Você está na página 1de 9

Corrosion Science 52 (2010) 2489–2497

Contents lists available at ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

The behaviour of second phase particles during anodizing of aluminium alloys


M. Saenz de Miera, M. Curioni *, P. Skeldon, G.E. Thompson
Corrosion and Protection Centre, School of Materials, The University of Manchester, P.O. Box 88, Manchester, M60 1QD, UK

a r t i c l e i n f o a b s t r a c t

Article history: Several model second phase particles and a practical alloy (AA7075-T6) have been anodized in chromic
Received 28 January 2010 and sulphuric acid to disclose the relation between particle composition, electrolyte nature and oxidation
Accepted 31 March 2010 behaviour. Generally, magnesium-containing second phases are readily oxidized, because the presence of
Available online 8 April 2010
sufficient magnesium hinders the formation of a stable oxide, while copper- and iron-containing particles
oxidize at reduced rates and support a relatively stable oxide. At low potential, in chromic acid, the oxi-
Keywords: dation rate of magnesium-containing particles is reduced, due to passivation induced by chromate
A. Aluminium
anions. Conversely, for particles containing only copper and/or iron, chromate anions increase the oxida-
A. Intermetallics
B. SEM
tion rate.
B. TEM Ó 2010 Elsevier Ltd. All rights reserved.
C. Anodic films
C. Oxide coatings

1. Introduction chromate species to provide corrosion inhibition to the aluminium


alloy at anodic and/or cathodic sites [13,14]. When chromate ions
High strength aerospace aluminium alloys are engineered for a are available in the vicinity of the alloy surface, they form a protec-
combination of mechanical properties achieved by tailoring the tive film, which heals defects present after surface treatment
metallurgical microstructure through addition of copper and zinc [15–17]. Additionally, chromate and dichromate oxyanions reduce
as primary alloying elements, and suitable heat treatment. How- the zeta potential of the anodic aluminium oxide film, restricting
ever, such alloys are susceptible to localized corrosion through chloride adsorption and associated weakening of the Al–O bonds
microgalvanic coupling between the different metallurgical phases [18]. Further, unlike sulphuric or phosphoric acid based anodizing
developed. Local electrochemical studies [1–3] on various alloys baths, CAA does not impact significantly on the fatigue life of the
indicate that second phase particles containing only elements no- aerospace components [19].
bler than aluminium are generally cathodic compared with the Localized phenomena that occur during anodizing, at and in
surrounding matrix and constitute preferential sites for corrosion the zone of influence of second phase particles, are important con-
initiation [4]. Second phase particles that also contain magnesium tributions to the performance of the filmed alloy [20] and have
are generally anodic with respect to the matrix, but their behaviour been investigated widely. Guminski et al. [21] generated various
may be affected by the electrolyte pH [5]. In neutral or acidic, chlo- large second phase particles by slow cooling of appropriate binary
ride-containing electrolytes, Al–Cu–Mg-containing particles may alloys and investigated their anodizing behaviour in oxalic acid.
undergo preferential dissolution or de-alloying [6]. Following They found that Mg2Si, CuAl2, b-AlMg and CrAl7 particles oxidized
de-alloying, copper-rich residues, generated on the alloy surface, faster than the surrounding alloy matrix, while FeAl3 particles oxi-
provide active sites for the cathodic reaction [6,7]. dized at a similar rate and TiAl3, NiAl3 and MnAl6 particles oxidize
In order to reduce corrosion susceptibility, the alloy is generally at reduced rate. Similar studies were performed by Cote et al. dur-
anodized in acidic electrolytes and, subsequently, painted with an ing potentiostatic anodizing in sulphuric acid [22], disclosing that
inhibitor-loaded primer and a polymeric topcoat. Concerning the MnAl6 particles were relatively inert and were sometime occluded
anodizing process, chromic acid anodizing (CAA) has proven to in the oxide, while FeAl3 particles oxidized approximately at the
be highly effective in generating porous oxide films with outstand- same rate as the matrix, and CuAl2 and AlZnMg particles oxidized
ing anticorrosion properties [8–10], which are also suited as a base faster than the matrix. Keller et al. found that CrAl7 and b-AlFeSi
layer for both paints and adhesives [11,12]. The outstanding anti- particles oxidized at approximately the same rate as the matrix
corrosion performance following CAA is due to the capability of when anodized galvanostatically in sulphuric acid [23,24]. Frati-
la-Apachitei et al. [25] studied the behaviour Al–Fe, Al–Fe–Si
* Corresponding author. Tel.: +44 1613065971; fax: +44 1613064865. and Al–Cu containing particles produced in cast Al 99.80 wt.%,
E-mail address: michele.curioni@manchester.ac.uk (M. Curioni). Al–10 wt.% Si and Al–10 wt.% Si–3 wt.% Cu alloys during anodizing

0010-938X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.corsci.2010.03.029
2490 M. Saenz de Miera et al. / Corrosion Science 52 (2010) 2489–2497

in relatively concentrated sulphuric acid at 0 °C. They found that If a porous anodic oxide is formed above a second phase
oxygen evolution accompanied the anodizing process [26] and particle, it is generally less regular than the film generated under
that Al–Fe and Al–Fe–Si containing particles were relatively inert similar conditions on relatively pure aluminium due to the for-
and protruded into or were occluded in the oxide; Al–Cu contain- mation of flaws within the oxide [31] associated with oxygen
ing particles oxidized preferentially to produce flawed oxides [26]. evolution within the barrier layer during anodizing [32–34].
Brown et al. [27] investigated the anodizing behaviour of the For this reason, the anodic oxides generated on second phase
Al3Ni second phase in phosphoric acid, observing major changes particles generally provide reduced protection compared with
in the film morphology on the second phase particle, compared oxides formed on the surrounding matrix. From the previous, it
with the surrounding matrix, and oxygen evolution. Dim- is evident that, for successful implementation of protective anod-
ogenontakis et al. related oxygen evolution observed during anod- izing treatments, knowledge of the anodic oxidation behaviour of
izing of an AA2024-T3 alloy in borate solutions with localized second phase particles with respect to the surrounding alloy
phenomena occurring on second phase particles. More recent matrix is critical as well as understanding of the properties of
studies [28–30] on aerospace alloys indicated that second phase the oxides covering the second phases. In the present study,
particles oxidized at a higher or lower rates than the matrix, the surface microstructural modifications induced by anodizing
depending on the anodizing potential, the particle composition model and practical alloys in two different anodizing electrolytes
and the electrolyte nature. Specifically, Al–Cu–Mg-containing par- have been studied. Specifically, the effect of the anodizing
ticles were selectively dissolved by relatively small anodic polari- electrolyte on the electrochemical and filming behaviour of the
zation in sulphuric acid electrolytes [30] , but additions to the second phase particles was investigated by anodizing model sec-
anodizing electrolyte, such as tartaric acid, significantly influenced ond phase particles obtained by magnetron sputtering, while
this process [29]. Conversely, Al–Cu and Al–Cu–Fe-containing variations of the distribution and surface area of second phase
particles, were relatively inert compared with the matrix at low particles present on the surface of a practical alloy (AA7075-
potential, but oxidized preferentially when the alloy was polar- T6) were examined by field emission gun-scanning electron
ized above 5–6 V(SCE) [28,30]. microscopy (FEG-SEM).

Table 1
Nominal composition of AA7075-T6 alloy in weight %.

Alloy Zn Cu Mg Cr Fe Si Mn Ti Other
7075-T6 5.1–6.1 1.2–2.0 2.1–2.9 0.18–0.28 0.50 0.40 0.30 0.20 0.5Zr + Ti

Fig. 1. Scanning electron micrographs of the AA7075-T6 alloy (a, c and e) after etching/desmutting and (b, d and f) after anodizing in 0.4 M sulphuric acid for 90 s at
5 mA cm 2 (e and f are reported from Ref. [13]). Elemental analyses of second phase particles are given in Table 2.
M. Saenz de Miera et al. / Corrosion Science 52 (2010) 2489–2497 2491

2. Experimental NaOH solution for 120 s at 313 K and subsequently desmutted in


4.7 M HNO3 for 15 s at room temperature. Anodizing of the
Aluminium AA7075-T6 alloy specimens, of nominal composi- AA7075-T6 specimens was undertaken at 5 mA cm 2 , in stirred
tion given in Table 1, were ultrasonically degreased for 3 min in 0.4 M sulphuric acid at 295 K (pH 1.51) or in 0.25 M chromic acid
acetone and subsequently for 3 min in ethanol (90 vol.%). After electrolyte at 313 K (pH 1.128) for 90 s in a 2-electrode cell with
degreasing, the specimens were individually etched in 20 g l 1 a large platinum mesh counter electrode. Following anodizing,
the specimens were rinsed and dried. The current density of
5 mA cm 2 ensured the generation of a well-developed porous
Table 2
anodic oxide in both electrolytes and enabled qualitative compar-
Elemental analyses of second phase particles after etching/desmutting (AE) and after
anodizing (AA) in sulphuric acid (particles 1–10) and in chromic acid (particles 11–18). ison of the anodizing behaviour. Field emission gun-scanning elec-
tron microscopy (Phillips XL-30 FEG-SEM) observation, with
IMC # Cu (at.%) Fe (at.%) Mg (at.%) Si (at.%) O (at.%)
energy dispersive X-ray analysis (EDX), was carried out on selected
AE AA AE AA AE AA AE AA AE AA areas of the alloy specimens prior to and after anodizing to evalu-
1 3.6 0.8 14.0 4.6 1.3 0.7 2.8 1.6 – 57.5 ate the changes in morphology and composition of the second
2 25.1 – – – 25.4 – – – – – phase material located at the alloy surface.
3 4.3 1.2 16.3 5.4 2.2 0.9 – – – 58.4
In order to simulate the behaviour of selected second phase
4 4.4 1.7 16.3 7.1 2.5 1.2 – – – 50.0
5 3.5 1.8 11.6 5.0 1.7 1.8 – – – 35.5
particles, macroscopic model specimens of specific compositions
6 3.2 1.7 14.2 6.2 2.4 1.2 – – – 49.7 were prepared by magnetron sputtering, as described elsewhere
7 30.2 – 16.4 – – – – – – – [28]; such specimens are termed model second phase particles.
8 30.0 – – – 16.1 – – – – – The selected compositions of the model second phase particles,
9 2.2 0.8 11.1 4.7 – – 1.8 0.9 – 57.5
determined by EDX and RBS analyses, are Al–30 at.% Cu, Al–17 at.%
10 – – – – 1.5 – 32.3 29.1 42.3 53.4
11 13.2 – 7.0 – – – – – – – Cu–7 at.% Fe, Al–25 at.% Cu–25 at.% Mg, Al–28 at.% Fe, Al–25 at.%
12 3.5 – 14.8 – 0.9 – – – 18.3 – Mg and Al–7 at.% Fe. A sputtered Al–5 at.% Cu alloy, representative
13* 1.1 1.8 14.3 13.2 0.9 – 4.1 3.9 – 12.7 of a copper-containing aluminium matrix, was also prepared. The
14 1.9 1.4 14.6 11.0 – – 4.2 3.0 – 25.8
first three model second phase particles are designated Al2Cu, Al7-
15 23.4 – – – 18.9 – – – 7.3 –
16 2.8 0.9 14.8 3.2 1.3 – – – – 63.9
Cu2Fe and Al2CuMg, respectively. Anodizing of the model second
17 – – – – 1.4 1.3 24.6 25.1 44.7 47.8 phases was undertaken under the conditions previously detailed
18 2.9 1.5 18.7 9.6 – – – – – 33.7 as well as potentiodynamically at 297 K, employing a Solartron
1280 potentiostat and a 3-electrode cell, with a potential scan from

Fig. 2. Scanning electron micrographs of the AA7075-T6 alloy after etching/desmutting in 20 g/l NaOH for 2 min followed by 15 s in 4.7 M HNO3. Elemental analyses of
second phase particles are given in Table 2.
2492 M. Saenz de Miera et al. / Corrosion Science 52 (2010) 2489–2497

0.4 V below the open circuit potential (OCP) to 12 V vs saturated 3. Results


calomel electrode (SCE). When commercial or model aluminium al-
loys are anodized under potentiodynamic conditions, potential 3.1. Modification of the surface microstructure after anodizing
dependent and time dependent processes may overlap [35], result-
ing in current responses dependent on the sweep rate [30,35]. In or- Characterization of second phase particles present in the
der to identify the characteristic potentials associated with AA7075-T6 alloy after etching and desmutting was undertaken
generation, modification, or disruption of oxide layers, low sweep by FEG-SEM, with the typical surface shown in Fig. 1(a, c and e).
rates are generally preferred because processes displaying some From energy dispersive X-ray analysis (EDX), on average (Table
time dependence can proceed in a near-to-equilibrium condition, 2), 52% of the revealed particles contained copper and iron, 42%
and the direct relationship between current and potential can be re- contained copper and magnesium and 7% displayed a composition
vealed. However, in the present work, the thickness of the model al- close to silicon oxide (SiO2). The modification of the surface micro-
loy layer was of a few hundred nanometers, due to the limited structure after anodizing in sulphuric acid is evident in Fig. 1(b, d
thicknesses of the alloy films produced by magnetron sputtering. and f), with the following general observations: all Al–Cu–Mg-con-
Therefore, based on previous works [28,30,35] and preliminary taining particles, such as particles 2, 7 and 8, were removed by
tests, a sweep rate of 2 V min 1 was selected for all the experiments anodizing in sulphuric acid (independent of their size); coarse
since it was sufficiently low to obtain reliable responses and suffi- Al–Cu–Fe-containing particles (e.g. particles 1 and 4–6) remained
ciently high to guarantee an appropriate potential scan prior to con- on the surface, although with reduced dimensions. After sulphuric
sumption of the sputtered layer due to the anodic current. Cross- acid anodizing, the surface area of the second phases particles was
sections of selected galvanostatically anodized specimens, pre- reduced to approximately 55% of their original area and the num-
pared by ultramicrotome to a nominal thickness of 15 nm, were ber of particles was reduced by about 53%. Before chromic acid
examined in a JEOL 2000 FX II transmission electron microscope. anodizing (Fig. 2), the second phase particles revealed on the spec-
Compositions of model second phase particles and anodic oxides imen surface again included Al–Cu–Fe-containing particles (64%),
were determined by RBS using 1.7 MeV He4+ ions supplied by the Al–Cu–Mg-containing particles (34%) and silicon oxide particles
Van de Graaff accelerator of the University of Paris. The ion beam (2%). The reduction in second phase surface area after CAA was
was incident normal to the specimen surface, with scattered ions about 80%, with approximately 82% of particles being removed
detected at 165° to the direction of the incident beam. The RBS data (Fig. 3). The behaviour in the two electrolytes is now investigated
were interpreted using the RUMP program [36]. further by study of the model second phase particles.

2
Fig. 3. Scanning electron micrographs of the AA7075-T6 alloy after anodizing in 0.25 M chromic acid at 5 mA cm for 90 s at 313 K. Elemental analyses of second phase
particles are given in Table 2.
M. Saenz de Miera et al. / Corrosion Science 52 (2010) 2489–2497 2493

Fig. 4. Potentiodynamic polarization curves (scan rate: 2 V min 1) for model Al2Cu,
Al7Cu2Fe, Al–28 at.% Fe and Al–7 at.% Fe second phase particles during anodic
oxidation in 0.25 M chromic acid (CAA) at 313 K and in 0.4 M sulphuric acid (SAA)
at 297 K to 12 V(SCE).

Fig. 5. Potentiodynamic polarization curves (scan rate: 2 V min 1) for model Al–
5 at.% Cu, Al2CuMg and Al–25 at.% Mg second phase particles during anodic
oxidation in 0.25 M chromic acid (CAA) at 313 K and in 0.4 M sulphuric acid (SAA)
at 297 K to 12 V(SCE).
Fig. 6. Transmission electron micrographs of anodized model (a) Al2Cu, (b)
Al7Cu2Fe and (c) Al–28Fe second phase particles in 0.4 M sulphuric acid (SAA) at
5 mA cm 2 for 90 s at 295 K.
3.2. Potentiodynamic polarization behaviour

Potentiodynamic polarization curves of model second phase cle. Conversely, the anodic oxidation behaviour of the magnesium-
particles, prepared by magnetron sputtering to give compositions containing model second phases was affected significantly by the
close to those identified in Table 2, namely Al2Cu, Al–28 at.% Fe, anodizing electrolyte, as evident from Fig. 5. Specifically, for the
Al7Cu2Fe, Al–25 at.% Mg, Al2CuMg and Al–5 at.% Cu, recorded in sulphuric acid electrolyte, the anodic current increased approxi-
0.25 M chromic acid at 313 K and 0.4 M sulphuric acid at 297 K, mately linearly with potential from the OCP until the sputtered
are presented in Figs. 4 and 5. In both figures, the potentiodynamic layer was fully consumed. Conversely, for chromic acid anodizing,
behaviour of a magnetron sputtered Al–5 at.% Cu alloy is reported a relatively low anodic current was measured for an extended po-
as representative of an alloy matrix containing copper as an alloy- tential region above the OCP and, subsequently, the current in-
ing element in solid solution. It is revealed immediately that all the creased rapidly to values higher than those revealed for the
model second phase particles display significantly higher anodic alloys anodized in sulphuric acid electrolyte.
currents compared with the alloy matrix. As revealed in Fig. 4,
the oxidation behaviour of the model second phases containing 3.3. Transmission electron microscopy
copper and/or iron was similar in chromic and sulphuric acids.
Generally, the current densities were slightly higher for anodizing Transmission electron micrographs of ultramicrotomed sec-
in chromic acid, with the difference between the values decreasing tions of the Al2Cu, Al7Cu2Fe and Al–28 at.% Fe model second phases
as the relative content of iron increased in the second phase parti- after galvanostatic anodizing at 5 mA cm 2 in sulphuric acid for
2494 M. Saenz de Miera et al. / Corrosion Science 52 (2010) 2489–2497

2
Fig. 7. Transmission electron micrographs of anodized model (a) Al2Cu, (b) Al7Cu2Fe and (c) Al–28Fe second phase particles in 0.25 M chromic acid (CAA) at 5 mA cm for
90 s at 313 K.

90 s displayed thin anodic films of relatively uniform thickness,


with an internal sponge-like appearance; fine pores passing per-
pendicularly to the alloy surfaces were not readily resolved
(Fig. 6). The anodic films appeared to be more compact as the rel-
ative content of iron in the model second phase particles increased.
Further, disruption of the anodic film and detachment of the film
from the Al2Cu substrate were frequently observed during TEM
observations. It is thus probable that oxygen gas generation and
evolution proceeding during anodizing weakened the oxide
through generation of lateral porosity and subsequent film rupture
and loss of oxide fragments to the anodizing electrolyte. The anodic
films developed on the selected model second phase particles after
anodizing in 0.25 M chromic acid at 313 K were similar to those
generated in sulphuric acid (Fig. 7). However, the anodic oxides
were generally thinner than those formed during sulphuric acid
for the same charge passed. The anodic film thicknesses developed
on the model Al2Cu, Al7Cu2Fe and Al–28 at.% Fe particles were
respectively 33, 29 and 60 nm for anodizing in chromic acid
and 67, 68 and 75 nm for anodizing in sulphuric acid.

3.4. Rutherford backscattering spectroscopy


Fig. 8. Experimental (dots) and simulated (solid line) RBS spectra of a model Al2Cu
second phase particles before (red spectrum) and after (blue spectrum) anodizing in
Rutherford backscattering spectroscopy (RBS) was undertaken 0.4 M sulphuric acid at 295 K. (For interpretation of the references to color in this
on the as-deposited model Al2Cu and Al7Cu2Fe second phases figure legend, the reader is referred to the web version of this paper.)
M. Saenz de Miera et al. / Corrosion Science 52 (2010) 2489–2497 2495

Table 3
2
TEM and RBS analyses of model Al2Cu and Al7Cu2Fe model second phase particles after anodizing at 5 mA cm in 0.4 M sulphuric acid (SA) at 295 K and in 0.25 M chromic acid
(CA) at 313 K.

Model alloy Average Charge used Charge Charge due to Density of Film stoichiometry Al2O3  xAl2(SO4)3  yCuO  zFe2O3
TEM for film due to oxygen anodic films
thickness formation losses (%)a generation (%)a (g cm3)
(nm) (%)a
SA CA SA CA SA CA SA CA SA CA SA CA
Al2Cu 67 33 12 4 12 13 76 83 1.5 1.0 Al2O3  0.1Al2(SO4)3  0.01CuO Al2O3
Al7Cu2Fe 68 29 15 9 3 8 82 83 2.1 2.2 Al2O3  0.08Al2(SO4)3  0.01CuO  0.07Fe2O3 Al2O3  0.24CuO  0.08Fe2O3

All values are within an accuracy of 5%.


a 2
Percentage with respect to the total charge at 90 s of anodizing, QT = 0.45 C cm .

and after anodizing in chromic acid and sulphuric acid electrolytes, Conversely, for the second phase containing metals capable of
with typical spectra for as-deposited and anodized Al2Cu alloys forming an adherent oxide with reduced solubility in the anodizing
shown in Fig. 8. Relevant information calculated from fitting of electrolyte, the rate of consumption of the second phase depends
the RBS spectra are presented in Table 3. Thus, from RBS, the com- on the properties of the anodic oxide. The simultaneous oxidation
position and thickness of the model second phase particles, ex- of aluminium and copper or iron, results in an anodic oxide con-
pressed as atoms cm 2 of aluminium, copper and/or iron, are taining Al–O units together with Cu–O and Fe–O units that display
obtained before and after anodizing. Subtraction of the previous modified electrical properties. Specifically, an anodic oxide with
values gives the number of atoms cm 2 that are oxidized from me- high content of non-aluminium atoms can be locally generated
tal to cations. RBS also provides the number of cations that consti- where the oxidation of the copper or irons, enriched at the alloy-
tute the anodic film material; subtracting the number of oxidized oxide interface, take place. Due to the semi-conductive properties
atoms cm 2 from metal to cations and those that comprise the of Fe–O and Cu–O units, preferential paths for the direct discharge
anodic film gives the losses to the anodizing electrolyte. Such cal- of the electrons of the oxygen anions are available. Therefore, a
culations quantify the charge densities consumed by the various proportion of the inward-migrating oxygen anions lose their elec-
electrode reactions, e.g. anodic film formation, losses to the elec- trons and form atomic oxygen within the barrier layer of the por-
trolyte and oxygen generation (Table 3). ous anodic film. Subsequently, the atomic oxygen combines to
As revealed in Table 3, the charge used in forming the anodic form oxygen gas that generates a gas-filled void, with the develop-
films on the Al2Cu and Al7Cu2Fe model second phase particles by ing oxygen gas pressure eventually increasing to levels sufficient to
chromic acid anodizing (CAA) was less (4% and 9%, respectively) rupture the oxide. The rupture of the oxide provides a path for
than for anodizing in sulphuric acid (12% and 15%, respectively). electrolyte penetration in close proximity of the underlying alloy.
Additionally, there was a higher presence of Cu–O units in the At this location, the high electric field present across the residual
anodic films formed in CAA compared with the films formed in sul- oxide promotes accelerated local growth that progresses until a
phuric acid. Further, typical concentrations of SO24 ions incorpo- new oxygen-filled void is generated and the process is cyclically re-
rated into the anodic films developed on all model second phases peated. Overall, this process results in accelerated second phase
were in the range 16–19 wt.%. The previous values are greater than particle consumption compared with conditions where a relatively
the levels reported in films formed on high purity aluminium, i.e. pure aluminium oxide is generated. According to this view, it is
12–14 wt.% sulphate [37], probably due to residual electrolyte possible to explain the relatively high anodic current observed
within the relatively disordered film morphology. for the oxidation of the model Al2Cu second phase. In this case,
the high amount of copper produces a film which supports vigor-
ous oxygen gas generation and evolution within the oxide, as con-
4. Discussion firmed by the calculations form the RBS data. Oxygen evolution
ruptures the oxide, contributing to the limited film thickness and
During anodizing of a practical alloy, a protective porous oxide to the sponge-like morphology observed by transmission electron
film is generated above the alloy matrix and the second phase par- microscopy. As the relative content of copper is reduced and the
ticles variously present on the alloy surface respond individually to relative content of iron is increased, the process of electron dis-
the electrochemical conditions. This may result in the formation of charge is less favoured and, therefore, less oxygen evolution takes
a chemically and morphologically modified oxide film above and in place. This results in a reduced anodic current and in a transition to
the zone of influence of the specific second phase. The oxidation a more defined porous oxide morphology, where a barrier layer can
behaviour depends on the composition of the second phase parti- be distinguished from the material constituting the cell walls.
cle and on the electrolyte. Specifically, second phase particles with The conclusions drawn for the model second phases are con-
high contents of metals with cations that migrate faster than alu- firmed by the anodizing experiments on the practical alloy. Thus,
minium cations across the oxide under the field (for example, Mg the magnesium-containing second phase particle are the most
or Li) and are highly soluble in the acidic anodizing electrolyte tend reactive to the anodizing condition and were fully oxidized in both
to form an oxide with very poor adhesion to the metal and voids at electrolytes, independent of their initial size. Conversely, the cop-
the metal/film interface [38]. In the case of the model second per/iron-containing second phases, supporting an oxide that was
phases, high amounts of magnesium results in a highly flawed chemically stable in the anodizing electrolyte, yet cyclically dis-
mixed oxide [38]. Further such an oxide may also be dissolved in rupted by oxygen generation and evolution during anodizing, were
the anodizing electrolyte due to the high solubility of magnesium partially or completely oxidized depending of their initial size.
oxide in acidic environment if the amount of magnesium cations From the previous, it is possible to understand the effect of the
is sufficient to prevent the formation of a continuous alumina net- electrolyte on the oxidation behaviour of the various second phase
work. Therefore, locally, the metal can be exposed directly to the particles. Specifically, variation in the behaviour due to electrolyte
anodizing electrolyte and the high potential available results in a composition is more marked for second phase particles that do not
high rate of consumption. support a stable oxide film. In this case, due to incomplete alumina
2496 M. Saenz de Miera et al. / Corrosion Science 52 (2010) 2489–2497

film formation, the electrolyte is in direct contact with the metal Acknowledgement
and determines the dissolution behaviour. From the potentiody-
namic experiments, it is evident that active dissolution of the EPSRC is acknowledged for provision of financial support
magnesium-containing intermetallics proceeds at all potentials through the LATEST Portolio Partnership.
during polarization in sulphuric acid. Conversely, in chromic acid
electrolyte, reduction of chromate may occur on the surface during
the early stages of immersion and, consequently, a protective outer References
layer is formed, below which a passive film develops readily during
the potentiodynamic polarization, since it does not contact the [1] F. Andreatta, M.M. Lohrengel, H. Terryn, J.H.W. De Wit, Electrochemical
characterisation of aluminium AA7075-T6 and solution heat treated AA7075
electrolyte. However, at relatively high potential, the chromate-
using a micro-capillary cell, Electrochim. Acta 48 (2003) 3239–3247.
containing passive layer on the second phase surface is disrupted [2] P. Campestrini, E.P.M. van Westing, H.W. van Rooijen, J.H.W. de Wit, Relation
under the high electric field and cannot prevent the active dissolu- between microstructural aspects of AA2024 and its corrosion behaviour
investigated using AFM scanning potential technique, Corros. Sci. 42 (2000)
tion, as evident from the potentiodynamic polarization behaviour.
1853–1861.
For second phase particles supporting a stable oxide, the effect of [3] R. Ambat, A.J. Davenport, G.M. Scamans, A. Afseth, Effect of iron-containing
the anodizing electrolyte is less marked. This is due to the more intermetallic particles on the corrosion behaviour of aluminium, Corros. Sci. 48
complex oxidation regime, which involves formation of oxide (2006) 3455–3471.
[4] V. Guillaumin, G. Mankowski, Localized corrosion of 6056 T6 aluminium alloy
and oxygen generation within it, leading to local rupture. There- in chloride media, Corros. Sci. 42 (2000) 105–125.
fore, there is no direct contact between the oxidising surface and [5] Y. Liu, M.A. Arenas, P. Skeldon, G.E. Thompson, P. Bailey, T.C.Q. Noakes, H.
the anodizing electrolyte and, consequently, the interaction is lim- Habazaki, K. Shimizu, Anodic behaviour of a model second phase: Al–20 at.%
Mg–20 at.% Cu, Corros. Sci. 48 (2006) 1225–1248.
ited. However, from observation of the practical alloy, where the [6] V. Guillaumin, G. Mankowski, Localized corrosion of 2024 T351 aluminium
reduction in intermetallic size and number is more marked after alloy in chloride media, Corros. Sci. 41 (1999) 421–438.
CAA than SAA, and consideration of the potentiodynamic polariza- [7] K.A. Yasakau, M.L. Zheludkevich, S.V. Lamaka, M.G.S. Ferreira, Role of
intermetallic phases in localized corrosion of AA5083, Electrochim. Acta 52
tion behaviour, where the currents recorded for polarization in the (2007) 7651–7659.
chromic acid electrolyte are generally higher than those recorded [8] F. Mansfeld, M.W. Kendig, Evaluation of anodized aluminum surfaces with
in the sulphuric acid electrolyte, it can be concluded that for electrochemical impedance spectroscopy, J. Electrochem. Soc. 135 (1988) 828–
833.
second phases supporting a relatively stable oxide film chromic [9] G.E. Thompson, L. Zhang, C.J.E. Smith, P. Skeldon, Boric/sulfuric acid anodizing
acid increases the oxidation rate and, therefore, acts as a mild oxi- of aluminum alloys 2024 and 7075: film growth and corrosion resistance,
dation promoter. Corrosion 55 (1999) 1052–1061.
[10] D. Chidambaram, C.R. Clayton, G.P. Halada, A duplex mechanism-based model
From the practical viewpoint, the nature of the electrolyte is not
for the interaction between chromate ions and the hydrated oxide film on
critical in determining the oxidation behaviour of magnesium-con- aluminum alloys, J. Electrochem. Soc. 150 (2003) B224–B237.
taining second phases on conventional alloys because the applied [11] P.G. Sheasby, R. Pinner, The Surface Treatment and Finishing of Aluminium and
voltage in industrial processes is generally sufficiently high to trig- its Alloys, ASM International, Materials Park, OH, 2001.
[12] N. Fin, H. Dodiuk, A.E. Yaniv, L. Drori, Oxide treatments of Al 2024 for adhesive
ger their preferential dissolution during the very early stages of bonding- surface characterization, Appl. Surf. Sci. 28 (1987) 11–33.
anodizing. Conversely, the capability of chromate to promote the [13] W.J. Clark, J.D. Ramsey, R.L. McCreery, G.S. Frankel, A galvanic corrosion
oxidation of copper and/or iron-containing second phases is bene- approach to investigating chromate effects on aluminum alloy 2024-T3, J.
Electrochem. Soc. 149 (2002) B179–B185.
ficial to the anticorrosion performance because it reduces the num- [14] E. Akiyama, G.S. Frankel, Influence of dichromate ions on aluminum
ber and extent of potentially active cathodic sites present on the dissolution kinetics in artificial crevice electrode cells, J. Electrochem. Soc.
alloy surface after anodizing. For this reason, the capability of pro- 146 (1999) 4095–4100.
[15] M. Kendig, S. Jeanjaquet, R. Addison, J. Waldrop, Role of hexavalent chromium
moting second phases oxidation should be maintained by candi- in the inhibition of corrosion of aluminum alloys, Surf. Coat. Technol. 140
date environmentally-friendly electrolytes. (2001) 58–66.
[16] J. Zhao, G. Frankel, R.L. McCreery, Corrosion protection of untreated AA-2024-
T3 in chloride solution by a chromate conversion coating monitored with
Raman spectroscopy, J. Electrochem. Soc. 145 (1998) 2258–2264.
5. Conclusions [17] J. Zhao, L. Xia, A. Sehgal, D. Lu, R.L. McCreery, G.S. Frankel, Effects of chromate
and chromate conversion coatings on corrosion of aluminum alloy 2024-T3,
Surf. Coat. Technol. 140 (2001) 51–57.
The behaviour of model and real second phase particles present
[18] M. Kendig, S. Jeanjaquet, R. Addison, J. Waldrop, Role of hexavalent chromium
in a pretreated alloy and the relationship between second phase in the inhibition of corrosion of aluminium alloys, Surf. Coat. Technol. 140
composition, electrolyte nature and oxidation behaviour have been (2001) 58–66.
investigated. Concerning the composition, magnesium-containing [19] L. Domingues, J.C.S. Fernandes, M. Da Cunha Belo, M.G.S. Ferreira, L. Guerra-
Rosa, Anodising of Al 2024-T3 in a modified sulphuric acid/boric acid bath for
second phases, are more readily oxidized compared with second aeronautical applications, Corros. Sci. 45 (2002) 149–160.
phases containing only aluminium and copper and/or iron. This [20] R. Akeret, H. Bichsel, E. Schwall, E. Simon, M. Textor, Influence of chemical
was attributed to the absence of a stable oxide above the magne- composition and fabrication procedures on the properties of anodised
aluminium surfaces, Trans. Inst. Met. Finish. 68 (1990) 20–28.
sium-containing second phases which results in direct metal disso- [21] R.D. Guminski, P.G. Sheasby, H.J. Lamb, Reaction rates of second-phase
lution. Further, the oxidation behaviour of second phases constituents in aluminium during etching, brightening and oxalic acid
supporting stable oxides was related to the oxide properties. Spe- anodizing processes, Trans. Inst. Met. Finish. 46 (1968) 44–48.
[22] J. Cote, E.E. Howlett, H.J. Lamb, Behavior of intermetallic compounds in
cifically, the rate of second phase particle consumption increased aluminium during sulfuric acid anodizing, Plating 57 (1970) 484–496.
with copper content due to increase of oxygen evolution and asso- [23] F. Keller, G.W. Wilcox, M. Tosterud, C.J. Slunder, Anodic coating of
ciated cyclic disruption of the anodic oxide. Concerning the effect aluminum – behaviour of alloy constituents, Metals Alloys 10 (1939) 219–
225.
of the electrolyte, the presence of a passivating ion, such as chro-
[24] F. Keller, G.W. Wilcox, Anodically oxidized aluminium alloys – metallographic
mate, significantly affects the oxidation behaviour of the magne- examination, Metals Alloys 10 (1939) 187–195.
sium-containing second phase particles, by producing an [25] L.E. Fratila-Apachitei, H. Terryn, P. Skeldon, G.E. Thompson, J. Duszczyk, L.
Katgerman, Influence of substrate microstructure on the growth of anodic
extended passive region prior to the onset of active dissolution.
oxide layers, Electrochim. Acta 49 (2004) 1127–1140.
Conversely, for the second phases containing only copper and iron, [26] L.E. Fratila-Apachitei, J. Duszczyk, L. Katgerman, Voltage transients and
the effect of the electrolyte composition was less marked. This is morphology of AlSi(Cu) anodic oxide layers formed in H2SO4 at low
attributed to the presence of a stable oxide film between the sec- temperature, Surf. Coat. Technol. 157 (2002) 80–94.
[27] G.M. Brown, K. Shimizu, K. Kobayashi, P. Skeldon, G.E. Thompson, G.C. Wood,
ond phase and the electrolyte which prevents direct contact be- The growth of a porous oxide film of a unique morphology by anodic oxidation
tween the chromate species and the oxidising metal. of an Al–0.5 wt.% Ni alloy, Corros. Sci. 40 (1998) 1575–1586.
M. Saenz de Miera et al. / Corrosion Science 52 (2010) 2489–2497 2497

[28] M. Saenz de Miera, M. Curioni, P. Skeldon, G.E. Thompson, Modelling the [33] X. Zhou, G.E. Thompson, H. Habazaki, K. Shimizu, P. Skeldon, G.C. Wood,
anodizing behaviour of aluminium alloys in sulphuric acid through alloy Copper enrichment in Al–Cu alloys due to electropolishing and anodic
analogues, Corros. Sci. 50 (2008) 3410–3415. oxidation, Thin Solid Films 293 (1997) 327–332.
[29] M. Curioni, P. Skeldon, E. Koroleva, G.E. Thompson, Role of tartaric acid on the [34] X. Zhou, G.E. Thompson, P. Skeldon, G.C. Wood, K. Shimizu, H. Habazaki,
anodizing and corrosion behavior of AA 2024 T3 aluminum alloy, J. Anodic oxidation of an Al–2 wt.% Cu alloy: effect of grain orientation, Corros.
Electrochem. Soc. 156 (2009) C147–C153. Sci. 41 (1999) 1089–1094.
[30] M. Curioni, M. Saenz De Miera, P. Skeldon, G.E. Thompson, J. Ferguson, Macro- [35] M. Curioni, P. Skeldon, G.E. Thompson, Anodizing of aluminum under
scopic and local filming behavior of AA2024 T3 aluminum alloy during anodizing nonsteady conditions, J. Electrochem. Soc. 156 (2009) C407–C413.
in sulfuric acid electrolyte, J. Electrochem. Soc. 155 (2008) C387–C395. [36] RUMP Program. Available from: <http://www.genplot.com/>.
[31] S.J. Garcia-Vergara, K. El Khazmi, P. Skeldon, G.E. Thompson, Influence of [37] G.E. Thompson, Porous anodic alumina: fabrication, characterization and
copper on the morphology of porous anodic alumina, Corros. Sci. 48 (2006) applications, Thin Solid Films 297 (1997) 192–201.
2937–2946. [38] X. Zhou, G.E. Thompson, P. Skeldon, G.C. Wood, K. Shimizu, H. Habazaki, Film
[32] L. Iglesias-Rubianes, S.J. Garcia-Vergara, P. Skeldon, G.E. Thompson, J. formation and detachment during anodizing of Al–Mg alloys, Corros. Sci. 41
Ferguson, M. Beneke, Cyclic oxidation processes during anodizing of Al–Cu (1999) 1599–1613.
alloys, Electrochim. Acta 52 (2007) 7148–7157.

Você também pode gostar