Investigation of the Trivalent Chromium Process Conversion Coating as a Sealant for Anodized AA2024-T3

All chemicals were analytical grade quality, or better, and were used without additional purification. Sodium sulfate (Na₂SO₄), sulfuric acid (H₂SO₄), and sodium chloride (NaCl) were purchased from Sigma-Aldrich (St. Louis, MO). Nitric acid (HNO₃) was purchased from Fisher Scientific (Hampton, NH) and 99.9% isopropyl alcohol was purchased from VWR chemicals (Ontario, Canada). The degreaser, Cleaner 1000, and TCP-HF (hereafter referred to as TCP) solutions were provided by CHEMEON Surface Technology. The chemical composition of the TCP bath has been discussed elsewhere.²⁷ The degreaser was prepared to a concentration of 45 g l⁻¹ and the deoxidizer, HNO₃, was used at 35 wt% when applying TCP. The pH of the TCP bath was adjusted to 3.85 using 0.25 M NaOH or 0.5 M H₂SO₄, as required. Nickel acetate salt was obtained from Sigma Aldrich. All solutions were prepared using ultrapure water (Barnstead water purification system) with a resistivity >17 MΩ-cm. No filtration step was employed to remove any suspended particulates from the TCP coating solution prior to sealing the anodized specimens.

Aluminum alloy 2024-T3 was procured as a 2 mm-thick sheet (www.metalsonline.com) and cut into 1in² squares for the experiments. The specimens were prepared for use by first abrading with 1500 grit wet aluminum oxide paper for 2 min This was followed by hand polishing with 0.3 μm diameter alumina grit (Buehler), slurried in ultrapure water, on a felt pad. A final hand polish was performed using 0.05 μm diameter alumina powder slurried in ultrapure water on a separate felt pad. A 20-min ultrasonic cleaning in ultrapure water followed by a 20-min ultrasonic cleaning in purified isopropyl alcohol was applied after each polishing step to remove debris and clean the surface. The polished specimens were then degreased in Cleaner 1000 at 55 °C for 10 min followed by a 2-min ultrapure water rinse. The degreased specimens were then deoxidized in 35 wt% HNO₃ for 2 min at room temperature followed by a 2-min ultrapure water rinse. The degreased and deoxidized specimens were dried with N₂ before being electrochemically anodized.

The degreased and deoxidized specimens were electrochemically anodized under controlled potential for a total of 23 min in 9.8 wt% H₂SO₄ under ambient conditions. The AA2024-T3 panels were the anode and a stainless-steel plate of the same dimension served as the cathode. The two electrodes were mounted vertically and parallel to one another using plastic alligator clips at a distance of 5 cm. The electrolysis was performed in a glass beaker. The aluminum alloy was connected to the positive terminal and the stainless-steel plate was connected to the negative terminal of a DC power supply (Tenma, 30 V and 5 A (150 W max.)). The electrodes were lowered into the acid to immerse three-quarters of their surface for anodization. Anodization was performed by applying a 1 V per 12 s ramp to 15 V DC for 3 min. This voltage was then maintained for 20 min resulting in a total anodization time of 23 min. The measured current density ranged from 25–35 mA cm⁻² during the electrolysis with a trend toward increasing values during the anodization period. The electrolyte was unstirred during the anodization. The anodized specimens were removed and first rinsed with ultrapure water for 2 min. This was followed by immersion in room temperature ultrapure water for 20 min to dissolve aluminum hydroxysulfate salts (anodization smut) from the surface. The cleaned specimens were dried with N₂ and stored overnight in a desiccator before further testing. The stainless-steel cathode was abraded with 1500 grit wet aluminum oxide paper to refresh the surface prior to each anodization run. Although the anodization was performed under ambient conditions, resistive heating caused the solution to warm during the process. The temperature rise was qualitatively observed and not quantitatively measured.

The anodized specimens were sealed by immersion with TCP. A sealing time of 10 min was used at room temperature. This sealing time was determined to be optimum based on electrochemical measurements of specimens sealed for different times in naturally aerated 0.5 M Na₂SO₄ + 0.01 M NaCl. Specimens were sealed using times from 1–20 min with 10 min providing the maximum corrosion resistance. Comparison studies were performed with anodized specimens sealed by immersion in hot water and with nickel acetate. Hot water sealing was performed using a 30-min immersion in ultrapure water at 90 °C. Nickel acetate sealing was performed by a 15-min immersion in a bath solution of 5 g l nickel acetate, 1 g l⁻¹ cobalt acetate, and 8 g l⁻¹ boric acid at 85 °C. The nickel acetate-sealed specimens received a final rinse with ultrapure water before further use.

The oxide coating weight was determined using the chromic acid stripping method described in ASTM B137. The anodized specimens were air dried for 24 h before chemically dissolving the oxide. The specimens were weighed dry before being placed in the stripping solution that consisted of 20 g CrO₃ and 35 ml of concentrated (85 wt%) phosphoric acid, both diluted with ultrapure water to 1 l. The specimens were immersed for 5-min periods at room temperature to dissolve the oxide. They were then rinsed with ultrapure water and dried with N₂ gas before remeasuring the weight. Typically, a total of 15 min was required to dissolve the oxide and to achieve an unchanging weight. The nominal oxide weight was found to be 1192 ± 43 mg ft⁻² (n = 3 anodized specimens). According to document MIL-A-8625F, the minimum oxide coating weight required for a Type II sulfuric acid anodization is 1000 mg ft⁻².⁷

Anodized AA2024-T3 specimens, TCP-sealed and unsealed, were exposed to a continuous 14-day (336 h) neutral salt-spray (NSS), according to ASTM B117 (5 wt% NaCl and 35 ± 1 °C), in a commercial 4 ft³ salt-spray chamber (MX 9204, Associated Environmental Systems). The salt solution in the chamber was replenished daily. The specimens were 1 in² in area and were mounted at a 20-degree angle relative to the vertical axis. The backside and edges of the specimens were masked off with corrosion protection tape (3 M Scotchrap™) to limit salt mist contact to just one surface with the anodic coating.

Anodized AA2024-T3 specimen surfaces, TCP sealed and unsealed, were continuously contacted with 3.5 wt% NaCl at room temperature for 5 days. Open circuit potential (OCP) and electrochemical impedance spectroscopy (EIS) measurements at the OCP were performed daily. The purpose for this test was to investigate the time dependence of any solution penetration through the sealed oxide to the underlying metal.

All electrochemical measurements were performed in a 1 cm² flat cell (Bio-Logic Science Instruments, France) using a computer-controlled electrochemical workstation (Gamry Instruments, Inc., Reference 600, Warminster, PA). An aluminum alloy specimen was mounted against a Viton® O-ring that defined the exposed geometric area, 1 cm². All currents reported herein are normalized to this geometric area. The counter electrode was a Pt flag and the reference was a home-made Ag/AgCl electrode (4 M KCl, E⁰ = +0.197 V vs NHE) that was housed in a Luggin capillary with a cracked glass tip. All measurements were made in naturally aerated 0.5 M Na₂SO₄ + 0.01 M NaCl at room temperature (23 ± 2 ^oC). The following electrochemical measurements were made on all specimens in this order: (i) a 30-min OCP vs time curves, (ii) EIS at the OCP using a 10 mV rms sine wave at frequencies from 10⁶ to 10⁻² Hz, and (iii) potentiodynamic polarization curves from the OCP (vs Ag/AgCl) to 0.5 V (anodic) and −1.2 V (cathodic) at a scan rate of 1 mV s⁻¹. Separate specimens were used for the individual anodic and cathodic potentiodynamic polarization curves, but all were first used in OCP and EIS measurements at the OCP. The polarization curve scans were initiated +20 mV vs OCP for the cathodic scans and −20 mV vs OCP for the anodic scans.

Scanning electron microscopy (SEM) was performed using a JEOL 7500 microscope (JEOL Ltd., Tokyo, Japan). Energy dispersive X-ray spectroscopy (EDS) was performed with an AZtec system detector (Oxford Instruments, UK) attached to the electron microscope. Spectral data were analyzed using the system software (version 3.1). Micrographs and spectra were recorded using a 15 keV accelerating voltage. The SEM with focused ion beam (FIB) milling was performed using an Auriga® dual column microscope (Composite Material and Structural Characterization Facility, MSU). A Ga ion beam was used to mill through the sample to obtain the cross sections for analysis. Sectioning was performed using 30 kV accelerating voltage at 4 nA. The ion beam current was applied at a 45^o angle relative to the normal axis. A 0.60 nA current was used for fine sectioning.

A digital optical microscope (Keyence VHX 600) was used to construct 3-dimensional surface topographical profiles of the specimens before and after the accelerated degradation testing. The depth resolution was ca. 0.2 μm. Areas 1000 × 1000 μm² were analyzed to determine the surface texture.

Glow discharge optical emission spectroscopy was performed to depth profile the chemical composition through the TCP-sealed oxide coating. The instrument used was a LECO Model GDS-850A. The spectral range probed was 120–800 nm. A sputter time of 2 min was used in the measurement with an apparent sputter rate of 0.5 μm min⁻¹.

The U.S. ARMY Corrosion Control and Prevention Group has established a scale for the assessment of corrosion damage and material degradation during accelerated degradation testing.³² This scale was used herein and is as follows: Stage 0 shows no visible corrosion; Stage 1 shows sample discoloration and staining; Stage 2 reveals loose isolated rust or corrosion product and early stage pitting of the surface along with minor etching; Stage 3 shows more extensive rust or corrosion product, minor etching, pitting and more extensive surface damage; and Stage 4 exhibits extensive rust or corrosion product formation, extensive etching, blistering, de-adhesion and pitting that has progressed to the point where the life of the specimen has been affected.

Figure 1 presents representative plan-view SEM micrographs of AA2024-T3 specimens (A) unanodized (degreased and deoxidized only), (B) anodized and unsealed, and (E) anodized and sealed with TCP. Micrographs of anodized specimen cross sections revealing the oxide layer thickness are presented in Figs. 1C and 1D. The primary features on the unanodized alloy are the rolling grooves that lay from the upper left to the lower right and etch pits decorating the surface (Fig. 1A). Many of these pits are formed during the deoxidation step. The micrograph for the anodized and unsealed specimen reveals a porous oxide coating (Fig. 1B). Electrochemical anodization of aluminum alloys produces a nanoporous oxide coating with pore diameters in the range of 10–50 nm, depending on the acid type and the process parameters.^33–38 The oxide coating thickness for these specimens is ca. 5 μm, as determined from SEM-FIB micrographs of cross sections like those presented in Figs. 1C and 1D. Specimens were milled at a 45^o angle relative to the vertical axis. SEM micrographs of the cross section were recorded at the same angle. The 5 μm oxide thickness was reproducibly achieved with the sulfuric acid anodization conditions employed and is in reasonable agreement with the estimated thickness of ca. 3 μm that calculated from the nominal coating weight and the density of aluminum oxide (3.95 g cm⁻³). Cross sectional analysis of the oxide coating (Fig. 1D) reveals a defect caused by a dislodged and entrapped second phase particle. Such micron-sized defects and heterogeneities have been previously observed in anodic coatings formed on 2xxx series aluminum alloys.^31,33,39,40 These defects have been attributed to the dissolution of copper-rich and Al₉FeNi intermetallic phases at the high potentials used for anodization in sulfuric acid and entrapment of particles within the oxide coating.³⁹ The SEM micrographs reveal that the TCP sealant forms across the surface of the oxide coating (Fig. 1E). There is evidence of cracking due to sealant dehydration in the vacuum environment of the SEM, so-called mud cracking. There is also evidence of coating bath precipitates on the surface of the sealant layer. These are the particulates seen on the surface of the TCP sealant layer and are demarked in the micrograph (Fig. 1E). These particulates have been confirmed by EDS analysis to be mixed deposits of ZrO₂ and Cr(OH)₃. The sealant cracking and coating precipitate particles are characteristic of other commercial TCP sealants as well (Fig. 1F).

Zoom In Zoom Out Reset image size

The SEM micrograph in Fig. 1B reveals the porosity of the anodic coating (ca. 20 nm pores), which is partially covered by aluminum hydroxysulfate deposits. These deposits form during the anodization and are incompletely removed by the initial 2-min immersion rinse in ultrapure water. EDS analysis revealed the deposits consist of Al, S and O. It was found that the deposits can be completely removed by a longer 20-min full immersion in room temperature ultrapure water with no agitation. Evidence for this removal is provided in the SEM micrographs presented in Fig. 2. For example, micrographs of an anodized specimen (A) after the 2-min initial rinse and (B) after the longer 20-min immersion soak reveal removal of the smut. Clearly, a 20-min ultrapure water soak removes the slow dissolving deposits and exposes the underlying porous oxide. Figure 2C presents a higher magnification micrograph of the deposit-free oxide coating in Fig. 2B that reveals the 10–20 nm diameter pore size.

Zoom In Zoom Out Reset image size

If not removed, the smut could cause weak adhesion with topcoats and incomplete sealing with TCP, or other sealants.

Figure 3 shows digital optical micrographs of an AA2024-T3 specimen anodized and sealed with TCP (A) after 16 h of laboratory atmosphere exposure and (B) after an additional 30-min vacuum exposure. The micrograph in Fig. 3 A reveals a crack-free sealant layer before the vacuum exposure. Figure 3B shows mud cracking of the sealant layer caused by dehydration in the vacuum. Figure 3 C presents an SEM-FIB micrograph in the cross section that reveals an interesting observation made on multiple specimens and that is a crack through the sealant layer, estimated to be in the 100–150 nm range, and into the outer oxide coating.

Zoom In Zoom Out Reset image size

The depth of the crack in the oxide coating is ∼1 μm. The cracking was observed only on the TCP-sealed specimens examined, not on unsealed anodic coatings. The cause for the oxide cracking is unclear but may be due to chemical degradation during the sealing process.

The micrographs in Figs. 1 and 3 reveal the sealant forms across the outer oxide surface. How deeply the sealant penetrates the oxide was assessed qualitatively from EDS elemental profiles of FIB-milled specimens in the cross section. The specimens were milled at a 45^o angle relative to the vertical axis to create the cross sections for analysis. Figure 4 shows elemental maps for an anodized AA2024-T3 specimen sealed with TCP. Presented are maps for Zr, Cr, F, Al with depth into the oxide (top to bottom in all panels). The elemental maps suggest that the sealant elements are located at the surface and within the oxide coating. Zr is detectable up to some 400 nm into the oxide. F and Cr are detected at a greater depth of some 800 nm. In addition to the conversion coating sealant (ZrO₂·nH₂O and Cr(OH)₃), it is reasonable to expect that hexafluoroaluminate and oxyfluoride chemical species, such as AlF₆³⁻, AlFO and AlF₂O, form due to reactions of fluoride with the oxide coating.

Zoom In Zoom Out Reset image size

Figure 5 presents preliminary GD-OES elemental profiles for Zr and Cr with depth into an anodized specimen sealed with TCP. The sputter depth is apparent and was not quantified directly by microscopy or profilometry. The profiles show that both sealant elements, Zr and Cr, extend from the surface into the oxide coating. The measurement was stopped at approximately 800 nm so these elements appear to extend deeper into the oxide. Future analysis will look this more completely. The Cr content is larger at the surface than in the oxide. The elemental ratio of Zr/Cr is ca. 2.0–2.5 with depth into the oxide. These results are consistent with the EDS data presented in the previous figure indicating that the sealant forms a layer on the outer surface of the oxide and penetrates the oxide pores up to depths at least approaching 1 μm.

Zoom In Zoom Out Reset image size

Figure 6A shows $\mathrm{log}\left|Z\right|\,{\rm{vs}}\mathrm{log}\left|f\right|$ plots for anodized AA2024-T3 specimens, unsealed and sealed with TCP, in naturally aerated 0.5 M Na₂SO₄ + 0.01 M NaCl. The measurements were made at the OCP. The low frequency impedance at Z_{0.01 Hz} is dominated by the charge transfer (R_ct) resistance; a direct measure of the corrosion resistance. The curves reveal a Z_{0.01 Hz} value (∼3 × 10⁶ ohm-cm²) for the anodized specimen sealed with TCP that is 5× greater than the value for the unsealed specimen (∼5 × 10⁵ ohm-cm²). Importantly, both are over 1000× higher than the value for a unanodized (degreased and deoxidized only) alloy.^{25–30,41–44} The remainder of the curves in the mid (capacitive) and high frequency (solution and electrode ohmic resistance) regions are overlapping with no distinct differences.

Zoom In Zoom Out Reset image size

Figure 6B shows phase angle ${\rm{vs}}\,\mathrm{log}\left|f\right|$ curves for the same specimens. The data indicate two time-dependent processes over the frequency range: one at ca. 1 kHz and a second at a lower frequency of ca. 1 Hz. The most distinct differences between the sealed and unsealed specimens are seen in the low frequency region. The phase angle for the TCP-sealed specimen at 0.01 Hz is ≥ −60° while the phase angle is considerably lower for the unsealed specimen with a value of −10°. The more negative phase angle for the TCP-sealed specimen is suggestive of greater capacitive behavior. The data suggest the lower frequency process is associated with the TCP sealant layer on the oxide surface and within the pores.

Figure 7 presents anodic and cathodic potentiodynamic polarization curves for anodized AA2024-T3 specimens, unsealed and sealed with TCP, in naturally aerated 0.5 M Na₂SO₄ + 0.01 M NaCl. The anodic curves shown in Fig. 7A reveal reduced current for the TCP-sealed specimen. The OCP values were ca. −0.28 V for the TCP- sealed and unsealed alloy specimens. The anodic current magnitude, for example at 0.1 V, is 10× lower for the TCP-sealed specimen, as compared to the unsealed anodized specimen. There is no clear onset potential for stable pit formation and growth in the curves for either specimen, at least to potentials as positive as 0.5 V. It should be noted that while the OCP values were similar for the separate specimens used for the anodic and cathodic polarization curves presented here, sometimes shifts were observed from the initial value when specimens were first used in the EIS measurements, usually in the negative direction.

Zoom In Zoom Out Reset image size

The cathodic curves in Fig. 7B reveal a decreased current for dissolved oxygen reduction at the TCP-sealed specimen, as compared to the unsealed anodized specimen. In fact, the current suppression across the entire potential range for oxygen reduction (−0.4 to −0.9 V) is 5× lower than the current for the unsealed anodized specimen.

Figure 8 presents a summary of the electrochemical parameters measured for anodized AA2024-T3 specimens, unsealed and TCP-sealed, measured in naturally aerated 0.5 M Na₂SO₄ + 0.01 M NaCl. Figure 8A shows that sealing the oxide layer with TCP coating shifts the nominal OCP towards more noble potentials by some 40 mV. There is, however, no statistically significant difference in the mean OCP values for the two specimen types given the magnitude of the variation in the data for both. The nominal anodic current density (J) at +0.1 V and the cathodic current density at −0.7 V, measured from the potentiodynamic polarization curves, are both suppressed by the TCP sealant. The similarity in the OCP values and the suppressed currents indicate the sealant provides both anodic and cathodic corrosion protection. Figure 8B reveals a 10× suppression of the anodic current density for the TCP-sealed specimens, as compared to the unsealed anodized controls. Figure 8C reveals a 5× suppression of the cathodic current density for the TCP-sealed specimens, as compared to the unsealed anodized controls. Notably, the variability in the measured values is less for TCP-sealed specimens. Finally, Fig. 8D reveals a nominal Z_0.01Hz increase by a factor of 5× after sealing the oxide with TCP.

Zoom In Zoom Out Reset image size

The electrochemical behavior of anodized (9.8 wt% H₂SO₄) specimens, unsealed and sealed with TCP, was compared with anodized specimens sealed in hot water and with nickel acetate. Table I presents a summary of the numerical data.

These preliminary data reveal the electrochemical behavior of the TCP-sealed specimens is similar to that of the nickel acetate-sealed ones. This is based on the similar magnitudes of the anodic and cathodic current densities, and the Z_{0.01 Hz} values. The current densities for the TCP-sealed specimens are nominally lower and the Z_{0.01 Hz} value is nominally higher than the values for the hot water-sealed specimens. This is suggestive of better corrosion resistance after TCP than after hot water sealing.

The electrochemical data presented above indicate the TCP sealant increases the corrosion resistance of the anodized specimens by a factor on the order of 10×. Accelerated degradation tests were conducted to assess the corrosion resistance of anodized AA2024-T3 specimens, unsealed and sealed with TCP, and to compare the results with the predictions from the electrochemical data. Figure 9 shows comparison of EIS data recorded for anodized AA2024-T3 specimens, unsealed and sealed with TCP, before and after 5-days of continuous contact with 3.5 wt% NaCl at room temperature. The test was performed in an electrochemical flat cell so that only the anodized surface was exposed to the solution. This test was performed to probe for solution penetration through the TCP sealant layer. Plots of $\mathrm{log}\left|Z\right|\,{\rm{vs}}\,\mathrm{log}\left|f\right|$ reveal that Z_{0.01 Hz} decreases by 10× for the unsealed specimen, actually, after just 3 days of exposure (∼10⁵ ohm-cm²) (Fig. 9A, top and left). The low frequency impedance for this specimen did not decrease any further through Day 5. The reduced low frequency impedance is likely due to Cl⁻ penetration through the oxide and attack of the barrier layer at the base of the pores as well as the underlying alloy. In contrast, Z_{0.01 Hz} is unchanged (∼10⁶ ohm-cm²) for the TCP-sealed specimen (Fig. 9B, top and right) during the 5-day period.

Zoom In Zoom Out Reset image size

Subtle differences in the phase shift ${\rm{vs}}\,\mathrm{log}\,\left|f\right|$ plots are seen for the sealed and unsealed specimens (Figs. 9A and 9B, bottom graphs). The unsealed specimen exhibited a decreased low frequency phase angle from −80 to −20^◦ after just 3 days of solution contact. There is also apparent loss of the two time-dependent processes resulting in one process at ca. 50 Hz. In contrast, the low frequency phase angle data for the TCP-sealed specimen remain largely unchanged. The two time-dependent processes are also unaffected by the solution exposure. The results indicate the TCP sealant provides a stable barrier layer that inhibits electrolyte solution penetration through the oxide to the base metal.

A 14-day NSS test was conducted, according to ASTM B117, to further assess the corrosion resistance provided by the TCP sealant on anodized AA2024-T3 specimens. Camera photographs of unsealed and TCP-sealed specimens, before and after 14 days, are presented in Fig. 10. The top photographs reveal significant corrosion damage on the unsealed specimen after 14 days. The arrows indicate regions with visible pitting and corrosion damage (Stage 2). In contrast, the TCP-sealed specimen shows no discoloration, pitting or corrosion damage (Stage 0). The unanodized region at the top of each specimen was covered with corrosion protection tape during the NSS test, so no corrosion is visible.

Zoom In Zoom Out Reset image size

Figure 11 presents (A) and (B) 3-dimensional and (C and D) 2-dimensional topographic maps of an unsealed (Figs. 11A and 11C) and TCP-sealed (Figs 11 and 11D) specimen after the 14-day salt-spray exposure. The profiles were generated from image analysis of digital optical micrographs obtained at different focal depths. Deeper regions (e.g., pits), relative to the surface, appear blue in color in the 2-dimensional representations. Second phase particles extending above the surface appear red in color. The unsealed specimen (Figs. 11A and 11C) has localized pitting and corrosion damage that developed in the area probed. This is evidenced by the presence of the 20–40 μm deep pits. The surface roughness (S_q, root mean square of height) over the 1000 × 1000 μm² image area increased ∼3× from 0.46 ± 0.04 to 1.18 ± 0.11 μm after the 14 days. In contrast, the 3- and 2-dimensional maps for the TCP-sealed specimen (Figs. 11B and 11D) show no indication of pitting or surface damage, and no increase in surface roughness. In fact, the surface roughness decreased ∼2× from 0.65 ± 0.08 to 0.37 ± 0.03 μm after the 14 days. This is partly because of the dissolution of the TCP coating precipitates that form on the sealant surface (white aggregates in Figs. 1E and 1F) and because of detachment of some of the sealant on the oxide surface during the test and the follow up cleaning.

Zoom In Zoom Out Reset image size

Evidence for the detachment of small areas of the sealant layer is presented in Fig. 12. An SEM micrograph of an anodized AA2024-T3 specimen, sealed with TCP, after the 14-day NSS test is presented. The micrograph reveals loss of the TCP sealant from the oxide surface during the test. Arrows mark where the sealant detached and where it remained intact. The damaged areas were small in dimension, generally on the order of a few μm². It is important to note that the specimen was subjected to post-processing cleaning treatments prior to performing the SEM analysis. The cleaning treatments included soaking the salt-spray exposed specimen in ultrapure water for 30 min to dissolve excess NaCl and then ultrasonic cleaning for 20 min in ultrapure water to remove any corrosion product. Control experiments were performed to determine if the sealant degradation was caused by the salt-spray exposure or by these post-processing steps. For example, SEM micrographs of an anodized specimen (data not shown) before and after the ultrapure water soak and ultrasonic cleaning revealed no damage to the sealant layer. Therefore, the loss appears to be due to degradation processes in the salt-spray environment. This needs to be investigated further. Even with this minor sealant degradation, there was no visible corrosion damage to the TCP-sealed specimens (see Fig. 10). This suggests that the sealant inside the oxide pores remains intact.

Zoom In Zoom Out Reset image size