in dissolution of Mg corrosion products.

• Corrosion volume loss measured using optical profilometry increased as a function of exposure time in all environments. The rate of increase of corrosion volume loss was higher in acidified ASTM B117 modified with SOW/UV environment and the least in field environments with moderate protection in ASTM B117. Compared to bare AA2024-T351, AA2024-T351/Pretreatment/MgRP had lower corrosion volume loss for all chosen pretreatments.

• The high-performance advanced coating topcoat was observed to significantly hinder the depletion of Mg pigment from the MgRP, as well as the galvanic protection capabilities, in all exposure environments studied as compared to identical environmental exposures of non-topcoated samples. The barrier properties’ degradation was also significantly reduced in presence of topcoat in all environments. Only the modified ASTM B117 in ASTM acidified substitute ocean water with UV environment resulted in coating degradation of MgRP systems with topcoat as a result of acidic environment. The topcoat suppresses the MgRP scribe protection by heavily mediating the galvanic protection capabilities and Mg transport. Therefore, topcoat systems have a lower fraction of corrosion products and relatively higher corrosion volume loss in the scribe compared to the AA2024-T351/Pretreatment/MgRP system in similar environments.

This work was supported by the U.S. DoD OUSD Corrosion University Pilot Program under the direction of Daniel Dunmire and by the National Science Foundation under NSF DMR #0906663. This material is based on research sponsored by the U.S. Air Force Academy under agreement number FA7000-13-2-0020. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation thereon. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of the U.S. Air Force Academy or the U.S. Government NRL/JA/7330-2014-217. Members of NAVAIR¹ and Battelle Memorial Institute² are acknowledged for the generous supply and preparation of MgRP-coated AA2024-T351 panels, specifically: Craig Matzdorf,¹ Frank Pepe,¹ Jerry Curran,¹ and William Abbott.² Jason Lee of the U.S. Naval Research Lab is acknowledged for helpful discussions regarding throwing power. Andrew King, Raymond J. Santucci, and Drew Wolanski are acknowledged for helpful discussions regarding design of experiments for MgRP.