Cooperative and synergistic corrosion inhibition of AA 7075-T6 by praseodymium and CaSO₄

2.1 Sample preparation

A large number of small polished samples were prepared for a full factorial free corrosion exposure study of Pr-CaSO₄ corrosion inhibition. Two-millimeter thick heat treated AA 7075-T6 sheet was cut into 5 × 15 mm strips. The short transverse face was used in this corrosion study because of the small grain size that orientation presents. The strips were mounted in the epoxy mold with the short transverse side facing down. The mold was ground with 120 grit silicon carbide paper until it was smooth and flat. Metallurgical polishing was carried out through a 0.05 micron finish in water-based diamond suspension. After being cleaned by ethyl alcohol and air dried, the polished surfaces were then protected with Kapton tape that left no residue after removal. Samples were then cut into a 2 mm thick wafer. The wafer was then broken off by hand. Tweezers were used to pick out a 2 × 2 × 5 mm sample from the broken epoxy mold. One mold produced 30–35 polished samples. This process was repeated until the desired number of samples was produced.

2.2 Free corrosion exposures

Free corrosion exposure experiments were carried out to characterize the effect of inhibitor concentration, pH and exposure time on pitting corrosion damage accumulation. Pr³⁺ was derived from reagent-grade PrCl₃^.7H₂O. Ca²⁺ was from reagent-grade CaSO₄^.2H₂O. The upper bound on the practical solubility of CaSO₄ has been reported to be 15–19 mM (Sawyer 1983), and the effective concentration of CaSO₄ used in the experiment was 5 mM. The Pr concentration used was 1.06 mM and was selected based on prior work (Chilukuri 2012).

For the experimental matrix, a 0.1 M NaCl solution was used as a basis solution to which three concentration levels of Pr were added; 0.71, 1.71, 3.55 mM, one concentration level of CaSO₄; 5 mM, four levels of pH; 3, 5, 8, 10, and three different times; 1, 10, and 30 days. In addition, 12 baseline control exposures with 0.1 M NaCl at four levels of pH and three different exposure times were included. A total of 48 samples were tested including baseline control samples. The pH adjustment was made by titrating concentrated HCl and NaOH until target pH was reached. Solutions were not buffered and the pH was not monitored during exposure. The polished samples were exposed in 20 ml vials. At the conclusion of the exposure experiments, the samples were removed, cleaned with 30% v/v nitric acid with 5 min in an ultrasonic bath, and dried. They were then characterized using Veeco ContourGT-K1 optical profilometer to study pitting damage accumulation.

2.3 Optical profilometry

A Veeco ContourGT-K1 optical profilometer (OP) was used to characterize pit morphology and quantify pit size and depth. The OP uses white light interferometry to detect surface topography with high spatial and depth resolution. (Deck and Degroot 1994, Frantziskonis et al. 2000). The on-board analysis software used was Vision64. The mode was vertical scanning interferometry (VSI). All images were scanned at 1x rate. The Multiple Region© tool was used to measure pit depth, number of pits, pit volume, pit area, and maximum area-equivalent diameter. Pit number density was calculated by counting the number of pits divided by surface area. The unit of pit number density is counts/mm². Pit area was calculated by determining total pit mouth area and divided by total area. The area-equivalent diameter of a pit is the diameter of a circle whose area is equal to that of the irregularly shaped pit mouth area. Three different areas were analyzed for each sample. Pits that were shallower than 0.5 µm and smaller than 10 µm² were excluded from the analysis in order to distinguish pits from scratches and surface defects.

2.4 Electrochemical experiments

All the electrochemical analysis was conducted using a Gamry Instruments^® potentiostat electrochemical impedance spectrometry (EIS) was performed by scanning from 0.1 MHz to 0.01 Hz with the sinusoidal voltage signal of 10 mV. Potentiodynamic polarizations were conducted using a 0.167 mV/s scan rate. All experiments were delayed until the open circuit potential (OCP) was stable. Cathodic polarization started the scan from OCP to 1 V below OCP. All reported potentials were referenced to the saturated calomel electrode (SCE). The counter electrode was platinum. Polarization experiments were repeated at least three times while EIS experiments were only done once due to time-sensitive measurement setup. The ones shown are the typical behavior for each condition.

2.5 X-ray photoelectron spectroscopy

X-ray photoelectron spectroscopy was conducted using a Kratos Axis Ultra with a monochromated Al X-ray source. Emission voltage to generate X-ray beam was 12 kV. The analyzer was set at 10 mA in spectrum mode. The lens mode was hybrid and the aperture setting was slot. To study the film on the surface, the analysis was focused on O, Ca, Al, and Zn. Samples were exposed in 0.1 M NaCl solution with or without inhibitor to compare oxidation states of elements on the surface. All the data analysis was carried out using CasaXPS software. All the calibration and peak fitting were done using the NIST XPS database (Naumkin et al. 2012). All data were calibrated to the adventitious C 1s peak at 284.5 eV. The binding energy reported in this study was rounded to the nearest tenth of an eV. The precision range of assigning binding energy was within ±0.2 eV.

2.6 Titration curves

Titration experiments were carried out using 100 mL of analyte solution. Unless indicated, the concentration of NaCl in solution was always 0.1 M. The Pr³⁺ concentration was 1.77 mM and the CaSO₄ concentration was 5 mM. A titrant solution comprising 0.1 M NaOH was used. A small increment of titrant was added to the analyte solution from a 25 mL burette while a pH meter was used to monitor the change of pH. All experiments were carried out in triplicate to ensure reproducibility.

3.1 Contour plots

Forty eight samples were analyzed by OP. Maximum pit depth, pit number density, pit volume, maximum area-equivalent diameter and pit area were extracted using the Multiple Region© tool. Comparing pit characteristics with the baseline control exposures in 0.1 M NaCl solution, inhibiting effects were characterized.

Contour plots are used to present pit characteristic data for clear comparison. It was formed by nine data points. In these plots, the y-axis is concentration of Pr. The x-axis is exposure time. The color-coded scale is the magnitude of designated pit parameter. Four graphs are shown in each figure and the layout is as follows: (a) pH 3 is in the top left corner. (b) pH 5 is in the top right corner. (c) pH 8 is in the bottom left, and (d) pH 10 is the bottom right. The baseline control condition (0.1 M NaCl only) is shown as 0 mM of Pr on the x-axis. The scale was identical for easy comparison unless indicated. The black dots indicate data points that were used to form the contour plots.

In this paper, “synergistic” inhibition means that one inhibitor alone is not effective, but when two inhibitors are together, it shows some inhibition. “Cooperative” inhibition means that one inhibitor alone is effective, but when two inhibitors are together, the effect of inhibition is greater.

3.1.1 The effect of inhibitors on maximum pit depth

Pitting was suppressed by the presence of Pr at all three concentrations examined. Across the range of pH investigated, Pr alone works the best at pH 8, showing the smallest maximum pit depth (Figure 1). CaSO₄ helps Pr in reducing maximum pit depth from more than 10 µm to less than 2 µm at pH 3, 5, and 8 especially at 30 days (Figure 2).

Figure 1:

Contour plots of maximum pit depth in the presence of Pr only in 0.1 M NaCl solution at pH 3 (a), pH 5 (b), pH 8 (c), pH 10 (d).

Figure 2:

Contour plots of maximum pit depth in the presence of Pr and CaSO₄ in 0.1 M NaCl solution at pH 3 (a), pH 5 (b), pH 8 (c), pH 10 (d).

Figure 1 shows that, at pH 3 with Pr alone, the deepest pit was around 12 µm and it occurred in a 3.55 mM Pr solution. At pH 3 and 5, maximum pit depth was not greater than 5 µm except at 3.55 mM Pr exposed at the long exposure times (10 and 30 days). Maximum pit depth was shallow at pH 8 at all Pr concentrations examined. At pH 10, no inhibition was observed. Maximum pit depth was greater than 20 µm, which was the deepest pit depth observed in these experiments.

Figure 2 shows contour plots of maximum pit depth with additions of both Pr and CaSO₄. CaSO₄ helped suppress pit growth at pH 3, 5, and 8. Maximum pit depth at those pH values was lower than 2 µm in the presence of both Pr and CaSO₄.

3.1.2 Pit number density

In the presence of Pr alone, overall lowest pit number density occurred at pH 5 (less than 50 mm⁻²). Inhibition by the combination of Pr and CaSO₄ was judged to be effective throughout the range of pH values examined because pit number density was less than 10 mm⁻². Pit number density (number of pits per unit area) characterizes the incidence of pitting across the surface of the sample. In general, the pit number density depended on the pH and the Pr concentration in solution. Baseline control exhibited pit number density as high as 200 to 750 mm⁻² at 30 day exposure. The lowest pit number density was observed at Pr concentration of 1.77 mM. Figure 3 shows the effect of Pr alone. It should be noted that Figure 3d is scaled differently from Figure 3a, b, and c because pit number density at pH 10 was very high and much larger than that observed at other pH values (more than 1000 mm⁻²). At pH 3, a high pit density was observed at a Pr concentration of 0.71 mM, but pits were shallow as indicated in Figure 1. At pH 5, pit density was reduced significantly compared to baseline control at 30 days, especially at a Pr concentration of 1.77 mM. Pit number density was lowest at pH 5. At pH 8, pit density increased with increasing Pr concentration. Figure 4 shows the effect of both Pr and CaSO₄. The presence of CaSO₄ lowered pit number density to be lower than 10 mm⁻² at all pH values examined, but at pH 10, the pit density was still very high, around 1000 mm⁻².

Figure 3:

Contour plots of pit density (counts/mm²) in the presence of Pr only in 0.1 M NaCl solution at pH 3 (a), pH 5 (b), pH 8 (c), pH 10 (d). The scale in (d) pH 10 was not normalized.

Figure 4:

Contour plots of pit density (count/mm²) in the presence of Pr and CaSO₄ in 0.1 M NaCl solution at pH 3 (a), pH 5 (b), pH 8 (c), pH 10 (d). The scale in (d) pH 10 was not normalized.

3.2 Visual observation

Immersion testing was carried out to cross correlate qualitative visual observations with quantitative characterization of localized corrosion damage made by OP. The OP data suggests that inhibition occurred at pH 5 and 8, therefore only these two pH values were investigated. Figure 5 shows optical micrographs of 7075-T6 samples that were exposed in 0.1 M NaCl solution with and without the inhibitors for 5 days at pH 5 and 8. Concentrations of inhibitors if present were 1.77 mM of Pr and 5 mM of CaSO₄. The concentration of Pr was 1.77 mM because pit parameters indicated that corrosion damage was the least at this concentration. One day exposure did not reveal much of the inhibiting effect, but the inhibition can be observed in 10-day and 30-day exposure samples. This implies that it takes time for corrosion inhibitor to take effect. Five-day exposures were performed to ensure that the inhibitors had time to act.

Figure 5:

Comparison of optical micrographs of AA 7075-T6 exposed in 0.1 M NaCl solution with or without inhibitors in pH 5 and 8 for 5 days. Concentration of Pr³⁺ was 1.77 mM and concentration of CaSO₄ was 5 mM.

As expected, the sample surface exposed to solution without the inhibitors showed a dark brown color with white cloudy powder observed on the surface after immersion. This is a typical appearance for aluminum alloys after corrosion damage occurred. Figure 5a and b shows significant pitting on the surface. Samples exposed to pH 5 in both Pr alone and Pr and CaSO₄ together showed a shiny and smooth finish similar to an as-polished surface. As illustrated in Figure 5c and e, optical micrographs and OP images do not suggest any significant localized corrosion. However, at pH 8, Pr alone showed severe pitting as shown in Figure 5d. With the presence of both Pr and CaSO₄, sample surface looked shiny and smooth. Figure 5f shows no pitting corrosion and a shiny finish.

3.3 Polarization measurements

3.3.1 Cathodic polarization

Cathodic polarization curves were collected in triplicate on 7075-T6 with and without the Pr and CaSO₄ inhibitors at pH 3, 5, 8, and 10 (Figure 6). Three conditions were studied: baseline control (0.1 M NaCl only), Pr alone (0.1 M NaCl and 1.77 mM Pr), Pr and CaSO₄ (0.1 M NaCl, 1.77 mM Pr, and 5 mM CaSO₄). The cathodic polarization response was similar to that of the baseline control for all cases except at pH 5 where Pr alone and Pr plus CaSO₄ reduced the cathodic current by an order of magnitude. At pH 8, Pr alone did not show cathodic inhibition, but Pr and CaSO₄ reduced cathodic current slightly. At pH 10, cathodic polarization curves do not suggest any inhibition.

Figure 6:

Cathodic polarization curves of 7075-T6 in 0.1 M NaCl baseline solution with or without inhibitors in various pH. (a) pH 3, (b) pH 5, (c) pH 8 and (d) pH 10.

3.3.2 Anodic polarization

Anodic polarization was carried out in aerated 0.1 M NaCl with and without inhibitor additions to reveal anodic inhibition, if any. At pH 3 and 10, evidence of anodic inhibition was not observed in the polarization response, so the experiments were not carried beyond a preliminary stage. Figure 7 shows anodic polarization curves of samples exposed to 0.1 M NaCl solution with or without the inhibitors at pH 5 and 8 because indications of inhibition were observed at these two pH values in free corrosion experiments. Anodic inhibition was not detected at pH 5, but at pH 8 with additions of both Pr and CaSO₄, an increase in the pitting potential and a decrease in the corrosion potential were occasionally, but not always observed. Figure 8 shows the cumulative probability distribution plot of E_pit from replicate experiments. The observed pitting potential ranged from −0.740 V_SCE, which is a value consistent with no inhibition of pitting, to −0.580 V_SCE, which is indicative of appreciable inhibition of pitting.

Figure 7:

Anodic polarization curves of 7075-T6 exposed in 0.1 M NaCl solution with or without inhibitors in (a) pH 5 and (b) pH 8.

Figure 8:

Cumulative distribution plot of pitting potentials of 7075-T6 in 0.1 M NaCl solution with Pr and CaSO₄ at pH 8.

3.4 Electrochemical impedance spectroscopy

Electrochemical impedance spectroscopy (EIS) was carried out using a 10 mV_rms sinusoidal voltage perturbation about the open circuit potential. The polarization resistance was obtained from fitting an equivalent circuit model to the EIS data (Mansfeld 1995). A simple equivalent circuit model was used to fit the data. The circuit was consisted of a solution resistance in series with a constant phase element (CPE) and a polarization resistance in parallel.

It is important to note that the Nyquist plots show an inductive loop in all cases except Pr and CaSO₄ at pH 8, which showed a synergistic effect. Figure 9 shows examples of Nyquist plots from EIS measurements collected as a function of exposure time. Only the Nyquist plot of the sample with Pr and CaSO₄ additions at pH 8 shows a diffusion response at low frequencies. The difference in appearance of the Nyquist plot at low frequency will be discussed in the next section.

Figure 9:

Examples of Nyquist plots at pH 5 and 8 with or without inhibitors. Inductance loops appear in all condition except (f) pH 8 Pr + CaSO₄ which show synergistic inhibition. R_p increased with time. The exposed area was 5.5 cm².

The EIS results showed that sustainable protective film was detected at pH 8 with Pr and CaSO₄ together. EIS measurements were performed during the first 24 h at pH 5 and 8 in 0.1 M NaCl solution with or without inhibitors. The results are shown in Figure 10a and b for pH 5 and 8 respectively. At pH 5, R_p of the case with Pr and CaSO₄ was 1.5 orders of magnitude higher than the baseline control condition initially, but a protection film broke down quickly within 12 h. Pr alone at pH 5 showed a slight inhibition, but R_p quickly fell to the same range as R_p of the baseline control. CaSO₄ alone did show a sign of inhibition at pH 5. Figure 10b shows R_p with different inhibitors at pH 8. Pr alone or CaSO₄ alone did not show inhibition at pH 8. Pr and CaSO₄ together showed some corrosion protection that increased over time. The EIS measurements were continued in the Pr and CaSO₄ case at pH 8 to see how long the protective film would last. This protective film broke down after 6 days of exposure when R_p dropped sharply as shown in Figure 10c.

Figure 10:

Polarization resistance of 7075-T6 in 0.1 M NaCl solution with various inhibitors at (a) pH 5 and (b) pH 8. (c) polarization resistance shows a breakdown of protective film after 6 days.

3.5 X-ray photoelectron spectroscopy

After 5 days of immersion, samples exposed in different conditions were analyzed by X-ray photoelectron spectroscopy (XPS). The concentrations in this experiment were 0.1 M of NaCl, 1.77 mM of Pr, and 5 mM of CaSO₄. The conditions tested were 1) NaCl at pH 5, 2) NaCl + Pr at pH 5, 3) NaCl + Pr + CaSO₄ at pH 5, 4) NaCl +Pr at pH 8, 5) NaCl + Pr + CaSO₄ at pH 8, 6) NaCl + Pr at pH 10.

3.5.1 Appearance of Cu and Zn on samples that showed no inhibition

Cu 2p and Zn 2p peaks were present only when no inhibition was observed. Cu and Zn are alloying elements in 7075-T6, which implies that the protective film was very thin or non-existent allowing these elements to enrich at the corroding surfaces. Two conditions were chosen to represent the uninhibited case. One condition was 0.1 M NaCl at pH 5 and the other was 0.1 M NaCl plus 1.77 mM PrCl₃ at pH 10. In both cases, no inhibition was observed during free corrosion experiments. The peak separation was 19.9 eV which corresponds to peak separation of Cu 2p. Additional evidence to support the presence of Cu was that rare earth metal oxides displayed characteristic satellite peak at 929.8 eV, which were not found in pH 5 NaCl (Burroughs et al. 1976, Ogasawara et al. 1991). Figure 11b shows that Zn 2p peak appeared at 1021.8 eV and corresponds to metallic Zn (Deroubaix and Marcus 1992). The Zn 2p peak was prominent in NaCl at pH 5 but was not found in samples exposed in 0.1 M NaCl solution with Pr alone or Pr and CaSO₄ at pH 5.

Figure 11:

XPS analysis shows that Zn and Cu peaks were detected in samples exposed in 0.1 M NaCl solution at pH 5 (labeled “pH 5 NaCl”) and 0.1 M NaCl plus 1.77 mM Pr³⁺ at pH 10 (labeled “pH 10 Pr”) where no inhibition was observed.

On the surface of the sample exposed to 0.1 M NaCl solution and 1.77 mM Pr at pH 10, the Pr 3d and Cu 2p and peaks were overlapped in Figure 11c. The peak separation was 19.8 eV which matched to Cu 2p peak separation. Pr 4d peaks were absent as shown in Figure 11d. There was some residual of Pr ions on the surface after exposure, but did not contribute to the inhibition. Figure 11d shows that, on the survey scan, Pr 4d peaks were not present on sample exposed to 0.1 M NaCl and 1.77 mM Pr at pH 10. Zn peaks were also prominent as shown in Figure 11e.

3.5.2 XPS spectra of samples exposed to pH 5 solutions

Figure 12 shows overlaid spectra of Pr 3d, O 1s, and Al 2p in pH 5 conditions with or without the inhibitors. In addition to the Cu 2p and Zn 2p peaks that appeared on sample exposed to 0.1 M NaCl at pH 5, an O 1s peak presented at 531.6 eV corresponding to Al₂O₃ (Nefedov et al. 1975). Al 2p peaks appeared at 74.1 eV which corresponded to Al₂O₃ (Wagner et al. 1982). Peaks that show in Figure 14a corresponds to Cu as discussed above.

Figure 12:

XPS spectra revealed a mixture of Al, O, Pr, species on the surface of 7075-T6 exposed in 0.1 M NaCl baseline solution with 1.77 mM Pr and 5 mM CaSO₄ at pH 5. Cooperative inhibition was observed.

In the case of the sample exposed to 1.77 mM Pr at pH 5, an O 1s peak was observed at 531.8 eV. The Al 2p spectra showed a peak at 74.2 eV corresponding to Al(OH)₃ (Lindsay et al. 1973, Nylund and Olefjord 1994). Pr 3d peaks showed up at 933.6 eV and a characteristic satellite peak appeared at 929.5 eV. The match with the literature suggests Pr-hydroxy carbonate (Lopez-Garrity and Frankel 2014).

On the sample surface exposed to 1.77 mM Pr and 5 mM CaSO₄ in 0.1 M NaCl solution at pH 5, an O 1s peak was found at 531.9 eV, which encompasses hydroxide or carbonate species (John et al. 1992). Two Al 2p peaks appeared at 74.5 and 71.3 eV which can be attributed to Al₂O₃ (Pashutski et al. 1989) or Al(OH)₃(Rotole and Sherwood 1999) and Al metal respectively(Carley and Roberts 1978). S 2p peaks appeared in conditions where CaSO₄ were present at 169.3 eV. S 2p peaks can be ascribed to sulfate (Siriwardane and Cook 1986).

3.5.3 XPS spectra of samples exposed at pH 8 solutions

Figure 13 shows overlaid spectra of samples exposed in pH 8 with Pr alone and Pr and CaSO₄ together. At pH 8 of sample exposed to Pr alone, an O 1s peak appeared at 531.3 eV which encompasses hydroxides, carbonates, and Al₂O₃ (John et al. 1992). Al 2p peaks appeared at 73.9 eV corresponding to Al₂O₃ (Barr 1983). Pr 3d peaks correspond to Pr₂O₃.

Figure 13:

XPS spectra of samples exposed in 0.1 M NaCl with 1.77 mM Pr³⁺ at pH 8 (labeled “pH 8 Pr”) and 0.1 M NaCl with 1.77 mM Pr and 5 mM CaSO₄ at pH 8 (labeled “pH 8 Pr + CaSO₄”).

In the case of the sample exposed to Pr and CaSO₄ at pH 8, an O 1s peak appeared at 531.9 eV which is likely to be hydroxide or sulfate. The binding energy of O 1s peak in pH 8 Pr and CaSO₄ was higher than all other conditions. Al 2p peaks appeared at 74.8 eV which is consistent with a literature value of Al(OH)₃ (Carley and Roberts 1978). Pr 3d peaks were found at 933.9 and 929.9 eV which are slightly higher than binding energy of Pr₂O₃, but corroborating reports in the literature were not found.

3.6 Predominance diagrams

Predominance diagrams in aqueous solutions were created using Medusa^® thermodynamic software in order to predict the tendency of precipitating compounds at some pH ranges. The input parameters were derived from the experimental conditions, which were 0.1 M NaCl, 1.77 mM PrCl₃, 5 mM CaSO₄, 0.04% atmospheric CO₂. For the Pr stability diagram, the precipitated compound depended on CO₂ concentration as shown in Figures 14 and 15, with and without CO₂ respectively. In the presence of CO₂, the Pr is predicted to form Pr₂(CO₃)₃ at pH 6 or higher. In the absence of CO₂, Pr is predicted to form Pr(OH)₃ at pH 7.7 or higher. Figure 16 shows predominant species diagrams for sulfate in function of pH. The labels “c” or “cr” in parenthesis denote solid crystalline forms suggesting precipitation in certain pH ranges. In pH 2.5–4.7, aluminum hydroxysulfate is predominant species.

Figure 14:

Chemical stability diagram for praseodymium in aqueous solution with the presence of CO₂.

Figure 15:

Chemical stability diagram for praseodymium in aqueous solution without the presence of CO₂.

Figure 16:

Chemical stability diagram for sulfate with the presence of Al³⁺ ions.

3.7 Titration curves

Titration experiments were conducted to verify the hypothesis that CaSO₄ helps with Pr precipitation. Figure 17 shows titration curves of four conditions that were tested. It should be noted that every solution has 0.1 M NaCl as a base solution. The conditions were: 1) 0.1 M NaCl only (baseline control), 2) 5 mM CaSO₄, 3) 1.77 mM Pr, 4) 1.77 mM Pr and 5 mM CaSO₄.

Figure 17:

Titration curves of solutions contain inhibitors compared to baseline control.

In the case of the baseline control, buffering was not observed. The pH value increased to about 10 and stabilized. Adding more bases only increased pH slightly. The solution was clear with no detectable precipitation. The CaSO₄-only solution was also clear and the curve was similar with NaCl-only curve. This showed that CaSO₄ does not precipitate when pH increases.

In the presence of Pr alone, there was a buffering effect when Pr³⁺ reacted with OH⁻ to precipitate. During the precipitation process, the pH increase was mitigated and a white solid precipitate was observed. The equivalence point was reached when all precipitation was formed, causing a sharp increase at around 4.5 mL of 0.1 M NaOH added before the second plateau was observed.

With the presence of both Pr and CaSO₄, the equivalence point was shifted to a lower value of NaOH added. The equivalence point was at around 4.0 mL of 0.1 M NaOH. Reproducibility of results was excellent as shown in Figure 18. This suggests that CaSO₄ plays a role in Pr precipitation. Adding more SO₄²⁻ did not shift the curves farther to the left when experiment was done with 10 mM Na₂SO₄ instead of 5 mM CaSO₄. A minimum of 5 mM CaSO₄ was needed to aid Pr precipitation according to the experiment. The titration experiment can be used to study precipitation by inhibitors.

Figure 18:

Reproducibility of the titration curves was very consistent.

4.1 Free corrosion exposure

To develop an overall assessment of the effectiveness of the Pr-CaSO₄ inhibitor pair, ratings of the extent of inhibition were developed from examination of the pit characteristics by optical profilometry. Five rating levels were assigned to each characteristic: very poor, poor, fair, good, and excellent. Ratings were assigned based on the overall appearance of the contour plots compared to other pH and baseline control and therefore include a degree of subjectivity. For example, for maximum pit depth, if the contour plot shows that majority of the area is dark blue, which means maximum pit depth is shallower than 5 µm, and the inhibition is judged to be “excellent”. If the contour plot shows some green which is in the middle of scale bar, the inhibition is judged to be “fair”. If the contour plot shows majority in red or grey, the inhibition is judged to be “very poor”. Table 1 summarizes the ratings of Pr alone and Pr and CaSO₄.

As mentioned above, the term “cooperative effect” refers to inhibition that appears to be the additive sum of inhibition by Pr and CaSO₄ each separately. The term “synergistic effect” refers to inhibition of corrosion by both Pr and CaSO₄ together, but inhibition does not exist when either Pr or CaSO₄ is present alone.

The quantitative OP characterization of corrosion damage and visual assessment correlate reasonably well. The contour plots and the visual observation at pH 5 and 8 both showed that Pr alone worked well at pH 5, but not at pH 8 and a synergistic inhibiting effect between Pr and CaSO₄ was detected only at pH 8. The contour plots collected at pH 5 and 8 with the presence of both Pr and CaSO₄ were rated “excellent” for all pit parameters. At pH 3, Pr alone does not inhibit localized corrosion, but when CaSO₄ was added, it shows some inhibition. This could be due to that fact that CaSO₄ passivates the surface and stifles pitting at pH 3 as shown in a separate report (Klomjit and Buchheit 2016). Samples exposed to pH 5 with Pr alone show some inhibition except maximum pit depth which showed a “fair” rating and a “good” pit density rating. When CaSO₄ was added, pitting corrosion was suppressed. All pit characteristics show “excellent” rating at pH 5 with Pr and CaSO₄. Samples exposed to pH 8 solutions with Pr alone show some inconsistency in pit parameters with some “good,” “fair,” and “poor” ratings. Interestingly, when CaSO₄ was added to Pr at pH 8, all pit parameters showed “excellent” rating. This was judged to be a synergistic effect of Pr and CaSO₄ because Pr alone does not suggest inhibition. At pH 10, the results do not suggest any inhibition in either Pr alone or Pr and CaSO₄. The visual observation at pH 5 and 8 has also confirmed the findings from the quantification of pitting damage. Cooperative inhibition was observed at pH 5 and synergistic inhibition at pH 8 according to all free corrosion exposure.

4.2 Electrochemical experiments

Electrochemical experiments were performed to substantiate the findings in free corrosion exposure. Polarization curves showed a slight reduction in cathodic reaction rates at pH 5 and elevation in the pitting potential when Pr and CaSO₄ were present at pH 8. Cathodic polarization curves at pH 3 and pH 10 did not suggest any reduction in cathodic kinetics. This agrees with free corrosion experiments in which inhibition was not observed at pH 10 and our earlier study that showed CaSO₄ did not reduce cathodic current at pH 3, though pitting was suppressed by passivation (Klomjit and Buchheit 2016). Synergistic inhibition was not observed at pH 5. Pr precipitation was likely responsible for the reduction in cathodic current in both cases. Anodic polarization curves showed that at pH 8, Pr and CaSO₄ together extend the passive region. This confirms some synergistic inhibition at pH 8 because Pr alone did not suggest inhibition in anodic and cathodic polarization.

Polarization resistance from EIS also confirmed the cooperative and synergistic effect of Pr and CaSO₄. A slight inhibition was observed at pH 5 with Pr alone. Polarization resistance of Pr and CaSO₄ at pH 5 was 1.5 orders of magnitude higher. Some synergistic inhibition was found at pH 8 Pr and CaSO₄. The shape of the Nyquist plot is particularly different than other conditions (the absence of an inductive loop at low frequency). An inductive loop was observed because the anodic part of the sinusoidal cause localized corrosion and creating current spikes. E_corr and E_pit of aluminum alloys in aerated condition are really close together, so the non-linearity in EIS measurement can cause the inductive loop (Mansfeld and Fernandes 1993). Diffusion tails were observed only in pH 8 Pr + CaSO₄ where synergistic inhibition was observed in the free corrosion experiment. The diffusion tails can be attributed to a passive film that blocks diffusion of ions and oxygen.

4.3 Origin of inhibition

By comparing with conditions that showed no inhibition, the origin of inhibition can be commented upon. Table 2 shows the binding energy of O 1s, Al 2p, and Pr 3d in all conditions measured. It should be noted that only the main peak of Pr 3d_5/2 was shown. It was also interesting to observe that O peaks shifted to higher binding energy when inhibition was observed.

The appearance of Zn and Cu peaks implies a comparative lack of inhibition because corrosion activity has occurred to expose intermetallic compounds. The samples that were exposed to 0.1 M NaCl without inhibitors showed Cu and Zn peak. The appearance of Cu 2p and Zn 2p peaks indicated corrosion damage on the samples or protective film was very thin or non-existent. The indication agrees with free corrosion experiment that concluded the ineffectiveness of Pr in pH 10 solution. Accompanied by O 1s binding energy that suggests metal oxides, Al 2p peaks at 74.0 ± 0.2 eV that were observed in pH 5 NaCl, pH 8 Pr alone, and pH 10 Pr alone can be matched to Al₂O₃ (Barr 1983). This form of passive film did not inhibit localized corrosion well according to free corrosion experiment that showed localized corrosion damage and electrochemical experiments.

In the presence of Pr alone, thin Pr-rich film protected 7075-T6 at pH 5. On the sample exposed in pH 5 solution with Pr alone, Pr 3d peaks showed up at 933.6 and 929.5 eV. Accompanying carbonate C 1s peak at 288.2 eV, the form of Pr could be praseodymium carbonate (Pr₂(CO₃)₃), praseodymium hydroxide (Pr(OH)₃) or praseodymium hydroxyl carbonate (PrCO₃.OH). The responsible form of inhibition was likely to be Pr hydroxide. The fact that the Al metal peak was visible indicates that the protective film was thin.

Cooperative inhibition was observed at pH 5. On the sample exposed in pH 5 solutions with Pr and CaSO₄ together, the XPS spectra showed S 2p peaks also appeared and exhibited the presence of sulfate (SO₄²⁻) on the surface. It is possible that samples exposed in 0.1 M NaCl with Pr and CaSO₄ at pH 5 could contain a mixture of three compounds: Al₂O₃/Al(OH)₃, Pr₂(CO₃)₃, and SO₄²⁻. The role of each compound is unknown. Predominant species diagrams suggest that Al(OH)SO₄ compound is responsible for surface passivation and it can suppress localized corrosion (Klomjit and Buchheit 2016). It is possible that aluminum hydroxysulfate works cooperatively with Pr compound to suppress localized corrosion in slightly acidic condition (pH 5–6).

Pr alone did not inhibit localized corrosion at pH 8. Praseodymium is known to be a cathodic inhibitor that blocks oxygen reduction reaction when local pH increases (Hinton 1992). Therefore, the ideal bulk pH should be lower than a point of precipitation. When the local pH rises, the inhibitor precipitates on the site and stifles oxygen reduction reaction. The reason why the inhibition at pH 8 was not observed is that, when the pH was adjusted using NaOH, availability of Pr³⁺ paired up with OH⁻ and formed precipitate. White cloudy substance was observed on the bottom during pH adjustment. The white cloudy substance during the experiment was likely to be Pr(OH)₃ (without CO₂) or Pr₂(CO₃)₃ (with CO₂), which tend to precipitate in basic condition. Therefore, adjusting the pH took away most of Pr³⁺ ion availability to inhibit especially at pH from 8 to 10. Polarization and EIS data did not suggest inhibition by Pr alone at pH 8. The XPS results showed that Pr 3d peaks can be attributed to Pr₂O₃ found on the sample exposed to Pr alone at pH 8. Pr₂O₃ was found only on sample exposed to Pr alone at pH 8 for which no inhibition was observed. According to the chemical stability diagram, Pr₂O₃ was not suggested to precipitate, thus, it is unlikely to be responsible for inhibition. Pr₂O₃ could have formed during transportation to the XPS chamber because there might have been unreacted Pr³⁺ on the surface after exposure.

Synergistic inhibition of Pr and CaSO₄ was observed at pH 8. A slight shift to higher binding energy of Pr 3d, O 1s and Al 2p compared to other conditions was observed with the presence of Pr and CaSO₄ together at pH 8. The O 1s binding energy was found to be higher than any condition tested. This shows that sulfate, which tends to have higher binding energy of O 1s peak than hydroxides and aluminum oxide, was more prominent and could contribute to the synergistic effect found only at pH 8. Similar to pH 5, a mixture of the three compounds was observed. Even though chemical stability diagram from Medusa^® or OLI Analyzer^® did not show any complex between Pr and CaSO₄, sulfate aids a limited amount of Pr in precipitation and gelation process to suppress localized corrosion as proven by a titration curve. A shift of the equivalence point when CaSO₄ was present was subtle, but the titration curves were very reproducible.