Acceleration of the Cathodic Kinetics on Aluminum Alloys by Aluminum Ions

TheeffectofAl 3 + onthecathodickineticsofAlalloysaswellasPtandSS316Lhasbeeninvestigatedbyanumberofelectrochemical techniques. It has been observed that the addition of Al 3 + into NaCl solution can signiﬁcantly increase the diffusion limited cathodic kinetics of Al alloys, and this increase is proportional to the [Al 3 + ]. The same phenomenon was also observed on Pt and SS316L, which indicates that this enhancement in cathodic kinetics is not related the surface structure of Al alloy, and it is the HER diffusion- limited kinetics that are increased rather than ORR kinetics as a result. Based on electrochemical studies on Pt, it is proposed that although the addition of Al 3 + can lead to the precipitation of an oxide/hydroxide ﬁlm on Pt there is a greatly enhanced proton diffusivity which overwhelms the barrier effect of the precipitate ﬁlm, leading to substantially increased cathodic kinetics. The results are interpreted in terms of an extension of the Grotthuss Theory in which Al 3 + can facilitate transport of the proton. an the terms of

High strength aluminum alloys (Al alloys) have received wide application in the aerospace and automotive industries for decades. However, these alloys are prone to localized corrosion due to their highly heterogenous microstructures, 1,2 with susceptibility to pitting and crevice corrosion, 3-6 intergranular corrosion, 7-9 exfoliation corrosion, 10,11 and environmental assisted cracking, [12][13][14] depending upon the alloy composition and temper. These localized corrosion modes are detrimental to the Al-alloy based structures, and there has been a great number of experimental studies performed aimed at interrogation of the underlying mechanisms as well as development of pertinent test protocols to better quantify these phenomena. Among these experimental protocols, accelerated corrosion test methods are designed to create corrosion morphologies that mimic those observed in service and to evaluate the interactions between exposure environment and the degree of localized corrosion in a timely fashion. For exfoliation corrosion, the available accelerated testing protocols include ASTM G34 15 (EXCO), ASTM G85, 16 and ANCIT. 17 For ANCIT test method 17 proposed by Lee and Lifka as an improvement over ASTM G34, there are four major metal ion species which were identified from the solution chemistry of AA7150 exposed in the EXCO environment after 24hrs: Al 3+ , Zn 2+ , Mg 2+ , and Cu 2+ . The corresponding elements are the dominant constituent ones in AA7150. Lee and Lifka found that Al 3+ was the cation that was most important in the success of the exfoliation corrosion test. However, the role of Al 3+ in causing severe exfoliation corrosion in susceptible tempers was not explained in their work.
From an electrochemical perspective, the localized corrosion rate of aluminum alloys under open circuit potential conditions can be controlled by either the passivity of the alloys which attenuates the dissolution kinetics, or the rate of the oxygen reduction reaction (ORR) or hydrogen evolution (HER) rate on the alloy surface. Some corrosion inhibitors focus on mitigation of the mass transfer limited cathodic kinetics on Al alloys. Metallic cation-based inhibitors (most of which belong to transition metal group such as cerium) form metal oxide/hydroxides on the noble intermetallic particles embedded in the Al matrix, which are the key cathodic sites, to achieve cathodic inhibition goal. [18][19][20][21][22][23][24][25] However, there has been limited literature pertinent to the effect of four major metallic cations in the ANCIT test on the cathodic kinetics of Al alloys, despite the fact that under corrosion conditions, these metallic ions will be in the highest concentration on the surface.
Recently, Parker and Kelly 26 parsed the electrochemical foundation of G34 and ANCIT tests, and discovered that the addition of AlCl 3 (∼0.224M) significantly increased the hydrogen evolution reaction (HER) rate on both AA2060-T3 and AA2060-T86, which in turn increased the localized corrosion of those Al alloys. However, the means by which the addition of Al 3+ caused the large increase in cathodic kinetics observed remains unclear. There are several potential applications of understanding the role of Al 3+ in the localized corrosion behavior of Al alloys: 1) during localized corrosion, Al 3+ will diffuse away from the localized corrosion site into the adjacent solution environment, which can further impact the corrosion behavior of Al alloys; 2) by utilizing the fact that Al 3+ can increase the cathodic kinetics which in turn increases localized corrosion rate, improved accelerated corrosion testing for Al alloys can be developed by the addition of Al salts; 3) if HER enhancement by Al 3+ can be extended to the other metal and alloys, one can make use of Al 3+ for the enhanced production of hydrogen.
The overarching goal of this study is to first examine the phenomenological effect of Al 3+ on the cathodic kinetics and corrosion rates on 2xxx, 5xxx and 7xxx series Al alloys, and then extend the Al 3+ effect study to pure platinum (Pt) and stainless steel 316L (SS), in order to identify the underlying mechanism of cathodic kinetics enhancement by Al 3+ .

Experimental
Materials preparation.-Three types of Al alloys: AA2024-T351, AA5083-H131 and AA7050-T451, all of which were obtained from ALCOA (Pittsburgh, PA), as well as SS316L (McMaster-Carr Supply Company, Elmhurst IL) were cut into 1" × 1" test coupons. A platinum rotating disk electrode (Pine Research Instrumentation, Inc., Durham, NC) with a diameter to 0.5 cm was used for the RDE test. The nominal compositions of the three types of Al alloys are listed in Table I. In addition to the regular Al alloy samples for electrochemical tests, sensitized AA5083-H131, DoS = 50 mg/cm 2 specimen were mounted in the epoxy and used for immersion exposure tests. All of these specimens were polished to a surface finish of 1200 grit with SiC paper, and subsequently degreased with ethanol followed by deionized water, prior to experiment.  Solution pH 0.6M NaCl 5.60 0.57M NaCl+0.01M AlCl 3 3.57 0.45M NaCl+0.05M AlCl 3 3.20 0.3M NaCl+0.1M AlCl 3 3.01 0.2M AlCl 3 2.87 Electrochemical tests.-A three-electrode electrochemical cell was used with the alloy test specimens as the working electrode (WE) with an exposure area of 1cm 2 (except that Pt disc electrode had surface area = 0.2 cm 2 ), a saturated calomel reference electrode (SCE) as reference electrode (RE), and a platinum-niobium mesh as the counter electrode (CE). A Bio-Logic SP-200 (Bio-Logic SAS, Claix, France) potentiostat was utilized with EC-Lab (Version 11.01) software to perform all of the electrochemical measurements. The details of each electrochemical technique will be described in below.  Table II. Two additional test solutions with a basis of 0.6 M NaCl with the pH adjusted by HCl were prepared to compare to solutions 2 and 4 at the same pH: 0.6 M NaCl with pH = 3.57 and 0.6 M NaCl with pH = 3.01. The purpose of performing cathodic polarization from OCP is to avoid the effects of prior anodic dissolution on the surface structure if the scan was to be started above OCP, and only focus on the cathodic characteristics of as-polished samples. To determine the dependence of the polarization resistance (R p ) and corrosion potential (E corr ) on the [Al 3+ ], full polarization measurements were performed on all Al alloy samples with a scan rate of 0.2 mV/s, which started at 0.3V below OCP and ended at −0.6 V SCE .
Electrochemical impedance spectroscopy (EIS).-A typical electrochemical impedance spectrum was acquired over the frequency range from 10 kHz to 10 mHz at 10 points per decade with a 10 mV AC amplitude around OCP after 1hr immersion time. AA7050-T7451 specimen were immersed in solutions 1-5 as described above to study the effect of Al 3+ concentration on interfacial electrochemical parameters. All the EIS data was fitted by the EC-Lab software.
Rotating disc electrode measurements.-A Pine Instrument ASR rotator (Pine Research Instrumentation, Inc., Durham, NC) was used with a Pt WE. Two types of experiments with RDE configurations were performed: (1) after 10-min stabilization, direct RDE cathodic polarization measurements from −0.20 V SCE to −1.10 V SCE with a scan rate of 0.20 mV/s and rotation speeds of 100, 200, 500, and 1,000 rpm, in 0.3M NaCl+0.1M AlCl 3 and 0.6M NaCl (pH = 3.01), respectively; (2) RDE cathodic polarization measurements starting from OCP to −1.00 V SCE after 1hr OCP measurement in 0.6 M NaCl, 0.6 M NaCl (pH = 3.01) and 0.3 M NaCl+0.1 M AlCl 3 (all three solutions were in a quiescent condition). One additional test in N 2purged environment was conducted in the last solution.

Results
A description of the corrosion kinetics of Al alloys in the absence and presence of Al 3+ , as well as the dependence of mass-transfer limited cathodic kinetics and corrosion resistance of Al alloys on the Al 3+ concentration will be presented here. Most of the results presented are for AA7050-T451.  Fig. 1a. Before introducing Al 3+ , the cathodic limiting current density (i lim ) of AA7050 was ORR-controlled in neutral 0.6 M NaCl (on the order of 1E-06 A/cm 2 ) and smaller than a typical i lim of SS/Pt in the same solution (on the order of 1E-05 A/cm 2 ). 27,28 Once [Al 3+ ] was added to the NaCl, i lim increased significantly, and i lim increased steadily as [Al 3+ ] increased. A comparison of i lim chosen at a reference potential −1.00 V SCE as a function of [Al 3+ ] is shown in Fig. 1b  It is clearly seen that, with addition of Al 3+ into NaCl solution, i lim was significantly enhanced for all three Al alloys, and the order of the increased i lim being AA2024>AA7050>AA5083 in both solutions in the presence and absence of Al 3+ . The fact that AA2024 had highest i lim can be attributed to its high Cu content which allows increased mass-transfer limited cathodic kinetics. 19 It is known that Al 3+ can be hydrated with six water molecules to form complex ion Al(H 2 O) 6 3+ in neutral solutions. 29 This complex ion can undergo a series of complicated hydrolysis reaction, with the dominant reaction being:

Mass transfer limited cathodic kinetics and polarization resistance of Al alloys as a function of [Al
The hydrolysis of hydrated Al 3+ results in an acidic pH. To distinguish the enhanced cathodic kinetics of Al alloys by adding Al 3+ from that due to a lowered pH, a comparison of cathodic kinetics of AA7050 in 0.6 M NaCl (pH = 5.60), 0.6 M NaCl (pH = 3.01), and 0.3 M NaCl + 0.1 M AlCl 3 (pH = 3.01) is displayed in Fig. 3. Using the value at −1.00 V SCE again for reference potential, the i lim for the Al 3+containing solution was about twice that for 0.6 M NaCl at the same pH, indicating the increase in i lim was not solely caused by low pH due to hydrolysis of Al 3+ .
As for the effect of [Al 3+ ] on the polarization resistance, a comparison of full polarization curves of AA7050 as a function of [Al 3+ ] is shown in Figure 4a. R p , E corr and i corr (calculated by the EC-lab software) for each solution in Fig. 4a are listed in Table III. It can be seen that R p fell dramatically from 1.98E+04 cm 2 to 3.79E+03 cm 2   EIS analyses.-In order to better understand the electrochemical mechanism of the effect of [Al 3+ ] on the corrosion behavior of Al alloys and evaluate passive oxide film quality as a function of [Al 3+ ], EIS analyses were performed on AA7050-T7451. Fig. 5 shows the corresponding Nyquist plots for the different solutions. It is clearly seen that an inductive loop appears in the low frequency regime in Al 3+ -containing solution, and as a result, two types of equivalent circuit models were used as shown in Fig. 6. The equivalent circuit model in Fig. 6a was used to fit data in 0.6 M NaCl, and that in Fig. 6b was used for the NaCl+AlCl 3 solutions to take in account the inductive loop, which might be caused relaxation of absorbed intermediate in the oxide film during HER. [30][31][32][33] Also, a Warburg impedance element W was used in both models to account for mass transfer of protons to the oxide/electrode interface in the low frequency regime. At high frequencies, both models have CPE behavior attributed to the passive oxide film, in order to estimate the effective capacitance values from the CPE following the method developed by Hirschorn et al. 34,35 They showed the mathematical relation between the constant phase element (CPE) and the effective capacitance for oxide film thickness to be C e f f = gQ(ρ δ εε 0 ) 1−α [2] Where C eff is the effective capacitance, ρ δ is the film resistivity, ε is the dielectric constant for the film, ε 0 is the vacuum permittivity (8.852E-12 F/m), and g is a dimensionless number which is a function of  α : g = 1 + 2.88(1 − α) 2.375 . For ρ δ and ε, it is assumed that the oxide film is mainly alumina with film resistivity = 1.00E+012 ohm m 36 and ε = 9. 37 All of the fitted EIS parameters are listed in Table IV. Based on these fitted parameters, two factors can be evaluated: the R film C film,eff product to evaluate oxide film quality, and R total = R film +R ct +R Loop. to evaluate total corrosion resistance. The reason the RC product can be used to evaluate oxide film quality is that: is a only a function of film resistivity assuming the dielectric constant is unchanged. Fig. 7 shows the dependence of R film C film,eff as well as R total on [Al 3+ ]. The R film C film,eff product declined dramatically with small additions of Al 3+ , which might be attributed to hydration of oxide film under anodic control when Al 3+ was introduced into NaCl. Higher Al 3+ concentrations led to small increases in R film C film,eff , possibly due to the accumulation of aluminum oxide/hydroxide on the alloy surface with increased Al 3+ concentration. The total resistance decreased with increasing Al 3+ concentration with the largest decrease occurring at the smallest Al 3+ concentrations, echoing the tendency of polarization resistance calculated previously, indicating that the Al alloy is more prone to localized corrosion with increasing [Al 3+ ]. Fig. 8 shows the optical micrographs of cross-sections of sensitized AA5083-H131 samples after exposure to solutions containing 1M Cl − and 0.05M K 2 S 2 O 8 (with varying Al 3+ ion concentration and pH = 3). Qualitatively, the micrographs show an increase in corrosion damage with an increase in the Al 3+ ions in solution. ImageJ software 38 was used to analyze the images to quantitatively evaluate the difference between the corrosion damage obtained in each case. The images were converted into 8-bit and processed for binary (black and white) color composition. Table V shows the results from ImageJ analysis. As evident qualitatively, the analysis confirmed that overall damage increased with an increase in Al ion concentration in solution.

The effect of Al 3+ on cathodic kinetics of Pt and SS.-Several
previous studies have investigated the effect of cathodic polarization on the corrosion of pure Al. Lin and Hebert 39,40 found that the hydrated oxide film transformed into hydroxide and became an ohmic conductor with high conductivity upon polarization to a cathodic potential ∼−1.70 V SCE , resulting in higher cathodic charge and enhanced corrosion rate. Moon and Pyun 41,42 studied the electrochemical behavior of Al after prior cathodic polarization, showing that the observed enhanced corrosion rate was caused by the dissolution of native oxide film due to the prior cathodic polarization (and alkalinization of the surface). However, both of these studies indicate that cathodic  charge changed the property of oxide film which in turn accelerated the corrosion rate. Here, we show that the enhanced cathodic kinetics or corrosion rate is not limited to the Al alloy system, but can be also extended to the other metals and alloys, of which Pt and SS were utilized as two examples in this study. Cathodic polarization curves of SS316L in three different solutions are shown in Fig. 9a. Similar to the phenomenon observed with the Al alloys, SS316L showed enhanced mass transfer limited cathodic kinetics in the presence of Al 3+ compared to that in 0.6M NaCl with natural pH, and 0.6M NaCl adjusted to the same pH, proving again that this enhancement was contributed to the Al 3+ ion, not only the low solution pH due to hydrolysis of Al 3+ . However, it is still difficult to distinguish the ORR limiting current density from HERone. Another group of cathodic polarization curves of Pt in the same three solutions as used in the SS 316L tests, plus one additional test of 0.3M NaCl+0.1 M AlCl 3 in deaerated solution, are displayed in Fig. 9b. Because the i lim in 0.6 M NaCl (pH = 5.60) for Pt is attributed to ORR only, this curve can be utilized to distinguish the HER diffusion limited potential region from that of ORR in both 0.6M NaCl pH = 3.01, and 0.3 M NaCl+0.1M AlCl 3 quiescent solutions. Such an analysis led to an assignment of 0.20 V SCE ∼−0.30 V SCE for ORR diffusion limited kinetics, and −0.50 V SCE ∼−0.90 V SCE for HER diffusion limited kinetics. It is clearly seen that the i lim pertinent to ORR in both 0.6M NaCl (pH = 3.01) and 0.3M NaCl+0.1 M AlCl 3 solutions were almost identical to that in 0.6 M NaCl with natural pH, proving that Al 3+ has negligible effects on ORR related diffusional cathodic kinetics. By comparing i lim in quiescent and deaerated 0.3 M NaCl+0.1 M Table V. IGC damage measured using ImageJ analysis for the exposures shown in Fig. 8. The total damage reported here corresponds to anodic dissolution of both the grain boundaries and the matrix.

100-hr full immersion exposure in solution (pH)
Damage (μm 2 ) 1M NaCl 5.49E+04 0.93M NaCl+0.022 M AlCl 3 7.79E+04 0.85M NaCl+0.050 M AlCl 3 8.69E+04 0.70M NaCl+0.100 M AlCl 3 1.28E+05 ) unless CC License in place (see abstract). ecsdl.org/site/terms_use address. Redistribution subject to ECS terms of use (see 207.241.231.82 Downloaded on 2019-04-27 to IP AlCl 3 solutions, one can see that i lim in deaerated solution was only about 40% of that in quiescent solution, because the reduced ORRrelated current density resulted in a lower total current density due to deaeration, whereas i lim pertinent to HER was still about 40 times higher than that in 0.6 M NaCl with same pH, proving that Al 3+ is the dominant factor contributing significant enhancement in HER diffusional cathodic kinetics. This finding also implies that the interaction between Al 3+ and the native oxide film on the Al alloy surface is not the only contributing factor to the enhancement of cathodic kinetics, but the change in solution property must occur when introducing Al 3+ into NaCl system. It was also observed that a very thin, white film was formed after cathodic polarization of SS/Pt in Al 3+ -containing solution, and SEM/EDS was then utilized to look at the film morphology and composition. Typical SEM and EDS images of Pt after cathodic polarization in 0.3 M NaCl+0.1 M AlCl 3 are displayed in Fig. 10. In the SEM image, the film consists of loosely packed, irregularly shaped flakes, and the composition of film consists of Al and O, implying some type of aluminum oxide/hydroxide was formed on the Pt surface. Lin and Hebert 39 found that aluminum hydroxide can exist on the Al surface even in an acidic environment (0.1 M HCl). However, the detailed composition and chemical structure of this film is beyond the study scope in this work.

Discussion
According to Fick's 1 st Law, the theorical diffusion limited current density assuming one dimensional diffusion can be expressed as where n is the number of transferred during electrochemical reaction, D is the diffusivity of reacting species, C bulk and C surface are the concentration of reacting species at the bulk solution and electrode surface respectively, and δ is the diffusion layer thickness. In this section, a deconstruction of Eq. 3 will be performed to target the dominant parameter controlling the increased HER related diffusion current density.

Formation of aluminum oxide/hydroxide film on the surface as diffusional barrier to HER.-A thin white film was formed on
the Pt surface after cathodic polarization in Al 3+ solution as indicated previously. To discern the role of this film in mass transfer limited cathodic kinetics, continuous cathodic polarization measurement was performed on Pt in 0.3 M NaCl+0.1 M AlCl 3 , in which after one scan, Pt was allowed to rest in the test solution for 5 mins, and then a following scan from the OCP to the final potential was conducted. This procedure was repeated three times. The measured cathodic polarization scans are shown in Fig. 11. After the formation of aluminum oxide/hydroxide film on the surface during the 1 st scan, the i lim in both the ORR region and the HER region were decreased, which indicates that the formed film is diffusion barrier to both ORR and HER. According to Eq. 3, the effective diffusion layer thickness is δ = δ natural_convection + δ f ilm , where δ natural_convection 27,28,43 is the critical natural convection boundary layer thickness in quiescent solution, and δ f ilm is the film thickness. Once the film formed on the surface, the diffusion layer thickness increased from δ natural_convection to δ natural_convection + δ f ilm , resulting in a lower limiting current density. However, both i lim, ORR and i lim, HER were almost the same in 2 nd and 3 rd scan, which implies that the film covered Pt surface might not be a good substrate for future film deposition such that the accumulation of film thickness was negligible after the 1 st scan.

Mechanism of enhanced diffusivity of proton (Grotthuss mechanism)
.-To investigate the diffusivity of the proton in the Al 3+ solution and compare it with that in purely acidified NaCl solution, RDE tests were performed on Pt in both 0.6M NaCl pH = 3.01 and 0.3 M NaCl+0.1 M AlCl 3 (pH = 3.01) with rotation speeds of 100, 200, 500 and 1,000 rpm (Fig. 12). It can be clearly observed that addition of Al 3+ did not alter ORR diffusional cathodic kinetics but increased the HER diffusional cathodic kinetics as rotation speed increased. To exclude the effect of charge-transfer limited kinetics from the total cathodic kinetics, Koutecky-Levich (K-L) analysis 44 was applied which can be expressed as [4] where i total,cathodic is the total cathodic current density, i lim is the true mass-transfer limited current density, i ct is the charge-transfer limited kinetics, B L is Levich constant, and ω is the rotation speed of the RDE. An example of K-L plots at potential = −0.7, −0.8 and −0.9 V SCE in Al 3+ solution is displayed in Fig. 13a. From this figure, one can obtain the values of i ct at each potential and then calculate the average i lim at each rotation speed based on three different potentials. As a result, modified Levich analyses can be obtained in NaCl, and NaCl+AlCl 3 solutions respectively (shown in Fig. 13b) i lim = 0.620 · nF D H + 2/3 υ −1/6 C H + ω 0.5 [5] where D H + is the diffusivity of proton at 25°C, υ is the kinematic viscosity of solution, C H + is the bulk concentration of proton assuming fast HER at the electrode/electrolyte interface such that C surface, H + is zero, and ω is the rotation speed of the RDE. Assuming two solutions have the same υ and C H + , the ratio of the slopes of the two Levich expressions is equal to Hence D H + (Al 3+ ) : D H + (no Al 3+ )is about 108, which means diffusivity in Al 3+ containing solution was two orders of magnitude larger than pure NaCl solution under the same pH value. To roughly estimate the value of proton diffusivity in Al 3+ solution, one can start from calculating D H + in acidified NaCl: assuming that the viscosity of 0.6M NaCl (pH = 3.01) is equivalent to that of neutral 0.6M NaCl solution (9.20E-07 m 2 /s), 28 and considering that the Levich equation has a slope = 0.620 · nF D H + 2/3 υ − 1 6 C H + , one can calculate the proton diffusivity in acidified 0.6 M NaCl without any dissolved aluminum ions to be ∼5.03E-09 m 2 /s. Then, based on the Eq. 6, the estimated D H + (0.1M Al 3+ ) would be ∼5.43E-07 m 2 /s. A mechanism based on Grotthuss Theory 45,46 is proposed to explain the large increased proton diffusivity in Al 3+ -containing solutions. In this theory, proton transport occurs along chains of hydrogen bonds in water involving hopping or tunneling of the proton from one molecule to the next. In pure acidified NaCl solution, the transport of proton can be illustrated in Fig. 14a. Protons from the bulk solution, move along the proton chains in water system, and reach the electrode surface to undergo reduction. The Grotthuss Theory can be also applied into Al 3+ -containing solution. 47 When Al 3+ is introduced into the system (Fig. 14b), for example, the Al 3+ coordinates with O at location 1), such that the center of electron cloud moves from the O toward Al 3+ , lowering the bond energy between O and H relative to Fig. 14a. As a result, the proton will transport more easily in Al 3+ -containing solution, leading to a higher diffusivity. Figure 14. Proton Transport in: a) purely acidified NaCl solution and b) Al 3+ containing solution based on Grotthuss theory. 45,46 The diagrams are revised based on Grozovski et al. 46 's work.

Conclusions
The effect of Al 3+ on the cathodic kinetics of Al alloys has been explored. It has been found that the addition of Al 3+ into NaCl solutions can significantly increase the diffusion limited cathodic kinetics of Al alloys, and this increase is proportional to the [Al 3+ ]. The same phenomenon was also observed on Pt and SS316L, which indicates that this enhancement in cathodic kinetics is not related the surface structure of Al alloy, and HER diffusion limited kinetics are increased rather than ORR kinetics as a result. Based on electrochemical studies on Pt, it is proposed that, a mechanism similar to that described by the Grotthuss Theory is operative in which Al 3+ can facilitate transport of the proton, even through a precipitated oxide/hydroxide film on Pt which one might expect to act as a diffusional barrier to HER during cathodic polarization. This enhanced proton diffusivity due to Al 3+ overwhelms the barrier effect of precipitate film, and then significantly increases cathodic limiting current density.