Electrochemical behaviour of ZrO2 sol-gel pre-treatments on AA6060 aluminium alloy

Article abstract of DOI:10.1016/j.electacta.2006.12.065

ZrO₂ pre-treatments applied with the sol-gel technique are a possible replacement of chromium based pre-treatments on aluminium alloys. The thickness and homogeneity of the films deposited on AA6060 alloy are strongly related to the process parameters like preparation of the surface, number of dips and thermal treatment of the film. ZrO₂ films were prepared using the dip-coating technique in sol obtained from metal-organic precursors in an organic solvent (0.1 M Zr(OBuⁿ)₄ in anhydrous n-butanol with addition of acetic acid as complexing agent). Different layers were applied on AA6060 changing number of dips and thermal treatment (150 °C for 1 h or 250 °C for 4 min). The typical thickness of the deposited layers was in the range 70-180 nm depending on process parameters. The electrochemical behaviour of the pre-treated alloy in diluted Harrison solution (0.05 wt% NaCl + 0.35 wt% (NH₄)₂SO₄) was investigated by means of potentiodynamic polarization, open circuit potential measurements, and electrochemical impedance spectroscopy. In addition, the electrochemical behaviour of ZrO₂ sol-gel films was compared with that of chromatized AA6060 and fluotitanated/fluozirconated AA6060. In order to evaluate the adhesion properties of the films, ZrO₂ pre-treated AA6060, chromatized AA6060 and fluotitanated/fluozirconated AA6060 were painted with a polyester resin and subjected to thermal cycles in 0.05 wt% NaCl. Each thermal cycle consisted of heating the samples at 90 °C, permanence at 90 °C for 6 h, cooling at room temperature and permanence at room temperature for 18 h. Impedance measurements were performed at the end of each cycle. Potentiodynamic polarization curves and impedance spectra indicate that ZrO₂ pre-treatments have similar barrier properties to those of chromatized AA6060. However, no self-healing ability is observed for ZrO₂ films. The barrier properties of ZrO₂ films are strongly dependent on process parameters. In particular, the number of dips determines the amount of defects in the film and its homogeneity. The electrochemical behaviour strongly improves increasing the number of dips in the deposition bath. Thermal aging cycles evidence good adhesion properties for ZrO₂ pre-treatments.

Full text of DOI:10.1016/j.electacta.2006.12.065

Electrochimica Acta 52 (2007) 7545–7555
Electrochemical behaviour of ZrO2 sol–gel pre-treatments
on AA6060 aluminium alloy
F. Andreattaa,∗, P. Aldighieria, L. Paussaa, R. Di Maggiob, S. Rossib, L. Fedrizzia
a Department of Chemical Science and Technology, University of Udine, Via del Cotonificio 108, 33100 Udine, Italy
b DIMTI, University of Trento, Italy
Received 17 July 2006; received in revised form 29 November 2006; accepted 17 December 2006
Available online 23 January 2007
Abstract
ZrO2 pre-treatments applied with the sol–gel technique are a possible replacement of chromium based pre-treatments on aluminium alloys. The
thickness and homogeneity of the films deposited on AA6060 alloy are strongly related to the process parameters like preparation of the surface,
number of dips and thermal treatment of the film.
ZrO2 films were prepared using the dip-coating technique in sol obtained from metal-organic precursors in an organic solvent (0.1 M Zr(OBun)4
in anhydrous n-butanol with addition of acetic acid as complexing agent). Different layers were applied on AA6060 changing number of dips and
thermal treatment (150 ◦C for 1 h or 250 ◦C for 4 min). The typical thickness of the deposited layers was in the range 70–180 nm depending on
process parameters. The electrochemical behaviour of the pre-treated alloy in diluted Harrison solution (0.05 wt% NaCl + 0.35 wt% (NH4)2SO4)
was investigated by means of potentiodynamic polarization, open circuit potential measurements, and electrochemical impedance spectroscopy.
In addition, the electrochemical behaviour of ZrO2 sol–gel films was compared with that of chromatized AA6060 and fluotitanated/fluozirconated
AA6060. In order to evaluate the adhesion properties of the films, ZrO2 pre-treated AA6060, chromatized AA6060 and fluotitanated/fluozirconated
AA6060 were painted with a polyester resin and subjected to thermal cycles in 0.05 wt% NaCl. Each thermal cycle consisted of heating the samples
at 90 ◦C, permanence at 90 ◦C for 6 h, cooling at room temperature and permanence at room temperature for 18 h. Impedance measurements were
performed at the end of each cycle.
Potentiodynamic polarization curves and impedance spectra indicate that ZrO2 pre-treatments have similar barrier properties to those of
chromatized AA6060. However, no self-healing ability is observed for ZrO2 films.
The barrier properties of ZrO2 films are strongly dependent on process parameters. In particular, the number of dips determines the amount of
defects in the film and its homogeneity. The electrochemical behaviour strongly improves increasing the number of dips in the deposition bath.
Thermal aging cycles evidence good adhesion properties for ZrO2 pre-treatments.
© 2007 Elsevier Ltd. All rights reserved.
Keywords: ZrO2 pre-treatments; Sol–gel pre-treatments; Chromium-free pre-treatments; Corrosion behaviour; Paint adhesion; Aluminium
1. Introduction
obtained with sol–gel technique might replace chromate con-
version coatings proven that these sol–gel coatings can provide
good adhesion and barrier properties. The sol–gel technique is
widelyusedfortheproductionofceramicmaterialsstartingfrom
metal-organic precursors or soluble metal salts. This technique
is employed for deposition of mixed metal oxides and metal
oxide-organic composites [4–6].
Oxide layers deposited by dip-coating and spin-coating from
sol–gel systems improve the corrosion resistance of aluminium
alloys [1,7]. Beccaria et al. [8,9] showed that silane treatment
improves the resistance of aluminium to general and local-
ized corrosion. Yang et al. [10] reported that mixed SiO2 and
ZrO2 oxide layers deposited on AA2024 with sol–gel technique
Adhesion of an organic coating to a metal substrate is very
important for the corrosion protection of painted aluminium
alloys like AA6060. Chromate conversion coatings provide
very good adhesion properties associated with good barrier
properties [1]. Moreover, chromate conversion coating exhibit
self-healing ability [1]. However, this type of conversion coating
is carcinogenic and not environmental friendly [1–4]. Coatings
∗
Corresponding author. Tel.: +39 0432 558839; fax: +39 0432 558803.
E-mail address: francesco.andreatta@uniud.it (F. Andreatta).
0013-4686/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2006.12.065

7546
F. Andreatta et al. / Electrochimica Acta 52 (2007) 7545–7555
Table 1
improve the resistance to localized attack in diluted Harrison
solution.
Composition of AA6060 aluminium alloy
Besides, the sol–gel technique presents the possibility of
introducing inhibitors in the oxide layer. Zheludkevich et al.
[11] recently showed that the addition of ZrO2 nanoparticles
in sol–gel coatings could improve the corrosion resistance of
AA2024.
Al
Mg
Si
Cr
Mn
Ti
Cu
Zn
Fe
Max 97.8
0.35–0.6
0.3–0.6
Max 0.05
Max 0.1
Max 0.1
Max 0.1
Max 0.15
0.1–0.3
Fedrizzi et al. [12] deposited zirconia amorphous films on
AA1050 with dip-coating technique. They showed that control
of hydrolysis leads to the formation of pre-treatments, which
promote adhesion of organic coatings [12]. Besides, the electro-
chemical behaviour of AA1050 coated with amorphous zirconia
films is strongly dependent on the deposition parameters.
Different techniques might be employed for the characteriza-
tion of the electrochemical behaviour of pre-treated aluminium
alloys. Potentiodynamic polarization has been employed for
the characterization of conversion layers [13,14] and hybrid
sol–gelcoatings[15,16]. Cathodicpotentiodynamicpolarization
enables assessment of possible inhibition of cathodic reactions
due to the deposition of a layer on the alloy surface. Anodic
potentiodynamic polarization curves are useful for the study of
the breakdown of the film deposited on the alloy surface (barrier
properties). Besides, open circuit potential measurements can be
used to study dynamic recovery of passivity (self-healing). Elec-
trochemical impedance spectroscopy (EIS) has been extensively
employed for the study of corrosion behaviour of chromate con-
version coatings [17] and for hybrid sol–gel films [18]. This
technique enables to follow the electrochemical properties of
pre-treatments as a function of the immersion time in an aggres-
sive solution. This is an important aspect in order to study barrier
properties and durability of pre-treatments.
2. Experimental procedure
2.1. Material and sample preparation
The substrate for the deposition of ZrO2 films was AA6060
aluminium alloy. The alloy composition is given in Table 1.
The surface preparation of the substrate consisted in degreasing
in acetone for all samples. The samples underwent degreas-
ing and pickling using commercial products from Chemetall
(Gardoclean 515, Gardoclean 620 HP and Gardacid 65) before
deposition of ZrO2 on the sample surface.
The deposition of the ZrO2 layer was performed using
dipping sol–gel technique. The starting solution was 0.1 M
Zr(OBun)4 in anhydrous n-butanol with addition of acetic acid
as complexing agent where [AcH]/[Zr] = 0.6. The deposition of
the film was performed by dipping the samples in the sol–gel
solution. The samples were withdrawn with controlled rate of
1 mm/s. The dipping step was repeated one, two or three times
in order to deposit overlapping zirconia layers on the samples.
Aftereachdipinthesol–gelsolution, thesamplesweresubjected
to thermal treatment. The thermal treatment was performed at
150 ◦C for 1 h and at 250 ◦C for 4 min. This step was followed
by drying at room temperature. Table 2 reports the sample codes
used in this work for ZrO2 pre-treated AA6060 specimens.
The surface preparation for the application of conven-
tional conversion coatings (chromatization and fluotitana-
tion/fluozirconation) was the same as for the deposition of
sol–gel pre-treatments. The conversion coatings were applied
by dipping in commercial baths from Chemetall (Gardobond
C714 for chromatization and Gardobond X4705 and Gardolene
D6800 for fluotitanation/fluozirconation).
Fedrizzi et al. [19,20] have investigated adhesion of organic
coatings to ZrO2 pre-treatments deposited with the sol–gel
technique by means of EIS. They have shown that electrolyte
ingress in the top coat due to cyclic thermal aging of painted
galvanized steels can be monitored by evolution of coating
capacitance.
This work investigates thin ZrO2 pre-treatments deposited
with sol–gel technique on AA6060. The small thickness of the
layers is an important requirement for the application of the
ZrO2 sol–gel pre-treatments as a replacement of chromate con-
version layers on aluminium alloys. Indeed, one of the factors
limiting the application of sol–gel layers is that these are usu-
ally rather thick coatings. Due to the limited thickness of the
layers deposited in our work (70–180 nm), it is very impor-
tant to evaluate the ability of ZrO2 sol–gel pre-treatments to
provide a good barrier against corrosion. Moreover, the self-
healing ability of sol–gel layers must be evaluated since this
is an important aspect of chromate conversion layers, which
need to be replaced. Besides, adhesive properties of sol–gel
layers should also be studied since these layers are usually
painted.
2.2. Electrochemical measurements
The electrochemical behaviour of the samples coated with
oxide was investigated by means of potentiodynamic polar-
ization measurements, open circuit potential measurements
Table 2
Deposition parameters for ZrO2 pre-treated AA6060 samples
The barrier properties, self-healing ability and adhesion
properties of ZrO2 pre-treatments are evaluated by means of
electrochemical techniques. Besides, this work compares the
behaviour of ZrO2 pre-treatments with that of chromatized
AA6060 and fluotitanated/fluozirconated AA6060.
Sample
Surface preparation
Number of dips
Thermal treatment
1dip-250
3dip-250
3dip-150
Degreasing + pickling
Degreasing + pickling
Degreasing + pickling
1
3
3
250 ◦C for 4 min
250 ◦C for 4 min
150 ◦C for 1 h

F. Andreatta et al. / Electrochimica Acta 52 (2007) 7545–7555
7547
and electrochemical impedance spectroscopy (EIS) in diluted
Harrison aqueous solution (0.05 wt% NaCl and 0.35 wt%
(NH4)2SO4). The counter electrode was a platinum wire (diam-
eter 1.5 mm) and the reference electrode was Ag/AgCl for all
measurements. A Pt mesh was used as counter electrode for
impedance measurements. The measurements were carried out
at room temperature in a standard electrochemical cell and the
area exposed was 3.6 cm2 (30 ml electrolyte) for potentiody-
namic polarizations and open circuit potential measurements.
The test area was 14.8 cm2 (80 ml electrolyte) for impedance
measurements. The scan rate was 0.2 mV/s for potentiodynamic
polarization measurements. Five potentiodynamic polarization
measurements were repeated for each sample investigated in this
work. Impedance measurements were carried out at open circuit
potential with ac voltage amplitude of 10 mV and frequency
range from 10 mHz to 100 kHz (10 points were measured for
each decade of frequency). All measurements were performed
using a AUTOLAB PG-STAT 12 potentiostat equipped with a
Frequency Response Analyser module.
dense uniform coverage of the alloy surface without large scale
defects. Large defects and cracks are sometimes observed for
single depositions (one dip) leading to a discontinuous ZrO2
film on the alloy surface. Moreover, highly concentrated sol–gel
solutions can produce a spongy like structure of the deposited
film [22]. TOF-SIMS depth profiles show that the thickness of
the layer is generally very thin (in the range of 70–90 nm after
the first dip). The thickness of the layer does not increase pro-
portionally to the number of dips reaching a maximum value of
about 180 nm after three dips.
3.2. Barrier properties
Fig. 1 shows cathodic potentiodynamic polarization curves
in diluted Harrison solution for as received AA6060, chroma-
tized AA6060, fluotitanated/fluozirconated AA6060 and sol–gel
pre-treated AA6060 (3dip-250). The potentiodynamic polariza-
tion curve for ZrO2 pre-treated AA6060 (3dip-250) exhibits
an open circuit potential of −0.447 V versus Ag/AgCl. The
open circuit potential for ZrO2 pre-treated AA6060 is strongly
dependent on the quality of pre-treatment. Therefore, it is rather
difficult to obtain reproducible values for the open circuit poten-
tial of ZrO2 pre-treated AA6060. Conversion coated AA6060
(chromatized and fluotitanated/fluozirconated) displays similar
open circuit potential to as received AA6060 (−0.660 V ver-
sus Ag/AgCl). However, fluotitanated/fluozirconated AA6060
exhibits cathodic current density similar to as received AA6060
over the entire cathodic polarization curve. In contrast, chroma-
tized AA6060 exhibits a strong reduction of the cathodic current
density indicating that chromate conversion layer inhibits
cathodic reactions [23,24]. The cathodic reactions taking place
on the alloy surface are oxygen reduction and water reduction.
The first reaction is the most likely to occur during potentiody-
namic polarization while the second takes place only at rather
In order to compare the barrier properties of samples coated
with ZrO2, potentiodynamic polarizations, open circuit potential
measurements and EIS were performed also on chromatized
AA6060 and fluotitanated/fluozirconated AA6060.
2.3. Adhesion
In order to test the adhesion properties of ZrO2 deposited
with sol–gel technique, AA6060 covered with ZrO2 films, flu-
otitanated/fluozirconated AA6060 and chromatized AA6060
samples were painted with polyester type coating (the thick-
ness of the top coat was 20 ␮m for all samples). Coated samples
underwent thermal cycles in 0.05 wt% NaCl solution (the area
exposed was 3.6 cm2). Each thermal cycle consisted of heating
the samples at 80 ◦C (above the glass transition temperature of
the paint), permanence at this temperature for 6 h, cooling and
permanence at room temperature for 18 h. Impedance measure-
ments were performed at the end of each cycle. The impedance
results were fitted by Zview Software, in order to obtain the
admittance modulus of the related constant phase element of the
coating. This provides information about the adhesion behaviour
of the coating [20].
3. Results and discussion
3.1. Structure of ZrO2 pre-treatments
StructureandthicknessoftheZrO2 layeronaluminiumalloys
is strongly affected by process parameters like number of depo-
sitions, thermal treatment applied to the layer, withdraw rate and
composition of the sol–gel solution [12,21]. The structure and
chemical composition of ZrO2 layers deposited with sol–gel
technique on AA6060 is reported in another paper [22]. The
layer mainly consists of zirconium oxide but organic residuals
may still be present depending on time and temperature of ther-
mal treatment. In general, the amount of organic residuals in
the layer decreases as time or temperature increases. Succes-
sive depositions (two or three dips) of ZrO2 layers lead to a
Fig. 1. Cathodic potentiodynamic polarization curves in diluted Harrison solu-
tion for as received AA6060, chromatized AA6060, fluotitanated/fluozirconated
AA6060 and ZrO2 pre-treated AA6060 (3dip-250).

7548
F. Andreatta et al. / Electrochimica Acta 52 (2007) 7545–7555
negative potentials [25]. Therefore, the increase of current den-
sity visible for all curves in Fig. 1 at potentials more negative
than −1.4 V versus Ag/AgCl is probably due to water reduction.
According to results reported in literature [23,24], the existence
of a chromium oxide/hydroxide film on the alloy surface limits
the kinetics of the oxygen reduction reaction. Moreover, it is
reported that chromate conversion coatings strongly reduce the
oxygen reduction rate on intermetallics [26], which are the main
initiation sites for localized corrosion of aluminium alloys. ZrO2
pre-treated AA6060 (3dip-250) exhibits a significant reduction
of cathodic current density as compared to as received AA6060.
This decrease is quite similar to that observed for chromatized
AA6060. The rather low cathodic current density observed in
cathodic polarization for ZrO2 pre-treated AA6060 (3dip-250)
might indicate that the ZrO2 layer deposited on the surface
inhibits the cathodic reaction similarly to chromatization. It is
possible that the existence of an inert ceramic film on the surface
might reduce the rate of oxygen transport to the alloy surface.
Moreover, the ZrO2 layer strongly reduces the area available for
cathodic reactions, which are expected to take place only in the
areas of the film containing defects.
The anodic polarization curves are strongly influenced by the
existence of defects in the pre-treatment both for conventional
conversion coatings and sol–gel pre-treatments. This behaviour
is particularly important for ZrO2 pre-treated AA6060, for
which the continuity of the film is strongly related to process
parameters. In particular, one can expect that the defects in
the ZrO2 layer are a preferential path for access of aggres-
sive species to the alloy surface favouring localized attack.
This behaviour is shown in Figs. 2 and 3 that show respec-
tively anodic polarization curves for pre-treatments with good
performance and for pre-treatments containing defects. In both
figures anodic polarization curves are reported for as received
Fig. 3. Anodic potentiodynamic polarization curves in diluted Harrison solution
for as received AA6060, chromatized AA6060 and ZrO2 pre-treated AA6060
(1dip-250). These potentiodynamic polarization curves were performed on sam-
ples containing defects in the pre-treatment.
AA6060, chromatized AA6060 and ZrO2 pre-treated AA6060
(fluotitanated/fluozirconated AA6060 is shown only in Fig. 2).
Fig. 2 shows the anodic polarization curve for sample 3dip-250
while Fig. 3 displays the curve for sample 1dip-250. As con-
firmed by SEM characterization (not shown here), the former
sample presents a dense and continuous ZrO2 layer on the alloy
surface with almost no defects since successive dipping in the
sol–gel solution leads to the formation of overlapping films that
cover defects of underlying ZrO2 layers. In contrast, sample
1dip-250 exhibits several defects in the ZrO2 film mainly in the
formofcracksorsmallfilmportionsdetachedfromthesubstrate.
The potentiodynamic polarization curve for fluoti-
tanated/fluozirconated AA6060 in Fig. 2 does not show a
passive range and the breakdown of the passive film takes place
at the open circuit potential, as for as received AA6060. This
indicates that the fluotitanation/fluozirconation pre-treatment
performed on AA6060 does not have significant barrier proper-
ties. In contrast, the anodic polarization curves for chromatized
AA6060 and sample 3dip-250 (sol–gel pre-treated AA6060)
exhibit an extended passive range characterized by very low
anodic current density. For these curves the breakdown takes
place at very positive potentials (+1.445 V versus Ag/AgCl
for chromatized AA6060 and +1.285 V versus Ag/AgCl for
sample 3dip-250). This indicates that the deposited films are
able to protect the substrate providing a good barrier against
localized attack. In particular, ZrO2 pre-treated AA6060
show similar corrosion behaviour to chromatized AA6060
in Fig. 2. This applies in the case of sol–gel pre-treatment,
which uniformly covers the surface and is not defected, as in
the case of sample 3dip-250 in Fig. 2. However, the strong
variability of electrochemical behaviour should be taken into
account for this type of pre-treatment, as stated above. This
is demonstrated by the anodic potentiodynamic polarization
Fig. 2. Anodic potentiodynamic polarization curves in diluted Harrison solu-
tion for as received AA6060, chromatized AA6060, fluotitanated/fluozirconated
AA6060 and ZrO2 pre-treated AA6060 (3dip-250).

F. Andreatta et al. / Electrochimica Acta 52 (2007) 7545–7555
7549
Fig. 4. Anodic potentiodynamic polarization curves in diluted Harrison solution
for different ZrO2 pre-treatments on AA6060 (1dip-250, 2dip-250 and 3dip-
250).
Fig. 5. Open circuit potential in diluted Harrison solution for as received
AA6060, chromatized AA6060 and ZrO2 pre-treated AA6060 (1dip-250 and
3dip-250).
curve in Fig. 3 for sample 1dip-250 (ZrO2 pre-treated AA6060)
that exhibits behaviour similar to as received AA6060 due to a
large amount of defects present in the ZrO2 layer (not shown
here). Also the curve for chromatized AA6060 (Fig. 3) displays
a less extended passive range and higher anodic current density
as compared to the curve for chromatized sample in Fig. 2.
The passivity breakdown occurs at about −450 mV versus
Ag/AgCl. This behaviour can be attributed to the existence of
defects in the pre-treatment and is related to a large variability
of the protective properties of industrially produced chromate
conversion coatings.
Fig. 4 exhibits anodic potentiodynamic polarization curves
for sol–gel pre-treatments on AA6060 obtained with different
number of dips in the sol gel bath. These curves exhibit a passive
range with different extension (or no passive range as in the
case of sample 1dip-250) and different anodic current density
depending on the number of dips. This confirms further that
the properties of ZrO2 pre-treatments on AA6060 are strongly
affectedbyprocessparametersandconsequentlybythepresence
of randomly distributed film flaws.
solution. This is possibly related to activation of anodic reactions
on the alloy surface and initiation of localized attack. Afterwards
the open circuit potential increases and becomes more noble than
that of as received AA6060 after 6 h immersion. This trend con-
tinues even after an immersion time of 10 h. This ennoblement
of the open circuit potential is most likely associated to repair
of the conversion layer, which is a well-known phenomenon for
chromate conversion coatings [23,27,28].
The open circuit potential of sample 3dip-250 shows a large
scatter during the first 3 h of immersion and thereafter reaches a
value more positive as compared to chromatized AA6060. The
initial scatter shown by the open circuit potential is probably
associated to the ceramic nature of the thin sol–gel film. Since
ZrO2 pre-treated AA6060 does not show a clear self-healing
ability, it is very important to characterize the durability of this
type of pre-treatment by EIS measurements.
Fig. 6 shows Bode plots of the impedance modulus for dif-
ferent immersion times in diluted Harrison solution measured
for as received AA6060, chromatized AA6060 and sol–gel pre-
treated AA6060 (1dip-250 and 3dip-250). Fig. 6A displays the
Bode plot immediately after immersion in solution (0 h). Chro-
matized AA6060 exhibits the highest impedance module at low
frequency indicating that this pre-treatment provides good bar-
rier properties. Similarly, sample 3dip-250 (ZrO2 pre-treated
AA6060) shows a rather high impedance modulus at low fre-
quency although about one order of magnitude lower than that
of chromatized AA6060. Sample 1dip-250 (ZrO2 pre-treated
AA6060) exhibits similar behaviour as sample 3dip-250 at high
and intermediate frequencies. However, at low frequency the
measurement becomes unstable for this sample probably due to
initiation of localized attack, which causes fluctuations of the
open circuit potential. The existence of a protective oxide on the
treated sample is also already visible in the behaviour at high
frequency as compared to that of as received aluminium sample.
3.3. Self-healing and durability
Fig. 5 shows the open circuit potential measured in
diluted Harrison solution for as received AA6060, chromatized
AA6060 and ZrO2 pre-treated AA6060 (1dip-250 and 3dip-
250). Fluotitanation/fluozirconation conversion coatings were
not considered as reference samples for this test because they
already failed in giving barrier properties. The sample 1dip-
250 shows similar open circuit potential to as received AA6060.
After an initial decrease due to surface activation, the open cir-
cuit potential becomes stable at a value of about −600 mV versus
Ag/AgCl. The chromatized AA6060 exhibits an initial decrease
of the open circuit potential during the first 2 h of immersion in

7550
F. Andreatta et al. / Electrochimica Acta 52 (2007) 7545–7555
Fig. 6. Bode plots of the impedance modulus for different immersion times in diluted Harrison solution measured for as received AA6060, chromatized AA6060
and ZrO2 pre-treated AA6060 (1dip-250 and 3dip-250): 0 h (A), 9 h (B), 24 h (C), and 48 h (D).
After 9 h immersion in solution (Fig. 6B), the impedance
modulus at low frequency is in the same range for chroma-
tized AA6060 and sample 3dip-250 (ZrO2 pre-treated AA6060).
Since the impedance modulus is rather high (4 × 106 ꢀ cm2) for
these samples, it can be stated that their barrier properties remain
good after 9 h immersion in diluted Harrison solution, although
an impedance diagram with two time constant appears for the
sol–gel treated sample.
Fig. 6C shows the impedance spectra after an immersion time
of 24 h. The impedance modulus at low frequency of sample
3dip-250 (ZrO2 pre-treated AA6060) does not decrease with
respect to the measurement for 9 h immersion revealing good
durability. Besides, the impedance modulus at low frequency
slightly increases with respect to the value measured after 9 h
immersion for chromatized AA6060. This might be a first sign
of dynamic recovery of the chromate oxide layer (self-healing).
The impedance modulus at low frequency for sample 1dip-250
(ZrO2 pre-treated AA6060) matches the value measured for as
received AA6060. This shows that this ZrO2 pre-treatment is
not able to protect the substrate. The trend observed in Fig. 6C
becomes more evident in Fig. 6D (48 h immersion time), where
the impedance modulus at low frequency further increases for
chromatized AA6060 and that for 3dip-250 sample (ZrO2 pre-
treated AA6060) remains rather constant.
In order to try to better explain the trend of the impedance
diagrams as a function of time, it is necessary to model the elec-
trochemical behaviour using equivalent electric circuits. Fig. 7
reports the equivalent circuits reported in literature [29,30] for

F. Andreatta et al. / Electrochimica Acta 52 (2007) 7545–7555
7551
Fig. 8. Equivalent circuit from Campestrini et al. [30] (A) used to model
impedance spectra reported as Bode plot for impedance modulus (B) and phase
(C) for chromatized AA6060.
Fig. 7. Equivalent circuits reported by Campestrini et al. [30] and Goeminne
et al. [29] for fitting impedence data relative to chromatized aluminium alloys
depending on type of chromate film.
circuit reported by Campestrini et al. [30]. The circuit in Fig. 8A
is electrically equivalent to the circuit shown in Fig. 7C. The con-
stant phase element CPEox and the resistor Rox in Fig. 8A are
used to model the non-ideal capacitive behaviour of the layer
at the location of large defects like cracks [29,30]. The double
layer capacitance is modelled with the constant phase element
CPEdl and Rct and Rel are respectively the charge transfer resis-
tance and the electrolyte resistance. The fit obtained with the
equivalent circuit proposed by Campestrini et al. [30] is very
good both for the impedance module (Fig. 8B) and for the phase
angle (Fig. 8C).
The equivalent circuit in Fig. 9A was proposed by Zhelud-
kevich et al. [31] to model the protective behaviour of sol–gel
hybrid coatings on AA2024 and by Zhu and van Ooij [32] to
study silane films on aluminium alloys. This circuit is based on
the assumption that the ZrO2 pre-treatment is constituted by an
outer layer containing large pores (indicated as sol–gel layer
with defect in Fig. 9A) and by an intermediate layer that might
also contain defects (indicated as oxide in Fig. 9A). The con-
stant phase element CPEcoat models the intact layer while the
resistance Rp is associated to large pores in the outer layer. The
intermediate oxide is modelled with the constant phase element
CPEox and the resistance Rox. Fig. 9B and C shows respectively
the Bode plot for the impedance modulus and for the phase angle
together with the fitted data. Although the fit is generally good,
these results evidence that the equivalent circuit in Fig. 9A does
not fit perfectly our electrochemical impedance spectra. Indeed,
fitting parameters relative to the CPEox and to the Rox (Table 3)
are in good agreement with data reported in literature for similar
sol–gel coatings [31]. However, the fitting error for Rox is rather
high.
modelling impedance spectra for chromatized aluminium alloys
depending on type of chromate film and immersion time in
aggressive solution. Campestrini et al. [30] stated that the equiv-
alent circuit in Fig. 7A corresponds to the formation of a small
amount of chromate corrosion products (ccp), which form a
thin, dense and compact layer (with finite length of the dif-
fusion path). This layer is modelled by Rccp and CPEccp. Rpl
models conductive paths due to existence of nanopores in the
chromate conversion layer. CPEbl describes the non-ideal capac-
itivebehaviourofthebarrierlayeratthebottomoflargecracksor
defects, while CPEdl is related to the portion of surface involved
in the corrosion process. According to the physical interpreta-
tion given by Campestrini et al. [30], Rpl and CPEbl are related
to the large pores in the film rather than to the film itself. More-
over, Rct and CPEdl represent the electrochemical processes
inside the defects. The circuit in Fig. 7B is associated to the
existence of a thick, porous and less protective film (with infi-
nite length of the diffusion path) due to the formation of some
corrosion products. This behaviour is modelled by means of a
Warburg element W. The equivalent circuit shown in Fig. 7C cor-
responds to a chromate layer with a large amount of corrosion
products resulting in a less adherent film onto the alloy surface.
Circuits shown in Fig. 7B and C were found to better repre-
sent industrial chromate conversion layer applied on AA2024
[30].
An example of experimental and fitted data for chromatized
AA6060 is shown in Fig. 8. This data were fitted with the equiv-
alent circuit shown in Fig. 8A, which is based on the equivalent

7552
F. Andreatta et al. / Electrochimica Acta 52 (2007) 7545–7555
Fig. 10. Alternative equivalent circuit (A) used to model impedance spectra
reported as Bode plot for impedance modulus (B) and phase (C) for ZrO2 pre-
treated AA6060 (3dip-150).
Fig. 9. Equivalent circuit from Zheludkevich et al. [31] (A) used to model
impedance spectra reported as Bode plot for impedance modulus (B) and phase
(C) for ZrO2 pre-treated AA6060 (3dip-150).
which is covered with a thin sol–gel layer. The parameter Rox is
affected by nanopores in this layer. Therefore, this is the most
important parameter of the equivalent circuit as far as corrosion
protection is concerned. Table 4 displays the fitting parameters
obtained for ZrO2 pre-treated AA6060 (sample 3dip-150) using
the equivalent circuit in Fig. 10A. The fitting error for Rox is
strongly reduced indicating a rather good fit for this parameter.
Fig. 11 shows the oxide resistance Rox as a function of immer-
sion time in diluted Harrison solution for chromatized AA6060
and ZrO2 pre-treated AA6060 (sample 1dip-250 and 3dip-250).
Rox is calculated using the equivalent circuit in Fig. 8A for chro-
matized AA6060 and from the circuit in Fig. 10A for sol–gel
pre-treated AA6060. Rox for chromatized AA6060 exhibits an
initial decrease during the first 10 h immersion followed by an
increase until a plateau region is reached between 72 and 150 h
immersion. Rox for sample 3dip-250 (ZrO2 pre-treated AA6060)
In order to improve the fitting result for ZrO2 pre-treatment
on AA6060, the equivalent circuit in Fig. 10A is proposed. This
circuit includes the equivalent circuit proposed by Campestrini
et al. [31] for chromate conversion coatings with the addition of a
mesh modelling an outer sol–gel layer with a defect. In this case
the constant phase element CPEcoat describes the behaviour of
the intact ZrO2 layer and the resistance Rp is related to the defect,
as in the equivalent circuit proposed by Zheludkevich et al. [31].
In contrast with the circuit used by Zheludkevich et al. [31],
the mesh describing the substrate (CPEdl and Rct) is in series
with CPEox in the equivalent circuit in Fig. 10A, as reported by
Campestrini et al. [30] for chromate conversion layers. CPEox
describes a network of electrolyte resistances and double layer
capacitors in the pores of the layer [29,30], while Rox is related
to the barrier properties of the film. Based on the circuit shown in
Fig. 10A, charge transfer processes take place on the substrate,
Table 3
Table 4
Fitting parameters relative to the sol–gel layer for ZrO2 pre-treated AA6060
(3dip-150) (these values were obtained using the equivalent circuit from Zhe-
ludkevich et al. [31])
Fitting parameters relative to the sol–gel layer for ZrO2 pre-treated AA6060
(3dip-150) (these values were obtained using the equivalent circuit shown in
Fig. 10A)
Value
Error
Error %
73.7
Value
Error
Error %
0.7
Rox (ꢀ)
2.4 × 105
1.8 × 105
Rox (ꢀ)
2.2 × 105
1.5 × 103
CPEox
Y0/F
n
CPEox
Y0/F
n
2.3 × 10−5
1.3 × 10−6
5.8
1.3
3.9 × 10−5
8.3 × 10−8
0.2
0.1
0.9
0.01
0.9
0.001

F. Andreatta et al. / Electrochimica Acta 52 (2007) 7545–7555
7553
Fig. 11. Oxide resistance (Rox) as a function of immersion time in diluted
Harrison solution for chromatized AA6060 and for ZrO2 pre-treated AA6060
(1dip-250 and 3dip-250).
exhibits an initial decrease of about one order of magnitude
during the first 9 h of immersion followed by a limited recov-
ery. In contrast, sample 1dip-250 (ZrO2 pre-treated AA6060)
shows a marked decrease of Rox. Moreover, Rox for this sample
is significantly lower than for sample 3dip-250 and chromatized
AA6060 for the entire duration of the test. A decrease of Rox was
observed after more than 150 h immersion time for chromatized
AA6060.
Fig. 12. Bode plot for impedance modulus for chromatized and top coated
AA6060 subjected to thermal cycles in 0.05% NaCl solution.
rather high(about1011 ꢀ cm2) for26 thermal cycles, whileatthe
27th thermal cycle there is a decrease of the impedance modulus.
Fig. 13 shows the Bode plot for the impedance modulus for
sol–gel treated and painted AA6060 (sample 3dip-250) after
the same number of thermal cycles as in Fig. 12. The impedance
modulusdoesnotshowanysignificantdecreaseafterperforming
27 thermal aging cycles.
Electrochemical impedance data obtained after thermal aging
cycles are modelled using the equivalent circuit in Fig. 14. Qc
represents the non-ideal capacitance of the top coat, Rc is the
The trend observed for Rox of chromatized AA6060 indi-
cates that the conversion layer is able to dynamically recover
the barrier properties of the layer after an initial damage. This
corresponds to the well-known self-healing ability of chromate
conversioncoatings. ThebehaviourofZrO2 pre-treatedAA6060
is strongly dependent on the quality of the film formed on the
alloy surface. Indeed, sample 1dip-250 exhibits a fast deteriora-
tion of the barrier properties that could be related to the existence
of large defects in the ZrO2 layer and to film detachment due
to corrosion processes. In contrast, sample 3dip-250 exhibits
barrier properties comparable to those of chromatized AA6060.
This is related to the very limited number and small size of
defects present in the ZrO2 layer after a three-dip deposition.
Besides, the 3dip-250 sample displays a limited recovery of the
barrier properties after the initial deterioration during the first
hours of immersion in solution. This limited recovery might be
caused by the formation of corrosion products that could plug
the defects or pores existing in the coating rather than by self-
healing ability of the layer. This phenomenon is expected to be
effective only for small defects in the ZrO2 pre-treatment, as in
the case of sample 3dip-250.
3.4. Adhesion properties
Fig. 12 shows the Bode plot for the impedance modulus mea-
sured in 0.05% NaCl solution at the end of each thermal aging
cycle for chromatized and painted AA6060. Thermal cycle num-
ber 0 is an impedance measurement performed in solution before
performing the first thermal cycle. The impedance modulus is
Fig. 13. Bode plot for impedance modulus for ZrO2 pre-treated (3dip-250) and
top coated AA6060 subjected to thermal cycles in 0.05% NaCl solution.

7554
F. Andreatta et al. / Electrochimica Acta 52 (2007) 7545–7555
Fig. 14. Equivalent circuit to model pre-treated samples top coated with
polyester spray resin (the thickness of the top coat was 20 ␮m for all samples).
ionic resistance of the top coat through its intact or damaged
structure and the constant phase element CPEox and the resis-
tance Rox describe the response of the pre-treatment together
with that of the substrate. This equivalent circuit was previously
employed to describe the behaviour of painted galvanized steel
under thermal aging [20].
Fig. 15 reports the coating capacitance Qc normalized versus
the initial capacitance as a function of the thermal aging cycle
number for chromatized AA6060, fluotitanated/fluozirconated
AA6060 and sol–gel pre-treated AA6060 (sample 3dip-250,
3dip-150 and 1dip-250). Three different phases can be distin-
guished for the behaviour of Qc of all samples characterized.
An initial increase of Qc (i) after the first thermal cycle is
associated to not reversible solution ingress in the top coat.
This first increase of Qc is significantly higher than the value
reached at saturation when water uptake occurs at room tem-
perature. Indeed, water can fill many microstructural and defect
Fig. 16. Micrograph obtained with a stereo microscope on the surface of a
chromatized and top coated AA6060 after 27 thermal cycles.
voids within paints including interfaces between pigments and
polymer binder. Afterwards, Qc progressively increases due to
further water uptake in the paint (ii). The last phase corresponds
to a remarkable increase of Qc (iii). This could correspond to the
decrease of the electrochemical impedance modulus observed
in Fig. 12. With reference to our previous works [20,22], it is
possible to correlate this clear increase of Qc (iii) with blister for-
mation and local loss of adhesion. As an example, Fig. 16 shows
a blister on the surface of chromatized and painted AA6060
taken immediately after the electrochemical impedance mea-
surement corresponding to thermal aging cycle 27. Besides, it
was shown that samples that exhibit the large increase of Qc
after a relatively small number of thermal aging cycles are char-
acterized by poor wet adhesion of the paint on the pre-treated
substrate [20].
It is well known that chromatized aluminium provides
good adhesion to organic coatings. Therefore, the adhesion
properties of sol–gel pre-treatments and other conversion coat-
ings have to be compared to that of chromatized AA6060.
Fluotitanated/fluozirconated AA6060 exhibits a progressive
degradation of Qc. This is associated by a progressive increase
of the slope of Qc that is already evident after 18 thermal cycles.
The morphology of fluotitanated/fluozirconated and painted
AA6060 shows several blisters after 27 thermal aging cycles (not
shown here). Hence, adhesion properties provided by fluotita-
nation/fluozirconation are less good than those of a chromate
conversion coating, as well established in literature. Sample
3dip-250 and sample 3dip-150 do not exhibit a marked increase
of Qc, after 27 thermal cycles and no blisters are visible on
the surface of these samples. This behaviour can be attributed
to the limited number of defects observed for samples that
underwent three successive depositions. This means that ZrO2
pre-treatments have a rather high resistance to water uptake
and that there is very good wet adhesion of the top coat to
the substrate. Besides, the trend observed for sample 3dip-250
and sample 3dip-150 indicates that the wet adhesion proper-
ties are slightly better for ZrO2 pre-treatments that underwent
Fig. 15. Coating capacitance Qc (normalized vs. the initial capacitance Q0)
as a function of thermal cycle number for chromatized AA6060, fluoti-
tanated/fluozirconated AA6060 and ZrO2 pre-treated AA6060 (1dip-250,
3dip-150 and 3dip-250).

F. Andreatta et al. / Electrochimica Acta 52 (2007) 7545–7555
7555
thermal treatment at 150 ◦C than at 250 ◦C. This behaviour
could be most probably attributed to the existence of a higher
content of organic and hydroxyl residuals in the sample heat
treated at the lower temperature. This results in a structure
with a higher content of bonding sites between ZrO2 layer and
top coat.
pean Community as contract no. NMP3-CT-2005-011783 under
the 6th Framework Programme for Research and Technological
Development.
References
In contrast to samples that underwent three dips in the sol–gel
solution (sample 3dip-250 and sample 3dip-150), sample 1dip-
250 shows a marked increase of Qc after only after eight thermal
cycles (Fig. 15) associated to the existence of a high amount of
blisters and severe delamination (not shown here). The poor
adhesion of the top coat to the ZrO2 pre-treatment is caused by
the defected structure of films obtained with only one dip in the
sol–gel bath, as discussed above.
[1] T.L. Metroke, R.L. Parkhill, E.T. Knobbe, Prog. Org. Coat. 41 (2001) 233.
[2] R. Di Maggio, L. Fedrizzi, S. Rossi, J. Adhesion Sci. Technol. 15 (2001)
793.
[3] M. Bethencourt, F.J. Botana, J.J. Calvino, M. Marcos, M.A. Rodriguez-
Chacon, Corros. Sci. 11 (1998) 1803.
[4] J.H. Osborn, Prog. Org. Coat. 41 (2001) 280.
[5] M. Guglielmi, J. Sol–Gel Sci. Technol. 8 (1997) 443.
[6] R.L. Ballard, J.P. Williams, J.M. Njus, B.R. Kiland, M.D. Soucek, Eur.
Polym. J. 37 (2001) 381.
[7] N.N. Voevodin, N.T. Grebasch, W.S. Soto, F.E. Arnold, M.S. Donley, Surf.
Coat. Technol. 140 (2001) 24.
4. Conclusions
[8] A.M. Beccaria, G. Padeletti, G. Montesperelli, L. Chiaruttini, Surf. Coat.
Technol. 111 (1999) 240.
[9] A.M. Beccaria, L. Chiaruttini, Corros. Sci. 41 (1999) 885.
[10] X.F. Yang, D.E. Tallman, V.J. Gelling, G.P. Bierwagen, L.S. Kasten, J.
Berg, Surf. Coat. Technol. 140 (2001) 44.
[11] M.L. Zheludkevich, R. Serra, M.F. Montemor, I.M. Miranda Salvado,
M.G.S. Ferreira, Electrochim. Acta 51 (2005) 208.
[12] L. Fedrizzi, R. Di Maggio, S. Rossi, L. Leonardelli, Benelux Metall. 43
(2003) 15.
[13] O. Lunder, J.C. Walmsley, P. Mack, K. Nisancioglu, Corros. Sci. 47 (2005)
1604.
ZrO2 pre-treatments applied with the sol–gel technique on
AA6060 have been investigated by means of complementary
electrochemical techniques like potentiodynamic polariza-
tion, open circuit potential measurements and electrochemical
impedance spectroscopy. Besides, ZrO2 pre-treated and top-
coated AA6060 has been subjected to thermal aging in order
to induce blistering of the top coat. The adhesion properties of
the top coat have been evaluated by means of electrochemical
impedance spectroscopy.
[14] O. Lunder, C. Simensen, Y. Yu, K. Nisancioglu, Surf. Coat. Technol. 184
(2004) 278.
The barrier properties of ZrO2 pre-treatments are strongly
dependent on process parameters. The electrochemical
behaviour exhibits a strong variability depending on the qual-
ity of the ZrO2 pre-treatment. In particular, ZrO2 pre-treated
AA6060 exhibits barrier properties similar to chromatized
AA6060 when a continuous layer is deposited on the surface
through successive dipping steps in the sol–gel solution (three
dips). AA6060 coated with a single deposition (one dip) does
not show barrier properties since the ZrO2 layer exhibits several
defects. The number of defects in the layer is strongly reduced
by deposition of overlapping ZrO2 layers.
The ZrO2 pre-treatments on AA6060 do not exhibit self-
healing ability as chromatized AA6060. However, a limited
recovery of the barrier properties is observed for ZrO2 pre-
treatments with a small number of defects. This is attributed to
the formation of corrosion products that might plug the defects
in the layer.
[15] N.N. Voevodin, N.T. Grebasch, W.S. Soto, L.S. Kasten, J.T. Grant, F.E.
Arnold, M.S. Donley, Prog. Org. Coat. 41 (2001) 287.
[16] H.M. Wang, R. Akid, J.E.P. Metcalf, Proceedings of the International
Conference on Environmental Friendly Pre-treatments for Aluminium and
Other Metals ICEPAM, Oslo, Norway, 2004, http://www.sintef.no.
[17] P. Campestrini, E.P.M. van Westing, J.H.W. de Wit, Electrochim. Acta 46
(2001) 2631.
[18] M.L. Zheludkevich, M.G.S. Ferreira, I.M. Miranda Salvado, J. Mater.
Chem. 48 (2005) 5099.
[19] L. Fedrizzi, F.J. Rodriguez, S. Rossi, F. Deflorian, R. Di Maggio, Elec-
trochim. Acta 46 (2001) 3715.
[20] L. Fedrizzi, A. Bergo, M. Fanicchia, Electrochim. Acta 51 (2006) 1864.
[21] R. Di Maggio, P. Aldighieri, S. Rossi, L. Fedrizzi, Proceedings of the Euro-
pean Corrosion Congress EUROCORR2005, Lisbon, Portugal, September
4–8, 2005.
[22] L. Fedrizzi, R. Di Maggio, P. Aldighieri, F. Andreatta, F. Girardi, Proceed-
ings of the Fourth International Symposium on Aluminium Science and
Technology, in press.
[23] J. Zhao, G.S. Frankel, R. McCreery, J. Electrochem. Soc. 145 (1998) 2258.
[24] J.D.B. Sharman, Proceedings of the First International Symposium on
Aluminium Science and Technology, Antwerp, Belgium, 1997, p. 118.
[25] M. Pourbaix, Atlas of Electrochemical Equilibria in Acqueous Solutions,
NACE International Cebelcor, 1974, p. 101.
[26] G.O. Ilevbare, J.R. Scully, J. Electrochem. Soc. 148 (2001) B196.
[27] L. Xia, E. Akiyama, G.S. Frankel, R. McCreery, J. Electrochem. Soc. 147
(2000) 2556.
ZrO2 pre-treated and top-coated AA6060 shows high resis-
tance to blistering under thermal aging cycling. This indicates
that there is good adhesion of the top coat to the pre-treatment.
The adhesion of a top coat onto a ZrO2 pre-treatment obtained
through a single deposition is poor due to the existence of a large
number of defects in the ZrO2 layer.
[28] J.D. Ramsey, R. McCreery, J. Electrochem. Soc. 146 (1999) 4076.
[29] G. Goeminne, H. Terryn, J. Vereecken, Electrochim. Acta 40 (1995) 479.
[30] P. Campestrini, H. Terryn, J. Vereecken, J.H.W. de Wit, J. Electrochem.
Soc. 151 (2004) B370–B377.
Acknowledgement
[31] M.L. Zheludkevich, R. Serra, M.F. Montemor, I.M. Miranda Salvado,
M.G.S. Ferreira, Surf. Coat. Technol. 200 (2006) 3084.
[32] D. Zhu, W.J. van Ooij, Corros. Sci. 45 (2003) 2177.
This paper and the work it concerns were generated in the
context of the MULTIPROTECT project, funded by the Euro-