Electrochemical testing of exfoliation corrosion sensitivity of 7XXX

Aluminum alloys

T. Marlaud1,2, B. Malki1, A. Deschamps1,M. Reboul1, B. Baroux1

1Institut National Polytechnique de Grenoble, ENSEEG/LTPCM 1130 rue de la piscine, 38402 Saint Martin d’Hères, France

2 Alcan Centre de Recherches de Voreppe, Parc Economique Centr’Alp, 725 rue Aristide Berges BP27 38341 Voreppe Cedex, France

This paper presents an investigation of exfoliation corrosion (EFC) of 7000 series aluminum alloys using a new electrochemical test. For this study two metallurgical states, namely T6 (peak hardness) and T76 (over-aged) of a new high strength 7XXX alloy have been investigated. The experimental protocol consists of a galvanostatic test in an optimized NaCl-NaNO3-AlCl3 electrolyte at pH~4, chosen to be less aggressive than that used in the standard EXCO test1. The real time fluctuations of the measured potentials, which are expected to contain relevant information on the exfoliation corrosion processes, have been analyzed. SEM observations in the early stages suggest that the measured potential transients are directly related to fracture of surface blisters, commonly evoked in the literature as the predominating mechanism in exfoliation corrosion. This new test opens a new way for quantifying the susceptibility to exfoliation corrosion in aluminum alloys.

Introduction

High strength 7000 series aluminum alloys, used in aircraft industries, are known to be sensitive to structural corrosion, particularly to exfoliation corrosion depending on the metallurgical state, and particularly at peak strength2. The usual way to reduce the corrosion susceptibility is to over age the alloy2 at the expense of 15% decrease of its mechanical strength.

Exfoliation corrosion can be seen as a form of intergranular corrosion that occurs at

the surface of high strength aluminum alloys with an elongated grain structure parallel to the plate surface3. In strongly demanding environments, the corrosion along intergranular active paths produces hydrated-chlorurated aluminum precipitates having higher molar volumes than the original metal4. This results in wedging stresses that lift the surface grains, giving rise to a layered appearance. Numerous procedures have been proposed to assess the susceptibility of aluminum alloys to this type of corrosion (e.g. ASSET, MASTMAASIS, EXCO...)5,6. Most of them, even widely accepted, give only qualitative

ECS Transactions, 3 (31) 285-293 (2007)10.1149/1.2789235, copyright The Electrochemical Society

285

286

results. Among the most common of these techniques is the EXCO test (ASTM G34). This test is essentially based on a visual examination of the alloys surface, which are compared to standard corrosion morphologies. However, this test does not provide any quantitative measurement (e.g. on corrosion kinetics), which are required for corrosion lifetime predictions and for a better understanding of the underlying corrosion mechanisms. An attempt to determine EFC kinetics was made using deflection techniques based on mechanical compliance measurements under four-point bending7. However, this technique has proven to be valid only during uniform corrosion.

In this context, the objective of the present work is to develop a new test based on the quantitative analysis of electrochemical transients. For this end, two metallurgical states of a 7XXX aluminum alloy have been studied. First, EXCO tests were carried out to asses the different susceptibility of these two states to EFC. Afterwards, both galvanostatic and potentiostatic measurements have been conducted in a new chloride based electrolyte, with a monitoring of the electrochemical transients.

Materials and methods

The studied samples were taken from a 25 mm industrial plate of the 7XXX alloy produced at the Alcan Issoire plant. The approximate composition is given in Table 1. Two metallurgical states were obtained with industrial two step heat treatments: T6 (peak hardness) and T76 (over-aged). These two metallurgical states, coming from the same plate are found to have the same elongated grain structure, which is one of the main parameters leading to EFC behavior.

Zn Mg Cu

Composition (%wt) ~10 ~2 ~1.5 Table 1: Chemical composition of the studied 7XXX alloy. Only the major alloying elements are reported.

Exfoliation corrosion susceptibility of each metallurgical state was first assessed

using the standard EXCO test (ASTM G-34) on a 50mm×100mm surface machined from the T/2 section of the plate and polished to 1200 Grit. In this test, the so-called EA-ED classification describes a range from superficial to severe exfoliation (ASTM-G34 1974). During the test, both pH and electrochemical potential time evolution were recorded. The EFC kinetics during the EXCO test were then evaluated by stopping the test and quoting the visual rating of corroded specimens.

The galvanostatic and potentiostatic measurements were performed on cylinders 15mm in diameter cut from the industrial plate and polished up to 1200 grit. The electrochemical instrumentation was a VoltaLab PGZ301 potentiostat. The optimized electrolyte contains: 1M NaCl, 0,25M NaNO3 and 0,033M AlCl3. The solution, not de-aerated, was kept at room temperature. A Pt counter electrode and a saturated calomel

ECS Transactions, 3 (31) 285-293 (2007)

287

reference electrode (SCE) were used. Finally, SEM characterization was performed on corroded samples using an LEO STEREOSCAN 440 SEM, operating at 20kV.

Results and discussion

EXCO test

The EXCO rating results for the two tempers as a function of time are reported in Figure 1. The results reveal clearly the influence of the metallurgical heat treatments. As expected, the over-aged temper T76 (final quotation EB) is less sensitive to exfoliation corrosion than the T6 temper (final quotation ED). However, this difference appears only at long corrosion times; at short times both tempers show a similar susceptibility, with even a slight advantage for the T6 temper.

The corresponding Open Circuit Potential (OCP) and pH evolution are shown in

Figure 3. This Figure reveals that the time evolution of these two parameters is approximately the same for the two tempers, which tend to prove that these two parameters are not suitable for discriminating exfoliation corrosion susceptibility of aluminum alloys. Their time evolution is detailed below.

In a first approximation, considering mainly the aluminum dissolution (<92 at %),

one can outline a simple corrosion mechanism following the two electrochemical reactions:

i) anodic reaction:

Al → Al3+ + 3e- (1)

with: [ ]+++ +−=+=+ 3

Al

Al30

3 Allog0.0666.1a

aln

nFRT/Al)(AlE/Al)E(Al 3 ;

ii) cathodic reaction: H+ + e- → ½ H2 (2)

with: pH0.06HaH

aln

nFRT)/H(HE)/HE(H

2

202 −=+=+

++ .

In the first twenty hours, the pH increases from 0.4 to 3.4 (see Figure 3) due to proton consumption at the cathodic reaction. Subsequently, precipitation of Al(OH)3 hydroxides occurs, which stabilizes the pH of the solution according to the chemical reaction:

Al3+ + 3H20 → Al(OH)3 + 3H+ KS = 10-32,34 (3)

As for the OCP evolution, we suggest that it results directly from anodic and cathodic

processes as schematically illustrated in Figure 3. The increase of pH will shift the hydrogen couple potential E0 (H+/H2) towards lower potential, resulting in an OCP diminution. Nevertheless, this same pH enhancement can decrease the anodic process kinetics and increase the OCP. So the OCP evolution depends in a complex way on the pH enhancement during the EXCO test. In addition, the anodic process depends on the active surface which increases during the exfoliation, because of corrosion layer production.

ECS Transactions, 3 (31) 285-293 (2007)

288

The new test Solution:

The chosen solution is an electrolyte of 1M NaCl, 0,25M NaNO3 and 0,033M AlCl3. As in the EXCO test, the electrolyte is mainly composed of Cl- and NO3

- ions. Cl- ions are known to destabilize any formed passive film, enhancing uniform corrosion attacks, while NO3

- ions are known to be corrosion inhibitors, resulting in localized corrosion attacks. AlCl3 completes the solution in order to limit the aluminum Al3+ cation evolution and to stabilize the pH8.

Protocol:

The experimental protocol consists of a galvanostatic test conducted with an applied current density intensity of 2,5 mA/cm2. The resulting potential transients were recorded using a sampling frequency of 2.7 hertz, which is the maximum allowed by the experimental equipment.

Results: Figure 5 shows the corrosion morphologies of the two metallurgical states after a 12 h of galvanostatic test. Similarly to the EXCO results, only the T6 temper developed evident EFC. On the T7 sample, no apparent corrosion layers have been observed.

Now, if we examine the potential fluctuations response (see Figure 6), we can see that the steady state potential is constant during the galvanostatic test but differs for the two tempers: ET7 < ET6. Here again, this potential could be taken as a criterion for EFC susceptibility, however, because of its dependence on alloy composition it seems difficult to use it to compare different aluminum alloys. The difference between the two tempers indicates only their difference in matrix chemical composition. Nevertheless, if we look in more detail to potential fluctuations, well defined transients can be observed for the T6 sample (see Figure 6). We expect that these elementary transients are relevant for EFC activity. Indeed, in a galvanostatic polarization mode, any potential decrease toward more negative values can be seen as the response of the triggering-setup to an increase of the current density. During EFC, corrosion layer detachment and/or fracture of blisters create permanently new fresh active surfaces, and this will necessarily accelerate the anodic dissolution, increasing the current. In this case, the transients may be considered as the signature of EFC kinetics.

In order to confirm this observation, potentiostatic tests have been performed for the

two tempers at steady state potentials measured during galvanostatic tests: -620mV/SCE for the T6 state and -680mV/SCE for the T76 state. The intensity fluctuations for the two states are reported in Figure 7. As expected, the recorded intensity is the same for the two tempers (2,5mA). Moreover, positive intensity transients are only observed for the T6 state (see Figure 7). This demonstrates clearly that the observed electrochemical activity is closely related to EFC growth processes. Finally, SEM observations in the early stages

ECS Transactions, 3 (31) 285-293 (2007)

289

of the test after 50 minutes show surface blisters fracturing, which is characteristic of exfoliation corrosion in aluminum alloys (Figure 8).

Conclusions

With the objective of better understanding the EFC of aluminum alloys, a new electrochemical test has been developed, and tested on two metallurgical states showing different exfoliation corrosion susceptibility. This test, based on the measurement of potential transients during a galvanostatic experiment in a solution at stable pH, is shown to reproduce the EFC mechanisms observed during the standard EXCO test. The observed potential transients are shown to be related to individual exfoliation corrosion events. Therefore it is expected that the quantification of these events in magnitude and frequency can lead to a better understanding of the EFC susceptibility as a function of alloy and temper, as well as a better understanding of the underlying physico-chemical mechanisms.

Acknowledgements

Thanks are due to Gaëlle Pouget and Christine Henon for their work on the EXCO

test, as well to Alcan CRV for permission to publish.

References

[1] ASTM-G34 (1974). Standard test method for exfoliation corrosion susceptibility in 2xxx and 7xxx series Aluminum alloys (EXCO test). [2] R. Develay, «Traitement thermiques des alliages d'aluminium» in Techniques de l'ingénieur, M1290, M1291. [3] M. J. Robinson and N. C. Jackson, Corr. Sci. 41, 1013-1028 (1999). [4] D. J. Kelly and M. J. Robinson, Corrosion, 49, 787-795 (1993). [5] B. W. Lifka and D. O. Sprowls, Corrosion, 22, (1966). [6] D. O. Sprowls and J. D. Walshal, "Simplified exfoliation testing of aluminum alloys, Localized Corrosion-Cause of Metal Failure", ASTM STP, 516 (1972). [7] E. A. G. Liddiard and J. A. Whittaker, J. Inst. Met., 89, 377-384 (1960). [8] R. T. Foley and T. H. Nguyen, Aluminum Corrosion, 129, 464-467 (1980).

ECS Transactions, 3 (31) 285-293 (2007)

290

Figure 1: EXCO rating evolution with time for the two metallurgical states T6 (peak age) and T76 (over-aged) of the 7XXX aluminum alloy.

T6 : ED T76 : EB

Figure 2: Visual EXCO rating results for the 2 metallurgical states T6 and T76 of the 7XXX aluminum alloy.

a) b)

Figure 3: Evolution of (a) the OCP and (b) the pH during EXCO test for two metallurgical states of the 7XXX aluminum alloy (T6 and T76). These two evolutions are qualitatively the same for the two tempers.

0

5

10

15

20

25

30

0 10 20 30 40 50

temps (h)E

XCO

ED

EC

EB

EA

P

Time (h) Time (h)

T6

T76

ECS Transactions, 3 (31) 285-293 (2007)

291

log i

E

Anodic oxydation

Cathodic reduction

E(H+/H2) pH = 0,4

E(Al3+/Al) (invariant with pH)

EOCP

icor

E(H+/H2) pH = 3,4

pH ↑ pH ↑

E1 E2

Figure 4: Schematic evolution of the Open Circuit Potential (OCP) with pH during EXCO test based on a simplified exfoliation mechanism.

Figure 5: Corrosion face of the 7XXX T6 (peak aged) and T76 (over-aged), after 12h of the galvanostatic test. Visual EFC are observed for the T6 state, with layers. Only powder is present on the T76 sample.

Decrease of the anodic process kinetics

Powder on the surface

ECS Transactions, 3 (31) 285-293 (2007)

292

-0,700

-0,670

-0,640

-0,610

0 10000 20000 30000 40000 t (s)E(V)

T6

T76

-0,690

-0,660

-0,630

8000 8200 8400 t(s)E(V)

3 min

~50 mV

1.5E-03

2.0E-03

2.5E-03

3.0E-03

0.00E+00 5.00E+03 1.00E+04 1.50E+04 t [s]

I(A)for 1 cm2 T6 state E= -620mV

T76 state E= -680mV

1.8E-03

2.3E-03

2.8E-03

1.220E+04 1.225E+04 1.230E+04 1.235E+04 1.240E+04 t [s]

I (A) for 1 cm2

T6 stateT76 state

1,5 min

Figure 6: Potential fluctuations during the galvanostatic test for the 7XXX alloy having two metallurgical states T6 and T76. Big potential transients are observed for the T6 state (sensitive to EFC), given a electrochemical signature of EFC.

Figure 7: Current density fluctuations during the potentiostatic test for the 7XXX T6 and T76 alloy. For the T6 state, local intensity transient which is associated in potentiostatic mode to EFC growth kinetics (blister fractures..).

ECS Transactions, 3 (31) 285-293 (2007)

293

Figure 8: SEM micrograph of exfoliation blisters observed for 7XXX T6 alloy after 50min of galvanostatic test.

ECS Transactions, 3 (31) 285-293 (2007)