Progress in Organic Coatings 97 (2016) 99–109

Contents lists available at ScienceDirect

Progress in Organic Coatings

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

Thermal curing of para-phenylenediamine benzoxazine for barrier

coating applications on 1050 aluminum alloys
Marc Poorteman a , Alexis Renaud a , Julien Escobar a , Ludovic Dumas a,b , Leïla Bonnaud b ,
Philippe Dubois b , Marie-Georges Olivier a,∗
a
Department of Materials Science, Materials Engineering Research Center (CRIM), University of Mons, Place du Parc 20, B-7000 Mons, Belgium
b
Laboratory of Polymeric and Composite Materials, Center of Innovation and Research in Materials and Polymers (CIRMAP), Materia Nova Research Center
& University of Mons, Place du Parc 20, B-7000 Mons, Belgium

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

Article history: Polybenzoxazine coatings have been applied by spin coating of a solution, prepared from a laboratory
Received 3 February 2016 synthesized phenol, para-phenylenediamine benzoxazine (P-pPDA), on 1050 aluminum alloy. A heat
Accepted 28 March 2016 treatment was performed on the coating to polymerize the benzoxazine monomer by a ring opening
Available online 8 April 2016
mechanism attested for by Fourier Transform Infrared spectroscopy (FT-IR) and followed by Differential
Scanning Calorimetry (DSC). Dielectric Analysis (DEA) showed curing to proceed at a temperature range
Keywords:
up to 230 ◦ C, without thermal decomposition taking place, as has already been reported in the literature
Benzoxazine coating
for commercial bisphenol-a benzoxazines (Ba-A). The barrier protection has been evaluated by Electro-
Aluminum substrate
Thermal curing
chemical Impedance Spectroscopy (EIS) for 30 days in a sodium chloride solution (0.1 M). The results were
Barrier properties fitted with equivalent electrical circuits in order to understand the electrochemical behavior of such a
system. The initial barrier properties of both precursor and polymerized coatings appear to be excellent
for coatings having a thickness of only 5 ␮m. Though delamination occurs in the polymerized coating
during saline tests after several days of immersion, such kind of coatings have remarkable properties
with respect to epoxy coatings as they show a high geometrical stability during curing and a very low
water uptake during saline tests.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction resin coatings are currently used as organic coatings to prevent

Aluminum alloys, such as the 2xxx (Cu) and 7xxx (Ti) series,
are widely used in the aircraft industry as they offer improved
mechanical properties compared to pure aluminum (1050). How-
ever, these alloys are strongly sensitive to localized corrosion [1–3]
and, therefore, need to be protected. In order to improve their corro-
sion resistance, several types of coatings have been developed and
are commonly used at an industrial level such as anodizing layer
[4], conversion coating [5] and organic coatings [6]. These coatings
are often chromate-containing layers [4–6], showing excellent anti-
corrosive performance, but due to the high bio-toxicity of chromate,
the use of such kind of compounds will be banished in the aircraft
sector by 2017. As a consequence, several new technologies and
alternatives, avoiding the use of chromate, have been investigated:
sol-gel coatings [7–11], new anodizing layers [12–15] and the elab-
oration of organic-, inorganic- or hybrid coatings [16–19]. Epoxy
∗ Corresponding author.
E-mail address: marjorie.olivier@umons.ac.be (M.-G. Olivier).
corrosion of aeronautic structures. However, they suffer from some
drawbacks such as water uptake, which can be at the origin of cor-
rosion, blistering and delamination processes [20,21] when water
diffuses to the metal-coating interface. Moreover, most applied
epoxy coatings show a limited thermal stability and therefore can-
not be applied on engine parts.
Recently, polybenzoxazines have gained an increasing interest
due to their potential of combining the excellent properties of tra-
ditional epoxy- and phenolic resins. This class of polymers offers
highly attractive properties such as high glass transition tempera-
tures, low absorption of water, near zero shrinkage [22] and good
dielectric properties [23].
Benzoxazine monomers can be easily synthesized by the
reaction of three basic compounds: a phenol, an amine and
formaldehyde. By changing the different groups of the compo-
nents, several types of benzoxazines can be elaborated with a large
panel of different properties [24,25]. Amongst the few studies in the
literature about the use of this new polymer family as an alterna-
tive for epoxy coatings and their potential performances as barrier
protective coatings, the authors have recently published a paper

http://dx.doi.org/10.1016/j.porgcoat.2016.03.026
0300-9440/© 2016 Elsevier B.V. All rights reserved.
100 M. Poorteman et al. / Progress in Organic Coatings 97 (2016) 99–109

Fig. 2. Polymerized structure of Ba-A (left) and P-pPDA (right).

Fig. 1. One pot synthesis of P-pPDA precursor.

showing the barrier properties of a commercial bisphenol A, ani-

line benzoxazine (Ba-A) coating applied on aluminum 1050 [26].
Although a performant barrier coating could be fabricated by apply-
ing an in acetone dissolved Ba-A monomer solution by dip coating
onto the aluminum substrate and subsequent step curing succes-
sively at 180 and 200 ◦ C, the thermal polymerization of this kind
of monomer appeared to be accompanied by an irreversible degra-
dation with the release of volatile imino species [27]. Consequent
out-gassing can then be responsible for the formation of voids in
the cured materials and decrease the performance and reliabil-
ity of the synthesized coatings. Alternatively, a thermally stable
benzoxazine precursor based on phenol and p-phenylene diamine
benzoxazine (P-pPDA), according to a reaction represented in Fig. 1,
can be synthesized by an easy and scalable one pot process [28].
The increased thermal stability of P-pPDA compared to Ba-A can
be explained by the fact that the aminophenyl moiety is trapped
between two benzoxazine rings limiting its cleavage during the
network formation (Fig. 2), hence resolving the outgassing draw- Fig. 3. Coatings obtained by spin coating for different concentrations of P-pPDA
back. Moreover, bulk cured P-pPDA materials show excellent fire solutions, before and after curing.
resistant properties with LOI values of 38% [28] making it a very
attractive alternative as nonflammable material to replace halo-
genated materials for instance for coatings inside an airplane. <0.25%Si, <0.07%Zn, <0.05%Cu, <0.05%Mg, <0.05%Ti, <0.05%Mn and
In this work, a diamine based benzoxazine was synthesized Al accounts for the remainder.
according to a procedure presented elsewhere [28], leading to an Each sample surface (60 mm × 45 mm × 1 mm) was first cleaned
obtained precursor consisting of a mixture of P-pPDA monomers and degreased in acetone. A consecutive etching is needed before
and some oligomers. In this paper, this precursor mixture has been benzoxazine deposition in order to improve the surface wettabil-
used without carrying out any further purification to prepare a neat, ity and coating adhesion. This was performed as follows: first, an
thick coating (from 4 to 7 ␮m) on an aluminum substrate. A 1050 immersion in a 1 M NaOH bath maintained at 50 ◦ C for 60 s, fol-
aluminum alloy was chosen because of its intrinsic corrosion resis- lowed by rinsing with deionized water and, finally, a neutralization
tance performance, which is an advantage if the barrier properties in HNO3 (50 vol%) carried out for 60 s at ambient temperature.
and the organic water uptake are to be highlighted. This alloy is The monomer solution was prepared by dissolving laboratory syn-
often used as clad layer of a 2024 aluminum alloy in order to obtain thesized P-pPDA precursor in chloroform (from 50 to 200 g/L),
a good compromise between corrosion resistance and mechanical obtaining a yellow-orange solution after reflux for 4 h at 60 ◦ C. After
properties for aeronautic applications. The initial benzoxazine was cooling and stirring during 12 h the solution is deposited onto the
dissolved in a solvent and deposited by spin coating. After room substrate by a spin-coating process at 3000 rpm for 30 s.
temperature drying, a final heat treatment was applied to polymer- The drying process of the coated aluminum consisted in leaving
ize the coating. The structural transformation after thermal curing the sheets at room temperature for 24 h.
was followed by Fourier Transform Infrared spectroscopy (FT-IR) After deposition and drying, a thermal treatment was carried
and Dielectric Analysis (DEA). The polymerization degree was esti- out to obtain the cross-linked polymer by successive step curing
mated from data obtained by Differential Scanning Calorimetry at 140 ◦ C during 60 min, 180 ◦ C and 200 ◦ C, with each time a dwell
(DSC). Finally, the barrier properties and the water uptake of the of 2 h, 220 ◦ C during 30 min and, finally, at 230 ◦ C during 30 min.
polybenzoxazine coatings applied on aluminum substrate were Thereafter, samples were allowed to slowly cool down to room
assessed by Electrochemical Impedance Spectroscopy (EIS) versus temperature.
immersion time in a 0.1 M NaCl neutral solution. The coating per-
formances were determined before and after curing treatment. 2.2. Characterization

2. Experimental The surface roughness of the substrate (Ra and Rz), before and
after pretreatment, was measured by surface profilometry (Nano-
2.1. Preparation process Jura High Precision Surface Metrology System—NJHP 505).
FT-IR measurements were carried out in the 370–2000 cm−1
The selected substrate material was a 1050 aluminum alloy range in order to follow the structural curing of the polymer by
(AA). Its chemical composition in weight percent is: <0.40%Fe, using a Perkin Elmer Spectrum 2000 apparatus and the results were
 M. Poorteman et al. / Progress in Organic Coatings 97 (2016) 99–109 101

Monomer coating
Polymer coating
100

80

Reflexion (%) 60

40

20 1035 cm
-1

-1
1240 cm
0
2000 1800 1600 1400 1200 1000 800 600 400
-1
Wavenumber (cm )

Fig. 4. FTIR spectra between 370 cm−1 and 2000 cm−1 of the P-pPDA coating before (M200b) and after (P200b) thermal treatment.

4 240

220

2 200

180

Temperature (°C)
0 160
Heat flow (mW)

140

-2 120

100

-4 80

60

-6 heat flow 40
temperature 20

-8 0
0 60 120 180 240 300 360

Time (min)

Fig. 5. Heat flow versus the time during the heat treatment of bulk P-pPDA benzoxazine.

analyzed by considering the evolution of the reflection as a function
of the wavenumber.
The coating thicknesses were determined by Foucault current
(ELCOMETER 456) by performing 20 measurements on different
locations of the coated aluminum.
The following thermal analyses were carried out:
- DSC measurements (Q200 from TA Instruments) under N2
(50 mL/min) to determine the extent of reaction ␣ and the glass
transition temperature Tg . An Indium standard was used for cal-
ibration.
- For curing monitoring through DEA, a 230/2 epsilon instru-
ment from Netzsch was used. Heating was carried out within a
small lab furnace with an integrated J type thermocouple. The
external interface box used was a mid-conductivity interface,
enabling to measure electrical conductivities between 10−3 and
10−13 S cm−1 .
An inter-digitized electrode sensor was used (IDEX 036S from
Netzsch) with an electrode spacing of 130 ␮m and an A/D ratio
(cm) of 80. The thickness applied by bar coating of the benzoxazine
solution on top of the sensor was at least 115 ␮m.
The log loss domain covered by this type of sensor is typically
between 1 and 4.
The samples were thermally treated in a Heraeus Instruments
LUT 6050 oven in horizontal position to avoid the flow and loss of
benzoxazine during the treatment.
A conventional three electrode cell was used for the electro-
chemical tests. The working electrode was the investigated sample
(exposed area of 7 cm2 ), the counter electrode was a platinum
plate and all potentials were measured with respect to an Ag/AgCl
102 M. Poorteman et al. / Progress in Organic Coatings 97 (2016) 99–109

300
Log Loss 1 Hz
Log Loss 10 Hz
4 Log Loss 100 Hz 250
Log Loss 1000 Hz
Temperature

Temperature (°C)
200

Log loss
2
150

100
0

50

-2 0
0 60 120 180 240 300 360

Time (min)

Fig. 6. Plot of loss factor of P-pPDA benzoxazine coating during consecutive isothermal treatments from 140 ◦ C up to 230 ◦ C.

(+0.205 V vs SHE) reference electrode. The cell was placed in a Table 1

Thickness of coatings with different pPDA concentrations applied before and after
Faraday cage in order to minimize external electromagnetic inter-
curing.
ference on the system. The impedance measurements were carried
out over frequencies ranging from 100 kHz to 10 mHz, at ambi- Applied coating Before heat treatment (␮m) After heat treatment (␮m)
ent temperature. The impedance spectra were acquired by using [pPDA] = 50 g/L 4.7 ± 1.3 3.8 ± 0.5
a potentiostat coupled with a frequency response analyser (Parstat [pPDA] = 100 g/L 3.9 ± 1.2 4.4 ± 0.9
2273 from Ametek), computer-controlled with Powersuite® soft- [pPDA] = 200 g/L 6.9 ± 1.1 6.0 ± 0.8

ware. Electrochemical impedance measurements were performed

after different immersion times in 0.1 M NaCl solution on the ben-
zoxazine (not cured) and polybenzoxazine (cured) samples. The cant thickness difference is observed after curing as expected for a
signal amplitude voltage was 30 mV rms. benzoxazine resin [22].

3. Results and discussions 3.2. Thermal analysis of the phenol pPDA benzoxazine coating

3.1. Application and curing of the pPDA benzoxazine coatings
Application of P-pPDA from chloroform solutions onto the pre-
treated aluminum surface by spin coating leads to neat coatings
(Fig. 3). The observed cohesion between monomers in the solid
state – at the level of powder agglomerates or after application
of the monomer – can be attributed to intermolecular interac-
tions, most likely to be dipolar in nature and/or by ␲-␲ interactions
[26]. After spin coating of the solution, the cohesion between the
monomer and the substrate is likely to be mostly mechanical, con-
sidering the roughness of the aluminum surface (Rz = 3.17 ␮m for
Al1050 and 4.17 ␮m after pretreatment).
In order for polymerization or pre-polymerization to take place,
heating is necessary as the ring opening mechanism is a thermally
activated chemical reaction [22]. Fig. 4 shows the FTIR spectra
between 370 cm−1 and 2000 cm−1 of the P-pPDA coating before and
after thermal treatment. Characteristic absorption bands of ben-
zoxazine appear at 1240 cm−1 and 1035 cm−1 (antisymmetric and
symmetric stretching of C O C respectively) [29]. After heat treat-
ment, these bands disappear, which can be attributed to the ring
opening mechanism leading to the polymerization of benzoxazine
monomer in agreement with previous studies [28,30].
The thickness of the coatings performed by spin coating and
dried depends on the concentration of the P-pPDA dissolved in the
solvent and has a slight tendency to increase with the monomer
concentration (Table 1). It appears that the dispersion of results
is small before and after curing indicating a coating with uniform
thickness throughout the whole process. Furthermore, no signifi-
One of the reasons for using P-pPDA as benzoxazine precursor
is the fact that no significant thermal degradation – less than 1 wt%
weight loss – has been observed for bulk materials in the litera-
ture [28] for temperatures up to the curing temperature (230 ◦ C),
whereas for a previous used commercial Ba-A precursor weight
losses as high as 9–10 wt% have been reported in a temperature
range up to 200 ◦ C [31,32].
The degradation of Ba-A and the related reaction kinetics have
only been recently described in the literature [32–40]. The ther-
mal initiation of the polymerization leads to the formation of
two species in equilibrium, a carbocation and an iminium ion.
The carbocation is the active species leading to the formation
of the polymer network whereas the latent iminium ion can
be dissociated as a function of temperature, liberating a volatile
aromatic imine. This proposed mechanism is a possible expla-
nation for the weight loss occurring during polymerization at
high temperatures. This mechanism is less likely to occur when
para-phenylenediamine is used as a precursor because this would
implicate a cleavage at each side of the central aromatic ring.
The results of differential scanning calorimetry performed on
the coating obtained from a P-pPDA benzoxazine solution (200 g/L)
by using the step curing conditions as described in the experimental
part (starting at 140 ◦ C up to 230 ◦ C) are shown in Fig. 5.
At first, a small exothermic reaction appears to occur during
the first curing step (140 ◦ C). During the next temperature dwells
(180 ◦ C, . . ., 230 ◦ C) the heat flow each time increases and reaches a
maximum and then decreases until the end of each step indicating
that the curing of benzoxazine takes place.
 M. Poorteman et al. / Progress in Organic Coatings 97 (2016) 99–109 103

(a) 1011

1010 1h
24h
109 3 days
7 days
108 14 days
21 days
107

|Z| (Ohm cm2)

28 days

106

105

104

103

102

101

100
10-2 10-1 100 101 102 103 104 105

Frequency (Hz)

(b)
80
- Phase angle (°)

60

40
1h
24h
3 days
7 days
20 14 days
21 days
28 days

0
10-2 10-1 100 101 102 103 104 105

Frequency (Hz)

Fig. 7. Evolution of the impedance modulus (a) and phase angle (b) versus frequency of monomer P-pPDA benzoxazine coated aluminum 1050 as a function of immersion
time in 0.1 M NaCl.

At the end of the heat treatment, no residual enthalpy was
observed which means that the polymerization reaction is maximal
under the investigated thermal cycle. The calculated glass transi-
tion temperature, is close to 220 ◦ C, when this value is only 45 ◦ C
for the precursor.
Thermal curing of the applied coating was also followed up by
dielectric analysis (DEA).
A phenomenological approach was adopted to model the dielec-
tric behavior of the chosen benzoxazine during curing.
When free charge migration occurs in a dielectric material, elec-
trical conductivity (␴) as well as dipole polarization contribute to
the loss factor (ε (ω))REFNUMLINK[41]:
ε = εd + εi (3)
• ε = ωε
the ionic polarization contribution, ␻ the frequency of
i 0
the electrical field and ε0 is the dielectric permittivity of vacuum
[ε0 = 8.54 10−12 AsV−1 m−1 ] [40]
• ε the dipolar polarization contribution (for instance given by the
d
Debye equation in the simple case of a spherical dipole) [41]
Of particular interest for curing monitoring is the direct current
(DC) conductivity, also known as the inverse ion viscosity, because
its relation to a fluid dynamic viscosity (␩). This is a measure of the
mobility of free ions through a medium under the influence of an
electric field.
By its definition, the DC conductivity ␴ can be expressed as
follows:
 = qn (4)

With: Where:
104 M. Poorteman et al. / Progress in Organic Coatings 97 (2016) 99–109

(a) 1011

1010 1h
24h
109 3 days
7 days
108 14 days
21 days
107
|Z| (Ohm cm2)
28 days
106

105

104

103

102

101

100
10-2 10-1 100 101 102 103 104 105

Frequency (Hz)

(b)
80

60
-Phase angle (°)

40
1h
24h
3 days
20 7 days
14 days
21 days
28 days
0
10-2 10-1 100 101 102 103 104 105

Frequency (Hz)

Fig. 8. Evolution of the impedance modulus (a) and phase angle (b) versus frequency of poly pPDA benzoxazine coated aluminum 1050 as a function of immersion time in
0.1 M NaCl.

• q = the charge of free ions [C] In that case, an increase of ␻ with factor 10 leads to a decrease
• ␮ = the free ion mobility [cm2 V−1 s−1 ] of logε with one unit.
• n = the free ion concentration [cm−3 ] Fig. 6 shows the temperature dependence of the loss factor for
a coating of the selected P-pPDA benzoxazine, measured at fre-
Using also the Einstein and the Stokes-Einstein relations, we quencies between 1 and 1000 Hz, the frequency being increased
obtain a proportionality between 1/␴ and ␩: successively by a factor 10 during the whole heat treatment.
At the first temperature dwell, the loss factor appears to slightly
6r decrease, confirming the slight curing taking place at this tempera-
1/ = (5)
q2 n ture as observed by DSC analysis. As the temperature further rises,
In Eq. (5), r is the ion radius. the loss factor increases proportionally, which can be related to the
For the interpretation of dielectric spectra (for instance ε (ω)), decrease in viscosity with temperature leading to an increase in DC
it is important to notice that the ionic current (DC current) can conductivity ([Eq. (5)]) and therefore an increase in ionic polariza-
be considered to be almost frequency independent, whereas the tion contribution [Eq. (3)]. The log loss reaches a first maximum,
energy loss due to reorientation of dipoles varies with frequency. independent of frequency, close to 180 ◦ C. At this moment the loss
Under the condition εd  εi , [Eq. (3)] becomes ε ≈ ωε

and is determined by the ionic polarization contribution as an increase
 
0
of ␻ with a factor 10 leading to a decrease of logε of about one
therefore logε ≈ log ε − log ω
0
 M. Poorteman et al. / Progress in Organic Coatings 97 (2016) 99–109 105

Fig. 9. Equivalent electrical circuits used to fit impedance data.

(a) 1011 (b) 1011

1010 80 1010 80

109 109

- Phase angle (°)

60

- Phase angle (°)

60
|Z| (Ohm cm2)

|Z| (Ohm cm2)

108 108

107 107
40 40
106 106

Measured |Z|
105 Calculated |Z| 105 Measured |Z|
20 Calculated |Z| 20
Measured phase angle
Calculated phase angle Measured phase angle
104 104 Calculated phase angle

103 0 103 0
10-2 10-1 100 101 102 103 104 105 10-2 10-1 100 101 102 103 104 105

Frequency (Hz) Frequency (Hz)

(c) 1011 (d) 1011

1010 80 1010 80
Measured |Z|
109 109 Calculated |Z|
Measured phase angle

- Phase angle (°)

60
- Phase angle (°)

60 Calculated phase angle

|Z| (Ohm cm2)
|Z| (Ohm cm2)

108 108

107 107
40 40
106 106
Measured |Z|
105 Calculated |Z| 105
Measured phase angle 20 20
Calculated phase angle
104 104

103 0 103 0
10-2 10-1 100 101 102 103 104 105 10-2 10-1 100 101 102 103 104 105

Frequency (Hz) Frequency (Hz)

Fig. 10. Plots of comparison between experimental impedance data and fitted impedance; (a) monomer coating/1 h of immersion/Equivalent Circuit a; (b) monomer
coating/28 days of immersion/Equivalent Circuit c; (c) Polymer coating/1 h of immersion/Equivalent Circuit a; (d) polymer coating/28 days of immersion/Equivalent Circuit
c.

unit. During the isothermal dwell, the loss factor decreases during
the whole dwell. A decrease in loss factor indicates a decrease in
DC conductivity and, therefore, an increase in viscosity at constant
temperature [Eq. (5)] which can be related to the curing of the ben-
zoxazine monomer. These results are in good agreement with the
DSC data. For each successive temperature increase and temper-
ature dwell, the same observations can be made, confirming that
curing takes place during the whole temperature cycle. This kind
of behavior was not observed when the curing of Ba-A was studied
by the same analyses [26]. In that case, the DEA curves were not
in accordance with the DSC data at 180 ◦ C. This was discussed by
the authors as being related to the creation of latent intermediary
ionic iminium species, leading to an increase of ionic charges due
to the slowing down of the diffusion rate of ionic species within an
increasingly cross-linked and hindered system. The ␩/n ratio could
therefore decrease and, according to [Eq. (5)], the DC conductivity,
as well as the loss, could increase as was experimentally observed.
Such a behavior has not been observed for P-pPDA and this is likely
to be related to the fact that no or less volatile iminium species are
created during curing of this benzoxazine.
3.3. Corrosion resistance of the coating before and after heat
treatment
As one of the objectives of this study is to determine the poten-
tial of poly P-pPDA benzoxazine to offer barrier properties to 1050
aluminum alloy, the barrier behavior and the corrosion at the
metal-coating interface of polybenzoxazine coatings were assessed
by electrochemical impedance spectroscopy (EIS). The evolutions
of impedance modulus and phase Bode diagrams with immersion
time in 0.1 M NaCl solution are given in Figs. 7 and 8 for P-pPDA
precursors and polymer coatings for a thickness of 6 ␮m, which
appeared to give the best barrier properties.
The value of the impedance modulus at low frequencies can be
used to evaluate the global barrier behavior of the coatings. Differ-
ent contributions have to be considered: the barrier properties of
106 M. Poorteman et al. / Progress in Organic Coatings 97 (2016) 99–109

1011

1010

109

108

107
Rp (Ohm cm2)
106

105

104

103 Monomer coating

Polymer coating
102

101

100
0 200 400 600

Immersion time (h)

Fig. 11. Evolution of calculated Rp over immersion time for both monomer and polymer coatings.

1.610-9

-9
1.410

1.210-9

-9
Qc (Fcm-2sn-1)

1.010

8 10-10

6 10-10

-10
4 10

Monomer coating
2 10-10
Polymer coating

0 200 400 600

Immersion time (h)

Fig. 12. Evolution of calculated Qc over immersion time for both monomer and polymer coatings.

the organic coating, the charge transfer resistance and the barrier
due to the passive nature of the aluminum in neutral medium. For
the uncoated 1050 aluminum alloy immersed in the sodium chlo-
ride solution, the low frequency modulus value is determined and
equal to 3 104  cm2 . In that case, no protection is due to an organic
layer.
At the beginning of the immersion in the NaCl solution, both
monomer and polymer coatings exhibit very good barrier prop-
erties. They present one capacitance covering a large range of
frequencies, and their resistance (estimated from the value of
the Bode modulus at 0.01 Hz) is of an order of magnitude of 1
1010  cm2 . This behavior is maintained during the first 24 h of
measurements, but we observe a significant decrease after several
days of immersion in the NaCl solution. As the Bode modulus at
0.01 Hz decreases, the time constant related to the coating is shifted
to higher frequencies while a new time constant, attributed to
appearance of corrosion, is emerging. This evolution can be related
to the progressive penetration of the aggressive electrolyte through
the coating to reach the metal surface, causing delamination and
corrosion processes at the metal/coating interface and leading to
the loss of barrier properties.
The evolution of physical properties of benzoxazine coatings
over immersion time can be estimated by fitting impedance data
by using equivalent electrical circuits. Fig. 9 shows the three dif-
ferent model circuits that have been used to analyze impedance
data, depending on the progress of the coating degradation. In
these circuits, Rs represents the electrolyte resistance, which can
be negligible compared to the rest of the system and considering
the good conductivity of the 0.1 M NaCl solution. Qc and Rp are the
capacitance and the resistance of the coating respectively, when
 M. Poorteman et al. / Progress in Organic Coatings 97 (2016) 99–109
Monomer coating
3
Polymer coating
Log(fb(Hz))
1
-1
-2
-3
0 200
Immersion time (h)
Fig. 13. Evolution of breakpoint frequency over immersion time for both monomer and polymer coatings.
Qdl and Rct refer to the capacitance of the double-layer and the
resistance to charge transfer. CPE elements were used instead of
pure capacitance in order to get more precise fitting, to take into
account possible non-ideal behavior of the coatings. In the liter-
ature, the impedance of a constant phase element is written as
−n
ZQ = 1/Y0 (iω) where n is the frequency dispersion factor varying
from 0 to 1. If n = 0, the CPE behaves as a resistor. If n = 1, the CPE is
a pure capacitance [42]. A Warburg element was added in order to
take into account the diffusion of electrolyte through the corrosion
products [43–47]. Fig. 10 gives some examples of comparison in
Bode plot between measured and fitted impedance points showing
an excellent correspondence between the fitted and experimen-
tally measured values when using appropriate models. Moreover,
we observe an evolution from an equivalent electrical circuit (EEC)
with one time constant (Fig. 9 circuit a) to an EEC showing two time
constants (Fig. 9 circuit c) as the electrolyte goes though the coating
and reaches the coating/substrate interface and in accordance with
the Bode diagrams (Figs. 7 b and 8 b).
Figs. 11 and 12 show plots of Rp and Qc respectively over immer-
sion time for both coatings. The values of n are close to 1 (in the
range of 0.97–0.99) meaning that the monomer and polymer coat-
ings each behave practically as a pure capacitance. At the same
thickness, the polymer coating appears to have a higher capaci-
tance than the monomer coating, and therefore a higher dielectric
constant: about 8 for the polymer, 4 for the monomer. However,
for both systems, the capacitance shows a slight increase during
the first hours of immersion, followed by a relatively good stabil-
ity of the capacitance after days of immersion, especially for the
monomer. Such a behavior suggests a very low water permeability.
From the plot of Rp over immersion time, two periods can
be identified in the decrease of the coating resistance. In the
first period, up to two weeks of immersion, Rp decreases from
1010  cm2 down to 108  cm2 for both coatings. This decrease is
mainly attributed to the progressive penetration of the electrolyte
through coating porosity. But in a second period of time, while
the monomer coating maintains an Rp value of about 108  cm2 ,
the polymer coating resistance further drops down to 105  cm2
after 28 days. This sharp decrease can be the consequence of
apparition of either coating cracks and/or the delamination at the
metal/coating interface.
107
400 600
The water uptake determination is a good indicator to assess the
resistance of organic coating and to evaluate its possible degrada-
tion in an aqueous medium. This parameter can be determined from
coating capacitance measurements at high frequency following the
Brasher and Kingsbury relation for short immersion times in aque-
ous solution [48]. According to this equation, the volume fraction
of water (Xv ) in organic polymers can be calculated as follows:
log Ct ⁄C0
Xv = (6)
log εw
where Ct is the coating capacitance at immersion time t, C0 the
“dry” coating capacitance, tested at t0 immersion time, and ␧w the
dielectric constant of water (80).
The capacitance of the film Ct was determined at a fixed fre-
quency of 12.66 kHz, which corresponds to the more capacitive part
of the spectra in Figs. 7 and 8 (phase angle close to −90 ◦ C)
In the case of the benzoxazine monomer coating, the capaci-
tance is about 4.5 10−10 F cm−2 and in the same order of magnitude
during the whole immersion time. The water uptake reaches a value
of only 0.1% after 24 h. For the polymer coating, the capacitance is
about 9 10−10 F cm−2 , and the calculated water intake is only 0.4%
within the first 24 h of immersion. Such a low absorption of water
is characteristic for benzoxazine systems [22]. The slightly higher
water intake in the case of polymer coating could be explained
by the presence of OH hydroxyl groups in its molecular structure,
providing more affinity to water molecules.
The break point frequency (fb ) can also be used as an estimation
of the degradation of a coating adherence to the substrate [49]. The
evolution of this parameter has been related to the growth of the
delaminated area. The breakpoint frequency corresponds to the fre-
quency at which the Bode phase angle plot attributed to the time
constant of the coating crosses 45◦ . At frequencies higher than fb ,
the coating has a mostly capacitive behavior, but acts resistive at
lower frequencies. The capacitance of the coating remains quite
constant over immersion time, so we can assume the dielectric
constant and the resistivity of the coating also to remain constant.
Therefore, a relative increase in fb indicates an increase in delami-
nated area. Fig. 13 compares the evolution of fb for both monomer
and polymer coatings. It is clear that the polymer coating suffers a
higher delamination speed compared to the monomer coating. This
108 M. Poorteman et al. / Progress in Organic Coatings 97 (2016) 99–109
difference can be explained by some coating defects in the polymer
that could have occurred during curing. A weaker interface with
the aluminum substrate due to stresses generated during curing
attributed to the expansion mismatch between the rigid polymer
coating and the aluminum substrate could also be a valuable expla-
nation. At the same time this could explain the better performance
of the monomer coating with time, as there are no thermal stresses
generated in this last coating and therefore a better adherence with
the substrate can be expected.
4. Conclusions
A laboratory synthesized para-phenlylenediamine benzoxazine
mixture consisting of mainly monomers and some oligomers, dis-
solved in chloroform, has been successfully applied by spin coating
on a 1050 aluminum substrate.
Curing of the coatings through a ring opening mechanism pro-
ceeded by thermal treatment, for temperatures up to 230 ◦ C and
was followed up by DEA and DSC. P-pPDA coatings appear to stable
during heat treatment, showing almost no weight loss, no shrink-
age and a very low water uptake (<0.5%). The results obtained by
DEA and DSC measurements of P-pPDA benzoxazine are in good
agreement during the whole thermal cycle proving the high ther-
mal stability of the P-pPDA formed films, which was not the case for
Ba-A benzoxazine as reported in the literature, due to the creation
of volatile imino species in the last case.
In the case of the benzoxazine monomer coating, the capac-
itance was about 4.5 10−10 F cm−2 and in the same order of
magnitude during the whole immersion time whereas for the
polymer coating, the capacitance is higher, about 1 10−9 F cm−2 ,
compared to the monomer.
The barrier properties of these thin (about 6 ␮m) P-pPDA coat-
ings, are characterized by a low frequency modulus in the 1 108
–1 1010  cm2 region, at initial time but polymerized P-pPDA suf-
fers some delamination after a couple of days. This drawback was
attributed to the interfacial characteristics between the substrate
and the coating leading to mechanical stresses generated at the
interface during cooling after curing at 230 ◦ C.
Finally, compared to epoxy coatings, [50] these coatings show a
very low water uptake and a very high geometric stability during
curing.
Acknowledgments
The authors wish to thank the “Région Wallonne” for the finan-
cial support in the framework of the ECOTAC project from the
“Pôle de compétitivité” SKYWIN. The authors also wish to thank
the “‘Région Wallonne”’ in the framework of the “Programme
d’Excellence: OPTI2MAT and FLYCOAT”. Finally, the authors want
to thank M. Poelman from Materia Nova (B) for her advices and
participation to discussions.
References
[1] N. Murer, R. Oltra, B. Vuillemin, O. Néel, Numerical modelling of the galvanic
coupling in aluminium alloys: a discussion on the application of local probe
techniques, Corros. Sci. 52 (2010) 130–139.
[2] A.-P. Romano, M.-G. Olivier, A. Nazarov, D. Thierry, Influence of crosslinking
density of a cataphoretic coating on initiation and propagation of filiform
corrosion of AA6016, Prog. Org. Coat. 66 (2009) 173–182.
[3] Geraint Williams, Andrew J. Coleman, H. Neil McMurray, Inhibition of
Aluminium Alloy AA2024-T3 pitting corrosion by copper complexing
compounds, Electrochim. Acta 55 (2010) 5947–5958.
[4] E. Harscoet, D. Froelich, Use of LCA to evaluate the environmental benefits of
substituting chromic acid anodizing (CAA), J. Clean. Prod. 16 (2008)
1294–1305.
[5] H.A. Katzman, G.M. Malouf, R. Bauer, G.W. Stupian, Corrosion-protective
chromate coatings on aluminum, Appl. Surf. Sci. 2 (1979) 416–432.
[6] P. Campestrini, E.P.M. van Westing, A. Hovestad, J.H.W. de Wit, Investigation
of the chromate conversion coating on Alclad 2024 aluminium alloy: effect of
the pH of the chromate bath, Electrochim. Acta 47 (2002) 1097–1113.
[7] D. Wang, G.P. Bierwagen, Sol–gel coatings on metals for corrosion protection,
Prog. Org. Coat. 64 (4) (2009) 327–338.
[8] M.E. Druart, J.B. Richir, C. Poirier, F. Maseri, N. Godeau, L. Langer, M.-G. Olivier,
Influence of sol–gel application conditions on metallic substrate for optical
applications, Corros. Eng. Sci. Technol. 46 (2011) 677–684.
[9] A.-P. Romano, M. Fedel, F. Deflorian, M.-G. Olivier, Silane sol–gel film as
pretreatment for improvement of barrier properties and filiform corrosion
resistance of 6016 aluminium alloy covered by cataphoretic coating, Prog.
Org. Coat. 72 (2011) 695–702.
[10] J.B. Bajat, I. Milošev, Ž. Jovanović, R.M. Jančić-Heinemann, M. Dimitrijević, V.B.
Mišković-Stanković, Corrosion protection of aluminium pretreated by
vinyltriethoxysilane in sodium chloride solution, Corros. Sci. 52 (3) (2010)
1060–1069.
[11] I. Recloux, M. Debliquy, A. Baroni, Y. Paint, A. Lanzutti, L. Fedrizzi, M.-G.
Olivier, Optimization of synthesis parameters of mesoporous silica sol–gel
thin films for application on 2024 aluminum alloy substrates, Appl. Surf. Sci.
277 (2013) 201–210.
[12] G.D. Sulka, K.G. Parkoła, Temperature influence on well-ordered
nanoporestructures grown by anodization of aluminium in sulphuric acid,
Electrochim. Acta 52 (2007) 1880–1888.
[13] G. Boisier, N. Pébère, C. Druez, M. Villatte, S. Suel, FESEM and EIS study of
sealedAA2024 T3 anodized in sulfuric acid electrolytes: influence of tartaric
acid, J. Electrochem. Soc. 155 (2008) C521–C529.
[14] L. Domingues, J.C.S. Fernandes, M. Da Cunha Belo, M.G.S. Ferreira, L.
Guerra-Rosa, Anodising of Al 2024-T3 in a modified sulphuric acid/boric acid
bath for aeronautical applications, Corros. Sci. 45 (2002) 149–160.
[15] G.D. Sulka, W.J. Stepniowski, Structural features of self-organized nanopore
arrays formed by anodization of aluminum in oxalic acid at relatively high
temperatures, Electrochim. Acta 54 (2009) 3683–3691.
[16] F. Khelifa, M.-E. Druart, Y. Habibi, F. Bénard, Ph. Leclère, M.-G. Olivier, Ph.
Dubois, Sol–gel incorporation of silica nanofillers for tuning the anti-corrosion
protection of acrylate-based coatings, Prog. Org. Coat. 76 (5) (2013) 900–911.
[17] F. Khelifa, M.-E. Druart, Y. Habibi, R. Rioboo, M.-G. Olivier, J. De Coninck, Ph.
Dubois, Effect of photo-crosslinking on the performance of silica
nanoparticle-filled epoxidized acrylic copolymer coatings, J. Mater. Chem. A 1
(35) (2016) 10334–10344.
[18] V.B. Mišković-Stanković, M.R. Stanić, D.M. Dražić, Corrosion protection of
aluminium by a cataphoretic epoxy coating, Prog. Org. Coat. 36 (1999) 53–63.
[19] J. Singh-Beemat, J.O. Iroh, L. Feng, Mechanism of corrosion protection of
aluminum alloy substrate by hybrid polymer nanocomposite coatings, Prog.
Org. Coat. 76 (2013) 1576–1580.
[20] A. Foyet, T.H. Wu, L. van der Ven, A. Kodentsov, G. de With, R. van Benthem,
Influence of mixing ratio on the permeability of water and the corrosion
performance of epoxy/amine coated un-pretreated Al-2024 evaluated by
impedance spectroscopy, Prog. Org. Coat. 64 (2009) 138–141.
[21] N.N.A.H. Meis, L.G.J. van der Ven, R.A.T.M. van Benthem, G. de With, Extreme
wet adhesion of a novel epoxy-amine coating on aluminum alloy 2024-T3,
Prog. Org. Coat. 77 (2014) 176–183.
[22] H. Ishida, D. Allen, Physical and mechanical characterization of near-zero
shrinkage poly benzoxazines, J. Polym. Sci. B: Polym. Phys. 34 (1996)
1019–1030.
[23] M.C. Tseng, Y.L. Liu, Self-assembled benzoxazine-bridged polysilsesquioxanes
exhibiting ultralow-dielectric constants and yellow-light photoluminescent
emission, Polymer 51 (2010) 5565–5567.
[24] T. Takeichi, T. Kawauchi, T. Agag, High performance polybenzoxazines as a
novel type of phenolic resin, Polym. J. 40 (2008) 1121–1131.
[25] G. Lligadas, A. Tüzün, J.C. Ronda, M. Galià, V. Càdiz, Polybenzoxazines: new
players in the bio-based polymer arena, Polym. Chem. 5 (2014) 6636–6644.
[26] J. Escobar, M. Poorteman, L. Dumas, L. Bonnaud, Ph. Dubois, M.-G. Olivier,
Thermal curing study of bisphenol A benzoxazine for barrier coating
applications on 1050 aluminum alloy, Prog. Org. Coat. 70 (2013) 53–61.
[27] H. Oie, A. Sudo, T. Endo, Acceleration effect of N-allyl group on thermally
induced ring-opening polymerization of 1,3-benzoxazine, J. Polym. Sci. Part A
96 (2011) 708–718.
[28] L. Dumas, L. Bonnaud, M.-G. Olivier, M. Poorteman, Ph. Dubois, High
performance benzoxazine/CNT nanohybrid network—An easy and scalable
way to combine attractive properties, Eur. Polym. J. 58 (2014) 218–225.
[29] T. Agag, T. Takeichi, Synthesis and characterization of novel benzoxazine
monomers containing allyl groups and their high performance thermosets,
Macromolecules 48 (23) (2010) 5357–5363.
[30] T. Takeichi, Y. Guo, S. Rimdusit, Performance improvement of
polybenzoxazine by alloying with polyimide: effect of preparation method on
the properties, Polymer 46 (2005) 4909–4916.
[31] L. Dumas, L. Bonnaud, M.-G. Olivier, M. Poorteman, Ph. Dubois, Facile
preparation of a novel high performance benzoxazine—CNT based
nano-hybrid network exhibiting outstanding thermo-mechanical properties,
Chem. Commun. 49 (2013) 9543–9545.
[32] A. Sudo, L.-C. Du, S. Hirayama, T. Endo, Substituent effects of N-alkyl groups
on thermally induced polymerization behavior of 1,3-benzoxazines, J. Polym.
Sci. A: Polym. Chem. 48 (2010) 2777–2782.
[33] S. Bagherifam, T. Uyar, H. Ishida, J. Hacaloglu, The use of pyrolysis mass
spectrometry to investigate polymerization and degradation processes of
methyl amine-based benzoxazine, Polym. Test. 29 (2010) 520–526.
 M. Poorteman et al. / Progress in Organic Coatings 97 (2016) 99–109
[34] S. Bagherifam, T. Uyar, H. Ishida, J. Hacaloglu, Investigation of polymerization
of benzoxazines and thermal degradation characteristics of polybenzoxazines
via direct pyrolysis mass spectrometry, Polym. Int. 61 (2012) 1532–1541.
[35] X. Wang, F. Chen, Y. Gu, Influence of electronic effects from bridging groups on
synthetic reaction and thermally activated polymerization of bisphenol-based
benzoxazines, J. Polym. Sci. A: Polym. Chem. 49 (2011) 1443–1452.
[36] D. Allen, H. Ishida, Effect of phenol substitution on the network structure and
properties of linear aliphatic diamine-based benzoxazines, Polymer 50 (2009)
613–626.
[37] A. Sudo, R. Kudoh, H. Nakayama, K. Arima, T. Endo, Selective formation of poly
(N,O-acetal) by polymerization of 1,3-benzoxazine and its main chain
rearrangement, Macromolecules 41 (2008) 9030–9034.
[38] P. Chutayothin, H. Ishida, Cationic ring-opening polymerization of
1,3-benzoxazines: mechanistic study using model compounds,
Macromolecules 43 (2010) 4562–4572.
[39] V.M. Russell, J. Koenig, H. Low, H. Ishida, Study of the characterization and
curing of benzoxazines using 13C solid-state nuclear magnetic resonance, J.
Appl. Polym. Sci. 70 (1998) 1413–1425.
[40] C. Liu, D. Shen, R.-M. Sebastiaı́n, J. Marquet, R. Schönfeld, Mechanistic studies
on ring-opening polymerization of benzoxazines: a mechanistically based
catalyst design, Macromolecules 44 (2011) 4616–4622.
[41] Friedrich Kremer, Andreas Schönhals (Eds.), Broadband Dielectric
Spectroscopy, Springer, 2013, p. 359.
[42] P. Zoltowski, On the electrical capacitance of interfaces exhibiting
constantphase element behaviour, J. Electroanal. Chem. 443 (1998) 149–154.
109
[43] M. Dornbush, S. Kirsch, C. Henzel, C. Deschamps, S. Overmeyer, K. Cox, M.
Wiedow, U. Tromsdorf, M. Dargatz, U. Meisenburg, Characterization of the
water uptake and electrolyte uptake of organic coatings and the consequence
by means of electrochemical impedance spectroscopy and UV–vis
spectroscopy, Prog. Org. Coat. 89 (2015) 332–343.
[44] Hongwei Shi, En-Hou Han, S.V. Lamaka, M.L. Zheludkevich, Fuchun Liu, M.G.S.
Ferreira, Cerium cinnamate as an environmentally benign inhibitor pigment
for epoxy coatings on AA 2024-T3, Prog. Org. Coat. 77 (2014) 765–773.
[45] D. Zhang, H. Qian, L. Wang, X. Li, Comparison of barrier properties for a
superhydrophobic epoxy coating under different simulated corrosion
environments, Corros. Sci. 103 (2016) 230–241.
[46] F. Cotting, I.V. Aoki, Smart protection provided by epoxy clear coating doped
with polystyrene microcapsules containing silanol and Ce(III) ions as
corrosion inhibitors, Surf. Coat. Technol. (2015), available online 22.11.15.
[47] Zongxue Yu, Haihui Di, Yu Ma, Liang Lv, Yang Pan, Chunli Zhang, Fabrication of
graphene oxide-alumina hybrids to reinforce the anti-corrosion performance
of composite epoxy coatings, Appl. Surf. Sci. 351 (2015) 986–996.
[48] D.M. Brascher, A.H. Kingsbury, Electrical measurements in the study of
immersed paint coatings on metal I. Comparison between capacitance and
gravimetric methods of estimating water-uptake, J. Appl. Chem 4 (1954) 62.
[49] A. Amirudin, D. Thierry, Application of electrochemical impedance
spectroscopy to study the degradation of polymer-coated metals, Prog. Org.
Coat. 26 (2) (1995) 1–28.
[50] S. Morsch, S. Lyon, S.D. Smith, Mapping water uptake in an epoxy-phenolic
coating, Prog. Org. Coat. 86 (2015) 173–180.