Characterization of ceramic sol-gel coatings as an alternative chemical conversion treatment on commercial carbon steel

Article abstract of DOI:10.1016/j.electacta.2008.11.023

Sol-gel yttria-stabilized zirconia (YSZ) thin films were prepared on commercial carbon steel sheets by dip-coating technique followed by a low temperature heat treatment (473, 573, and 673 K for 1 h) in order to improve both corrosion properties and adhes

Full text of DOI:10.1016/j.electacta.2008.11.023

Electrochimica Acta 54 (2009) 2932–2940
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Characterization of ceramic sol–gel coatings as an alternative chemical
conversion treatment on commercial carbon steel
M.A. Domínguez-Crespoa,∗, A. García-Murilloa, A.M. Torres-Huertaa,
F.J. Carrillo-Romoa, E. Onofre-Bustamanteb, C. Yán˜ez-Zamorac
a Instituto Politécnico Nacional, Grupo de Ingeniería en Procesamiento de Materiales CICATA-IPN, Unidad Altamira,
km 14.5, Carretera Tampico-Puerto Industrial Altamira, C.P. 89600 Altamira, Tamps, Mexico
b Universidad Nacional Autónoma de México, Edificio D Facultad de Química, Departamento de Metalurgia,
Ciudad Universitaria, C.P. 04510 México D.F., Mexico
c Alumna del postgrado en Tecnología Avanzada del CICATA-IPN, Unidad Altamira IPN,
km 14.5, Carretera Tampico-Puerto Industrial Altamira, C.P. 89600 Altamira, Tamps, Mexico
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 4 June 2008
Received in revised form 8 November 2008
Accepted 9 November 2008
Available online 21 November 2008
Sol–gel yttria-stabilized zirconia (YSZ) thin films were prepared on commercial carbon steel sheets by
dip-coating technique followed by a low temperature heat treatment (473, 573, and 673 K for 1 h) in order
to improve both corrosion properties and adhesion. For comparison, zirconia (ZrO2) coatings have been
also analyzed. Electrochemical techniques, Fourier Transform Infrared (FT-IR), X-ray diffraction (XRD)
and scanning electron microscopy (SEM) were used to characterize the anticorrosion behavior of the
coatings in a 3.5 wt% NaCl solution. The adhesion with a polyester organic coating was evaluated by the
pull-off technique. The typical thickness of the deposited layers ranged from 1 to 1.3 ␮m depending on
process parameters. The obtained results indicated that sol–gel ZrO2 and YSZ coatings without an organic
coating can act as protective barriers against wet corrosion during the first hours, but they fail when the
time exposure is longer than 1 day. However, when synthesized films were used as a pre-treatment
and an organic coating was added (top-coated), the anticorrosive and adhesion properties were strongly
affected by the temperature of the treatment, and an increase in both properties was observed at higher
temperatures. The structural and morphological characteristics of the coating provide an explanation of
the role of each film in the electrochemical behavior in this aggressive medium. Comparing both systems,
YSZ displayed greater protective and adhesion values than exhibited for ZrO2 which can be correlated
with the stabilization of the cubic phase.
Keywords:
Sol–gel
Adhesion
ac impedance
Ceramic coatings
YSZ
© 2008 Elsevier Ltd. All rights reserved.
1. Introduction
that naturally form a passive layer on their surface [8–10], although
sol–gel deposition on active corroding materials, such as iron, with
The sol–gel method is one of the most useful technologies
to produce amorphous or crystalline oxide coatings [1,2]. Very
homogeneous films are obtained at low temperature, and many
compositions have been produced to enhance some properties or
to give new ones to different substrates [3–5]. Attractive applica-
tions of these materials are expected in several fields such as optics
[6] and catalysts [7]. An interesting and relatively recent possibility
is the realization of protective films on metals and, particularly, the
fabrication of coating films starting from metal alkoxides. Several
coatings produced by sol–gel processing have been studied exten-
sively for corrosion prevention in stainless steel and other metals
a wide variety of coatings remains unexplored [11–13].
In the corrosion field, the protection of metals against envi-
ronmental attack is usually achieved by using an adherent and
protective coating deposited onto the metal surface. Adhesion to the
metal substrate can be considered in some practical situations as
the most important property of organic and inorganic coatings. This
property has a particular significance when painted metals work
in humid environments or in electrolytes, because the presence
of polar molecules such as water can greatly affect the chemical
bonds between the metal and the coating. In order to improve
the adhesion properties of organic coatings on steel, chemical pre-
treatments have been used for a long time: particularly, chromate
or phosphate have proved to be quite useful [14,15]. Some of the
new chemical conversion treatments, containing titanium, zirco-
nium or molybdenum salts, have shown promising results when
applied onto different substrates [16,17].
∗
Corresponding author. Tel.: +52 55 57 29 6000x87512.
E-mail addresses: mdominguezc@ipn.mx, adcrespo2000@yahoo.com.mx
(M.A. Domínguez-Crespo).
0013-4686/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2008.11.023

M.A. Domínguez-Crespo et al. / Electrochimica Acta 54 (2009) 2932–2940
2933
Ceramics are, in general, more resistant than metals to oxida-
tion, corrosion, erosion and wear [18]. Zirconia films are used as
thermal barrier coatings, wear resistance coatings, diffusion bar-
riers, oxygen sensors, underlayers for superconducting films, and
also for their corrosion–resistant properties [19–24]. By the addi-
tion of foreign cations (Y3+, Mg2+, Ca2+, etc.), or by the choice
of a suitable precursor during the synthesis, the tetragonal or
cubic forms can be stabilized at low temperatures, greatly improv-
ing the chemical or physical properties of the material. In these
stabilized systems, yttria-stabilized zirconia (YSZ) is the most
common system. YSZ has high ionic conductivity and thermal
stability offering an improvement in the adhesion or corrosion
resistance properties. In view of the above findings with regard to
improve in protection against corrosion and in adhesion proper-
ties, chemical conversion treatments of ceramic systems could be
incorporated either as a bond coat or on top of a commercial carbon
steel.
The current study evaluates the adhesion properties and electro-
chemical behavior of commercial carbon steel (AISI-1012), coated
with YSZ by a sol–gel technique and followed by a low temperature
heat treatment (473, 573 and 673 K), in order to study their per-
formance and behavior as a barrier against wet corrosion in NaCl
solution. The corrosion characteristics have been evaluated through
potentiodynamic polarisation curves, ac impedance, and analyses
of the relevant electrochemical parameters. For comparison, ZrO2
coatings have been also prepared.
sor sol was obtained. Before the deposition process, the precursor
solution was filtered using a 0.22 ␮m filter, then, carbon steel
sheets were coated by dip-coating technique with a withdrawal
constant rate of 1 mm/s, thereby obtaining thin films of yttria-
stabilized zirconia (ZrO2–8%Y2O3, YSZ). The thickness of the film
could be increased by repeating the deposition run. After film depo-
sition, the samples were treated in an oven at 473, 573 and 673 K
for 60 min in order to densify the coating and remove most of
the organic load. The final thickness of the zirconia layers var-
ied in the range of 1.0–1.3 ␮m. A similar experimental procedure
was used to prepare the ZrO2 system. The YSZ and ZrO2 thin
films were analyzed as either a bond coat or on top of a com-
mercial carbon steel; therefore, some specimens were painted
after the sol–gel process with a commercial polyurethane var-
nish using a micrometric aluminium film applicator. Dry film
thickness was measured with a surface profiler (digital coating
thickness gauge, Elcometer 345). These samples showed a thick-
ness between 40 and 50 ␮m, with a standard deviation about
1.2 ␮m.
2.2. Experimental techniques
The evolution and decomposition of the organic compounds
up to the crystallization process were determined by FT-IR spec-
troscopy. A Siemens 670 instrument equipped with a mercury
cadmium telluride A (MCTA) detector, XT-KBr beam-splitter, and
OMNIC software was used to obtain the spectra of the sample,
which were placed on a KBr pellet. Two scans at a resolution of
4 cm−1 were taken and averaged over the 4000–450 cm−1 region.
A background was collected before each sample then subtracted
from the sample spectra prior to further analysis. The analysis of
the coating structure was carried out by XRD patterns using a D8
Advance Bruker diffractometer, operating at 35 kV and 25 mA, with
a curve graphite crystal monochromator and broad focus copper
source. The samples were scanned at 2◦ min−1 between 20◦ and
100◦ in order to study the crystal structure. Morphological and
chemical aspects of coatings before electrochemical measurements
were studied by scanning electron microscopy and energy disper-
sive X-ray spectrometry (SEM/EDS, JEOL JSM-6300 equipped with
a EDS NORAN).
The electrochemical behavior of the samples was investigated
by means of open circuit potential (Eocp), Tafel, and electrochem-
ical impedance spectroscopy (EIS) in a 3.5 wt% NaCl solution. The
analyses were carried out at room temperature in a standard elec-
trochemical cell using an AUTOLAB 30 potentiostat equipped with
a frequency response analyzer module. The counter electrode was
a large-area graphite bar. The reference electrode was a satu-
rated calomel electrode SCE (0.2415 V vs. SHE). The cell consisted
of an acrylic rectangular box (60 mm × 80 mm × 100 mm) and the
exposed area of the sample was 0.985 cm2 (150 mL electrolyte). The
specimens were introduced by moderate pressure against an o-ring,
avoiding localized damage in the ceramic layers. Tafel plots were
acquired in the region from −250 mV vs. Eocp (open circuit poten-
tial) to +250 mV vs. Eocp. EIS measurements were carried out from
open circuit potential with an ac voltage amplitude of 10 mV and a
frequency range of 400,000 to 0.01 Hz (ten points were measured
for each decade of frequency). In order to prevent any influence
from previous polarisation, each measurement was taken from a
new area of the sample.
2. Experimental procedure
2.1. Preparation of the sols and coatings
Different conversion coatings were obtained on AISI-1012 car-
bon steel samples of 50 mm × 60 mm in size and 10 mm in
thickness. The chemical composition of this substrate in weight-
percentage terms is shown in Table 1. Before coating, the metal
surface was mechanically grinded starting from coarse 240- to
fine 600-grade emery papers, according to the standard metallo-
graphic techniques. In order to remove any remaining of water or
SiC in the samples due to the previous polishing, they were flushed
with acetone and left to dry at room temperature for 10 min. Sol
ZrO2–8%Y2O3 preparation was obtained using two different pre-
cursors, zirconium propoxide, yttrium chloride in ethanol presence
and 2,4-pentanedione as complexing agent, as follows: zirconium
propoxide was well mixed with ethanol and 2,4-pentanedione
was immediately added to the precursor solution in order to con-
trol the reaction. Thereafter, a convenient quantity of ethanol was
incorporate to this solution. Finally, yttrium chloride was mixed
in the solution; after 1 h a yellow, stable, and transparent precur-
Table 1
Chemical composition of the substrate.
Element
wt%
0.1210
C
Si
Mn
P
0.0321
0.4231
0.0401
0.0052
0.0067
0.0377
0.0134
0.0412
0.0027
0.0250
0.0026
0.0056
99.2436
S
Cr
Mo
Ni
Al
Co
Cu
Ti
The adhesion strength of the organic coatings on AISI-1012
commercial steel substrates, with and without ceramic coating,
was measured with a C. C. Elcometer 106 series, according to the
ASTM D-4541 [25] standard. 21-mm diameter aluminium studs
were glued onto the treated specimens for each test with epoxy
resin, followed by a de-aeration process at room temperature
over 24 h.
Pb
Fe

2934
M.A. Domínguez-Crespo et al. / Electrochimica Acta 54 (2009) 2932–2940
Fig. 2. Diffraction patterns of ZrO2 and YSZ after thermal treatment.
existing information on alkoxide-stabilizing complex formation,
these compounds are regarded as metal acetylacetonates with the
formula (C5H7O2)3Y for the yttrium complex and (C5H7O2)4Zr for
the zirconium complex. Differences in wave number associated
with functional groups in each case could be attributed to the
precursor nature, such as their reactivity with acetylacetone as a
function of their ionic radius. This is closely associated with the
coordination number of the metal and its capability to coordinate a
specific number of acetylacetone molecules. In the case of yttrium
and zirconium complexes, zirconium can coordinate more acac-H
molecules than yttrium, due to atomic size difference.
3.2. X-ray diffraction
Fig. 2 shows the diffraction patterns of both ZrO2 and YSZ after
being thermally treated. From a macroscopic point of view, each
heat treated specimen is covered with a thin white film, corre-
sponding to ZrO2 and YSZ materials. It seems that these coatings
completely cover the substrate and the film exhibits a homoge-
neous surface. XRD analyses on the coated substrates indicated
that all non-annealed (not shown here) samples are amorphous.
For heat treated samples, zirconia coatings are crystalline in the
cubic phase, and the crystallinity increases with the temperature
of the treatment, while, the intensity of XRD patterns of YSZ thin
films are weak, broadened, and also amorphous at 473 and 573 K;
which may be due to both the organic compounds remaining from
sol–gel process and the nanosize effect. It was also observed that
the formation of crystallized yttria-stabilized zirconia started above
573 K. The main reflections of the cubic YSZ (cubic fluorite phase)
were identified, and the broadening of the diffraction peaks was
attributed to the small crystallite size. The preliminary crystal size d
of the as-densified sample was estimated to be about 10 nm, accord-
ing to Scherrer’s equation d = 0.90ꢀ/ˇ cos ꢁ. Control of the adequate
sintering temperature allowed the formation of crystalline YSZ thin
films that could offer a suitable response as protective barriers.
Fig. 1. Infrared spectra and band assignments of (a) ZrO2 and (b) YSZ thin films.
3. Results and discussion
3.1. FT-IR analysis
Fig. 1a and b are FT-IR representative spectra of the ZrO2–8%Y2O3
and ZrO2 thermally treated coatings (473, 573 and 673 K), respec-
tively. At low temperatures for both ceramic compounds, the
spectra recorded intensity absorption bands of the O–H stretch-
ing at 3400–3456 cm−1, which disappeared at temperatures higher
than 573 K. Signals ranging from 1193 to 1796 cm−1 show the forma-
tion of a bidentate complex with keto-enolic equilibrium behavior.
According to Cueto et al. [26], this denotes evident ring formation
and coordination of the metal with acetylacetone carbonyl groups
with resonant character.
Furthermore, this behavior was confirmed by peaks at
620–1193 cm−1, which corresponded to M O and C C–C vibrations
and gave important information about the formation of a bidentate
species, a cycle with an acetylacetone framework in coordination
with the metal. An important difference observed in the spectra is
that from 573 K, the peaks arising from the formation of a bidentate
complex with keto-enolic equilibrium have almost missed. Peaks
below 610 cm−1 indicated M–O bond vibrations [26,27]. Based on
3.3. Surface morphology
The SEM technique was used in order to observe the morphol-
ogy of the different coatings. Fig. 3a–f illustrates the micrographical
comparisons between the morphological properties of the ZrO2 and
YSZ compounds during the heat treatment processes. SEM charac-
terizations of the series have shown that each deposit cracked after

M.A. Domínguez-Crespo et al. / Electrochimica Acta 54 (2009) 2932–2940
2935
Fig. 3. SEM images for coated samples and thermally treated at (a) ZrO2-473 K, (b) ZrO2-573 K, (c) ZrO2-673 K, (d) YSZ-473, (e) YSZ-573 and (f) YSZ-673.
heat treatment. This cracking phenomenon is partly due to the quick
densification of the ZrO2 and YSZ materials during the thermal
treatment and to the large difference of the expansion coefficients
between the ceramic coating and the ferrous substrate.
shape. These may be characterized by the low internal pressure due
to the decomposed organic materials, relatively filled cracks, and
micro-space volume (see inset figures). Structures almost disap-
pear with the temperature of treatment, so they can be attributed
to the elimination of organic compounds (Fig. 3c–f). In addition, the
thickness of the coating is strongly dependent on the thermal pro-
cess affecting the solidification, which provokes low fragmentation
during film growth.
Calcination leads to the diminishing of the cracks after the
removal of the organic compounds. Heterogeneities (length and
thickness) were also more important during heat treatments
at 473–673 K. Some appreciable differences were emphasized
between the two systems; for example, at low temperatures (Fig. 3a
and b), the topographic structure of the surface coating mate-
rial with ZrO2 produced numerous micro-cracks, which seemed
to affect the uniformity and adhesion of the thin film while the
YSZ system produced both micro-cracks and micro-globule struc-
tures on the surface texture (see inset and Fig. 3d). However, from
macroscopic observations, no difference was in evidence concern-
ing the adherence of the coating, and the samples are easy to handle.
The observed globules seem equally sized, with a slightly elongated
3.4. Electrochemical measurements
3.4.1. Ecorr vs. time
The ceramic systems were analyzed by electrochemical mea-
surements in order to determine the protective properties, first as
top coat and then as bond coat, of commercial carbon steel. The
variation of free corrosion potential (Ecorr) as a function of immer-
sion time for the bare steel and sol–gel coated specimens (ZrO2 and

2936
M.A. Domínguez-Crespo et al. / Electrochimica Acta 54 (2009) 2932–2940
Fig. 4. Variation of corrosion potential vs. time in a 3.5 wt% NaCl solution at different
Fig. 5. Tafel plots for the bare steel and the sol–gel coated specimens with ZrO2 or
thermal treatments.
YSZ sintered at 473, 573 and 673 K in a 3.5 wt% NaCl solution at room temperature.
YSZ), sintered at 473, 573 and 673 K in 3.5% NaCl solution at room
temperature, is presented in Fig. 4. It was observed that both sys-
tems were modified during immersion time. The observed trend of
variation of the Ecorr vs. time is almost similar for all the sol–gel
coated specimens. The Ecorr reduced gradually with the exposure
time and attained a steady state after 500 s of immersion which
indicated a good chemical stability for the oxide coatings. Compar-
ing coated samples with bare steel, a noble shift can be observed
in all YSZ coatings, while as-coated samples of ZrO2 show a nega-
tive displacement at temperatures of 473–573 K, and at 673 K shift
towards more positive potentials than bare steel. More positive
Ecorr values for the sol–gel coated specimens indicate their supe-
rior corrosion resistance in the open circuit potential conditions.
Occurrence of significantly higher positive Ecorr value for the sol–gel
coated specimens sintered at 673 K (YSZ) attributes its higher cor-
rosion resistance across specimens YSZ-473, YSZ-573, ZrO2-673,
ZrO2-573 and ZrO2-473. Ecorr was found to increase with an increase
in the temperature and a maximum shift was observed for YSZ-
673 K coating with a coating thickness of 1.0 ␮m. The enhanced
values of Ecorr and the attainment potential can be attributed to
the bifunctional nature of zirconia and yttria-stabilized zirconia
(cubic phase) coatings on the metal surface. Zirconia is a well known
solid electrolyte which transports oxygen ions material as a bar-
rier against oxidation. However, it has the advantage of a thermal
expansion coefficient close to that of stainless steel. Miyazawa et al.
[28] reported that the oxidation rate of stainless steel coated with
ZrO2–Y2O3 layers decreased with increasing Y2O3 content. They
suggested that the effect could be explained by the trapping of oxy-
gen by yttria, or by the diffusion of yttrium, known to be an element
for chromia scale-forming steels that reacts into the substrate. The
YSZ cubic phase can avoid oxygen diffusion, shifting the corrosion
potential to more positive values.
coatings, whereas the cathodic branches appear reduced by the
control of diffusion. The reduction of Tafel slopes in coated sam-
ples no longer dominated by oxygen reduction, with the shape of
the curve resembling the hydrogen evolution reaction, suggests
that the presence of yttria and/or zirconia coatings have a sig-
nificant effect on the cathodic kinetics controlling the corrosion
mechanism. In particular, YSZ coatings displayed a displacement
of corrosion potential Ecorr toward to more noble values as well as
lower current densities than that for bare steel, while ZrO2 shifted
in a negative direction at the range of 473–573 K. At 673 K, the
coated sample exhibited a response close to bare steel. Corrosion
current densities on the order of 10−3.5 to 10−3 mA cm−2 could
indicate the formation of resilient films with good barrier prop-
erties for YSZ, which are comparable with other studies [24,29]. It
can also be noticed that YSZ coated specimens sintered at 673 K
showed a minimum icorr value among the sol–gel coated samples.
This value was found to be of an order of magnitude lower than
that exhibited for bare steel. This behavior was shown by YSZ-573
and YSZ-473 specimens. The current measurement, as presented in
this figure, is the result of the overall current flow due to the dif-
fusion of oxidized metal ions from the substrate to the electrolyte
via metal/coating interface. Cracks and defects in the coating may
have increased the diffusion rate in the terms of increased current
density [29]. In this manner, the morphology of the coated samples
sintered at 473 and even at 573 K revealed the presence of several
cracks that may enhance the diffusion of the electrolyte through
the coating at the metal/coating interface. On the contrary, the YSZ
coated samples heat treated at 673 K shows a denser microstruc-
ture originated by organic compounds elimination. We believe that
this process reduces micro-cracks, increases crystallization of the
cubic YSZ phase, and, as a consequence, enhances its barrier prop-
erties. This conclusion is supported by FT-IR measurements (Fig. 1),
where it was observed that at higher temperatures the organic com-
3.4.2. Tafel plots
The corrosion current, icorr, and corrosion potential, Ecorr, were
measured on the polarisation curves represented in Fig. 5 by using
Tafel extrapolation methods. These parameters were obtained in
the potential range of 150 mV from OCP and are listed in Table 2.
The polarisation measurements were started after the attainment
of steady state at OCP conditions. As a reference, the curve cor-
responding to untreated sample has been included. It is worth
emphasizing the change in the cathodic slope form in the bare
steel samples, apparently controlled by diffusion, to values close
Table 2
Electrochemical parameters from Tafel curves for the different coating electrodes.
Electrode
Ecorr (mV)
icorr (A cm−2
)
Corr. rate (mm/year)
Bare steel
ZrO2-473 K
ZrO2-573 K
ZrO2-673 K
YSZ-473 K
YSZ-573 K
YSZ-673 K
−712.01
−805.04
−808.35
−653.52
−452.43
−429.89
−375.58
6.45E−06
2.57E−05
1.86E−05
2.04E−05
2.45E−06
3.54E−06
8.70E−07
0.07553
0.30071
0.21784
0.23886
0.02871
0.04151
0.01018
to −150 mV dec−1, while anodic slopes were around 110 mV dec−1
.
The anodic curves are almost unaffected by the presence of ceramic

M.A. Domínguez-Crespo et al. / Electrochimica Acta 54 (2009) 2932–2940
2937
pounds almost disappeared favoring a denser film, which can help
to enhance the anticorrosion properties of the samples. Interest-
ingly, the surface coated with ZrO2 displayed very poor protection
against corrosion, with values even lower than those observed for
bare steel. Based on this observation, it can be suggested that sur-
face treatments through ZrO2 have a detrimental effect on corrosion
resistance. This behavior may be attributed to the metastability of
zirconia, which transforms from cubic to tetragonal and to a mono-
clinic phase when it has not been stabilized which leads to volume
changes [30], thus affecting the resistance to corrosion. Cubic YSZ
acts as a barrier to the metal ion release from the metallic sub-
strates. This is attributed to the substantial role of the formation
of strong, intact, and pore free YSZ coatings [31]. In addition, it is
known that in most cases the coatings act as barriers to diffusion,
lowering the oxidation rate of the substrate. Therefore, thicker lay-
ers could offer better protection; however, it has been found that
when the thickness exceeds a certain value the behavior reverses
[32]. This is certainly due to cracking of the coating that occurs
above a critical thickness. Subsequently, the increase in the oxida-
tion rate due to cracks is likely to be proportional to the area of
the unprotected substrate surface. Thus, the presence of tiny and
localized cracks should not significantly change the overall protec-
tion offered by the layer. Instead, this should dramatically decrease
when extensive cracking and delamination is present in the film
or occurs during exposure to high temperature, which may explain
our electrochemical results.
Fig. 6. Nyquist plot and equivalent circuit for the bare steel and the sol–gel coated
specimens sintered at different temperatures in a 3.5 wt% NaCl solution at room
temperature.
3.4.3. Electrochemical impedance spectroscopy (EIS)
EIS analysis was carried out for bare steel and both sol–gel sys-
tems (YSZ and ZrO2) sintered at 473, 573 and 673 K, after 10 min
of immersion in order to detect the evolution of their protective
properties, as well as the corrosion mechanism. This is a powerful,
non-destructive method for investigating corrosion and electro-
chemical phenomena on metal samples coated by sol–gel. The
observed data for the above specimens are presented in Nyquist
plots (Fig. 6). The inset figure also shows enlarged diagrams in the
high frequency range. Different parameters related to impedance
measurement, derived by curve fitting method, are given in Table 3.
An equivalent circuit, consisting of resistors connected in a series
to a parallely connected resistor and capacitor, has been used for
data analysis. A simplistic circuit was applied in these conditions to
extract the impedance data because of the observation of a single
semicircle in the Nyquist plot (Fig. 7a). In this equivalent circuit,
Rs is the solution resistance, Rcracks is the resistance of the sol–gel
coating with cracks while CPEcoat models the intact layer, Rox and
CPEox are the resistance and capacitance of an intermediate layer.
Here, a constant phase element (CPE) is commonly used to rep-
resent capacitance because it is hardly pure capacitance in the real
electrochemical process. It is defined by the admittance Y and pow-
der index number n, given by Y = Yo(jω)n; this is a general formula,
if n = 0 it stands for resistance, while it is capacitance if n = 1, repre-
senting the capacitive behavior of the interface. Generally, ceramic
treatments improve the corrosion resistance due to the formation
of protective oxides that act as a barrier to oxygen diffusion to the
Fig. 7. Equivalent circuit of (a) ceramic coatings after 10 min immersion in NaCl
solution and (b) after 1–3 days’ immersion in NaCl solution.
Table 3
EIS fitting parameters relative to the sol–gel layers.
Electrode
Rs (ꢂ cm2)
CPEcoat
Rcracks (ꢂ cm2)
CPEox
Rox (ꢂ cm2)
CPEdl
Rct (ꢂ cm2)
Yo (F cm−2
)
n
Yo (F cm−2
)
n
Yo (F cm−2
)
n
Bare steel
ZrO2-473 K
ZrO2-573 K
ZrO2-673 K
YSZ-473 K
YSZ-573 K
YSZ-673 K
23.1
19.4
19.9
20.1
26.5
27.3
28.3
–
–
–
–
–
–
1225
2.23E−05
0.60
–
–
–
–
1245
1.34E−5
0.51E−5
0.35E−4
0.39E−4
6.30E−5
3.40E−4
0.59
0.75
0.84
0.65
0.71
0.82
603.6
391.3
817.3
433.2
332.1
195.2
1.51E−5
0.31E−4
0.13E−4
3.35E−4
0.39E−4
1.50E−3
0.61
0.62
0.64
0.68
0.80
0.87
–
–
–
–
–
–
–
–
–
–
–
–
956.2
1304.9
1837.1
3525.1
4145.8
–
–

2938
M.A. Domínguez-Crespo et al. / Electrochimica Acta 54 (2009) 2932–2940
metal surface. From the Nyquist plot and fitting parameters, it is
clear that the samples coated with YSZ and thermal treated at 573
and 673 K, respectively, displayed impedance values three and four
times higher than those presented by the bare steel, with an Rp
ranged from approximately 3857–4338 ꢂ cm2. However, the visual
aspect of YSZ-573 K specimens indicates that the film did not have
as good adherence as YSZ-673 K coatings. The arc in the Nyquist plot
of 10 min immersion relates to the explanation of the outer layer
ceramic coating and its defects. The information for the inner layer
ceramic-substrate is not shown in this case. The reason may be that
the outer layer of ceramic coating is not corroded and the inner layer
ceramic-substrate possesses a high resistance. According to the EIS
results, the evolution of YSZ as a top coat showed that samples
sintered at 673 K displayed an improvement in protection against
corrosion. Occurrence of lower Rcracks and high capacitance for ZrO2
and YSZ-473 K, as compared to the YSZ-673 samples, indicates the
presence of a defective layer of corrosion product on its surface,
which may be attributed to the continuous diffusion of chloride ion
at the metal/layer interface and subsequently causes a considerable
decrease of Rcracks and increase of CPEcoat. In Bode plots (not shown
here), three different frequency regions were observed with respect
to impedance (Z) and phase angle (ꢁ): a high frequency region
(>103 Hz) indicating the resistance solution properties, a medium
frequency region (1–103 Hz) indicating the capacitive behavior, and
a low frequency region (<1 Hz) explaining the charge transfer pro-
cess occurring at the solution/coating interface [33]. In our study
samples heat treated at 673 K displayed a capacitive behavior with
a high value of phase angle (−68◦) between the low and middle
frequency regions, which can be correlated with the insulating
and compactness properties of the coating [34,35]. In spite of the
obtained results at this temperature, the specimens failed as con-
tinuous immersion time increased shown in the impedance spectra
of Fig. 8. After the first day of testing, the response behavior was
different, with the EIS results containing a depressed semicircle
followed by what appeared to be a second semicircle, indicating
that the electrolyte had reached the metal base (while in the initial
period only one semicircle was observed). The first arc is relatively
small and the second relatively large. The electrolyte’s diffusion to
the substrate increased with the immersion time, and after 3 days
the outer layer of ceramic coating could be seen to have lost its
anticorrosive property.
The presence of cracks, as observed for the sol–gel coated speci-
mens sintered at the different temperatures, increases the diffusion
of electrolytes at the metal/solution interface through coating when
exposed in continuous immersion, resulting in the decrease of cor-
rosion resistance. In this case, the equivalent circuit consists of Rs,
resistance of solution, CPEcoat which describes the behavior of the
intact ceramic layer, and Rcracks is related to the defect; a mesh
describing the substrate (CPEdl and Rct) is in series with CPEox as has
been reported by Zheludkevich, Andreata and coworkers [36,37].
CPEox describes a network of electrolyte resistances and double
layer capacitors in the pores of the layer [38,39], while Rox is related
to the barrier properties of the film. Based on the circuit shown
in Fig. 7b, charge transfer processes take place on the substrate,
which is covered with sol–gel layer. As a result of the proposed
model being based on the physical model of the corrosion processes,
the sol–gel films produced in this work, being thin and having a
cracked microstructure, cannot protect the steel substrate as a bar-
rier when used as a top coat, and their efficiency could be regarded
as merely an improvement to steel coating adhesive bonding [24].
The present EIS results are in good agreement with the corrosion
rate measured by the Tafel plots and also in good accordance with
the microstructural studies.
Fig. 8. Impedance spectra of YSZ thin films sintered at 673 K and tested in continuous
immersion in a 3.5 wt% NaCl solution.
this reason, EIS measurements were then performed on samples
protected with a commercial polyurethane varnish in order to eval-
uate the adhesion between polyester and pre-treated substrates
and to correlate these data with those obtained from adhesion tests
[40,41]. The characterization of these systems is very important,
since two aspects can be analyzed: the barrier effect of the paint
and the role of the pre-treatment on the corrosion behavior of the
coated substrate. It is known that, in the absence of defects, the coat-
ing essentially behaves as a physical barrier between the aggressive
electrolyte and the metal. Usually a very good coating (uniform
and thick) behaves as an insulator showing very high resistance
(Gꢂ m2) [43]. Moreover, very low capacitances are usually obtained
and the phase angle of the impedance plots is around −90◦ over the
measured frequency range. In this situation, a long exposure time
is necessary to allow corrosion onset.
For this reason, in the present study a low quality organic coating
is used in order to reduce the experiment time and to facilitate the
access of aggressive species to the metallic substrate, and thus to
better understand the role of the pre-treatments in the presence of
paint coating.
After continuous immersion, the impedance spectra for YSZ
coatings reveal the presence of a one-time constant in the fre-
quency range 4 × 103–0.1 Hz, which can be ascribed to the organic
coating in combination with the intermediate layer capacitance
(sol–gel film); see Fig. 9a–c. This intermediate layer may be com-
posed of natural iron oxide and Fe–O–Zr and/or Fe–O–Y and/or
Fe–O–Zr–O–Y bonds. In spite of the low quality of organic paint
used in this work to allow us to evaluate differences existing among
EIS has also been proposed for several projects as a useful tool for
assessing both corrosion resistance and coating adhesion [40–42].
Moreover, this is the main interest in industrial applications. For

M.A. Domínguez-Crespo et al. / Electrochimica Acta 54 (2009) 2932–2940
2939
Table 4
Pull-off results.
Sample
Bond strength values (MPa)
Bare steel
ZrO2-473 K
ZrO2-573 K
ZrO2-673 K
YSZ-473 K
YSZ-573 K
YSZ-673 K
2.2
2.8
3.2
8.3
2.5
2.8
12.4
3.5. Pull-off tests
Because too long a testing time was required to obtain accurate
qualitative results in the samples thermally treated at higher tem-
peratures, it was decided to evaluate the adhesion of the organic
top coat in combination with sol–gel coatings to the metallic sub-
strates by performing pull-off tests. At the molecular level, high
adhesion strength is possible with sol–gel film systems bonding at
the interface between the organic molecules and the free hydroxyl
groups at the metal oxide surface. In this way, adhesion is quan-
tified in terms of the forces employed to detach the test dollies
glued to the paint film from the underlying metal. The adhesion
values determined in the pull-off tests are listed in Table 4. As seen
in the table, the average pull-off strength of the polymer varnish
to the bare steel is about 2.2 MPa, and it increased up to 8.3 and
12.4 MPa when the commercial carbon steel (AISI-1012) was coated
with ZrO2 and YSZ, respectively, and followed by sintering at 673 K.
In general, the failure took place at the coating-substrate interface,
and the paint was almost completely removed from the substrate
(around 90% of detached area). Samples without treatment showed
failures of about 70% of detached area. The results suggest that
the adhesion between polymer varnish and metallic substrate was
improved by using sol–gel coatings, but the temperature of treat-
ment, together with the oxide thickening processes, plays a vital
role in both the mechanism of corrosion protection in aggressive
media and the adhesion properties of carbon steel. The observed
mechanical strength of YSZ depends on the phase transformation
from the tetragonal to the monoclinic or cubic-monoclinic phase
affecting the morphology achieved during the sol–gel process, even
though this morphology could also cause a loss in the anticorrosion
properties, as has been discussed in the electrochemical measure-
ments.
4. Conclusions
The present study shows that sol–gel YSZ coatings on commer-
cial carbon steel provide corrosion protection and may be used
as alternative pre-treatments for commercial carbon steel. On the
basis of the present investigations, the following conclusions can
be drawn:
•
Fig. 9. Total impedance of YSZ coatings thermal treated at different temperatures
and organic top coat, tested in continuous immersion in a 3.5 wt% NaCl solution until
the barrier properties failed.
The sol–gel coatings are strongly dependent on the heat treat-
ment conditions, because the process requires the condensation
reaction between substrate and sols, and also among the sols,
which affects the diffusion of electrolytes at the metal/solution
interface through coating and can result in the decrease or
increase of corrosion resistance.
the pre-treatments and the influence of the temperature of treat-
ment, a capacitive behavior in all the samples was observed during
long testing times. However, comparison of the pre-treated speci-
mens with bare steel indicates that the difference in the capacitive
behavior can be correlated with the sol–gel coatings increasing
the barrier properties. Interestingly, YSZ coatings sintered at 573
and 673 K displayed total impedance data close to 1010 ꢂ cm2 and
remained fairly constant over 19 days, while for pre-treated sam-
ples with YSZ at 473 K, after only 4 days an abrupt decaying in the
protective properties was observed.
•
The complete removal of the organic groups from the sol–gel coat-
ings sintered at 673 K, as demonstrated by FT-IR analysis and the
mismatch of expansion coefficients, appears to be the reasons for
the development of a high crack density and heterogeneities in
the coating at low temperatures. Decrease of sintering tempera-
ture beyond 573 K appears to be detrimental to the corrosion and
adhesion performance of the coating, due to the organic com-
pounds that remain in the coating. This is the main reason that

2940
M.A. Domínguez-Crespo et al. / Electrochimica Acta 54 (2009) 2932–2940
the sol–gel ZrO2 coating sintered at 673 K has exhibited a very
poor corrosion resistance and adhesion, near to that of a bare
metals substrate.
[12] A.R. Di Giampaolo, M. Medina, R. Reyes, M. Velez, Surf. Coat. Technol. 89 (1997)
31.
[13] M. Fallet, H. Mahdjoub, B. Gautier, J.P. Bauer, J. Non-Cryst. Solids 293–295 (2001)
527.
•
•
Sol–gel ZrO2 and YSZ thin films used as a top coat have not
improved corrosion resistance to aggressive environments by bar-
rier action because the protective properties failed after a few
hours in continuous immersion.
The electrochemical and adhesion measurements with organic
coatings suggest that increases in the corrosion and adhesion
properties of sol–gel YSZ coatings on carbon steel sintered at
greater temperatures also may be due to the combination of
the stabilization of the cubic phase, the complete elimination of
organic compounds, and the crystallinity of the film. Therefore,
sol–gel coatings using YSZ and sintered at 673 K have shown max-
imum total impedance, as compared to sol–gel coatings sintered
at 573 and 473 K.
[14] G. Li, L. Niu, J. Lian, Z. Jiang, Surf. Coat. Technol. 176 (2004) 215.
[15] Y.F. Cheng, J.L. Luo, Electrochim. Acta 44 (1999) 4795.
[16] L.M. Nikolicˇı, L. Radonjicˇı, V.V. Srdic, Ceram. Int. 31 (2005) 26.
[17] A. Pardo, M.C. Merino, A.E. Coy, F. Viejo, R. Arrabal, E. Matykina, Corros. Sci. 50
(2008) 780.
[18] H. Endo, E. Marui, Wear 257 (2004) 80.
[19] G. Shanmugavelayutham, S. Yano, A. Kobayashi, Vacuum 80 (2006) 1336.
[20] J.H. Ouyang, S. Sasaki, K. Umeda, Surf. Coat. Technol. 154 (2002) 131.
[21] A. Morikawa, T. Suzuki, T. Kanazawa, K. Kikuta, A. Suda, H. Shinjo, Appl. Catal.
B: Environ. 78 (2008) 210.
[22] A.M. Torres-Huerta, J.R. Vargas-García, M.A. Domínguez-Crespo, Solid State Ion-
ics 178 (2007) 1608.
[23] E. Celik, Y. Islamoglu, Y. Akin, Y.S. Hascicek, Mater. Des. 24 (2003) 543.
[24] L. Fedrizzi, F.J. Rodriguez, S. Rossi, F. Deflorian, R. Di Maggio, Electrochim. Acta
46 (2001) 3715.
[25] Standard Practice for: Conducting Potentiodynamic Polarization Resistance
Measurements, ASTM G59-97, American Society for Testing and Materials, West
Conshohocken, PA, 1997.
•
Finally, adhesion results indicate that thin YSZ films can behave
better than commercial pre-treatments at promoting adhesion.
[26] L.F. Cueto, E. Sánchez, L.M. Torres-Martínez, G.A. Hirata, Mater. Charact. 55
(2005) 263.
[27] Z.W. Quan, L.S. Wang, J. Lin, Mater. Res. Bull. 40 (2005) 810.
[28] K. Miyazawa, K. Suzuki, M.Y. Wey, J. Am. Ceram. Soc. 78 (1995) 347.
[29] G. Ruhi, O.P. Modi, A.S.k. Sinha, I.B. Singh, Corros. Sci. 50 (2008) 639.
[30] R. McPherson, Surf. Coat. Technol. 39/40 (1989) 173.
[31] A. Balamurugan, G. Balossier, S. Kannan, J. Michel, J. Faure, S. Rajeswari, Ceram.
Int. 33 (2007) 605.
Acknowledgements
The authors wish to thank the financial support from SIP-IPN
and SEP- CONACYT through the projects 20080805, 20080654 and
2006-61354, respectively.
[32] S.L. Ishaug, M.J. Yaszemski, R. Bizios, A.G. Mikos, J. Biomed. Mater. Res. 28 (1994)
1445.
[33] J. Gallardo, A. Duran, J.J. de Danberenea, Corros. Sci. 46 (2004) 795.
[34] M. Metikos-Hukovic, E. Tkalcec, A. Kwokal, J. Piljac, Surf. Coat. Technol. 165
(2003) 40.
References
[35] G. Ruhi, O.P. Modi, A.H. Yegneswaran, A.K. Jha, I.B. Singh, Surf. Coat. Technol.
201 (2006) 1866.
[36] M.L. Zheludkevich, R. Serra, M.F. Montemor, I.M. Miranda Salvado, M.G.S. Fer-
reira, Surf. Coat. Technol. 200 (2006) 3084.
[37] F. Andreatta, P. Aldighieri, L. Paussa, R. Di Maggio, S. Rossi, L. Fedrizzi, Elec-
trochim. Acta 52 (2007) 7545.
[38] P. Campestrini, H. Terryn, J. Vereecken, J.H.W. de Wit, J. Electrochem. Soc. 151
(2004) B370.
[39] G. Goeminne, H. Terryn, J. Vereecken, Electrochim. Acta 40 (1995) 479.
[40] L. Fedrizzi, F. Deflorian, P.L. Bonora, Electrochim. Acta 42 (1997) 969.
[41] L. Fedrizzi, F. Deflorian, C. Ceol, Corrosion protection by organic coatings III, in:
The Electrochemical Society 97-41, 1998, p. 122.
[42] P. Innocenzi, M. Guglielmi, M. Gobbin, P. Colombo, J. Eur. Ceram. Soc. 10 (1992)
431.
[1] J.R. Martínez, S. Palomares-Sánchez, G. Ortega-Zarzosa, F. Ruiz, Y. Chumakov,
Mater. Lett. 60 (2006) 3526.
[2] A. Subramania, T. Saradha, S. Muzhumathi, J. Power Sources 165 (2007) 728.
[3] M. Luo, K. Cheng, W. Weng, Ch. Song, P. Du, G. Shen, G. Xu, G. Han, Mater. Lett.
62 (2008) 1965.
[4] P. Du, S.S. Wang, H. Chen, Z. Wang, J.C. Sun, Z. Han, W. Schmidt, H.W. Neumuller,
Phys. C: Superconduct. 463–465 (2007) 580.
[5] M. Seo, Y. Akutsu, H. Kagemoto, Ceram. Int. 33 (2007) 625.
[6] H.M.H. Fadlalla, C.C. Tang, E.M. Elssfah, F. Shi, Mater. Chem. Phys. 109 (2008)
436.
[7] Y. Rezgui, M. Guemini, Appl. Catal. A: Gen. 335 (2008) 103.
[8] S.V. Lamaka, M.F. Montemor, A.F. Galio, M.L. Zheludkevich, C. Trindade, L.F. Dick,
M.G.S. Ferreira, Electrochim. Acta 53 (2008) 4773.
[9] J.G. Che˛cmanowski, B. Szczygieł, J. Non-Cryst. Solids 354 (2008) 1786.
[10] H. Yun, J. Li, H.-B. Chen, Ch.-J. Lin, Electrochim. Acta 52 (2007) 6679.
[11] A. Pepe, P. Galliano, M. Aparicio, A. Durán, S. Ceré, Surf. Coat. Technol. 200 (2006)
3486.
[43] M.G.S. Ferrerira, R.G. Duarte, M.F. Motremor, A.M.P. Simoes, Electrochim. Acta
49 (2004) 2927.