Corrosion Inhibition of Pure Aluminium and Alloys AA2024-T3 and AA7075-T6 by Cerium(III) and Cerium(IV) Salts

Aluminium (Al > 99.0%) as 1.0 mm thick sheet was supplied by GoodFellow, Cambridge Ltd., UK. The aluminium alloys AA2024-T3 and AA7075-T6, in the form of 0.5 mm thick sheets and distributed by Kaiser Aluminum, USA, were used as substrates. The compositions of the aluminium and its alloys are given in Table I. Discs of diameter 14 mm were cut from the sheets, then water-ground successively up to 4000-grit SiC emery papers (Struers, Denmark), rinsed thoroughly with deionized water, cleaned ultrasonically in ethanol for 10 minutes and dried with a stream of nitrogen.

Aluminium was pre-treated in distilled water for one hour to obtain a thicker oxide layer, which improves the corrosion resistance and assists the incorporation of cerium ions into the oxide layer.

The corrosive medium, 0.1 M NaCl, was prepared with analytical grade NaCl (Carlo Erba 99.5 – 100.5%, pH = 5.5) and Milli-Q Direct water with a resistivity of 18.2 MΩ cm at 25°C (Millipore, Billerica, MA). Inhibitory cerium salts were dissolved in this solution at 1 to 5 mM.

The following cerium(III) and cerium(IV) salts were used as corrosion inhibitors: cerium(III) chloride, (CeCl₃ × 7H₂O, 99.9%, Aldrich), cerium(III) nitrate, (Ce(NO₃)₃ × 6H₂O, 99.5%, Fluka), cerium(IV) sulfate (Ce(SO₄)₂ × 2H₂O, 99.9%, Carlo Erba) and cerium(III) acetate, (Ce(CH₃COO)₃ × nH₂O, 99.9%, Aldrich). For the latter the amount of crystal-bound water was calculated from thermogravimetric analysis to give Ce(CH₃COO)₃ × 2H₂O.

Field emission scanning electron microscopy (SEM), in combination with energy dispersive X-ray spectrometry (EDS) was used to analyze the microstructure and composition of aluminium alloys. AA2024-T3 alloy was analyzed using a JEOL JSM-7600F SEM equipped with an Oxford Instruments Inca 400 EDS. AA7075-T6 was analyzed using a Carl Zeiss SUPRA 35VP SEM and an Oxford Instruments Inca 400 EDS. SEM analysis was performed in a secondary electron (SE) and back-scattered (BSE) mode with energy 5 kV.

Electrochemical measurements were performed in a three-electrode standard corrosion cell (Parstat Corrosion Cell Kit model K0047, volume 1 L) at 25°C. Substrate discs embedded in a Teflon holder (model K0105 Flat Specimen Holder Kit), leaving an area of 0.95 cm² exposed to the corroding solution, served as a working electrode. A saturated calomel electrode (SCE, Hg/Hg₂Cl₂, 0.242 V vs. saturated hydrogen electrode) placed in a Luggin capillary salt bridge was used as a reference electrode. Carbon rods served as a counter electrode. All potentials in the text refer to the SCE scale. Electrochemical experiments were carried out with an Autolab PGSTAT 12 (Metrohm Autolab, Utrecht, The Netherlands) potentiostat/galvanostat and controlled by Nova 1.10 software.

Prior to measurements, samples were allowed to stabilize under open circuit conditions for approximately 1 hour. During that time the open circuit potential was measured every 2 seconds. The stable, quasi-steady state potential reached at the end of the stabilization period is denoted as the open circuit potential, E_oc. Electrochemical measurements were carried out following stabilization. Linear polarization was measured over a potential range of ± 10 mV vs. E_oc, using a potential scan rate of 0.1 mV/s. Values of polarization resistance, R_p, were deduced from the slope of fitted current density vs. potential lines, using Nova software. Potentiodynamic measurements were performed using a 1 mV/s potential scan rate, from −250 mV to E_oc. The potential was then increased in the anodic direction. This type of measurement is appropriate for comparison between different inhibitors; for a more accurate assessment of anodic and cathodic kinetics, separate anodic and cathodic scans starting at the E_oc should be performed.

For each sample, measurements were performed in at least triplicate. A representative measurement was chosen and presented in figures. In tables, the results are presented as mean value ± standard deviation. The following electrochemical parameters were determined: corrosion current density (j_corr), corrosion potential (E_corr), pitting potential (E_pit) and the difference ΔE = E_pit – E_corr.

The corrosion inhibition effectiveness, IE, was calculated according to Equation 1

where R_p^* is the polarization resistance for the inhibited, and R_p that for the non-inhibited system. The values are given in %.

Short-term (24 h) immersion tests were carried out in 0.1 M NaCl solution, with and without the addition of 3 mM CeCl₃ and Ce(Ac)₃. Tests were performed in 100 mL polyethylene vials at room temperature. After immersion test the samples were analyzed using a confocal microscope (Axio, CDM 700, Zeiss, Gottingen, Germany) with 50 × objective zoom.

The heterogeneous microstructure of aluminium alloys results in a variety of types and a range of numbers of intermetallic particles (IMs). The IMs also differ in their electrochemical activity with respect to the aluminium matrix.³ The AA2024-T3 from 2xxx series contains the following main IMs: Al-Cu-Mg (S-phase), binary Al-Cu (θ phase) and Al-Cu-Fe-Mn-Si.^43–48 The SEM/EDS analysis of AA2024-T3 is presented in Fig. 1. The Mg-containing IMs are spherical and several micrometres in diameter. Binary Al-Cu particles are also spherical and small (1-2 μm). Larger, Fe-rich, Al-Cu-Fe-Mn IMs are usually irregular in shape and are noble (more positive corrosion potential relative to matrix) (Fig. 1). Hughes et al. recently analyzed 18,000 IMs in AA2024-T3 and identified nine different compositions, including matrix and a periphery composition around S-phase/θ-phase composites as well as around other IMs containing Al, Cu, Fe, Mn and Si.^48,49 The most common compositional domains were those containing Al, Cu, Fe, Mn and Si, constituting 47% of the total domain count. The S-phase and θ-phase constitute around 39% of the total. The remainder of the particles contained predominantly Al, Cu, Mn and Fe. 77% of the IMs domains were designated noble character.⁴⁸

Zoom In Zoom Out Reset image size

The main IMs in AA7075-T6 from 7xxx series are active (less positive corrosion potential) (Al-Cu-Mg-Zn) or noble (Al-Fe-Cu-Mn-Zn) relative to matrix.^43,50–54 SEM/EDS analysis of AA7075-T6 is presented in Fig. 2. The most common particles are Al₂CuFe, Al₇Cu₂Fe, Al₂CuMg and Mg₂Si particles; the Mg₂Si particles are active in respect to matrix and dissolve preferentially.^43,50 AA7075-T6 also contains precipitate particles of MgZn₂ that are formed by nucleation and growth from a supersaturated solid solution. These particles are dispersed within the matrix.

Zoom In Zoom Out Reset image size

It is noteworthy that sample preparation process, namely grinding and polishing, can create an altered surface layer on Al alloys that has been shown to influence corrosion properties.^44,45,55,56

Ground metal and alloy substrates have different appearances because of the presence of secondary phases, i.e. intermetallic particles of various shapes and sizes (Fig. 3, upper row), as described above. After 24 h immersion in chloride medium aluminium corroded slowly and the corrosion products were observed as grey-brown spots over an area more than 50 μm in diameter (Fig. 3, lower row). The corrosion was initiated at some weak points such as impurities, e.g. Cu, Fe, Mn, Si, Zn, present in aluminium, which may act as initiation sites for corrosion within the aluminium oxide surface.³⁵ In contrast, corrosion on AA2024-T3 and AA7075-T6 surfaces was localized on intermetallic particles. The corrosion process was more pronounced on AA7075-T6 where almost all the IM particles were corroded, corroded areas coalesced and corrosion spread around the dark areas around inclusions. The localized corrosion of different types of IMs in these two alloys was described in details and includes the preferential dissolution of Mg-rich particles leaving behind copper-rich particles, and dissolution of matrix around primarily noble Al-Cu-Fe-Mn particles leading to so called trenching.^{44–46,48,51,53}

Zoom In Zoom Out Reset image size

Following stabilization at the E_oc, values of polarization resistance, R_p, were measured in 0.1 M NaCl solution (Table II). The values suggest that AA7075-T6 exhibits the lowest corrosion resistance in chloride solution. Al and AA2024-T3 have more than 10-fold larger values of R_p. However, all three materials are subjected to corrosion and pitting in chloride solution (Fig. 3), and further protection is required.

Polarization curves were recorded for Al, AA2024-T3 and AA7075-T6 in NaCl solution (Figs. 4–6), together with the corresponding electrochemical parameters (Table II). Cathodic curves are related to reduction of dissolved oxygen which can proceed through 2e⁻ or 4e⁻ reactions.^17,57 Values of E_corr were between −0.68 V and −0.70 V, while those of j_corr differed considerably between the substrates, AA2024-T3 and AA7075-T6. The rapid increase in anodic current density with increasing potential is associated with the pitting corrosion related to dissolution of IM particles.⁵⁸ No passive region was established before pitting. At more positive potentials, one or two breakdown potentials may occur depending on the type of substrate, surface preparation and deaeration of solution.^{44,45,55,56,58} For AA7075 in deaerated solution the first breakdown is related to transient dissolution of surface layer enriched in Zn formed during polishing.⁵⁵ The second breakdown is associated with combined intergranular and selective grain attack.⁵⁶ For AA2024 two breakdowns are related to transient dissolution of S phase and initiation and initiation of intergranular corrosion.^44,45,58

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The span of ΔE (E_corr−E_pit) is taken as the measure of initiation of pit nucleation.¹¹ Values of ΔE ranged only from 10 to 30 mV, indicating the high susceptibility these alloys to pit nucleation (Table II). From the appearance of the surface (Fig. 3) and the electrochemical parameters (Figs. 3–6, Table II) it is evident that, of the three substrates, AA7075-T6 is the most susceptible to both uniform and localized corrosion.

In general, immersion in solutions containing Ce(III) salts improved corrosion properties, resulting in increased R_p, reduced j_corr, a shift in E_corr and the appearance of a passive region (Table II). The Ce(IV) salt, Ce(SO₄)₂, did not act as a corrosion inhibitor. As presented below, the polarization behavior in the presence of various Ce salts is dependent on their type and also on the type of substrate.

Ce(NO₃)₃ and Ce(Ac)₃ shifted the E_corr of Al to more negative values, indicating their effect on the blocking of cathodic sites and the reduction of oxygen (Fig. 4, Table II). On the anodic side a quasi-passive range was established in the presence of both salts, indicating that aluminium oxide formation/thickening is stimulated. The higher value of ΔE was noticed in the presence of Ce(Ac)₃ (210 mV) than that of Ce(NO₃)₃ (20 mV). On addition of CeCl₃ the shift of E_corr was more positive. It acted as a mixed inhibitor; however, no passive range was established as in the case of nitrate and acetate. Ce(SO₄)₂ accelerated the corrosion of Al.

The effect of Ce salts on AA2024-T3 is presented in Fig. 5. Ce(Ac)₃ acted in a similar way to its effect on Al, causing a negative shift of E_corr. It affected both the cathodic and anodic reactions, with a passive region established up to −0.58 V. The action of Ce(NO₃)₃ on AA2024-T3 however, differed from that on Al, with E_corr shifted ∼100 mV more positive. The established passive range is narrower than for Al but lower current densities were achieved. The action of CeCl₃ was similar to that on Al. Ce(SO₄)₂ increased the cathodic current density by 30 times.

Zoom In Zoom Out Reset image size

Polarization curves were determined for AA7075-T6 in the presence of various Ce salts in chloride solution (Fig. 6). Here CeCl₃ and Ce(Ac)₃ acted as classical cathodic inhibitors, shifting E_corr to more negative values by more than 200 mV. The cathodic current density was significantly reduced. At E>E_corr the increase in current density slowed down, however the abrupt increase in current density was similar as that in uninhibited solution. This behavior was described by Hinton et al.¹¹ for AA7075-T6 in the presence of CeCl₃. The absence of pitting was ascribed to displacement of E_corr away from E_pit in the negative direction. The value of ΔE can be considered as a relevant indicator of pitting resistance.¹¹ The values of ΔE of 210 mV and 230 mV for CeCl₃ and Ce(Ac)₃ indicate that these two salts act as effective cathodic inhibitors for AA7075-T6. The effect of Ce(NO₃)₃ was similar to that on AA2024-T3, i.e. it changed the value of E_corr to more positive and slowed down the dissolution, due to the stimulation of oxide formation.²⁴ Ce(SO₄)₂ acted as a corrosion accelerator owing to increases in both the cathodic and anodic current density.

From the values of electrochemical parameters and of inhibition effectiveness, IE, the most effective inhibitor for all three materials is seen to be Ce(Ac)₃ as it shifted values of E_corr to more negative values but affected both cathodic and anodic reactions (Table II). The passivation region appeared independently on the metal substrate. The broad ΔE region accounts for the good protection against pitting corrosion. CeCl₃ was less effective (Table II). It acted as a typical cathodic inhibitor on AA7075-T6, with values of E_corr displaced negatively but, for Al and AA2024-T3, the shift was positive. The related value of IE was less for AA7075-T6 than for Al (Table II). Ce(NO₃)₃ caused a positive shift of E_corr for AA2024-T3 and AA7075-T6 and affected the anodic reaction, as is typical for nitrate as a passivating inhibitor. In terms of IE, it is less effective than Ce(Ac)₃ and CeCl₃. The value of IE is highest for AA7075-T6 (Table II). CeCl₃ and Ce(NO₃)₃ acted similarly for AA2024-T3 but not for AA7075-T6. For Al, CeCl₃ and Ce(Ac)₃ acted similarly while nitrate showed an intermediate behavior.

The most effective cerium salts, Ce(Ac)₃ and CeCl₃, have been investigated in more detail by making electrochemical measurements as a function of concentration and by the immersion testing.

It has been reported that the inhibition effectiveness of various cerium salts on AA7075-T6^11,12 and on AA2024-T3^{16,23,24,26,34,35} is concentration dependent. For AA7075-T6 the optimal concentration of CeCl₃ is 1000 ppm (0.0027 M), while, for AA2024-T3, lower concentrations (e.g. 10⁻⁵ M) are more effective corrosion inhibitors than higher concentrations (10⁻² M).^16,24 Decreased inhibition effectiveness with increase in concentration is related to the decrease in solution pH with increasing salt concentration.^16,24 If the pH drops to values far below those at which Al₂O₃ is stable the corrosion protection is disabled. Herein, we present the behavior of Al and AA7075-T6 in the presence of various concentrations of CeCl₃ and Ce(Ac)₃ (Figs. 7–9). Different concentrations were studied to optimize the inhibition performance but all were in the lower range of 1–5 mM in order to avoid the dissolution of Al oxide at low pH.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

R_p values for Al in NaCl containing Ce(Ac)₃ were, at all concentrations, higher than those in non-inhibiting solutions (Table III). However, at the lowest concentration of Ce(Ac)₃ (1 mM) R_p was at its highest value (IE ∼92%) and decreased with increasing amount of cerium salt to an IE of 80%. At the same time E_corr shifted to more negative values, causing ΔE to increase to 280 mV, thus affecting the susceptibility to pit nucleation. The quasi passive range was established as being broader at higher concentrations (Fig. 7). The results thus show that it is not possible to predict the resistance against pit nucleation judging from the R_p and j_corr values alone. Although the optimal value of IE was observed at the lowest concentration (1 mM), only at higher concentrations did the modified surface of aluminium exhibit a stronger effect on the cathodic part as well as the anodic part, indicating that Ce(III) acetate is capable of protecting Al against both general and localized corrosion.

Potentiodynamic measurements were made on AA7075-T6 in 0.1 M NaCl containing Ce(Ac)₃ in the concentration range of 1 mM to 5 mM (Table IV, Fig. 8). With increasing concentrations of Ce(Ac)₃, E_corr shifted to more negative values. The broadening of ΔE was related only to shift of E_corr in the negative direction. Due to the larger negative shift of E_corr, the values of ΔE were larger, indicating greater protection against pit nucleation. Optimal conditions, large R_p, small j_corr and broad ΔE, were reached at 3 mM Ce(Ac)₃. At higher concentrations, ΔE broadened but values of j_corr increased and of R_p decreased. The higher IE for Ce(Ac)₃ on AA7075-T6 than on Al can be attributed to the deposition of cerium hydroxide on noble IMs in the alloy, which additionally increases the protection. As IMs are the main pitting sites in the alloy, protection of the alloy is dependent on their protection.

Zoom In Zoom Out Reset image size

Potentiodynamic measurements were made on AA7075-T6 in 0.1 M NaCl containing CeCl₃ in the concentration range from 1 mM to 5 mM (Table V and Fig. 9). The parameters obtained show that the optimal CeCl₃ concentration was 3 mM, resulting in the broadest ΔE, the largest R_p and the smallest j_corr. The protection mechanism is similar to that for Ce(Ac)₃ (Fig. 8).

Surface images of Al, AA2024-T3 and AA7075-T6 were obtained following their immersion in NaCl containing the most effective cerium salts, CeCl₃ and Ce(Ac)₃ (Fig. 10) and are to be compared to those of ground samples and those immersed in NaCl (Fig. 3). Following immersion in either of the two Ce salts, there was no evidence of corrosion on Al, the surfaces appearing similar to those of ground samples (Figs. 10a, 10d). Only a few, less deep, dark pits were observed on AA2024-T3 in the presence of CeCl₃ (Fig. 10b) while, in the presence of Ce(Ac)₃, even the surface between IMs appeared to be covered by particles (Fig. 10e). The effect of Ce salts is most significant for AA7075-T6. In CeCl₃ containing NaCl there were no dark areas of coalescent pits; IM particles appeared darker, probably due to deposition of Ce(III) hydroxide (Fig. 10c). In Ce(Ac)₃-containing NaCl the appearance of IM particles was similar although there were small particles deposited on the surface surrounding IMs (Fig. 10f); the overall effect was more emphasized than that on AA2024-T3. It is assumed that these particles are related to the deposition of Ce-rich particles on the Al matrix. This effect was observed only in the presence of Ce(Ac)₃.

Zoom In Zoom Out Reset image size

The above results indicate that several factors affect the inhibition effectiveness of a particular Ce salt for a particular substrate, viz. (i) the type of intermetallic particles, (ii) the particular Ce salt and (iii) its concentration. In all these cases the effect is pH dependent. The addition of 3 mM cerium salt to 0.1 M NaCl changed its pH from 5.5 to 6.8 for Ce(Ac)₃, 5.5 for Ce(NO₃)₃, 5.2 for CeCl₃ and 1.9 for Ce(SO₄)₂. The solution with acetate salt has the highest pH and that with sulfate salt the lowest. pH changes with the salt concentration. Addition of Ce(Ac)₃ increases the pH from 6.5 for 1 mM to 6.9 for 5 mM. Increasing the concentration of CeCl₃ and of Ce(NO₃)₃ in 0.1 M NaCl from 1 to 5 mM does not affect the pH significantly. Since the pH range of the stability of Al₂O₃ is between 4.5 and 8.5,¹ it can be assumed, judging from the pH alone, that the oxide would be most stable in acetate solution. However, the stability of the alloy is not determined solely by that of Al oxide but also by those of the most common IMs.^59,60 MgZn₂ is unstable at any pH within the range of 2.5 to 12.5; Al₂CuMg dissolves at lower pH but decreasingly so with increasing pH; Al₇Cu₂Fe is spontaneously passive at pH values between 2.5 and 6 while AlFe₃ and Al₂Cu show well-defined regions of passivity at higher pH as a result of the passivity of Fe.⁶⁰ Local morphology is therefore dependent on the solution pH: stochastic pitting is observed at acid pH, deterministic-type pitting at near-neutral pH, and general corrosion at an alkaline pH.⁶⁰ The Al-Cu-Mg IMs in AA2024-T3 are rapidly dissolved at E_oc, whereas for those in Al-Cu-Fe-Mn the dissolution is several orders of magnitude slower.⁵⁹ The differences in performance of the Ce salts observed could be related to the dependence on pH of the behavior of the different IMs in the two alloys. In AA2024-T3, the concentration of Cu governs the behavior of the alloy while, in AA7075-T6, Cu concentration is lower and the alloy contains more Zn and Mg (Table I). As MgZn₂ is not passivated at any pH,⁶⁰ AA7075-T6 may be even more affected by lowering the pH in the presence of Ce salt, due to higher dissolution of IMs. The type of protection of AA7075-T6 by Ce salts is reported to be identical to that of AA2024-T3, i.e. precipitation of Ce oxide/hydroxides at the cathodic sites in the alloy.²⁴ However, the type of salt affected the inhibition effectiveness as well as the mechanism of inhibition, as observed in the present work when using different substrates. Harvey et al. investigated a range of structurally-related organic compounds and noticed that the inhibition effectiveness is different depending whether AA2024 or AA7075 substrate was used, despite their corrosion rate in 0.1 M NaCl being similar.⁸ Furthermore, for AA2024 the addition of sodium acetate produced an acceleration effect, whilst for AA7075 the inhibition effect was only 15%.⁸ Considering the high inhibition effect produced by cerium(III) acetate in the present work it can be stated that cation (i.e. Ce³⁺) plays a major role in the inhibition process. The most effective inhibitors for AA7075-T6 are CeCl₃ and Ce(Ac)₃ which act by shifting E_corr in the negative direction and by reducing the range not susceptible to pitting in the same direction. Different mechanisms of protection were observed for Al and AA2024-T3 as CeCl₃ and Ce(Ac)₃ affected the anodic process as well. Since E_corr were shifted more positive, conditions were reached under which formation and thickening of aluminium oxide took place.

Enrichment of copper at Al-Cu-Mg precipitates occurs after selective dissolution of Mg and Al which leaves Cu-rich remnant particles. If copper dissolves, it can then redeposit as thin copper layer onto alloy matrix around the IMs.^16,44,45,61 Redeposited copper increases the effective surface area of the cathodic zone.^16,44,45,61 NO₃⁻ anions reduce to provide a higher cathodic current. In the presence of NO₃⁻ anions, the potential of the alloys shifts in the anodic direction, reaching the conditions for formation of an oxide film, which inhibits corrosion (Figs. 5, 6), as noticed previously.²⁴ Nitrate ions can be reduced electrochemically from very dilute solutions with a sufficiently high rate in the presence of cuprous ions.⁶² This process enhances the passivation ability of nitrates in acidic media − higher concentrations would otherwise be required for reliable passivation.⁶² Ce(SO₄)₂, as a Ce(IV) salt, proved ineffective as a corrosion inhibitor due to its low solubility and to the low pH of its aqueous solution, as well as to the fact that the formation of precipitates requires the formation of Ce(III) ions.