Mechanism of Corrosion Inhibition of AA2024 by Rare-Earth

J. Phys. Chem. B 2006, 110, 5515-5528

5515

Mechanism of Corrosion Inhibition of AA2024 by Rare-Earth Compounds Kiryl A. Yasakau,† Mikhail L. Zheludkevich,*,† Sviatlana V. Lamaka,† and Mario G. S. Ferreira†,‡ Department of Ceramics and Glass Engineering, UniVersity of AVeiro, CICECO, 3810-193 AVeiro, Portugal, and Department of Chemical Engineering, Instituto Superior Te´ cnico, ICEMS, AV. RoVisco Pais 1049-001 Lisboa, Portugal ReceiVed: October 23, 2005; In Final Form: January 22, 2006

The mechanism of corrosion protection of the widely used 2024-T3 aluminum alloy by cerium and lanthanum inhibitors in chloride media is described in detail in the present work. The corrosion process was investigated by means of scanning Kelvin probe force microscopy (SKPFM), in situ atomic force microscopy, and scanning electron microscopy coupled with energy dispersive spectroscopy. Employment of the high-resolution and in situ techniques results in a deep understanding of the details of the physical chemistry and mechanisms of the corrosion processes. The applicability of the SKPFM for mechanistic analysis of the effect of different corrosion inhibitors is demonstrated for the first time. The inhibitors under study show sufficient hindering of the localized corrosion processes especially in the case of pitting formation located around the intermetallic S-phase particles. The main role of Ce3+ and La3+ in the corrosion protection is formation of hydroxide deposits on S-phase inclusions buffering the local increase of pH, which is responsible for the acceleration of the intermetallics dealloying. The formed hydroxide precipitates can also act as a diffusion barrier hindering the corrosion processes in active zones. Cerium nitrate exhibits higher inhibition efficiency in comparison with lanthanum nitrate. The higher effect in the case of cerium is obtained due to lower solubility of the respective hydroxide. A detailed mechanism of the corrosion process and its inhibition is proposed based on thermodynamic analysis.

1. Introduction Localized corrosion is a very complex process involving many heterogeneous and homogeneous reactions. The information on the mechanism and the physicochemistry of localized corrosion is an issue of prime importance, since this knowledge can lead to the development of novel effective corrosion protection systems. The important role of intermetallic particles for the initiation and propagation of localized corrosion on the aluminum alloys is well-known and was discussed in a great number of works.1-10 The highest attention was justly paid to the localized corrosion of 2024 aluminum alloy, which is extensively used in the aerospace industry due to the excellent weight-tostrength ratio.3 The intermetallic particles segregated in the grain boundaries confer enhanced mechanical properties, but at the same time they increase the susceptibility of the alloy to a localized corrosion attack. Thus, AA2024 is one of the alloys in commercial applications most prone to localized corrosion attack. Nowadays many groups worldwide attempt to develop effective corrosion protection systems for this alloy. Two main types of localized corrosion occur on AA2024 in the neutral chloride solutions, namely, pitting corrosion and corrosion along the grain boundaries. Therefore the anodic polarization plot for this alloy exhibits two breakdown potentials. The first current increase, at more negative potential, was assigned to the dissolution of the intermetallic particles, while the nobler one is a result of intergranular corrosion.11,12 * Corresponding author: tel, +351-234-378146; fax, +351-234-425300; e-mail, [email protected] (M. L. Zheludkevich). † University of Aveiro. ‡ Instituto Superior Te ´ cnico.

Thus, the first place where corrosion starts is the zone of the intermetallic particles. Different kinds of intermetallic particles were found in the structure of AA2024. The most predominant type (above 60%) corresponds to the S-phase precipitates with Al2CuMg composition. These intermetallics cover almost 3% of the geometrical surface area of the alloy.13 A16(Cu, Fe, Mn) intermetallics were revealed as the second largest type constituting about 12% of all the precipitates. Minor concentrations of Al20Mn3Cu2, Al2Cu, Al7Cu2Fe, and (Al, Cu)6Mn are present in the alloy as well.11,13,14 All kinds of intermetallics excepting S-phase are composed of metals nobler than aluminum, thus showing cathodic character. However, the S-phase is composed of nobler copper as well as of active magnesium. Several assumptions were described in the literature concerning the electrochemical character of the S-phase. Many authors assert that the Al2CuMg particles have an anodic potential respective to the alloy matrix based on the fact that the dissolution of the S-phase occurs at the first stage of pitting corrosion.2,15-17 However, the Volta potential measurements clearly show cathodic character of Al2CuMg particles when compared to the alloy matrix.4,5,9,18 This contradiction raises the question: why do the cathodic intermetallics exhibit high anodic dissolution activity? Many works in the literature were devoted to the mechanism of pitting corrosion of AA2024 in chloride-containing media and clarifying details of different stages of this process. At the first stage, chloride ions attack a passive oxide film6 causing its breakdown in the places of the intermetallic precipitates. The dissolution of magnesium and aluminum from the copper-rich particle leads to the formation of copper remnants with “Swiss cheese”-like morphology. The fast dealloying of the S-phase

10.1021/jp0560664 CCC: $33.50 © 2006 American Chemical Society Published on Web 02/17/2006

5516 J. Phys. Chem. B, Vol. 110, No. 11, 2006 starts immediately after contact of solution with the surface of the intermetallic.5,7,8,13,14 Simultaneously the deposition of a thin copper layer occurs around the pits. The redeposited copper layer and the porous copper remnants become a good cathode for oxygen reduction that promotes the dissolution of the surrounding alloy matrix and the further propagation of pits. A balance between metal dissolution and mass transfer in the electrolyte control the kinetics of pit propagation.7 However, several details of the pitting corrosion mechanism are not yet completely clear and do not reveal a clear picture of the full process. Taking into account the high susceptibility of AA2024 to localized corrosion attack the effective corrosion protection appears as an issue of prime importance. The carcinogenic chromate pretreatments are currently used to hinder the localized corrosion of the AA2024. However, stricter environmental regulations and the needs of industry stimulated an intense research effort to develop novel environmental-friendly pretreatments and inhibitors. The salts of rare earth (RE) elements were found to provide an effective corrosion inhibition effect to the aluminum alloys.19-21 They control the cathodic reaction by precipitating metal hydroxide (Ln(OH)3) at local regions, which are associated with increase of pH due to oxygen reduction.21-23 Cerium shows maximum corrosion protection efficiency as compared with other RE compounds. An important role in superior efficiency of cerium can be played by Ce4+, which can be formed at high pH values in aerated chloride environments.24,25 RE compounds can be introduced in corrosion protection systems for the aluminum alloy using different strategies. Formation of a conversion coating composed by hydrated oxide layer on top of the aluminum alloy confers an enhanced corrosion protection.4,26 Another approach is use of the cerium conversion coating technique to seal the porous film of anodized aluminum alloy.27 The cerium-based inhibitors can be also introduced in the thin hybrid coatings used as pretreatment for aluminum alloys and exhibit promising results.28-30 However, introduction of cerium compounds can decrease the stability of the hybrid polymer matrix with decrease of the barrier properties. Introduction of zirconia nanoparticles doped with cerium ions into hybrid sol-gel matrix was found to avoid the negative effect of the cerium cations on the film and to provide a prolonged release of cerium inhibitor in the places of localized corrosion attack.30,31 While many works are dedicated to the investigation of the corrosion inhibition with RE-based inhibitors, there are still many contradictions and ambiguities concerning the mechanism of inhibition. The present work is devoted to the investigation of the mechanism of localized corrosion on AA2024 and to the inhibition of corrosion with cerium and lanthanum inhibitors in chloride media. The experimental techniques used were scanning Kelvin probe force microscopy (SKPFM), in situ atomic force microscopy (AFM), and scanning electron microscopy (SEM) coupled with energy dispersive spectroscopy (EDS). dc polarization was used as a supplemental method as well. 2. Experimental Section 2.1. Substrate Preparation. The 2024-T3 aluminum alloy with elemental composition as shown in Table 1 was used under study. Aluminum plates used for dc polarization tests were treated using the following procedure. Cleanup was with acetone and then immersion in a water solution of the alkaline cleaner TURCO 4215, 50 g/L, for 35 min at 65 °C, followed by rinsing

Yasakau et al. TABLE 1: Composition of 2024-T3 Aluminum Alloy, wt % Cu

Cr Fe

Mg

Mn

Si

Ti

Zn other

Al

concn 3.8-4.9 0.1 0.5 1.2-1.8 0.3-0.9 0.5 0.15 0.25 0.15 balance

with distillated water and then immersion in a 20% solution of nitric acid for 10 min at 35 °C following rinsing with distilled water and drying. The specimens used for AFM, SKPFM, and SEM analysis after treatment according to the above-mentioned procedure were polished with nonaqueous diamond paste down to 2 µm and then cleaned in acetone. 2.2. Testing Solutions. High grade reagents Ce(NO3)3‚6H2O and La(NO3)3‚6H2O were used as inhibitors. The aluminum substrates were immersed in 0.05 M sodium chloride testing solutions doped with different concentrations of the inhibitors for 10 min, 1 h, and 2 h prior to the dc polarization tests. Additionally, the polished aluminum samples were immersed in 0.005 M NaCl solution with different concentrations of Ce or La inhibitors for 1 or 2 h. After immersion, the samples were rinsed with distillated water and dried in a desiccator. For in situ AFM measurements, testing solutions 0.005 M and 0.5 NaCl with 0.5 wt % of Ce and La inhibitors were used. 2.3. Techniques. Potentiodinamic polarization curves were obtained in chloride solutions using a VoltaLab PGZ 100 potentiostat. A three-electrode cell was used, consisting of a saturated calomel reference electrode (SCE), a platinum foil counter electrode, and the AA2024 specimen as a working electrode with a surface area of 3.4 cm2. Polarization measurements were performed in the anodic direction from -200 mV vs SCE to 250-300 mV vs SCE versus the open circuit potential, at a sweep rate of 1 mV/s. Prior to the polarization measurements, the samples were immersed in NaCl solution for 10, 60, and 120 min. A commercial AFM Digital Instruments NanoScope III system with an Extender electronic module was used for scanning Kelvin probe force microscopy (SKPFM). The AFM operated in lift mode with two pass scans. Lift scan height was 100 nm. Drive amplitude in interleave control for the second pass scan was 800 mV. For all measurements silicon probes covered with PtIr5 were used. The in situ AFM measurements of the alloy topography evolution in a fluid cell containing NaCl solution were conducted in contact mode with silicon nitride tips. 3. Results 3.1. Corrosion of the AA2024 in Chloride Solution. In view of the fact that the main goal of the present work is to study the details of corrosion inhibition of 2024 aluminum alloy by cerium and lanthanum compounds, knowledge of the intimate details of the corrosion mechanism of this alloy appears as an issue of prime importance. The different scanning methods were employed to study the corrosion mechanism of alloy in the chloride solution. A typical electron micrograph of the AA2024 surface immersed in aerated 0.005 M NaCl solution for 2 h is depicted in Figure 1. Corrosion can already be seen even after relatively short immersion in such diluted electrolyte. At the beginning the localized corrosion preferentially starts in the places of intermetallic particles as was discussed above. The strongest corrosion attack appears in the region of the bright round inclusions. Analysis of the chemical composition (Table 2) shows that such zones (zones 1 and 4) are enriched in copper and magnesium in comparison with the surrounding alloy matrix (zone 5). Thus these particles can be surely ascribed to the

Corrosion Inhibition

J. Phys. Chem. B, Vol. 110, No. 11, 2006 5517

Figure 1. SEM image of AA2024 immersed in 0.005 M NaCl solution for 2 h.

TABLE 2: Relative Atomic Ratios of the Elements on Different Areas of AA2024 after Immersion in 0.005 M NaCl for 2h Mn/Al Fe/Al Cu/Al Mg/Al

#1

#2

#3

#4

#5

#6

0.753 0.043

0.002 0.005 0.048 0.034

0.027 0.065 0.106 0.032

0.178 0.074

0.015 0.036

0.030 0.034

intermetallic S-phase. However the relative concentration of magnesium in the heavily corroded zone 1 is much lower than that at intact zone 4. It seems that the intact intermetallic lies deeper and is covered with a thin layer of aluminum and native oxide film. The lower relative concentration of Mg in the corroded intermetallic shows preferable dissolution of this active element from the intermetallic matrix. A slight increase of copper content around S-phase intermetallics (zone 6) can appear due to copper redeposition processes that occurred during the dissolution of intermetallic particles. The other kind of intermetallics, darker with elongated shape (zone 3), reveals lower corrosion activity. The EDS analysis demonstrates that such particles are composed of Al, Cu, Fe, and Mn and can be associated with the A16(Cu, Fe, Mn) phase. Atomic force microscopy coupled with a scanning Kelvin probe was used to study evolution of the surface during the corrosion tests and to reveal the electrochemical nature of the intermetallic particles in the AA2024 alloy. The distribution of the Volta potential along the alloy surface was measured. The difference of Volta potentials between different zones of surface is related to the local alloy composition in these origins since the Volta potential difference in accordance to the IUPAC terminology is defined as the electric potential difference between one point in the vacuum close to the surface of M1 and another point in the vacuum close to the surface of M2, where M1 and M2 are two uncharged metals brought into contact.32 It is assumed that measured in SKPFM mode Volta potential difference in air is related to the difference in the work function of the probe and analyzed surface. Previously it was shown that the SKPFM technique allows high-resolution scanning of the Volta potential difference between the conductive AFM tip and substrate surface mapping the Volta potential difference along the surface.33-35 Parts a and b of Figure 2 present topography of the polished alloy and the Volta potential map of corresponding area after 2 h of immersion in 0.005 M NaCl. Two well-defined kinds of

localized defects are formed on the metallic surface. After immersion of the aluminum specimens in NaCl solution, dissolution of small intermetallic particles occurs. The biggest round-shaped intermetallics seem also to be partially dissolved but an additional solubilization of the surrounding alloy matrix occurs (Figure 2a,e). These intermetallics were identified by the SEM/EDS as S-phase (Al2CuMg). Dissolution of this type of intermetallics suggests their anodic character agrees with some previous studies.2,15-17 However, analysis of the Volta potential distribution completely refutes this suggestion. All the corroded places exhibit well-defined cathodic potential when compared with the aluminum alloy matrix (Figure 2b). The maximal potential difference between these cathodic intermetallics and the matrix in this case is about 350 mV (Figure 2e). An important point is that the potential maximum on the corroded surface is spread and without a well-defined frontier in comparison with the potential maxima of the uncorroded surface where clear frontiers are present (Figure 2d). The maximal Volta potential difference in the case of the uncorroded alloy is about 170 mV, which is two times lower than in the case of the corroded sample. Increase of the Volta potential and broadening of the potential peaks can be explained in terms of intermetallic particles dealloying and copper redeposition. The S-phase contains two active elements aluminum and magnesium and copper that is nobler. The nobler Volta potential of the S-phase compared to the aluminum alloy matrix comes from the noble potential of the copper in the intermetallic phase. Increase of the Volta potential evidently demonstrates an enrichment of the intermetallic in copper during the corrosion tests. The dealloying of the intermetallics occurs due to selective dissolution of magnesium and aluminum from the S-phase leaving copper-rich remnants with higher Volta potential. The broadening of the cathodic potential peaks can happen due to the processes of copper redeposition onto the alloy matrix around the intermetallic particles, which is in good agreement with previous works.13 The formation of a thin copper layer around the corroded intermetallics is clearly demonstrated in the optical micrograph taken after 2 h of immersion in 0.005 M NaCl solution (Figure 3). The formation of wide cathodes around the intermetallics and the creation of “Swiss cheese”-like copper remnants with developed surface area lead to sufficient acceleration of the cathodic processes consequently increasing the rate of the anodic dissolution of the aluminum alloy matrix around the intermetallics. The results presented above clearly demonstrate that the first stage of AA2024 corrosion in chloride medium is the dissolution of the cathodic S-phase. It can be explained in terms of the formation of local galvanic couples inside each intermetallic. The magnesium depleted zones can play the role of cathodes whereas the rest of the intermetallic surface is an anode. The corrosion processes inside the S-phase inclusions arise although the potential of the whole intermetallic is nobler than the surrounding alloy. In situ AFM measurements in chloride electrolyte were performed to investigate the kinetics of the pitting formation due to local dissolution of the S-phase. The evolution of the surface topography in the places of relatively small intermetallic particles was recorded during immersion in 0.005 M NaCl (Figure 4). The first signs of localized corrosion appear already after only 17 min of immersion. The defects are growing during the contact with the electrolyte leaving holes about 3 µm in diameter and several hundreds nanometers deep. Calculating the dissolved volume in the defined time, then plotting it against

5518 J. Phys. Chem. B, Vol. 110, No. 11, 2006

Yasakau et al.

Figure 2. Topography (a) and Volta potential map (b) for AA2024 after immersion for 2 h in 0.005M NaCl. Topography (c) and Volta potential map (d) for AA2024 before immersion. (e) Height and Volta potential difference across the profile.

Figure 3. Optical micrograph of the AA2024 surface after immersion in 0.005 M NaCl for 2 h.

the time of immersion in NaCl solution, we obtained the kinetic curve of S-phase dissolution which is presented in Figure 4c.

The rate of this process at the initial stage can be described with a linear kinetic law. 3.2. Influence of RE Inhibitors on Corrosion Processes. 3.2.1. SEM/EDS Study. Different concentrations of cerium(III) and lanthanum(III) nitrates were added to the chloride solution to study the effect of RE cations on the corrosion process of AA2024. Scanning electron microscopy revealed much lower corrosion in the case of the alloy immersed during 1 h in chloride solution with 5% of cerium nitrate than that for undoped chloride solution. Mainly the alloy surface is free of pitting (not shown). However, in some places signs of corrosion activity appear on the alloy surface as depicted in Figure 5. White spherical deposits are locally formed on the alloy surface. The removal of such a deposit formed after 2 h of immersion shows the first signs of localized corrosion under the dome (inset of Figure 5).



Figure 4. In situ AFM images of AA2024 corrosion in 0.005M NaCl: (a) after 17 min of immersion; (b) after 84 min of immersion. The kinetic curve of the pitting formation (c).

Careful EDS analysis shows that white deposits are composed mainly by cerium oxide (hydroxide) and they are located only in places of S-phase intermetallics (zone 1). The Al (Fe, Cu, Mn, Si) intermetallics (zone 2) are free of deposits and do not show any signs of corrosion activity. Some S-phase particles which are not corroded also do not have the cerium hydroxide domes (zone 3). Decrease of Ce concentration in solution leads to less pronounced corrosion activity and lower amounts of spherical deposits (not shown). Stronger corrosion impact and deeper dissolution of the S-phase in the case of the higher concentration of inhibitor can be due to lower pH values originated by hydrolysis of cerium ions. The experimental pH values of 0.05, 0.5, and 5% solutions of Ce(NO3)3 are 5.53, 5.12, and 4.51, respectively. The low pH values of the electrolyte can lead to faster chemical reaction of water with magnesium or aluminum from the S-phase at the initial stages. Figure 6a shows the surface of AA2024 after 1 h of immersion in a chloride solution doped with 5% La(NO3)3. Features similar to the case of the cerium-doped electrolyte are revealed on the surface of the alloy immersed in the solution with lanthanum-based inhibitor. The white deposits are also formed in the places of S-phase intermetallics, and the pittinglike defects are revealed under the lanthanum hydroxide domes. The EDS analysis demonstrates (Figure 6c) that formation of lanthanum hydroxide deposits occurs only in the places of active S-phase intermetallics as in the case of cerium-doped electrolytes. No signs of corrosion activity were revealed around Al(Fe, Cu, Mn, Si) intermetallics as well. The decrease of La(NO3)3 concentration until 0.05% leads to lower corrosion impact as shown in Figure 6b. Higher localized corrosion activity in the case of the electrolytes heavily loaded with lanthanum inhibitor can also be caused by the lower pH as in the case of the cerium inhibitor.

3.2.2. SKPFM Results. The AFM and Volta potential distribution map presented in parts a and b of Figure 7 confirms selective deposition of cerium hydroxide precipitates in the cathodic places. Previous SKPFM studies also showed the cathodic nature of the intermetallic inclusions of the 2024 aluminum alloy.34,35 The deposits (sharp peaks on the AFM scan) have heights larger than half a micrometer (Figure 7c) after 1 h of immersion in chloride electrolyte doped with 5% of cerium nitrate. They are formed only in the places of the active intermetallics which show potential peaks broadened due to the copper redeposition processes described above. The Volta potential map of AA2024 tested in the lanthanum-containing electrolyte demonstrates similar behavior (Figure 8). The hydroxide deposits are also formed in the places of cathodic intermetallics. This suggests that the mechanism of corrosion inhibition can be similar in the cases of different RE salts. Always when the corrosion processes started the broadening of the cathodic potential peaks on AA2024 occurs due to the S-phase dealloying and the copper redeposition on the surrounding alloy. Therefore the degree of peak broadening directly depends on the level of corrosion impact and can be used as a qualitative characteristic of the corrosion protection efficiency at initial stages. Figure 9 demonstrates two Volta potential maps obtained on AA2024 after 1 h in 0.005 M NaCl doped with the different concentrations of cerium nitrate. The broadening of the Volta potential maxima is more pronounced in the case of the 5% NaCl electrolyte than in the case of the one diluted 10 times. These maps confirm once again that a too high concentration of cerium nitrate leads to a deeper localized corrosion in the places of the S-phase intermetallics. Profiles across the intermetallic particles were made on Volta potential maps for 2024 specimens after immersion in different electrolytes to compare the Volta potential difference between


Figure 5. (a) SEM image of the AA2024 after immersion for 1 h in 0.005 M NaCl with 5% of the Ce(NO3)3. Inset (b) presents micrograph after 2 h of immersion in 0.005 M NaCl with 5% of the Ce(NO3)3. (c) EDS analysis of the areas indicated.

the intermetallic particles and alloy matrix (Figure 10). The potential profiles were recorded for alloy substrates immersed for different periods in the chloride solution doped with different concentrations of cerium- or lanthanum-based inhibitors. Figure 10a presents potential profiles for the alloy immersed in the cerium-doped solutions. Profiles show the dependence between the time of the immersion on the smoothness of Volta potential transition from the maxima (intermetallics) to the bottom (aluminum matrix). The increase of the immersion time tends to define spreading of the transition region between the potential of the S-phase and that of the alloy matrix. Evolution of this parameter with time exhibits local corrosion in the places of cathodic S-phase intermetallics. As was mentioned above the concentration of cerium nitrate has a great influence on the inhibition efficiency. At the beginning of corrosion tests, the 0.5% solution shows minimal potential broadening and in turn maximal efficiency in comparison with other concentrations. After 2 h of immersion the effect of 0.5% and 0.05% of inhibitor is similar and much better than in the case when 5% cerium nitrate was used. The Volta potential peak does not have a welldefined frontier between the potential of the S-phase and that of the alloy matrix when AA2024 is immersed in 5% Ce(NO3)3 for 2 h. Careful analysis of Volta potential profiles (Figure 10a) clearly shows the higher corrosion inhibition efficiency when the electrolyte with concentration of cerium nitrate from 0.05% to 0.5% is used. The Volta potential profiles of the aluminum alloy immersed in the chloride solution doped with lanthanum-

Yasakau et al. based inhibitor (Figure 10b) demonstrate similar behavior to that reported above. Spreading of the Volta potential maxima increases with immersion time. The concentration of inhibitor also has strong influence on the shape of the potential profiles as in the case of cerium nitrate. However a maximal inhibition efficiency was revealed when the electrolyte with lower concentration (0.05%) of La(NO3)3 was used. The optimal concentration of the RE cation inhibitors obtained with the SKPFM method is in good agreement with results reported elsewhere obtained by other methods for other aluminum alloys.19,22 3.2.3. In Situ AFM Study. In situ AFM scans of the AA2024 surface during immersion in the electrolyte doped with different concentrations of cerium- or lanthanum-based inhibitors were performed in order to study the kinetic aspects of hydroxide deposit formation on the top of the cathodic S-phase particles. Figure 11 presents the evolution of the alloy topography in the place of cathodic intermetallics during immersion in La-doped chloride solution. The topography of the alloy exhibits a weakly defined hill after 24 min of immersion (Figure 11a), which can be ascribed to the precipitate on the intermetallic. The relatively high deposit (about 800 nm) is formed on this intermetallic after 6 h and 20 min of continuous immersion (Figure 11b). At the beginning the substrate was immersed during 5 h and 23 min in the electrolyte containing 0.005 M NaCl and 0.5% of La(NO3)3. Then the electrolyte with higher concentration of chlorides (0.5 M) and the same concentration of inhibitor was introduced in the AFM cell instead of the diluted one and scanning of topography was continued for about 1 h. Figure 11c presents the evolution of the profile across the hydroxide deposit formed on the surface in the places marked by black lines in the AFM scans. Relatively slow growth of the deposit occurs during the first period of immersion (in weak electrolyte). A 2 orders of magnitude increase in chloride concentration leads to higher speed of lanthanum hydroxide deposition. Figure 12 shows the kinetics of the La(OH)3 precipitate formation at the place presented in the previous figure. A linear kinetic law can be used to describe the rate of the deposit formation. The estimated rate of the precipitate growth is about 0.4 nm/min in a weak chloride-based electrolyte. However, the rate is increased by almost 50 times reaching 20 nm/min when concentrated solution is used. These results indicate a very important role of the chloride ions in the kinetics of the hydroxide deposits formation. Obviously the increase of the chloride concentration leads to enhanced corrosion activity which consequently increases the pH in the places of active intermetallics leading to faster local deposition of hydroxide precipitates. Thus, the RE salts works as an “intelligent” inhibition system with an active feedback to the corrosive medium. 3.2.4. dc Polarization. Potentiodynamic polarization measurements were carried out in the potential range from -0.25 to 0.2 V vs open circuit potential to estimate the effect of the different inhibitors on the anodic and cathodic partial electrode reactions. To clarify kinetic features of the inhibition process, the polarization curves were recorded after different periods of immersion at the open circuit potential in 0.05 M chloride electrolyte doped with 0.05% RE-based inhibitors. Figure 13a presents the dc polarization curves for AA2024 in the undoped electrolyte and in the cerium-doped solution. The anodic branch shows two well-defined breakdown potentials corresponded to the S-phase dissolution (lower polarization) and to the beginning of the intergranular corrosion (higher polarization).12 Doping of chloride electrolyte with cerium nitrate leads



Figure 6. SEM image of the AA2024 after immersion for 1 h in 0.005 M NaCl with 5% of La(NO3)3 (a) and for 2 h in 0.005M NaCl with 0.05% of La(NO3)3 (b). (c) EDS analysis of the areas indicated.

Figure 7. Topography (a), Volta potential map (b), and profile (c) for AA2024 immersed in 0.005 M NaCl solution with 5% Ce(NO3)3 for 1 h.

to remarkable decrease of cathodic and anodic currents. After only 10 min of immersion in doped solution at the open circuit

potential, the cathodic current during polarization drops almost by 1 order of magnitude. Increase of the immersion time leads


Yasakau et al.

Figure 8. Topography (a), Volta potential map (b), and profile (c) for AA2024 immersed in 0.005 M NaCl solution with 5% La(NO3)3 for 1 h.

Figure 9. Volta potential maps for AA2024 immersed in 0.005 M NaCl solution with 0.5% Ce(NO3)3 (a) and 5% Ce(NO3)3 (b) for 1 h.

to further decrease in cathodic current showing that interaction of the cerium cations with the cathodic centers is a relatively slow process which can take several hours to achieve maximal inhibition of the cathodic reaction. The oxygen reduction may be the main cathodic process in this potential region.36,37 The anodic processes are also influenced by the cerium-based inhibitor. A remarkable decrease of anodic current between the two breakdown potentials already occurs after only 10 min of immersion. This current is related with dissolution of Al and Mg from Al2CuMg intermetallics. The longer immersion prior to the polarization tends to decrease of the anodic current in this region almost by 2 orders of magnitude when compared with that of the undoped chloride solution. An important shift of the second breakdown potential toward positive potentials occurs due to increase of the immersion time. Decrease of cathodic and anodic currents is originated from the formation of cerium hydroxide deposits in the places of S-phase intermetallics. This confirms that anodic and cathodic reactions at low polarizations occur mainly at the Al2CuMg precipitates in the AA2024 matrix. The longer immersion time before polarization

leads to formation of larger deposits on the S-phase hindering the cathodic reduction of oxygen on Cu-rich intermetallics as well as anodic dissolution of Al and Mg from these intermetallics. However, the addition of cerium nitrate almost does not influence the limiting current originating from intergranular corrosion at the high anodic polarizations confirming that the inhibition effect is mainly focused on the pitting corrosion in the S-phase zones. The influence of lanthanum nitrate on the corrosion of AA2024 was also studied by dc polarization (Figure 13b). The behavior of the alloy in the La-doped chloride solution is similar to that of the Ce-doped one. However, the inhibition effect of lanthanum cations for cathodic reactions is lower than that in the case of the cerium-based inhibitor. The hindering of anodic dissolution of the S-phase is also less effective when compared with the cerium-doped electrolyte, while the inhibition effect appears almost immediately after starting the immersion (10 min) and keeps a constant value even after 2 h, demonstrating a faster kinetics of the hydroxide deposits formation.


J. Phys. Chem. B, Vol. 110, No. 11, 2006 5523 bubbles. The simple chemical redox reaction of hydrogen evolution is a preferable way when compared with the electrochemical one since the electrochemical reduction of water can occur only at the cathodic polarizations. The most likely cathodic process is the reaction of oxygen reduction due to very high overpotential of the electrochemical hydrogen evolution on copper-rich cathodes.36,37 Thus the hydrogen evolution evidently proves the occurrence of reactions 1 and 2. However, simultaneously, the electrochemical dissolution can take place at this stage as well causing faster dealloying. The local increase of pH around the S-phase particles as well as enrichment of their surface with copper lead to shifting the dissolution processes from chemical to electrochemical. The hydrogen bubbles are removed from the alloy surface, and an electrochemical dissolution is starting. The cathodic reaction occurs on copper-rich remnants by the following equation:

O2 + 2H2O + 4e- f 4OH-

(3)

Simultaneously the oxidation of magnesium and aluminum occurs at the anodic parts of the S phase:

Figure 10. Evolution of the Volta potential profiles taken from center of maxima toward to the potential of alloy matrix for the AA2024 samples immersed in 0.005 M NaCl solution (a) with 5%, 0.5%, and 0.05% Ce(NO3)3 and (b) with 5%, 0.5%, and 0.05% La(NO3)3.

4. Discussion Analysis of the present results confers an opportunity to discuss deeper the intimate details of the corrosion inhibition mechanism of AA2024 by cerium- and lanthanum-based inhibitors. 1. Mechanism of Corrosion. Normally the 2024 aluminum alloy as well as any aluminum-based substrate is covered by a native oxide film. However, the intermetallics have a thinner and more defective oxide film due to the presence of different elements in an intermetallic phase.17 Immediately after contact of the alloy with a chloride-containing environment the chloride ions interact with the native oxide film (Figure 14a)6 leading to its breakdown especially in the weak places that originated from the S-phase intermetallic precipitates as shown in Figure 1. Water then comes in contact with the intermetallic surface after failure of the barrier oxide layer causing the chemical reaction of active aluminum and magnesium with water (Figure 14b):

2Al + 6H2O f 2Al3+ + 6OH- + 3H2v

(1)

Mg + 2H2O f Mg2+ + 2OH- + H2v

(2)

The chemical dealloying of the S phase occurs due to these processes leading to the formation of a copper-rich surface and local nonuniformities of the intermetallics composition. The possibility of chemical dealloying was never discussed before in the literature. However, clearly visible gas bubbles are formed at the initial stage of the corrosion processes at the location of pitting initiation as shown in Figure 15. The evolution of hydrogen can be the only a reason for formation of such gas

Al f Al3+ + 3e-

(4)

Mg f Mg2+ + 2e-

(5)

Dissolution of aluminum and magnesium leads to deeper dealloying of the S phase forming the porous copper remnants with “Swiss cheese”-like morphology.30 An additional increase of the pH occurs due to cathodic reaction 3 at the intermetallic surface. This leads to formation of a sufficient gradient of OH- ions between the bulk solution and the surface of the intermetallic particle. The aluminum and magnesium cations can immediately react with hydroxyl ions forming insoluble hydroxide sediments or soluble hydroxy complexes (Figure 14d) depending on pH values:

Al3 + + OH- f Al(OH)2+, log K ) 9.03

(6)

Al3+ + 2OH- f Al(OH)2+, log K ) 18.7

(7)

Al3+ + 3OH- f Al(OH)3V, log K ) 27

(8)

Al3+ + 4OH- f Al(OH)4-, log K ) 33

(9)

Mg2+ + OH- f Mg(OH)+, log K ) 2.58

(10)

Mg2+ + 2OH- f Mg(OH)2V, log K ) 11.33

(11)

Parts a, b, and f of Figure 16 present the fraction of the different aluminum, magnesium, and copper species in solution depending on the pH. These correlations were calculated using thermodynamic formation constants for metal hydroxo complexes, which were taken from refs 38 and 39. Such diagrams give a clear picture of the complexes distribution in comparison with the Pourbaix ones. The solubility of aluminum species in aqueous solutions is relatively high due to the possibility of formation of soluble complex compounds both in acidic and in basic conditions (Figure 16a). The pH value is very high in the cathodic zones where hydroxyl ions are generated. The aluminum cations originated from anodic dissolution of the S-phase, interact with the hydroxyl anions generated at the cathodic zones, and form Al(OH)4- complex ions, which diffuse to the bulk solution. On reaching the zones of the solution with lower


Yasakau et al.

Figure 11. In situ AFM scans of the AA2024 surface at the beginning of immersion in 0.005 M NaCl (a) and at the end of immersion in 3% 0.5 M NaCl (b) with 0.5% La(NO3)3 solution. (c) Evolution of topography on profile with time.

remnants.7,30 When the neck between a copper nanosized particle and a remnant is broken, the particle loses the electrical contact with the alloy.41 Then the chemical reaction of copper oxidation by dissolved oxygen is thermodynamically possible:

2Cu + O2 + 2H2O f 2Cu(OH)2, ∆G ) -271.23 kJ‚mol-1 38 (12)

Figure 12. Rate of the lanthanum hydroxide precipitation at the place showed in Figure 11 in 0.005 M NaCl at the beginning (black squares) and in 0.5 M NaCl (empty circles) at the end of in situ AFM measurement.

pH (about 8), the complex anions are transformed into the insoluble Al(OH)3, which is a thermodynamically preferable state in such conditions (see Figure 16a). Thus, aluminum hydroxide forms an insoluble deposit at a certain distance from the active intermetallic (Figure 14d). The longer corrosion leads to formation of a hydroxide dome covering the active pitting. At lower pH values on the periphery of the hydroxide sediments, aluminum hydroxychlorides are formed as well.1,30,40 The magnesium cations can also form hydroxide sediments; however a major part of magnesium is in solution due to the higher solubility of its hydroxide and the lower content of Mg in the alloy. The electrochemical dissolution of magnesium and aluminum from the S-phase leads to the formation of a very porous structure with the copper particles connected to the copper

The nanosize confers even an enhanced reactivity to the copper particles. Partial dissolution of the copper hydroxide occurs due to formation of hydroxo complexes as shown in Figure 16b. The complex ions then are electrochemically reduced again to metallic copper when contact with the surface of the aluminum alloy or with the copper remnants formed after dealloying as schematically shown in Figure 14e. The formation of chloride-based complexes can also assist the dissolution of copper particles. The reduced metallic copper redeposited on the copper remnants leads to copper “refining”.30 When redeposited on the alloy surface around pits, it forms a thin cathodic film as was demonstrated by optical microscopy (Figure 3). The redeposition of copper around the S-phase leads to broadening of the cathodic peaks on the Volta potential maps as demonstrated in Figures 2, 9, and 10. Therefore, the SKPFM technique can be used to detect the level of corrosion impact on the intermetallic particles. The effect of copper redeposition was previously discussed in the literature.8,13,41 However, another mechanism for this process was proposed ascribing migration of Cu particles from dealloyed intermetallic over the pit.13 The thin layer of copper redeposited around pits can play a role of an additional cathode accelerating the anodic processes of aluminum dissolution from the S-phase and from the surrounding alloy matrix (Figure 2a), which is depleted in copper and thus is anodically active.34 The localized corrosion process becomes thus autocatalytic. Therefore, redeposition of copper



Figure 15. Optical photograph of the surface of the AA2024 at the beginning of the immersion test in NaCl solution. On the left and right sides bubbles are seen.

Figure 13. Potentiodynamic polarization curves for the AA2024 after immersion in 0.05 M NaCl solution without (1 h) and with (a) Ce inhibitor (10 min, 1 h, and 2 h) and (b) La inhibitor (10 min, 1 h, and 2 h).

is achieved at the initial period of immersion due to a balance between the Faradaic processes and the diffusion of reagent and corrosion products. The present results and the results discussed in the literature demonstrate the prime importance of the S-phase in the localized corrosion of AA2024. The results presented here clearly prove that the S-phase intermetallics not only function as a cathode or as an anode in the corrosion processes but also play a complex role combining cathodic and anodic activity together in the same particle. Therefore, different strategies to stop the activity of these intermetallics can be used to hinder the localized corrosion of AA2024 in chloride-containing environments. 2. Mechanism of Corrosion Inhibition with RE Compounds. Introduction of lanthanum or cerium nitrate into the system sufficiently changes the situation. Figure 17 presents the pH range at which the different lanthanum and cerium species in the solution can be formed. The relative compositions of water-based solutions doped with cerium and lanthanum salts were calculated using thermodynamic equilibrium constants for the following reactions:

La3+ + OH- f La(OH)2+, log K ) 2.7 -

(13)

+ 3OH f La(OH)3V, log K ) 22

(14)

Ce3+ + OH- f Ce(OH)2+, log K ) 5.57

(15)

Ce3+ + 2OH- f Ce(OH)2+, log K ) 10.4

(16)

Ce3+ + 3OH- f Ce(OH)3V, log K ) 14.71

(17)

Ce4+ + OH- f Ce(OH)3+, log K ) 14.76

(18)

Ce4+ + 2OH- f Ce(OH)22+, log K ) 28.04

(19)

Ce4+ + 3OH- f Ce(OH)3+, log K ) 40.53

(20)

Figure 14. Schematic representation of corrosion mechanism of AA2024 in chloride solution.

Ce4+ + 4OH- f Ce(OH)4V, log K ) 51.86

(21)

appears as a very important factor responsible for fast propagation of pits around intermetallics. However, Figure 4c demonstrates a linear kinetic law of small intermetallic dissolution. This linear dissolution kinetics suggests that a stationary state

Formation constants for lanthanum hydroxo complexes were taken from refs 38 and 39. Formation constants for cerium species were calculated using Gibbs free energies of formation of cerium hydroxo complexes, which were taken from ref 42.

La

3+


Yasakau et al.

Figure 16. Fraction of different metal species in the solution depending on pH.

Figure 17. Diagram of the pH range of the respective hydroxides existence.

In 5-0.05% solutions of La(NO3)3, lanthanum is mainly present in the form of free hydrated La3+ ions. When the S-phase is activated with chloride-containing solution, the pH value is locally growing up. Increase of pH values up to 9 leads to formation of lanthanum hydroxide (14) (Figure 16c). The lanthanum hydroxide is deposited immediately on the top of the active copper-rich intermetallics as shown in Figure 6.

Formation of sediments occurs directly on the cathodic zones (Figure 8) since much higher pH cannot be achieved far from the source of the hydroxyl ions due to fast diffusion processes, which equalize the concentration of OH- at the cathode surface and in the bulk solution. The continued cathodic reaction generates hydroxyl ions causing further growth of lanthanum hydroxide deposits (Figure 11). The hydroxides formed on the top of copper-rich intermetallics significantly hinder the corrosion processes in the active S-phase. Decrease of the corrosion activity is confirmed by the linear polarization measurements, which show a sufficient decrease of the cathodic and anodic currents at low polarization when the lanthanum nitrate is added to the electrolyte (Figure 13b). The anodic current at low polarization is related with dissolution of magnesium and aluminum from the copper-rich intermetallics. Decrease of the anodic current for 1 order of magnitude confirms an effective suppression of the S-phase dealloying due to the lanthanum inhibitor. However, the anodic processes that originated from intergranular corrosion12 at higher polarization are not influenced by the lanthanum(III) compounds.



Analysis of the Volta potential distribution shows that 0.05% of La(NO)3 is an optimal concentration for inhibition of the S-phase dissolution in AA2024 immersed in the chloride solution, which is in a good accordance with literature data for the inhibiting action of La compounds.19,22 The high concentration of La3+ in solution leads to decrease of its pH due to a hydrolysis reaction. The lower pH of electrolyte in turn can be responsible for faster chemical dealloying of intermetallics (eqs 1 and 2). Decrease of pH leads also to a delay in the formation of lanthanum hydroxide deposits. The cerium(III) shows a similar effect to that of the lanthanum-based inhibitor. The relative composition of Ce(III) species in solution at different pH values is shown in Figure 16e. Cerium(III) is present mainly in the form of free Ce3+ in neutral solutions. Increase of pH leads to formation of hydroxo complexes (Figure 16e) at first and when the pH reaches 10 the hydroxide sediments can be formed as in the case of lanthanum. The formation of hydroxide domes occurs also selectively on the top of the S-phase intermetallics (Figure 5) hindering corrosion. Both cathodic and anodic processes on the S-phase intermetallics are suppressed in accordance to the dc polarization results (Figure 13a). The suppression of cathodic processes in the case of cerium is higher than that for lanthanumdoped electrolytes. The anodic current due to S-phase dealloying is also 1 order of magnitude lower than in the case of lanthanum. The more effective inhibition of the corrosion processes in the case of the cerium inhibitor can be related with the relatively lower solubility of cerium(III) hydroxide when compared with the solubility of lanthanum(III) hydroxide. However, an additional important feature related with the kinetics of deposits formation was found when cerium nitrate was used as a corrosion inhibitor. The dc polarization curves presented in Figure 13a clearly demonstrate that the corrosion inhibition efficiency increases with the increase of the immersion period before the polarization tests. It seems that deposits with excellent protective characteristics are not immediately formed on the S-phase particles in contrast to the case when lanthanum nitrate was used as corrosion inhibitor. This relatively long process can be related with the ability of trivalent cerium to be oxidized to the tetravalent state. The tetravalent cerium forms extremely insoluble hydroxides already at pH about 3 as exhibited in Figure 16d. As was discussed elsewhere25,42 the oxidation of cerium(III) can be caused by hydrogen peroxide, which was originated from the cathodic reaction of two-electron reduction of oxygen:

O2 + 2H2O + 2e- f H2O2 + 2OH-

(22)

The generated peroxide can in turn oxidize trivalent cerium by one of following reactions:

Ce3+ + H2O2 + 2OH- f Ce(OH)4 (CeO2‚2H2O)V (23) Ce(OH)2+ + H2O2 + OH- f Ce(OH)4 (CeO2‚2H2O)V (24)

Ce(OH)2+ + H2O2 f Ce(OH)4 (CeO2‚2H2O)V (25) Reactions 24 and 25 are preferable at high pH values near the cathode surface. On the other hand the reaction of cerium(III) oxidation proposed elsewhere24,31

Ce3+ + 1/2H2O2 + OH- f Ce(OH)22+

(26)

seems questionable since the Ce(OH)22+ complex ions can exist

only at extremely low pH (Figure 16d), which cannot be achieved in the system under study. The formed hydroxide of tetravalent cerium is also deposited on the cathodic parts providing additional hindering of the corrosion processes. 5. Conclusions The corrosion mechanism of AA2024 in chloride-based electrolytes was investigated in detail in the present work by means of dc polarization and local measurement methods. An important role for the intermetallics in the localized corrosion process is demonstrated. The pitting corrosion begins in the places of the S-phase intermetallics, which have a cathodic potential relative to the alloy matrix. S-phase dealloying starts by the chemical attack dissolving the active aluminum and magnesium and leaving copper-rich remnants. Then the electrochemical processes take place and the intermetallic precipitates play the role of active anodes and cathodes causing at the same time deeper dealloying. In situ AFM measurements demonstrate that the rate of the intermetallics dissolution can be described by a linear kinetic law. The process of copper redeposition, which leads to formation of refined copper remnants and copper film deposits around pits, plays an important role in further development of the local corrosion attack. The redeposited copper increases the effective surface area of the cathodic zone, which is confirmed by the spreading of cathodic peaks on the Volta potential maps. The addition of lanthanum(III) or cerium(III) leads to formation of the respective hydroxide deposits in the S-phase locations hindering the anodic and the cathodic processes. The formation of hydroxides occurs due to an increase of the pH in such locations resulting from the cathodic processes. The formation of the deposits decelerates the redeposition of copper and the broadening of cathodic Volta potential peaks. Thus, the SKPFM technique was found to be an effective method to estimate the protection performance of the RE salts and to find the optimal concentrations of inhibitor. Too high concentration of cerium or lanthanum nitrates leads to deeper localized corrosion in the places of the S-phase intermetallics. The growth of hydroxide deposits studied with the in situ AFM technique exhibit linear time dependence indicating that the corrosion processes are progressing in stationary state that results from a balance between diffusion and electrochemical reaction. The chloride ions strongly influence the rate of the corrosion processes and in turn the rate of the hydroxide deposits formation. The cerium nitrate shows superior inhibition properties in comparison with lanthanum nitrate probably due to consecutive formation of the extremely insoluble cerium(IV) hydroxide. The proposed mechanisms of pitting corrosion and corrosion inhibition with RE salts are supported by thermodynamic analysis. Acknowledgment. The authors acknowledge the sixth Framework Program of the European Community (Contract No. IP-011783) and the Portuguese Fundac¸ a˜o para a Cieˆncia e Tecnologia (Contract POCTI/CTM/59234/2004). The FCT Grants SFRH/BPD/12538/2003 and SFRH/BPD/20537/2004 are also gratefully acknowledged. References and Notes (1) Szklarska-Smialowska, Z. Corros. Sci. 1999, 41, 1743. (2) Liu, Z.; Chong, P. H.; Butt, A. N.; Skeldon, P.; Thompson, G. E. Appl. Surf. Sci. 2005, 247, 294. (3) Starke, E. A.; Staley, J. T. Prog. Aerospace Sci. 1996, 32,131.

5528 J. Phys. Chem. B, Vol. 110, No. 11, 2006 (4) Campestrini, P.; Terryn, H.; Hovestad, A.; de Wit J. H. W. Surf. Coat. Technol. 2004, 176, 365. (5) Zheludkevich, M. L.; Yasakau, K. A.; Poznyak, S. K.; Ferreira, M. G. S. Corros. Sci. 2005, 47, 3368. (6) Fonseca, I. T. E.; Lima, N.; Rodrigues, J. A.; Pereira, M. I. S.; Salvador, J. C. S.; Ferreira, M. G. S. Electrochem. Commun. 2002, 4, 353. (7) Dimitrov, N.; Mann, J. A.; Vukmirovic, M.; Sieradzki, K. J. Electrochem. Soc. 2000, 147, 3283. (8) Dimitrov, N.; Mann, J. A.; Sieradzki, K. J. Electrochem. Soc. 1999, 146, 98. (9) Campestrini, P.; van Westing, E. P. M.; van Rooijen, H. W.; de Wit J. H. W. Corros. Sci. 2000, 42, 1853. (10) Boag, A. P.; McCulloch, D. G.; Jamieson, D. N.; Hearne, S. M.; Hughes, A. E.; Ryan, C. G.; Toh, S. K. Nucl. Instrum. Methods Phys. Res., Sect. B 2005, 231, 457. (11) Zhang, W.; Frankel, G. S. Electrochim. Acta 2003, 48, 1193. (12) Guillaumin, V.; Mankowski, G. Corros. Sci. 1999, 41, 421. (13) Buchheit, R. G.; Grant, R. P.; Hiava, P. F.; Mckenzie, B.; Zender, G. L. J. Electrochem. Soc. 1997, 144, 2621. (14) Blanc, C.; Lavelle, B.; Mankowski, G. Corros. Sci. 1997, 39, 495. (15) Buchheit, R. G.; Montes, L. P.; Martinez, M. A.; Michael, J.; Hlava, P. F. J. Electrochem. Soc. 1999, 146, 4424. (16) Jin, F. L.; Zheng, Z. Q.; Jiang, N.; Tan, C. Y. Mater. Chem. Phys. 2005, 91, 325. (17) Shao, M. H.; Fu, Y.; Hu, R. G.; Lin, C. J. Mater. Sci. Eng., A 2003, 344, 323. (18) de Wit J. H. W. Electrochim. Acta 2001, 46, 3641. (19) Bethencourt, M.; Botana, F. J.; Calvino, J. J.; Marcos, M.; Rodriguez-Chacon M. A. Corros. Sci. 1998, 40, 1803. (20) Twite, R. L.; Bierwagen, G. P. Prog. Org. Coat. 1998, 33, 91. (21) Davo, B.; de Damborenea, J. J. Electrochim. Acta 2004, 49, 4957. (22) Aballe, A.; Bethencourt, M.; Botana, F. J.; Marcos, M. J. Alloys Compd. 2001, 323, 855. (23) Arenas, M. A.; Bethencourt, M.; Botana, F. J.; de Damborenea, J.; Marcos, M. Corros. Sci. 2001, 43, 157.

Yasakau et al. (24) Bilal, B. A.; Muller, E. Z. Naturforsch., A: Phys. Sci. 1992, 47, 974. (25) Aldykiewicz, A. J.; Davenport, A. J.; Isaacs, H. S. J. Electrochem. Soc. 1996, 143, 147. (26) Palanivel, V.; Huang, Y.; Ooij, W. J. Prog. Org. Coat. 2005, 53, 153. (27) Yu, X.; Cao, C. Thin Solid Films 2003, 423, 252. (28) Voevodin, N. N.; Grebasch, N. T.; Soto, W. S.; Arnold, F. E.; Donley, M. S. Surf. Coat. Technol. 2001, 140, 24. (29) Kasten, L. S.; Grant, J. T.; Grebasch, N.; Voevodin, N. N.; Arnold, F. E.; Donley, M. S. Surf. Coat. Technol. 2001, 140, 11. (30) Zheludkevich, M. L.; Serra, R.; Montemor, M. F.; Yasakau, K. A.; Miranda Salvado, I. M.; Ferreira, M. G. S. Electrochim. Acta 2005, 51, 208. (31) Zheludkevich, M. L.; Serra, R.; Montemor, M. F.; Ferreira, M. G. S. Electrochem. Commun. 2005, 7, 836. (32) IUPAC Compendium of Chemical Terminology, 2nd ed.; Royal Society of Chemistry: London, 1997; Vol. 55, p 1257. (33) Schmutz, P.; Frankel, G. S. J. Electrochem. Soc. 1998, 145, 2295. (34) Schmutz, P.; Frankel, G. S. J. Electrochem. Soc. 1998, 145, 2285. (35) Shiraishi, S.; Kanamura, K.; Takehara, Z. J. Phys. Chem. B 2001, 105, 123. (36) Cheng, Y. L.; Zhang, Z.; Cao, F. H. Corros. Sci. 2004, 46, 1649. (37) Kiourtsid, G.; Skolianos, S. M. Mater. Sci. Eng., A 1998, 248, 165. (38) Dean, J. A. Lange’s Handbook of Chemistry, 12th ed.; McGrawHill Book Co.: New York, 1979. (39) Yu Lur’e, Yu. Handbook of Analytical Chemistry, 6th ed.; Spravochnik po analiticheskoi himii, Himiya, 1989. (40) Traverso, P.; Beccaria, A. M. Surf. Interface Anal. 1998, 26, 524. (41) Buchheit, R. G.; Martinez, M. A.; Montes, L. P. J. Electrochem. Soc. 2000, 147, 119. (42) Hayers, S. A.; Yu, P.; O’Keefe, T. J.; Stoffer, J. O. J. Electrochem. Soc. 2002, 149, C623.

Mechanism of Corrosion Inhibition of AA2024 by Rare-Earth

Recommend Documents