Progress in Organic Coatings 76 (2013) 1273–1278

Contents lists available at SciVerse ScienceDirect

Progress in Organic Coatings

journal homepage: www.elsevier.com/locate/porgcoat

Performance evaluation of alkyd coatings for corrosion protection in urban and

industrial environments
F. Cadena a , L. Irusta b , M.J. Fernandez-Berridi b,∗
a
Centro de Investigaciones Aplicadas a Polímeros, Facultad de Ingeniería Química y Agroindustria, Escuela Politécnica Nacional, Quito, Ecuador
b
Polymat, Department of Polymer Science and Technology, Faculty of Chemistry, UPV/EHU, San Sebastian, Spain

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

Article history: The performance of alkyd based coatings exposed in two different sites in Ecuador, urban and indus-
Received 11 September 2012 trial locations, for corrosion protection was evaluated. Atmospheric test sites and corrosion resistance
Received in revised form 14 January 2013 of coatings were examined using mainly ISO and ASTM standards. The alkyd resin degradation was
Accepted 28 March 2013
characterized by different methods such as FTIR-ATR, DSC, TGA and SEM.
Available online 20 April 2013
The studies showed that the corrosion resistance and chemical structural changes were more evident
for those samples exposed in the industrial environment although the sun radiation was lower than that
Keywords:
measured in the urban location.
Alkyd coating
Corrosion © 2013 Elsevier B.V. All rights reserved.
Outdoor exposure
FTIR-ATR
DSC
TGA

1. Introduction of climates and the inherent variability of weather. Therefore, to

It is well known that atmospheric corrosion is responsible for
about 50% of the total economic loss when metal structures and
equipment are exposed outdoors.
Organic coatings are the most common, versatile and cost
effective mode of corrosion protection for metallic objects and
structures [1,2]. Some scientific aspects of corrosion protection by
coatings are understood but many are not, for example the exact
cause and nature of coating failure is still up for discussion [3–8].
There are many factors that determine the duration of the coating,
among which we can mention its nature, composition and thick-
ness as well as the nature and substrate preparation, the method
of coating and the environment [9–11].
Outdoor degradation of organic coatings is a complex phe-
nomenon influenced by humidity, solar radiation, temperature,
pollution and many other parameters [12,13]. Practical perfor-
mance of coatings is usually investigated by both laboratory tests
or/and outdoor exposure. Although there are significant advances
in the development of laboratory tests [14–21], natural weathering
is the most reliable exposure method for long-term performance
examination of organic coatings [13,19]. However, these types of
tests are extremely difficult to establish due to the wide variety
∗ Corresponding author. Tel.: +34 943018194; fax: +34 943015270.
E-mail address: mj.fernandezberridi@ehu.es (M.J. Fernandez-Berridi).
develop hypotheses linking performance with composition, it is
necessary to quantify the “environmental stress” test (meteorolog-
ical data, pollution and the corrosion categories of the test sites)
and detect and analyze the mechanism of formation of defects in
the coating. These data are a useful starting point for the design of
new coatings and can also be used to evaluate laboratory tests [19].
Advances in nanoscience and technology have led to improved
corrosion protection coatings, which is attributed to the small size
of the particles and interactions with binders. This provides an
important alternative to the employment of toxic anticorrosive
pigments [22–35].
This work analyzes the behavior of a pigment-free alkyd resin
and a coating formulated with the same resin containing TiO2
nanoparticles and fillers exposed at two different locations in
Ecuador under urban and industrial environments, Quito city
(78◦ 29 24 W, 0◦ 10 48 S) and Esmeraldas (79◦ 40 47 W, 0◦ 55 48 N)
respectively. The atmospheric test sites and the corrosion resis-
tance of coatings were examined using mainly ISO and ASTM
standards (also MEB-EDX). The alkyd resin degradation was char-
acterized by different methods such as FTIR-ATR, DSC, TGA and SEM
in order to elucidate its role in the failure of the coating.
An alkyd resin based coating was selected because these resins
are still common components in the formulation of paints [26,27]
and, particularly, because their own formulations involve the use of
natural oils that are relevant for developing countries like Ecuador
[2,28–36].

0300-9440/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.porgcoat.2013.03.017
1274 F. Cadena et al. / Progress in Organic Coatings 76 (2013) 1273–1278

2. Experimental parts Table 1

Categorization of test sites according to ISO standards.

2.1. Materials Parameter Quito Esmeraldas

Exp. ISO st. Exp. ISO st.

2.1.1. Metallic substrate
[SO2 ] (␮g/m3 ) 10.5 P0 37 P1
Low carbon steel specimens (100 × 150 × 0.9 mm), (ASTM 283,
[Cl− ] (mg/m2 day) – S0 16.4 S1
grade C) were cleaned with methyl ethyl ketone (MEK) and stored TOW (h) 3500 t4 6100 t4
in a dried atmosphere. A total of 12 samples was used, six of bare Steel corrosion rate (␮m/year) 14.6 C2 56.5 C4
metal and the other six to deposit the coatings.
Parameter Quito Esmeraldas
◦
Temperature ( C) 14 27
2.1.2. Coatings Sedimentary particles (g/m2 month) 3.75 1.76
Protective coatings were formulated from a commercial long oil Precipitation (mm/year) 2800 700
soya alkyd resin with pigments (1.5% TiO2 nanopowder <100 nm Radiation (w/m2 ) 260 190
[NO2 ] (mg/m3 ) 32 17
particle size, mixture of rutile and anatase) and extenders (8.5%
of calcium carbonate and silica). As reference, a coating without TOW, time of wetness.
pigments was prepared.
The coatings were air-sprayed onto both sides of the metal-
lic substrates following the manufacturer’s instructions. After the 2.6. Characterization of the degradation of coatings
application, the coatings were allowed to cure for seven days. The
edges of the specimens were protected with a thick epoxy resin Physicochemical characterization and coating degradation
coat in order to avoid the corrosion process starting at these points. studies were carried out by FTIR microspectroscopy (Nicolet 6700
The final thickness was 60 ± 7 ␮m measured with an Elcometer FTIR, Nicolet Continuum microscope) using the ATR technique (Ge
apparatus (model 345 Elcometers Instruments). crystal). The spectra were collected at a resolution of 4 cm−1 with
an averaging of 64 scans.
DSC studies were performed with a TA Instruments Q2000 under
2.2. Field-test procedure the following experimental conditions: Temperature range -40 to
200 ◦ C, with a ramp of 10◦ /min. Two scans were carried out for each
Two different sites in Ecuador were chosen to carry out the field sample.
tests. The first one is located in an urban environment of Quito Thermogravimetric data were collected on a thermobalance TA
(Metropolitan area 78◦ 29 24 W, 0◦ 10 48 S at 2835 meters above Instruments TGA-Q500 from 50 to 800 ◦ C under nitrogen atmo-
sea level). The second one was selected on the coast close to the sphere at a flow rate of 80 mL/min. For some samples, air was
Esmeraldas Petroleum Refinery (an industrial zone) (79◦ 40 47 W, introduced at 800 ◦ C and an isotherm for 5 min was employed. The
0◦ 55 48 N). heating rate was 10◦ /min.
For the study, three bare and three coated samples were allo-
cated at the stations with a 45◦ inclination in accordance with ASTM
G1 [37], G7 [38] and G50 [39] standards. 3. Results

3.1. Characterization of the test sites

2.3. Meteorological and pollution data
The meteorological and pollution data, measured with auto-
matic detection stations, was provided by the “Secretaría de
Ambiente del Municipio del Distrito Metropolitano de Quito” and
the “Instituto Nacional de Metorología e Hidrología” (INHAMI) from
Esmeraldas. The atmospheric chloride deposition rate was mea-
sured following the wet candle method ASTM G140 [40], and the
atmospheric SO2 was monitored using the sulfation plate technique
ASTM G91 [41].
2.4. Corrosion evaluation techniques for bare samples
Morphological aspects of rusting for bare samples were studied
by optical microscopy (Thomas Scientific 1230CM), and the weight
loss registered in accordance with ASTM D610 [42] standard.
The corrosion products were analyzed by X-ray powder diffrac-
tion (Siemens D-500, with Cu-Ka radiation of k 1/4 0.154 nm).
The results of the characterization of the test sites are summa-
rized in Table 1. In accordance with ISO 9223 standard [43], the
urban station environment corresponds to category 2 (weak corro-
sivity, between 1.3 and 25 ␮m), while the industrial station can be
classified as category 4 (high corrosivity between 50 and 80 ␮m). At
Quito station the specimen rusting was more uniform, compact and
adherent than in Esmeraldas. In both sites, the X-ray power diffrac-
tion results indicated that the specimens showed the presence of
lepidocrecite and goethite. However, the concentration of goethite
was higher for the sample exposed in Esmeraldas as a consequence
of a high corrosion rate in this site.
The higher NO2 concentration values and sedimentary particles
in Quito are due to the fact that the station is located in a very heavy
traffic zone. In addition, we must point out that the sun radiation
is higher in the city of Quito compared with that of Esmeraldas.
Although this factor does not have a direct influence on the corro-
sion rate, it does on the polymeric coating, as will be described later
in the article.

2.5. Corrosion evaluation techniques for coated samples 3.2. Corrosion resistance

The corrosion degree was measured using the ASTM D610
[42] standard and the morphological and chemical aspects of the
coatings were studied by optical microscopy (Thomas Scientific
1230 CM) and SEM in an electronic microscope Hitachi, model
S3000N.
The alkyd coatings showed decoloration, fissures and rusting,
without evidence of any other defects. According to the results
(Table 2), the sample upper faces in both sites showed a higher cor-
rosion attack. As in the bare metal corrosion process, Esmeraldas is
the station where alkyd coating resistance was more affected.
 F. Cadena et al. / Progress in Organic Coatings 76 (2013) 1273–1278 1275

Table 2
Degree of rusting in each station according to ASTM D 610 standard.

Test site Time of exposure Rusting (%)

(month)

Upper face ASTM D610 scale Lower face ASTM D610 scale

Quito 12 0.2 7-S 0.1 8-S

18 0.3 7-S 0.2 7-S
Esmeraldas 12 0.4 6-S 0.2 7-S
18 0.8 6-S 0.5 6-S

3.3. Resin degradation

3.3.1. FTIR spectroscopy

Fig. 1 shows the infrared spectra of the reference alkyd resin as
a function of the exposure time in Quito.
Comparing the registered spectra as a function of time
only slight changes are worth mentioning: the broadening of
the OH stretching vibration band (3400 cm−1 ), a shift toward
lower wavenumbers of the carbonyl stretching vibration band
(1720 cm−1 ) and the appearance of a very small intensity band
at 1630 cm−1 , possibly due to the formation of conjugated double
bonds along the oil chain.
These observations are in good agreement with those reported
in literature [44]. According to the authors, alkyd resin photoox-
idation occurs only at long exposure times through slow and
progressive oxidation of the aliphatic segments of the polymeric
chains giving rise to both crosslinking and chain scissions.
In order to determine if these spectral changes are due to chain
scissions or crosslinking reactions, the alkyd resin exposed 18
months in Quito was extracted with THF and the soluble fraction
was analyzed by FTIR. Fig. 2. FTIR spectra of the sample exposed 18 months in Quito (A) and that of the
soluble fraction (B).
Fig. 2 shows the FTIR spectrum of this soluble fraction together
with that of the sample before the extraction. The same bands
are visible in both spectra but in the spectrum of the soluble in the soluble fraction and the carbonyl stretching band shifts
part the absorptions attributed to the sp3 CH2 show a relatively toward higher wavenumbers, indicating a higher participation of
higher intensity in comparison with those of the ester bands. Such the aliphatic ester groups.
apparent enrichment in methylene groups suggests a process of If we compare the features of the spectra of the exposed samples
fragmentation in which the aliphatic segments are preferentially at both locations (Fig. 3), it is evident that the changes induced by
released. the irradiation are very similar in both places. However, the degree
This hypothesis can be confirmed by the fact that the band at of change is higher for the sample exposed in Esmeraldas. Thus,
740 cm−1 , due to the o.o.p. aromatic bending vibrations decreases in the spectrum of this sample the CH2 stretching vibration bands
decrease in intensity while the band due to the unsatured dou-
ble bonds (1630 cm−1 ) and that of primary alcohols (1050 cm−1 )
increase.

Fig. 1. FTIR spectra of the reference alkyd resin before and after outdoor exposure Fig. 3. FTIR spectra of the unexposed reference sample and samples exposed 18
in Quito for 12 and 18 months. months in Esmeraldas and Quito respectively.
1276 F. Cadena et al. / Progress in Organic Coatings 76 (2013) 1273–1278
Fig. 4. TG and DTG of the alkyd resin.
From the spectral changes, we can conclude the following:
The reduction of the CH2 band intensity indicates that chain
scissions have taken place during the exposure and low molecular
weight aliphatic fragments have been removed by metereological
agents or even volatilized.
The increase of the C C stretching mode evidences the existence
of an extended conjugation structure, which means a severe dam-
age in the coating, especially in the case of the Esmeraldas exposed
sample.
The increase of the band at 1050 cm−1 , assigned to primary
alcohols, can be explained by a hydrolytic process. Once again the
increase is higher in the sample exposed in Esmeraldas.
At first glance this result can be surprising taking into account
that the radiation degree in Esmeraldas is lower than that of Quito.
However, as shown in Table 1, [SO2 ] is higher in Esmeraldas and
according to literature [17], sulfur dioxide is readily converted to
sulfuric acid under the presence of humidity. As isophthalic struc-
tures show great vulnerability to hydrolytic backbone degradation
under acidic conditions, we assume that under Esmeraldas envi-
ronmental conditions, alkyd resins are prone to suffer a higher
extension of hydrolytic reactions. In addition, our results indicate
that under acidic conditions the catalysis of the photooxidation
process can also take place.
3.3.2. Thermogravimetric analysis (TGA)
Fig. 4 shows the TG and DTG of the reference resin. As can be
seen, the alkyd resin degrades in mainly three different steps with
a corresponding weight loss of 10, 68 and 11%, leaving a stable
residue of 11%. Taking into account that this formulation does not
have any mineral filler, the residue must arise as a consequence of
a carbonization process, especially from the aromatic groups.
The sample exposed in Esmeraldas also shows three degrada-
tion steps although the relative percentages of each of them change
slightly as a function of exposure time (Table 3). Thus, the major
changes are observed in the second step, which is the main degra-
dation step, where the weight loss reduces as the exposure time
increases, and in the remaining residue content, which increases
with the exposure time.
Table 3
TGA results of the unexposed and exposed alkyd resins for 12 and 18 months in
Esmeraldas.
Sample 1st step (%) 2nd step (%) 3rd step (%)
Unexposed 10 68 11
12 months 10 67 10
18 months 14 54 13
Fig. 5. DSC thermograms of samples exposed in Esmeraldas.
These results are similar to those of Ploeger et al. [45] who stud-
ied the behavior of different alkyd based paints by this technique
under air conditions. These authors observed a higher residue for
the older samples indicating a larger amount of inorganic filler, due
to the loss of volatile degradation species, and possibly, a higher
crosslinking density compared to unexposed samples. In our case,
as the samples studied do not contain any inorganic filler, the fact
that the residue content increases with exposure time must imply
the loss of volatile aliphatic fragments during the exposure time.
These results are in very good agreement with those obtained by
FTIR where the spectrum of the THF soluble fraction of the exposed
sample in Quito for 18 months showed an overall increase of the
bands attributable to the aliphatic groups.
3.3.3. DSC results
DSC analyses were also performed on the samples exposed in
Esmeraldas (Fig. 5). The unexposed resin showed a Tg at 25 ◦ C (onset
point) and an exothermic transition at about 155 ◦ C. According to
Mallégol et al. [46] this exothermic peak is due to peroxide decom-
position. Despite this decomposition is an endothermic reaction,
the numerous radicals formed in this process can initiate various
radical propagation reactions and recombination which results in
an exothermic total reaction.
The exposed samples show an increase in Tg as a function of
exposure time, while the peroxide peak reduces considerably. The
increase of Tg can be linked to the crosslinking of the exposed films,
which would lead to more rigid systems, and to the loss of aliphatic
moieties, which would decrease the chain mobility. The reduction
of the heat involved in the peroxide decomposition indicates com-
plete autoxidative crosslinking through the films.
3.4. Coating degradation
3.4.1. Infrared study
Fig. 6 shows the infrared spectra of the non-exposed coating
together with the exposed one (both faces) in Esmeraldas for 18
months.
As can be seen, the spectra of the lower face of the exposed sam-
ple and the non-exposed one are very similar, indicating that the
resin has not suffered an important change in its structure. How-
Residue (%) ever, the spectrum of the upper face of the exposed sample shows
11 a very important increase in the intensity of the band centered at
13 about 1000 cm−1 , which can be attributable to the Si O Si stretch-
19
ing vibration band. Taking into account that the spectrum has been
 F. Cadena et al. / Progress in Organic Coatings 76 (2013) 1273–1278 1277

Fig. 7. DSC thermograms for: (A) non exposed coating, (B) coating after 18 months
exposed in Quito and (C) coating after18 months exposed in Esmeraldas.

organic loss mass is higher for the upper face samples and for the
Fig. 6. FTIR spectra of (A) Non exposed coating, (B) Lower face of Esmeraldas exposed
(18 months) sample and (C) Upper face of Esmeraldas exposed (18 months) sample.
sample exposed in Esmeraldas.
These results are in good agreement with the FTIR spectrum
Table 4
of the sample exposed in Esmeraldas in the sense that the band
TG residue (wt.%) of the different coatings under N2 and air conditions. attributable to the presence of the inorganic filler was more intense
than in the case of the non-exposed sample.
Sample Residue (%) under Residue (%) under
N2 conditions air conditions

Non-exposed coating 16 9
3.4.3. DSC of coatings
18 months Quito lower face 17 11 Fig. 7 shows the thermograms for the non-exposed sample and
18 months Quito upper face 21 13 those exposed 18 months in Esmeraldas and Quito respectively.
18 months Esmeraldas lower face 19 13 As in the case of the resin, the Tg values increase with the expo-
18 months Esmeraldas upper face 23 17
sure time. Here again, the increase of the crosslinking reactions as a
consequence of the weathering degradation can be responsible for
obtained by means of the ATR accessory, it is clear that the inor- this result. However, only the non-exposed sample presents a clear
ganic filler is predominantly located at the surface of the coating. exothermic transition attributable to peroxide decomposition.
This observation evidences that the outdoor exposure of the sample
provokes a change in the relative content of the organic/inorganic 3.5. Corrosion mechanism
components in the formulation at least at the upper face surface.
These results corroborate those obtained for the reference resin Fig. 8 shows the micrographs of the non-exposed and exposed
degradation. coatings for 18 months in Esmeraldas. The original coating does not
present any failure evidence, while in the upper face of the exposed
3.4.2. TGA coating different size voids and filler aggregates can be clearly seen.
Table 4 summarizes the TGA results in relation to the remaining The lower face of this coating also presents some defects but in a
residue for all the coatings analyzed. much lower level, in accordance to the corrosion results presented
As can be seen, all the exposed samples, independently of the in Table 2.
thermogravimetric conditions, show a higher residue content than According to these results it is possible to establish a mecha-
the non-exposed coating. Taking into account that all samples were nism whereby the coating fails. The rupture of the covalent bonds
prepared from the same formulation, the residue should be the of the alkyd resin as a consequence of the UV irradiation simulta-
same for all systems, under air conditions. This fact evidences the neously with the diffusion of water, oxygen and pollutants across
loss of organic matter during the weathering process as it is known the coating surface can be considered the beginning of the corrosion
that silica is stable in the temperature range of TGA. In addition, the process. Consequently, zones with a poor adherence are generated,

Fig. 8. SEM Images of non-exposed sample (left), upper face (center) and lower face (right) of samples exposed 18 months in Esmeraldas.
1278 F. Cadena et al. / Progress in Organic Coatings 76 (2013) 1273–1278

which can be considered the prone points to be attacked by elec- References

trochemical corrosion reactions.
[1] C. Scholff, Prog. Org. Coat. 52 (2005) 21–27.
The lixiviation of the aliphatic part of the coating facilitates the [2] C. Deyá, G. Blustein, B. Amo, R. Romagnoli, Prog. Org. Coat. 69 (2010) 1–6.
pigment migration toward the coating surface, which in turn causes [3] H. Oehler, I. Alig, D. Lellinger, M. Bargmann, Prog. Org. Coat. 74 (2012) 719–725.
the development of defects and the enhancement of corrosion reac- [4] P.A. Sørensen, K. Dam-Johansen, C.E. Weinellb, S. Kiil, Prog. Org. Coat. 68 (2010)
283–292.
tions. In addition, the accumulation of pigments, especially on the [5] N.S. Sangaj, V.C. Malshe, Prog. Org. Coat. 50 (2004) 28–39.
upper surface (according to ATR results), makes them easier to [6] J.I. Iribarren, F. Cadena, F. Liesa, Prog. Org. Coat. 52 (2005) 151–160.
be removed by the direct washing action of rain and contaminant [7] D. Mills, S. Mabbutt, G. Bierwagen, Prog. Org. Coat. 46 (2003) 176–181.
[8] G.P. Bierwagen, Prog. Org. Coat. 28 (1996) 43–48.
attack.
[9] I. Alig, S. Tadjbachb, P. Krügerb, H. Oehler, D. Lellinger, Prog. Org. Coat. 64 (2009)
However, the UV radiation is not the only cause of the sample 112–119.
corrosion observed but a combination of both UV sunlight radia- [10] M. Morcillo, Prog. Org. Coat. 36 (1999) 137–147.
tion and the atmospheric agents. These two types of agents would [11] A. Amirudin, D. Thierry, Prog. Org. Coat. 28 (1996) 59–75.
[12] E. Almeida, D. Santos, F. Fragata, O. Rincón, M. Morcillo, Mater. Corros. 52 (2001)
explain the higher corrosion observed in the samples exposed in 904–919.
Esmeraldas owing to both a higher humidity and ClNa and SO2 [13] F. Deflorian, S. Rossi, L. Fedrizzi, C. Zanella, Prog. Org. Coat. 59 (2007) 244–250.
concentrations, which fosters resin hydrolysis. [14] S. Shreepathi, A.K. Guin, S.M. Naik, M.R. Vattipalli, J. Coat. Technol. Res. 8 (2)
(2011) 191–200.
This result is of paramount importance in the determination of [15] F.X. Perrin, C. Merlatti, E. Aragon, A. Margaillan, Prog. Org. Coat. 64 (2009)
the coating to be selected for each site and also for the type of 466–473.
lab test to be applied. In the case of alkyd resins an experiment [16] F.X. Perrin, M. Irigoyen, E. Aragon, J.L. Vernet, Polym. Degrad. Stab. 72 (2001)
115–124.
simulating the ambient SO2 action, as that of Prohesion, should be [17] F.X. Perrin, M. Irigoyen, E. Aragon, J.L. Vernet, Polym. Degrad. Stab. 70 (2000)
one of the selected ones. 469–475.
[18] J.I. Iribarren, F.L. Mestres, F. Cadena, L. Bilurbina Alter, Mater. Corros. 55 (2004)
689–694.
4. Conclusions [19] F. Deflorian, S. Rossi, M. Fedel, Corros. Sci. 50 (2008) 2360–2366.
[20] C. Merlatt, F.X. Perrin, E. Aragon, A. Margaillan, Prog. Org. Coat. 61 (2008) 53–62.
Protective alkyd coatings on both sides of metallic substrates [21] W. Funke, Prog. Org. Coat. 31 (1997) 5–9.
[22] N.N. Voevvodin, V.N. Balbyshev, M. Khobaib, M.S. Donley, Prog. Org. Coat. 47
were exposed in two different categorized (according to ISO9223 (2003) 416–423.
standard) sites of Ecuador during 18 months. [23] O.D. Lewis, G.W. Critchlow, G.D. Wilcox, A. de Zeeuwb, J. Sanderb, Prog. Org.
The exposed samples were characterized by different tech- Coat. 73 (2012) 88–94.
[24] A.K. Mishra, R.S. Mishra, R. Narayan, K.V. Raju, Prog. Org. Coat. 67 (2010)
niques in terms of corrosion resistance, chemical structural 405–413.
changes, thermal stability and resulting morphology. [25] M. Rashvand, Z. Ranjbar, S. Rastegar, Prog. Org. Coat. 71 (2011) 362–368.
As a consequence of the weathering, the alkyd coatings showed [26] R. Van gorkum, E. Bouwman, Coord. Chem. Rev. 249 (2005) 1709–1728.
[27] Ad. Hofland, Prog. Org. Coat. 73 (2012) 274–282.
decoloration, fissures and rusting, without evidence of any other [28] M. Martí, G. Fabregat, D.S. Azambuja, C. Aleman, E. Armelin, Prog. Org. Coat. 73
defects, the upper faces in both locations showing a higher corro- (2012) 321–329.
sion attack. As in the bare metal corrosion process, Esmeraldas is [29] D.T.C. Ang, S.N. Gan, Prog. Org. Coat. 73 (2012) 409–414.
[30] N.M. Ahmed, H. Tawfik, M. Abdel-Fatahb, E.A. Youssefa, Prog. Org. Coat. 73
the station where alkyd coating resistance was more affected.
(2012) 76–87.
FTIR studies evidenced the polymer degradation through chain [31] S.K. Dhoke, A.S. Khanna, T. Jai, M. Sinha, Prog. Org. Coat. 64 (2009) 371–382.
scissions and the lixiviation of oil segments. [32] J. Alam, U. Riaz, S. Ahmad, Curr. Appl. Phys. 9 (2009) 80–86.
DSC results showed that the exposed samples were more rigid as [33] S.K. Dhoke, A.S. Khanna, Corros. Sci. 51 (2009) 6–20.
[34] M.V. Popa, P. Drob, E. Kasilescu, J.C. Mirza-Rosca, A. Santana Lopez, C. Kasilescu,
a function of the exposure time as Tg values increased accordingly. J.I. Drob, Mater. Chem. Phys. 100 (2006) 296–303.
The loss of the resin aliphatic segments was confirmed by [35] D. De la Fuente, J. Simancas, M. Morcillo, Prog. Org. Coat. 46 (2003) 241–249.
TGA due to the increase of the remaining residue of the exposed [36] D. Baer, P. Burrows, A. El-Azab, Prog. Org. Coat. 47 (2003) 342–356.
[37] ASTM G1 Practice for preparing, cleaning, and evaluating corrosion test speci-
samples. mens.
[38] ASTM G7–05 Standard practice for atmospheric environmental exposure test-
ing of nonmetallic materials.
Acknowledgments
[39] ASTM G50–76 (Reapproved 2003) Standard practice for conducting atmo-
spheric corrosion tests on metals.
F. Cadena wants to mention a special recognition to the staff [40] ASTM G140-02 (2008) Standard test method for determining atmospheric chlo-
member of the Laboratory of Polymer Analysis and Characterization ride deposition rate by wet candle method.
[41] ASTM G91-11 standard practice for monitoring atmospheric SO2 deposition
of the Basque Country University, UPV/EHU, for their support dur- rate for atmospheric corrosivity evaluation.
ing his postdoctoral work and to Senescyt (Secretaria Nacional de [42] ASTM D610 standard test method for evaluation degree of rusting on painted
Educación Superior, Ciencia, Teconología e Innovación del Ecuador) steel surfaces.
[43] ISO, 1992, 9223 Corrosion of metals and alloys – Corrosivity of atmospheres –
for awarding the research fellowship. Classification.
L. Irusta and M.J. Fernandez Berridi acknowledge the University [44] M. Lazzari, O. Chiantore, Polym. Degrad. Stab. 65 (1999) 303–313.
of the Basque Country, UPV/EHU, (UFI 11/56) and the Ministerio de [45] R. Ploeger, D. Scalarone, O. Chiantore, Polym. Degrad. Stab. 94 (2009)
2036–2041.
Ciencia e Innovación (MAT2010-16171) for the funding received to [46] J. Mallégol, L. Gonon, S. Commereuc, V. Verney, Prog. Org. Coat. 41 (2001)
develop this work. 171–176.