- Email: [email protected]
On the development of polypyrrole coatings with self-healing properties for iron corrosion protection
Corrosion Science 47 (2005) 3216–3233 www.elsevier.com/locate/corsci
On the development of polypyrrole coatings with self-healing properties for iron corrosion protection G. Paliwoda-Porebska a,*, M. Stratmann a, M. Rohwerder a, K. Potje-Kamloth b, Y. Lu b, A.Z. Pich b, H.-J. Adler b a b
Max Planck Institute for Iron Research, Max-Planck-Strasse 1, 40237 D€ usseldorf, Germany Institute of Macromolecular Chemistry and Textile Chemistry, Technical University Dresden, D-01062 Dresden, Germany Available online 12 September 2005
Abstract This paper presents studies on the efficacy and on the limits of polypyrrole (Ppy) doped with either MoO24 or [PMo12O40]3 as self-healing corrosion protecting coatings. The kinetics of the cathodic delamination were studied by means of the Scanning Kelvin Probe (SKP). This method, in combination with cyclic voltammetry, UV–visible spectroscopy (UV–vis) and X-ray photoelectron spectroscopy (XPS), shows a potential driven anion release from the Ppy coating that results in an inhibition of the corrosion process taking place in the defect. Thus, an intelligent release of inhibitor occurs only when the potential at the interface decreases. Inhibitor anions are released only due to an active defect. However, the release mechanism can be easily negatively affected by the presence of small cations and/or by too high pH values at the buried interface. Hence, such a self-healing coating has to be carefully designed in order to ensure an effective performance. 2005 Elsevier Ltd. All rights reserved.
*
Corresponding author. Tel.: +49 211 6792538; fax: +49 211 6792218. E-mail address: [email protected] (G. Paliwoda-Porebska).
0010-938X/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.corsci.2005.05.057
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3217
Keywords: Scanning Kelvin Probe; Polypyrrole; Intelligent release; Corrosion inhibitors; Cathodic delamination; Self-healing coating
1. Introduction In the last 20 years, in numerous publications, intrinsically conducting polymers (ICPs) have been controversially discussed as possible corrosion preventing additions to organic layers [1–4]. One of the reasons for the reported discrepancies is that most papers are focused just on very restricted aspects of corrosion. For instance, an ICP containing coating might be very effective in preventing corrosion in pin holes, while the same coating might fail disastrously in the presence of larger defects [4–6]. Quite numerous possible corrosion protection mechanisms are proposed [7–15]. According to Barisci et al. [16] ICPs might be used as intelligent release systems for corrosion inhibitors on demand. In this work was for the first time pointed out that the polymer could be reduced and consequently dopants could be set free as a result of a galvanic coupling between the corroding metal (iron) and ICP (polyaniline). A possible mechanism for the case of polyaniline was also described by Kendig et al. [17]. Such a coating which releases anions only in the case of corrosive attack but prevents them from being leached out or being released by ion exchange processes, would be a self-healing coating. Nevertheless, no results showing successful dopant release, as a consequence of electrochemical ICP reduction through the corrosion in the defect have been reported until now. This paper shows that on the basis of polypyrrole coatings, which show a electrochemical release of inhibitor anions can be prepared. Even unusually large defects can be passivated because the switch for anions release is the decrease in potential at the metal/polymer interface during the delamination. If the defect is smaller less coating will delaminate, if it is larger more coating delaminates until successful passivation. Some results concerning a positive influence of aminotris(methylene phosphonic acid) (ATMP) anion doped polyaniline (PANI) on the corrosion behaviour of mild steel have been reported by Kinlen et al. [18,19]. Release of ATMP from the PANI and formation of stable adducts with iron ions have been observed. These adducts could lead to the passivation of pinholes. According to the proposed mechanism, the ATMP is released as a consequence of polyaniline–emeraldine salt (PANI–ES) reduction to polyaniline–leucoemeraldine salt (PANI–LS). The number of anions in the polymer chain does not change in the mentioned reaction [16,17,20] and this might be one of the critiques concerning this proposed mechanism. Additionally, it is worth of mentioning that the PANI–LS form is stable only at low pH and that during cathodic delamination strongly alkaline pH prevail. Hence, a high pH induced deprotonation of PANI–ES to the form of polyaniline–emeraldine base (PANI– EB) or reduction to the form of polyaniline–leucoemeraldine base (PANI–LB) is more likely to take place under these conditions. Also, in the most recent publications [21,22] concerning this subject, no observations clearly explaining the mechanism of anion release from PANI–ES in corrosion experiments have been reported.
3218
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
A defect, that penetrates the polymer coating to the bare metal, is a starting point for the electrochemically driven delamination. In the case of the here studied cathodic delamination the degradation at the metal/polymer interface is caused by radicals produced by oxygen reduction reaction at the metal/polymer interface (Fig. 1). Ions, water and oxygen can diffuse in sufficient quantities through any organic coating, thus providing the environment for oxygen reduction. At the isolated intact interface, however the anodic counter reaction is missing, i.e. no electrons are available for the oxygen reduction. Hence, the potential at the intact interface stabilizes at quite high potentials, where oxygen reduction is kinetically inhibited. However, in the presence of defect, a galvanic element is formed due to coupling between defect and delamination site. A closed electric circuit for the electron flow from the defect to the polymer/metal interface is established by cation migration from the defect to the buried interface (see e.g. [23]). The potential is pulled down and oxygen reduction fuelled by part of the iron dissolution in the defect can than take place, thus leading to a further degradation of the interface. This also facilitates cation transport and decreases the iR drop along the delaminated interface. Further details concerning the cathodic delamination process can be found in e.g. [24,25]. Results reported in this paper show that polypyrrole (Ppy), doped with [PMo12O40]3 , releases anions as a result of polymer reduction, which after decomposition yield MoO24 and HPO24 . Subsequently, mainly MoO24 as being an inhibitor of iron dissolution, is able to passivate the defect in the coating. The high pH arising at the metal/polymer interface during the delamination process might have an influence on the mechanism of anion release from Ppy. Numerous reports concerning OH interactions with Ppy can be found in the literature (see e.g. [26–28]). Some authors studied, for example with XPS, the changes induced by the high pH in the Ppy and their influence on a possible anion release or trapping reaction [26,29]. In this paper we also try to discuss the probable OH impact on the anion release process. Molybdate (VI) is well known as efficient and environmental friendly corrosion inhibitor for iron/steel [30,31]. Also in the screening corrosion inhibitor tests per-
Fig. 1. Scheme of the cathodic delamination process.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
[PMo12O40]3-
OHH+
[PMo11O39]7+ MoO42-
OHH+
[P2Mo5O23H]5+
OH-
HPO42-
H+
MoO42-
3219
MoO42-
Fig. 2. [PMo12O40]3 degradation reactions as a function of hydrogen ion concentration [34].
formed by ourselves MoO24 , next to PO24 , has shown the most positive corrosion inhibition effect. Application of Na3PMo12O40 as a corrosion inhibitor for stainless steel have also been reported in the literature [32,33]. One important aspect of [PMo12O40]3 anion chemistry has to be kept in mind while using it as a corrosion inhibitor in aqueous solutions, i.e. [PMo12O40]3 is stable only in highly acidic pH. At pH above 4 it undergoes several degradation reactions of which the final products are MoO24 and HPO24 (Fig. 2) [34]. This means that the inhibition effect attributed to [PMo12O40]3 is actually due to the MoO24 and HPO24 present in the system [35]. The Scanning Kelvin Probe (SKP) technique applied in this study is a non-contact, non-destructive technique that after suitable calibration allows the measurement of the electrode potential at the buried metal/polymer interface. For details concerning the technique see [15,36–38]. 2. Experimental 2.1. Materials and sample preparation Stable water dispersions of Ppy, doped with S2 O28 ; MoO24 and [PMo12O40]3 , have been prepared by chemical polymerization in the Institute of Macromolecular Chemistry and Textile Chemistry at the Technical University Dresden (Prof. H.-J. Adler) by Dr. A.Z. Pich and Dr. Y. Lu. The method has been described in detail elsewhere [39]. Ppy dispersion doped with the S2 O28 , which is not a corrosion inhibitor, has been used as a reference sample. All used dispersions are of the so-called core type, which means that the Ppy has polymerised in a form of sphere with diameter between 80 and 130 nm. For all experiments, iron samples (C 0.0053/Si 0.0071/Mn 0.0024/P <0.0001/S <0.0010/O 0.0029/Al 0.0015 wt%) 20 mm· 10 mm · 2 mm in size, were ground up to 1200 grid paper, cleaned in acetone and then, ethanol in ultrasound bath for 3 min. The Ppy dispersions have been mixed with a non-conducting matrix polymer provided by Chemetall GmbH (Frankfurt/Main). Prepared in this way the mixture was applied by use of a spin-coater on iron samples. Subsequently, the samples were cured at 100 C for 5 min. In the case of samples prepared for the SKP experiments an additional, polyacryl resin top coat (BASF) was applied with the use of the spin-coater. There are two reasons for applying the top coat. The first is that the studied Ppy coatings are meant as a primer in combination with further layers in a complex coating. Hence, by applying the top coat we get closer to the real situation. The second reason is that a very thin coating is lifted up by undercreeping electrolyte during the delamination [40], thus causing folds, which make
3220
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
the SKP measurement difficult, and also introducing an additional mechanical factor to the delamination. 2.2. Ppy films characterization SEM (LEO 1550 VEP) investigations on the prepared films revealed that after a single application no homogeneous film was obtained at the iron surface (Fig. 3a). (The iron surface is only partially covered by Ppy spheres.) A second application of Ppy dispersion did not cause washing away of the already existing layer but some surface areas were additionally filled (Fig. 3b). Application of the third layer filled the empty spaces that remained still available (Fig. 3c). Further spin-coating of Ppy dispersion showed no further changes in the SEM images. These observations indicate that the prepared Ppy composite coating is composed of only one monolayer of Ppy core particles, distributed quite evenly on the iron surface, the matrix polymer acting as glue between them. These observations are also confirmed by investigating a cross-section of the iron sample coated by PpyPMo composite (Fig. 4). The estimated thickness of the film is close to the diameter of the Ppy sphere i.e. from 80 to 130 nm. This conclusion is also supported by results of XPS (Quantum
Fig. 3. SEM images of Ppy dispersion coating morphology. After the dispersion has been applied (a) once, (b) twice and (c) three times.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3221
Fig. 4. SEM image of the perpendicular cut of iron sample coated with Ppy composite.
2000, Physical Electronics) sputtering and ellipsometric measurements (SE 800, Sentech Instruments). 2.3. SKP experiments The SKP used in this work was constructed in the Max Planck Institute for Iron Research. The electronic parts of the setup, consisting of a preamplifier, lock-inamplifier and an integrator were provided by commercial suppliers. A similar SKP setup has been successfully used in other research projects (see e.g. [41–43]). The used SKP tip has a diameter of 100 lm. 2.3.1. SKP delamination experiments In order to perform SKP experiments, part of iron sample was covered with the dispersion coating and additional top coat. On the other part of the sample an artificial defect has been prepared (Fig. 5). In the defect 0.1 M KCl solution has been injected. All measurements were performed in a humid atmosphere (RH = 93– 95%). After the SKP experiments, the delaminated Ppy coatings could be easily peeled off from the metal surface. XPS analysis of the delaminated Ppy films and of the surface of corroded iron samples was then performed. 2.3.2. SKP reduction experiments To evaluate the effect of the high pH present at the metal/polymer interface during the cathodic delamination, on the anion release from the Ppy composite coatings, reduction experiments in controlled N2 atmosphere have been performed. During these experiments no increase of the pH at the metal/polymer interface can take place and hence, the influence of OH on the release mechanism is excluded.
3222
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
Fig. 5. Sample preparation for the SKP experiments.
2.4. X-ray photoelectron spectroscopy To identify the chemical species at the delaminated iron samples X-ray photoelectron spectroscopy was used. All measurements were performed with a use of the Quantum 2000 Scanning ESCA Microprobe System from Physical Electronics. The X-rays employed in the used system, are Al Ka emitted by an aluminium anode. The take off angle, at which emitted electrons were analysed, was 45. All binding energies given here refer to the C(1s) signal of aliphatic hydrocarbon species with a value of 285.0 eV. 2.5. Electrochemical characterization Cyclic voltammetry and reduction (polarization) experiments were performed with a gold plate as counter electrode and an Ag/AgCl reference electrode in 50 ml of 0.1 M LiClO4, 0.1 M (C4H9)4NCl or 0.1 M NaClO4. A gold (evaporated on glass) substrate coated with the Ppy composite was used as the working electrode. No additional top coat was applied on these samples. All the experiments were carried out after the electrolyte was purged with N2 for 30 min. Cyclic voltammograms were performed by using a HEKA potentiostat with a scan rate of 50 mV/s. 2.6. Inhibitor tests Polarization experiments, at a sweep rate of 0.2 mV/s, were carried out in 10 ml aerated 0.1 M NaCl with 0.01 M Na2MoO4 or Na3PO4 solution or in 0.1 M NaCl by using a Zahner IM6d potentiostat. All the experiments were carried out with a gold plate as a counter electrode, Ag/AgCl reference electrode and iron plate as a working electrode. The iron samples were stored in the electrolyte for 3 h, until the open circuit potential became stable. As next, in separate experiments, anodic and cathodic polarization curves were recorded.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3223
3. Results 3.1. Molybdate (VI) and phosphate (V) as iron corrosion inhibitors Prior to the experiments with the Ppy coatings, polarisation experiments on iron in solutions containing different inhibitor anions were performed with the aim to assess the inhibition effects of prospective anions. The subsequent Tafel analysis was performed (Fig. 6). The obtained results indicated that especially molybdate (VI) and phosphate (V) seem to be promising. Both anions cause a large anodic shift of the corrosion potential (DEcorr = 600 mV) and lower the rate of iron dissolution significantly. The corrosion current density is two orders of magnitude smaller in the inhibitor containing solution compared to pure NaCl solution. From the shape of the cathodic curve it can be also seen that the oxygen reduction reaction is not influenced by the presence of MoO24 and PO34 ions. 3.2. Voltammetric characterization of PpyPMo12O40 and PpyMoO4 films Fig. 7 shows cyclic voltammograms (CV) of the PpyPMo12O40 (PpyPMo) and PpyMoO4 (PpyMo) films. Obviously, the prepared composite films are electrochemically active, i.e. the non-conducting matrix polymer does not inhibit the electrochemical activity of the Ppy. For both coatings the reduction and oxidation peaks positions are similar. Important difference is that in the case of PpyPMo coatings are always higher current densities measured. That is most probably connected with
3 2
(1)
log | i| / µ A cm-2
1 0 (2)
-1
(3)
-2 -3 -4 -0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
E / mV vs. AgACl
Fig. 6. Polarization curves of a iron electrode in 0.1 M NaCl (1) and in 0.1 M NaCl with addition of 0.01 M Na3PO4 (2) and 0.01 M Na2MoO4 (3).
3224
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
-5
1.0x10
i / A cm
-2
0.0
-5
-1.0x10
-5
-2.0x10
-5
-3.0x10
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
E / V vs. SHE Fig. 7. Cyclic voltammograms of PpyPMo (- - -) and PpyMo (—) performed in 0.1 M LiClO4.
the inherent electroactivity of [PMo12O40]3 anion in the applied potential range which may contribute to the measured current densities. In the first cycle of the CV, in both, PpyPMo as well as PpyMo containing coatings, significant reduction peaks can be observed, which decrease in the next cycle. As XPS analysis of the coating and UV–vis analysis of the electrolyte show, these peaks (at least some of them) are connected with Ppy reduction and the corresponding anion release process. After the first cycle, the peaks are much smaller thus indicating that the reduction–oxidation reactions of Ppy are not completely reversible in this electrolyte. Both coatings show a quite different behaviour in CVs performed in tetrabutylammonium chloride solution (Fig. 8), where the large size of the cation molecule makes its incorporation during the Ppy reduction rather improbable [44]. In the case of PpyPMo, the reduction–oxidation peaks change significantly their position after the few first cycles. This indicates that [PMo12O40]3 are systematically replaced by Cl anions, which are available in the solution in high concentration. Another important observation is that the size of the peaks does not change significantly with the number of cycles, in contrast to changes observed in CV performed in LiClO4. This observation indicates that in tetrabutylammonium chloride, where only anion exchange between the coating and electrolyte can take place, PpyPMo undergoes well reversible reduction–oxidation reactions. In the case of the PpyMo containing coating, peaks positions do not change, although probably, as in this case of the PpyPMo, the MoO24 =Cl anion replacement takes place. One explanation for this might be that these two anions do not differ so dramatically in structure as [PMo12O40]3 and Cl and hence, no significant
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3225
-5
2.0x10
-5
i / Acm
-2
1.0x10
0.0
-5
-1.0x10
-5
-2.0x10
-0.6
-0.4
-0.2
(a)
0.0
0.2
0.4
0.6
0.8
E / V vs. SHE -5
1.5x10
-5
1.0x10
-6
i / A cm
-2
5.0x10
0.0
-6
-5.0x10
-5
-1.0x10
-5
-1.5x10
-0.6
(b)
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
E / V vs. SHE
Fig. 8. Cyclic voltammograms of (a) PpyPMo and (b) PpyMo performed in 0.1 M tetrabutylammonium chloride: (—) first two cycles, (- - -) cycles after 15 min.
changes in the cyclic voltammogram with number of cycles have been observed, indicating that also here redox processes are reversible.
3226
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3.3. Polarization–reduction experiments To confirm and to quantify the anion release, the following polarization experiments have been performed, Ppy composite coatings were prepared on gold substrates and then samples were polarised to 0.4 V vs. SHE (iron corrosion potential) in NaClO4 for 10–15 min. After the experiments the electrolyte was analysed with the use of UV–visible spectroscopy to detect the presence of released anions. In the case of the electrolyte in which PpyPMo coatings were reduced, by taking the advantage of the inherent electrochemical activity of [PMo12O40]3 which makes it possible to identify the characteristic [PMo12O40]3 reduction–oxidation peaks, also cyclic voltammograms were performed. From the obtained cyclic voltammogram (Fig. 9) the peaks which can be assigned to [PMo12O40]3 , can be recognised. The current densities of these peaks are quite small. A separate reference measurement, with known amounts of Na3PMo12O40 which have been injected into the electrolyte, shows that they correspond to approximately 0.5 lM of [PMo12O40]3 concentration (equivalent to release of 0.025 lmole). The theoretically calculated number of [PMo12O40]3 anions, which should be released from a monolayer of PpyPMo spheres, is approximately two times smaller than the experimentally found. There were two important assumptions made during the theoretical calculation, which may result in the difference between the experimentally and theoretically found values. The first concerns the assumed doping rate, in this case
peaks after addition of 25µl of 10-3 M Na3PMo12O40 solution
-5
1.0x10
bigger peaks appear after next injection of 25µl
-6
5.0x10
i / A cm
-2
0.0
-6
-5.0x10
-5
-1.0x10
-5
-1.5x10
-5
-2.0x10
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
E / V vs. SHE Fig. 9. Cyclic voltammogram performed in 0.1 M LiClO4, (—) electrolyte in which PpyPMo reduction was performed and ( ) with addition of 10 3 M Na3PMo12O40.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3227
pyrrole:anion = 1:4, and the second the maximal volume fraction of cores in the monolayer is estimated to be 60%. However, the SEM investigation of the Ppy coating cross-section revealed that at some points the coating is thicker than a monolayer. A precise quantification is difficult to make, as this does not show up in the top view SEM images. In any case all particles seem to be active. In the UV–vis spectrum recorded for this electrolyte, an adsorption peak at approximately 210 nm and small adsorption peaks at 310 and 330 nm, corresponding to the [PMo12O40]3 anion adoption bands [45], can be found. The UV–vis spectrum obtained from the electrolyte in which PpyMo reduction has been performed shows a broad adsorption shoulder at 220–230 nm and a small peak around 260 nm, which can be assigned to MoO24 [46,47]. This indicates that the MoO24 was released from the polymer matrix. These results indicate that both [PMo12O40]3 and MoO24 can be released from Ppy coating by electrochemical reduction. When the PpyPMo or PpyMo sample has been just stored in the electrolyte (i.e. without electrochemical reduction) anions were not present in the solution. 3.4. SKP experiments
200
300
100
200 100
0
E / mV vs. SHE
E / mV vs. SHE
Fig. 10 shows the delamination behaviour of the different Ppy coatings. As can be seen, at the upper right side of each diagram, the potential of the intact Ppy coating
-100 -200 -300
0 -100 -200 -300 -400
-400 -500
(a)
-500 0
-600
200 400 600 800 1000 1200 1400 1600 1800
0
200 400 600 800 1000 1200 1400 1600 1800 2000
(b)
d / µm
d / µm
200 100
E / mV vs. SHE
0 -100 -200 -300 -400 -500
0
200 400 600 800 1000 1200 1400 1600 1800
(c)
d / µm
PpyS2 O28
Fig. 10. SKP delamination profiles: (a) -reference sample, (b) PpyMo and (c) PpyPMo. Intervals between profiles are in case of all samples the same and equal, 20 min.
3228
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
remains between 0.1 and 0.2 V vs. SHE. The border between defect and coating is at d = 0. Profiles moving towards the right side of the diagram represent the propagation of the delamination front, from the defect into the intact polymer coating on the right side (see also Fig. 5). Intervals between the profiles (lines) are in case of all samples the same and equal, 20 min. The reference sample as well as the Ppy coating doped with the MoO24 delaminates fast, i.e. hundreds of micrometers in a few minutes. No passivation of the defect could be observed. In the case of Ppy doped with [PMo12O40]3 (Fig. 9c) a different behaviour is observed. The first profile does not differ from the one measured in the case of the other samples, but already the second one shows a significant increase of the potential, DE = +0.17 V, in the delaminated area. Consequently, the potential difference between the delaminated area and the intact polymer coating, which is the driving force for a further delamination [23], is getting smaller. This induces a significant decrease in the delamination speed. The potential in the already delaminated area changes further and becomes even more positive with time. The change is in the case of the presented sample 0.28 V, and usually being between 0.15 V and 0.3 V. This observation shows a good agreement with a separate measurement of the open circuit potential at the defect, by using a micro-reference electrode [48], where also a significant anodic potential shift equal 0.17 V has been observed.
4. Discussion In the following section the results concerning delamination experiments on Ppy coatings will be discussed. 4.1. PpyPMo XPS analysis performed at the surface of the iron sample shows the presence of molybdate at the defect as well as at coating peeled off the delaminated area. The amount of molybdate varies in the range between 1% and 3% (atomic per cent) for individual experiments. Since from one released [PMo12O40]3 anion only 1 mole of phosphate (V) arises, compared to 12 moles of molybdate, it was usually impossible to identify it with XPS. Only in the case of two samples less then 1% of phosphate have been found at the metal surface. In the delaminated Ppy film K+ can be found, thus indicating that parallel to the anion release reaction, to some extent, cation incorporation takes place. It has been already reported [49,50] that during the reduction process of Ppy, not only anion release takes place but also cation incorporation is usually possible for the charge compensation. This parallel reaction is of course not desired, since it decreases the amount of released anions and thus the achievable inhibition effect. The XPS analysis shows that the Mo 3d peak found at the interface appears within the typical binding energy range for MoVI compounds [51,52]. The peak is very slender (Fig. 11) indicating the presence of only one molybdate oxidation state. Also the molybdate found in the defect is clearly MoVI.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3229
2
x 10 24 22 20
CPS
18 16 14 12 10 8 240
238
236
234
232
230
228
226
Binding Energy (eV) Fig. 11. XPS spectra of molybdenum (3d3/2, 3d5/2) found under PpyPMo polymer coating delaminated in O2 atmosphere.
According to observations reported in the literature [53–55], the role of the MoO24 as corrosion inhibitor is, next to other effects, strengthening of the native iron oxide layer by incorporating and healing its defects. Hence, it is possible that the molybdate released at the metal/polymer interface is, to some extent, already incorporated into the iron oxide. According to some authors [27,28,56] OH ions being strong nucleophiles can react with Ppy. As a consequence of this interaction, Ppy can be overoxidised and can release anions from the polymer chain (Fig. 12). Knowing that during the delamination experiments Ppy is exposed to OH attack, and the observed anion release might be due to the mentioned attack and not to the electrochemical reduction of the Ppy, additional experiments were performed. The PpyPMo coating was delaminated, more precisely said reduced, in controlled N2 atmosphere. Also in this case XPS analysis showed the presence of molybdate at the iron surface. This result shows that the [PMo12O40]3 anions are released by a truly electrochemical reduction and not due to interaction with OH ions. During the reduction experiments in N2 atmosphere, due to a lack of OH , no significant decomposition of the [PMo12O40]3 anion should take place. Indeed, a H AN +
N
OH-
N H
N
O
Fig. 12. Nucleophilic attack of OH on Ppy.
3230
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
Table 1 Binding energies (eV) of measured Mo 3d3/2 and Mo 3d5/2 peaks Description of measured Mo 3d peak
Molybdate oxidation state
Binding energy (eV) Mo 3d3/2
Binding energy (eV) Mo 3d5/2
Molybdate at the metal surface/delamination experiments in O2 atmosphere Molybdate in the defect area/delamination experiments in O2 atmosphere Molybdate in the Ppy[PMo12O40] film Molybdate in the Na3[PMo12O40] powder Molybdate at the metal surface/reduction experiments in N2 atmosphere
+6
234.7
231.5
+6
234.5
231.4
+6 +6 +6
235.8 235.1 235.0
232.7 231.9 231.9
significant shift of the Mo 3d peak, in comparison to the position of peak observed in the case of samples delaminated in O2 atmosphere, is observed (Table 1). This indicates that a different molybdate form exists at the metal/polymer interface. To assign molybdate spectra properly, reference XPS spectra of non-delaminated PpyPMo coatings on iron and Na3PMo12O40 powder have been measured. The comparison of the Mo 3d peak positions indicates, that in the case of coating reduction in N2 atmosphere [PMo12O40]3 anions are released and can be identified at the iron surface. According to these observations the following sequence of processes can be proposed: due to the potential decrease at the interface, caused by the delamination, i.e. more precisely by the related ohmic contact with the corroding defect, the [PMo12O40]3 anion is firstly released from the thusly reduced Ppy chain; subsequently this anion reacts with the OH ions available at the iron/polymer interface, yielding mainly MoO24 , which is responsible for the observed defect passivation. Although the ability of these coatings to passivate even quite large defects and to stop delamination is obvious, it is also obvious that it takes some time to release enough anions. Clearly, for technical applications the coatings need to be significantly improved. One approach would be to prepare thicker coatings. However, preliminary experiments suggest that thicker coatings cannot be prepared with such a high content of Ppy particles, and for lower contents the conductivity and electrochemical activity of the coatings decrease significantly. These questions are topics of further research. 4.2. PpyMo Also, one would assume that the direct use of the smaller MoO24 anion should result in a much better performance. There is no need for an additional decomposition reaction, and also the release and transport should be easier. This means that the competition with the parallelly occurring cation incorporation, which decreases the amount of releasable anions, should be favourably affected. But as already mentioned, the coatings containing Ppy doped with MoO24 show no inhibition of the delamination process. XPS analysis shows that after the delamination experiment no MoO24 anion is present at the iron/polymer interface or in the
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3231
defect, meaning that no measurable anion release takes place. Two hypotheses have been considered to explain this observation. The first hypothesis refers to the possibility that, contrary to the just discussed, cation incorporation is predominant in the PpyMo system, which would explain the remaining of MoO24 anions in the polymer matrix. However, XPS analysis shows that only small amounts of K+ are incorporated inside of the delaminated film. The second hypothesis considers the influence of OH on the release mechanism. A significant difference between the PpyMo and the PpyPMo release system is the anion decomposition reaction [34] that takes place after the [PMo12O40]3 anion is set free from PpyPMo. In this reaction OH is being consumed, which buffers the usually high pH present at the iron/polymer interface during the delamination process to lower values. The hypothesis proposed here is that, if the [PMo12O40]3 anion decomposition reaction provides this kind of buffering effect, in the case of PpyMo coating the polypyrrole is exposed to a much higher pH existing at the delaminating interface. That may cause a faster and more efficient Ppy overoxidation process. It is possible that depending on conditions and the used Ppy dopants, as a consequence of the overoxidation process the dopants will stay trapped in the non-conducting, overoxidised Ppy matrix [26]. To analyze if in the case of Ppy doped with MoO24 anions this undesired OH influence, resulting in anion immobilization in the coating, may be the reason for its inability to release the anions under ambient conditions, reduction experiments of the PpyMo coating in controlled N2 atmosphere were performed. The results of the corresponding XPS analysis of these samples clearly show that under these conditions the MoO24 anion is indeed released from the polymer coating. Analysis of the changes in the C 1s and O 1s detail spectra measured on the thusly reduced coating show a significantly lower amount of C@O and O–C@O groups, compared to the same coating delaminated in oxygen. This observation strongly supports the hypothesis that the PpyMo coating is overoxidised during the delamination in air, with involvement of OH , which results in the MoO24 anions being trapped in the Ppy matrix. One might think that for a further verification of this hypothesis a comparison of changes of the C 1s and O 1s spectra between the PpyPMo and PpyMo, both delaminated in oxygen atmosphere, would be informative. Although the [PMo12O40]3 decomposition reaction obviously provides the mentioned buffering effect, that lasts long enough for successful anions release, the buffering effect certainly decreases with time. But while the amount of anions responsible for the buffering effect is limited, this is not the case for the amount of OH that may be produced at the iron/polymer interface (unless the inhibition of the defect occurs much faster than is the case now). Hence, the XPS analysis of the C 1s and O 1s spectra on delaminated PpyPMo coatings also shows a presence of some amounts of C@O and O–C@O groups.
5. Conclusions The presented SKP results show that in the case of the Ppy doped with the [PMo12O40]3 anion, a significant inhibition of the delamination is provided by on
3232
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
demand functioning release. This result is supported by the observation of a partial passivation of the defect, where due to MoO24 anion accumulation the iron dissolution is inhibited. Additionally presented results confirm a truly electrochemical release of [PMo12O40]3 anions, not influenced by high pH at the metal/polymer interface. The second studied coating, doped with the MoO24 anions, showed no inhibition effect on the delamination process. According to the proposed mechanism this results from Ppy overoxidation, due to the highly alkaline pH at the buried metal/polymer interface. In the case of the PpyPMo coating, the [PMo12O40]3 anion decomposition reaction plays an essential role in the successful delamination inhibition. This reaction provides a significant buffering effect, preventing too high pH values at the metal/polymer interface. In the case of the Ppy doped with MoO24 coating this buffering effect cannot appear. Resuming, a Ppy release system, unaffected by leaching and pH changes, that safely stores the corrosion inhibitor ions can be used as an intelligent corrosion protecting coating. However, the influence of conditions during the cathodic delamination process on the release behaviour, especially the effect of small easy incorporable cations and the high pH, needs further investigations.
Acknowledgement The financial support from the Bundesministerium fu¨r Buildung und Forschung is gratefully acknowledged.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]
J.O. Iroh, W.C. Su, Synthetic Met. 114 (2000) 225. N.V. Krstajic, B.N. Grgur, S.M. Jovanovic, M.V. Vojnovic, Electrochim. Acta 42 (1997) 1685. W.K. Lu, R.L. Elsenbaumer, B. Wessling, Synthetic Met. 71 (1995) 2163. A.A. Pud, G.S. Shapoval, P. Kamarchik, N.A. Ogurtsov, V.F. Gromovaya, I.E. Myronyuk, Y.V. Kontsur, Synthetic Met. 107 (1999) 111. K. Idla, J. Reut, A. Opik, Synthetic Met. 102 (1999) 1392. G.M. Spinks, A.J. Dominis, G.G. Wallace, D.E. Tallman, J. Solid State Electron. 6 (2002) 85. D.W. Deberry, J. Electrochem. Soc. 132 (1985) 1022. V.T. Truong, P.K. Lai, B.T. Moore, R.F. Muscat, M.S. Russo, Synthetic Met. 110 (2000) 7. B. Wessling, Adv. Mater. 6 (1994) 226. N. Ahmad, A.G. MacDiarmid, Synthetic Met. 78 (1996) 103. T.D. Nguyen, M. Keddam, H. Takenouti, Electrochem. Solid St. 6 (2003) B25. T. Schauer, A. Joos, L. Dulog, C.D. Eisenbach, Prog. Org. Coat. 33 (1998) 20. P.J. Kinlen, Y. Ding, B.C. Silverman, Corrosion 58 (2002) 490. J.D. Stenger-Smith, Prog. Polym. Sci. 23 (1998) 57. R.J. Holness, G. Williams, D.A. Worsley, H.N. McMurray, J. Electrochem. Soc. 152 (2005) B73. J.N. Barisci, T.W. Lewis, G.M. Spinks, C.O. Too, G.G. Wallace, J. Intel. Mater. Syst. Str. 9 (1998) 723. M. Kendig, M. Hon, L. Warren, Prog. Org. Coat. 47 (2003) 183. P.J. Kinlen, D.C. Silverman, C.R. Jeffreys, Synthetic Met. 85 (1997) 1327. P.J. Kinlen, V. Menon, Y.W. Ding, J. Electrochem. Soc. 146 (1999) 3690.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233 [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56]
3233
D.E. Tallman, G. Spinks, A. Dominis, G.G. Wallace, J. Solid State Electron. 6 (2002) 73. A. Cook, A. Gabriel, D. Siew, N. Laycock, Curr. Appl. Phys. 4 (2004) 133. J.E. Pereira da Silva, S.I. Cordoba de Torresi, R.M. Torresi, Corros. Sci. 47 (2005) 811. M. Rohwerder, M. Stratmann, MRS Bull. 24 (1999) 43. W. Furbeth, M. Stratmann, Prog. Org. Coat. 39 (2000) 23. A. Leng, H. Streckel, M. Stratmann, Corros. Sci. 41 (1999) 579. F. Palmisano, C. Malitesta, D. Centonze, P.O. Zambonin, Anal. Chem. 67 (1995) 2207. Y.F. Li, R.Y. Qian, Electrochim. Acta 45 (2000) 1727. H.L. Ge, G.J. Qi, E.T. Kang, K.G. Neoh, Polymer 35 (1994) 504. C. Malitesta, I. Losito, L. Sabbatini, P.G. Zambonin, J. Electron Spectrosc. 76 (1995) 629. M.S. Vukasovich, J.P.G. Farr, Polyhedron 5 (1986) 551. V.A. Kumari, K. Sreevalsan, S.M.A. Shibli, Corros. Prevent. Contr. 48 (2001) 83. E.A. Lizlovs, J. Electrochem. Soc. 114 (1967) 1015. A. Weisstuch, C.E. Schell, Corrosion 28 (1972) 299. G.A. Tsigdinos, Topics Curr. Chem. 76 (1978) 1. D.M. Basher, J.E. Rhoades-Brown, Br. Corros. J. 8 (1973) 50. M. Stratmann, H. Streckel, Corros. Sci. 30 (1990) 681. M. Stratmann, H. Streckel, Corros. Sci. 30 (1990) 697. A. Leng, H. Streckel, M. Stratmann, Corros. Sci. 41 (1999) 547. Y. Lu, Polypyrrole-Containing Composite Particles: Preparation, Characterization and Application, TU/Dresden, Dresden, 2005. M. Rohwerder, E. Hornung, M. Stratmann, Electrochim. Acta 48 (2003) 1235. R. Hausbrand, M. Stratmann, M. Rohwerder, Steel Res. Int. 74 (2003) 453. M. Stratmann, R. Feser, A. Leng, Electrochim. Acta 39 (1994) 1207. M. Stratmann, H. Streckel, K.T. Kim, S. Crockett, Corros. Sci. 30 (1990) 715. G.A. Snook, G.Z. Chen, D.J. Fray, M. Hughes, M. Shaffer, J. Electroanal. Chem. 568 (2004) 135. L.F. Zhong, Y.M. Zhang, Y. Tang, Y. Bai, Polyhedron 22 (2003) 2647. B.E. Erickson, G.R. Helz, Geochim. Cosmochim. Acta 64 (2000) 1149. T. Ozeki, H. Adachi, S. Ikeda, J. Chem. Soc. Jpn. 69 (1996) 619. A.W. Hassel, K. Fushimi, M. Seo, Electrochem. Commun. 1 (1999) 180. J.Y. Lim, W.K. Paik, I.H. Yeo, Synthetic Met. 69 (1995) 451. S.J. Dong, G.H. Lian, J. Electroanal. Chem. 291 (1990) 23. W.E. Swartz, D.M. Hercules, Anal. Chem. 43 (1971) 1774. K.S. Kim, W.E. Baitinger, J.W. Amy, N. Winograd, J. Electron. Spectrosc. Relat. Phenom. 5 (1974) 351. M.A. Stranick, Corrosion 40 (1984) 296. A.K. Bairamow, S. Zakipour, C. Leygraf, Corros. Sci. 25 (1985) 69. A.K. Bairamov, Br. Corros. J. 27 (1992) 128. F. Beck, P. Braun, M. Oberst, Ber. Bunsen. Phys. Chem. 91 (1987) 967.
On the development of polypyrrole coatings with self-healing properties for iron corrosion protection G. Paliwoda-Porebska a,*, M. Stratmann a, M. Rohwerder a, K. Potje-Kamloth b, Y. Lu b, A.Z. Pich b, H.-J. Adler b a b
Max Planck Institute for Iron Research, Max-Planck-Strasse 1, 40237 D€ usseldorf, Germany Institute of Macromolecular Chemistry and Textile Chemistry, Technical University Dresden, D-01062 Dresden, Germany Available online 12 September 2005
Abstract This paper presents studies on the efficacy and on the limits of polypyrrole (Ppy) doped with either MoO24 or [PMo12O40]3 as self-healing corrosion protecting coatings. The kinetics of the cathodic delamination were studied by means of the Scanning Kelvin Probe (SKP). This method, in combination with cyclic voltammetry, UV–visible spectroscopy (UV–vis) and X-ray photoelectron spectroscopy (XPS), shows a potential driven anion release from the Ppy coating that results in an inhibition of the corrosion process taking place in the defect. Thus, an intelligent release of inhibitor occurs only when the potential at the interface decreases. Inhibitor anions are released only due to an active defect. However, the release mechanism can be easily negatively affected by the presence of small cations and/or by too high pH values at the buried interface. Hence, such a self-healing coating has to be carefully designed in order to ensure an effective performance. 2005 Elsevier Ltd. All rights reserved.
*
Corresponding author. Tel.: +49 211 6792538; fax: +49 211 6792218. E-mail address: [email protected] (G. Paliwoda-Porebska).
0010-938X/$ - see front matter 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.corsci.2005.05.057
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3217
Keywords: Scanning Kelvin Probe; Polypyrrole; Intelligent release; Corrosion inhibitors; Cathodic delamination; Self-healing coating
1. Introduction In the last 20 years, in numerous publications, intrinsically conducting polymers (ICPs) have been controversially discussed as possible corrosion preventing additions to organic layers [1–4]. One of the reasons for the reported discrepancies is that most papers are focused just on very restricted aspects of corrosion. For instance, an ICP containing coating might be very effective in preventing corrosion in pin holes, while the same coating might fail disastrously in the presence of larger defects [4–6]. Quite numerous possible corrosion protection mechanisms are proposed [7–15]. According to Barisci et al. [16] ICPs might be used as intelligent release systems for corrosion inhibitors on demand. In this work was for the first time pointed out that the polymer could be reduced and consequently dopants could be set free as a result of a galvanic coupling between the corroding metal (iron) and ICP (polyaniline). A possible mechanism for the case of polyaniline was also described by Kendig et al. [17]. Such a coating which releases anions only in the case of corrosive attack but prevents them from being leached out or being released by ion exchange processes, would be a self-healing coating. Nevertheless, no results showing successful dopant release, as a consequence of electrochemical ICP reduction through the corrosion in the defect have been reported until now. This paper shows that on the basis of polypyrrole coatings, which show a electrochemical release of inhibitor anions can be prepared. Even unusually large defects can be passivated because the switch for anions release is the decrease in potential at the metal/polymer interface during the delamination. If the defect is smaller less coating will delaminate, if it is larger more coating delaminates until successful passivation. Some results concerning a positive influence of aminotris(methylene phosphonic acid) (ATMP) anion doped polyaniline (PANI) on the corrosion behaviour of mild steel have been reported by Kinlen et al. [18,19]. Release of ATMP from the PANI and formation of stable adducts with iron ions have been observed. These adducts could lead to the passivation of pinholes. According to the proposed mechanism, the ATMP is released as a consequence of polyaniline–emeraldine salt (PANI–ES) reduction to polyaniline–leucoemeraldine salt (PANI–LS). The number of anions in the polymer chain does not change in the mentioned reaction [16,17,20] and this might be one of the critiques concerning this proposed mechanism. Additionally, it is worth of mentioning that the PANI–LS form is stable only at low pH and that during cathodic delamination strongly alkaline pH prevail. Hence, a high pH induced deprotonation of PANI–ES to the form of polyaniline–emeraldine base (PANI– EB) or reduction to the form of polyaniline–leucoemeraldine base (PANI–LB) is more likely to take place under these conditions. Also, in the most recent publications [21,22] concerning this subject, no observations clearly explaining the mechanism of anion release from PANI–ES in corrosion experiments have been reported.
3218
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
A defect, that penetrates the polymer coating to the bare metal, is a starting point for the electrochemically driven delamination. In the case of the here studied cathodic delamination the degradation at the metal/polymer interface is caused by radicals produced by oxygen reduction reaction at the metal/polymer interface (Fig. 1). Ions, water and oxygen can diffuse in sufficient quantities through any organic coating, thus providing the environment for oxygen reduction. At the isolated intact interface, however the anodic counter reaction is missing, i.e. no electrons are available for the oxygen reduction. Hence, the potential at the intact interface stabilizes at quite high potentials, where oxygen reduction is kinetically inhibited. However, in the presence of defect, a galvanic element is formed due to coupling between defect and delamination site. A closed electric circuit for the electron flow from the defect to the polymer/metal interface is established by cation migration from the defect to the buried interface (see e.g. [23]). The potential is pulled down and oxygen reduction fuelled by part of the iron dissolution in the defect can than take place, thus leading to a further degradation of the interface. This also facilitates cation transport and decreases the iR drop along the delaminated interface. Further details concerning the cathodic delamination process can be found in e.g. [24,25]. Results reported in this paper show that polypyrrole (Ppy), doped with [PMo12O40]3 , releases anions as a result of polymer reduction, which after decomposition yield MoO24 and HPO24 . Subsequently, mainly MoO24 as being an inhibitor of iron dissolution, is able to passivate the defect in the coating. The high pH arising at the metal/polymer interface during the delamination process might have an influence on the mechanism of anion release from Ppy. Numerous reports concerning OH interactions with Ppy can be found in the literature (see e.g. [26–28]). Some authors studied, for example with XPS, the changes induced by the high pH in the Ppy and their influence on a possible anion release or trapping reaction [26,29]. In this paper we also try to discuss the probable OH impact on the anion release process. Molybdate (VI) is well known as efficient and environmental friendly corrosion inhibitor for iron/steel [30,31]. Also in the screening corrosion inhibitor tests per-
Fig. 1. Scheme of the cathodic delamination process.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
[PMo12O40]3-
OHH+
[PMo11O39]7+ MoO42-
OHH+
[P2Mo5O23H]5+
OH-
HPO42-
H+
MoO42-
3219
MoO42-
Fig. 2. [PMo12O40]3 degradation reactions as a function of hydrogen ion concentration [34].
formed by ourselves MoO24 , next to PO24 , has shown the most positive corrosion inhibition effect. Application of Na3PMo12O40 as a corrosion inhibitor for stainless steel have also been reported in the literature [32,33]. One important aspect of [PMo12O40]3 anion chemistry has to be kept in mind while using it as a corrosion inhibitor in aqueous solutions, i.e. [PMo12O40]3 is stable only in highly acidic pH. At pH above 4 it undergoes several degradation reactions of which the final products are MoO24 and HPO24 (Fig. 2) [34]. This means that the inhibition effect attributed to [PMo12O40]3 is actually due to the MoO24 and HPO24 present in the system [35]. The Scanning Kelvin Probe (SKP) technique applied in this study is a non-contact, non-destructive technique that after suitable calibration allows the measurement of the electrode potential at the buried metal/polymer interface. For details concerning the technique see [15,36–38]. 2. Experimental 2.1. Materials and sample preparation Stable water dispersions of Ppy, doped with S2 O28 ; MoO24 and [PMo12O40]3 , have been prepared by chemical polymerization in the Institute of Macromolecular Chemistry and Textile Chemistry at the Technical University Dresden (Prof. H.-J. Adler) by Dr. A.Z. Pich and Dr. Y. Lu. The method has been described in detail elsewhere [39]. Ppy dispersion doped with the S2 O28 , which is not a corrosion inhibitor, has been used as a reference sample. All used dispersions are of the so-called core type, which means that the Ppy has polymerised in a form of sphere with diameter between 80 and 130 nm. For all experiments, iron samples (C 0.0053/Si 0.0071/Mn 0.0024/P <0.0001/S <0.0010/O 0.0029/Al 0.0015 wt%) 20 mm· 10 mm · 2 mm in size, were ground up to 1200 grid paper, cleaned in acetone and then, ethanol in ultrasound bath for 3 min. The Ppy dispersions have been mixed with a non-conducting matrix polymer provided by Chemetall GmbH (Frankfurt/Main). Prepared in this way the mixture was applied by use of a spin-coater on iron samples. Subsequently, the samples were cured at 100 C for 5 min. In the case of samples prepared for the SKP experiments an additional, polyacryl resin top coat (BASF) was applied with the use of the spin-coater. There are two reasons for applying the top coat. The first is that the studied Ppy coatings are meant as a primer in combination with further layers in a complex coating. Hence, by applying the top coat we get closer to the real situation. The second reason is that a very thin coating is lifted up by undercreeping electrolyte during the delamination [40], thus causing folds, which make
3220
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
the SKP measurement difficult, and also introducing an additional mechanical factor to the delamination. 2.2. Ppy films characterization SEM (LEO 1550 VEP) investigations on the prepared films revealed that after a single application no homogeneous film was obtained at the iron surface (Fig. 3a). (The iron surface is only partially covered by Ppy spheres.) A second application of Ppy dispersion did not cause washing away of the already existing layer but some surface areas were additionally filled (Fig. 3b). Application of the third layer filled the empty spaces that remained still available (Fig. 3c). Further spin-coating of Ppy dispersion showed no further changes in the SEM images. These observations indicate that the prepared Ppy composite coating is composed of only one monolayer of Ppy core particles, distributed quite evenly on the iron surface, the matrix polymer acting as glue between them. These observations are also confirmed by investigating a cross-section of the iron sample coated by PpyPMo composite (Fig. 4). The estimated thickness of the film is close to the diameter of the Ppy sphere i.e. from 80 to 130 nm. This conclusion is also supported by results of XPS (Quantum
Fig. 3. SEM images of Ppy dispersion coating morphology. After the dispersion has been applied (a) once, (b) twice and (c) three times.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3221
Fig. 4. SEM image of the perpendicular cut of iron sample coated with Ppy composite.
2000, Physical Electronics) sputtering and ellipsometric measurements (SE 800, Sentech Instruments). 2.3. SKP experiments The SKP used in this work was constructed in the Max Planck Institute for Iron Research. The electronic parts of the setup, consisting of a preamplifier, lock-inamplifier and an integrator were provided by commercial suppliers. A similar SKP setup has been successfully used in other research projects (see e.g. [41–43]). The used SKP tip has a diameter of 100 lm. 2.3.1. SKP delamination experiments In order to perform SKP experiments, part of iron sample was covered with the dispersion coating and additional top coat. On the other part of the sample an artificial defect has been prepared (Fig. 5). In the defect 0.1 M KCl solution has been injected. All measurements were performed in a humid atmosphere (RH = 93– 95%). After the SKP experiments, the delaminated Ppy coatings could be easily peeled off from the metal surface. XPS analysis of the delaminated Ppy films and of the surface of corroded iron samples was then performed. 2.3.2. SKP reduction experiments To evaluate the effect of the high pH present at the metal/polymer interface during the cathodic delamination, on the anion release from the Ppy composite coatings, reduction experiments in controlled N2 atmosphere have been performed. During these experiments no increase of the pH at the metal/polymer interface can take place and hence, the influence of OH on the release mechanism is excluded.
3222
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
Fig. 5. Sample preparation for the SKP experiments.
2.4. X-ray photoelectron spectroscopy To identify the chemical species at the delaminated iron samples X-ray photoelectron spectroscopy was used. All measurements were performed with a use of the Quantum 2000 Scanning ESCA Microprobe System from Physical Electronics. The X-rays employed in the used system, are Al Ka emitted by an aluminium anode. The take off angle, at which emitted electrons were analysed, was 45. All binding energies given here refer to the C(1s) signal of aliphatic hydrocarbon species with a value of 285.0 eV. 2.5. Electrochemical characterization Cyclic voltammetry and reduction (polarization) experiments were performed with a gold plate as counter electrode and an Ag/AgCl reference electrode in 50 ml of 0.1 M LiClO4, 0.1 M (C4H9)4NCl or 0.1 M NaClO4. A gold (evaporated on glass) substrate coated with the Ppy composite was used as the working electrode. No additional top coat was applied on these samples. All the experiments were carried out after the electrolyte was purged with N2 for 30 min. Cyclic voltammograms were performed by using a HEKA potentiostat with a scan rate of 50 mV/s. 2.6. Inhibitor tests Polarization experiments, at a sweep rate of 0.2 mV/s, were carried out in 10 ml aerated 0.1 M NaCl with 0.01 M Na2MoO4 or Na3PO4 solution or in 0.1 M NaCl by using a Zahner IM6d potentiostat. All the experiments were carried out with a gold plate as a counter electrode, Ag/AgCl reference electrode and iron plate as a working electrode. The iron samples were stored in the electrolyte for 3 h, until the open circuit potential became stable. As next, in separate experiments, anodic and cathodic polarization curves were recorded.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3223
3. Results 3.1. Molybdate (VI) and phosphate (V) as iron corrosion inhibitors Prior to the experiments with the Ppy coatings, polarisation experiments on iron in solutions containing different inhibitor anions were performed with the aim to assess the inhibition effects of prospective anions. The subsequent Tafel analysis was performed (Fig. 6). The obtained results indicated that especially molybdate (VI) and phosphate (V) seem to be promising. Both anions cause a large anodic shift of the corrosion potential (DEcorr = 600 mV) and lower the rate of iron dissolution significantly. The corrosion current density is two orders of magnitude smaller in the inhibitor containing solution compared to pure NaCl solution. From the shape of the cathodic curve it can be also seen that the oxygen reduction reaction is not influenced by the presence of MoO24 and PO34 ions. 3.2. Voltammetric characterization of PpyPMo12O40 and PpyMoO4 films Fig. 7 shows cyclic voltammograms (CV) of the PpyPMo12O40 (PpyPMo) and PpyMoO4 (PpyMo) films. Obviously, the prepared composite films are electrochemically active, i.e. the non-conducting matrix polymer does not inhibit the electrochemical activity of the Ppy. For both coatings the reduction and oxidation peaks positions are similar. Important difference is that in the case of PpyPMo coatings are always higher current densities measured. That is most probably connected with
3 2
(1)
log | i| / µ A cm-2
1 0 (2)
-1
(3)
-2 -3 -4 -0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
E / mV vs. AgACl
Fig. 6. Polarization curves of a iron electrode in 0.1 M NaCl (1) and in 0.1 M NaCl with addition of 0.01 M Na3PO4 (2) and 0.01 M Na2MoO4 (3).
3224
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
-5
1.0x10
i / A cm
-2
0.0
-5
-1.0x10
-5
-2.0x10
-5
-3.0x10
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
E / V vs. SHE Fig. 7. Cyclic voltammograms of PpyPMo (- - -) and PpyMo (—) performed in 0.1 M LiClO4.
the inherent electroactivity of [PMo12O40]3 anion in the applied potential range which may contribute to the measured current densities. In the first cycle of the CV, in both, PpyPMo as well as PpyMo containing coatings, significant reduction peaks can be observed, which decrease in the next cycle. As XPS analysis of the coating and UV–vis analysis of the electrolyte show, these peaks (at least some of them) are connected with Ppy reduction and the corresponding anion release process. After the first cycle, the peaks are much smaller thus indicating that the reduction–oxidation reactions of Ppy are not completely reversible in this electrolyte. Both coatings show a quite different behaviour in CVs performed in tetrabutylammonium chloride solution (Fig. 8), where the large size of the cation molecule makes its incorporation during the Ppy reduction rather improbable [44]. In the case of PpyPMo, the reduction–oxidation peaks change significantly their position after the few first cycles. This indicates that [PMo12O40]3 are systematically replaced by Cl anions, which are available in the solution in high concentration. Another important observation is that the size of the peaks does not change significantly with the number of cycles, in contrast to changes observed in CV performed in LiClO4. This observation indicates that in tetrabutylammonium chloride, where only anion exchange between the coating and electrolyte can take place, PpyPMo undergoes well reversible reduction–oxidation reactions. In the case of the PpyMo containing coating, peaks positions do not change, although probably, as in this case of the PpyPMo, the MoO24 =Cl anion replacement takes place. One explanation for this might be that these two anions do not differ so dramatically in structure as [PMo12O40]3 and Cl and hence, no significant
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3225
-5
2.0x10
-5
i / Acm
-2
1.0x10
0.0
-5
-1.0x10
-5
-2.0x10
-0.6
-0.4
-0.2
(a)
0.0
0.2
0.4
0.6
0.8
E / V vs. SHE -5
1.5x10
-5
1.0x10
-6
i / A cm
-2
5.0x10
0.0
-6
-5.0x10
-5
-1.0x10
-5
-1.5x10
-0.6
(b)
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
E / V vs. SHE
Fig. 8. Cyclic voltammograms of (a) PpyPMo and (b) PpyMo performed in 0.1 M tetrabutylammonium chloride: (—) first two cycles, (- - -) cycles after 15 min.
changes in the cyclic voltammogram with number of cycles have been observed, indicating that also here redox processes are reversible.
3226
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3.3. Polarization–reduction experiments To confirm and to quantify the anion release, the following polarization experiments have been performed, Ppy composite coatings were prepared on gold substrates and then samples were polarised to 0.4 V vs. SHE (iron corrosion potential) in NaClO4 for 10–15 min. After the experiments the electrolyte was analysed with the use of UV–visible spectroscopy to detect the presence of released anions. In the case of the electrolyte in which PpyPMo coatings were reduced, by taking the advantage of the inherent electrochemical activity of [PMo12O40]3 which makes it possible to identify the characteristic [PMo12O40]3 reduction–oxidation peaks, also cyclic voltammograms were performed. From the obtained cyclic voltammogram (Fig. 9) the peaks which can be assigned to [PMo12O40]3 , can be recognised. The current densities of these peaks are quite small. A separate reference measurement, with known amounts of Na3PMo12O40 which have been injected into the electrolyte, shows that they correspond to approximately 0.5 lM of [PMo12O40]3 concentration (equivalent to release of 0.025 lmole). The theoretically calculated number of [PMo12O40]3 anions, which should be released from a monolayer of PpyPMo spheres, is approximately two times smaller than the experimentally found. There were two important assumptions made during the theoretical calculation, which may result in the difference between the experimentally and theoretically found values. The first concerns the assumed doping rate, in this case
peaks after addition of 25µl of 10-3 M Na3PMo12O40 solution
-5
1.0x10
bigger peaks appear after next injection of 25µl
-6
5.0x10
i / A cm
-2
0.0
-6
-5.0x10
-5
-1.0x10
-5
-1.5x10
-5
-2.0x10
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
E / V vs. SHE Fig. 9. Cyclic voltammogram performed in 0.1 M LiClO4, (—) electrolyte in which PpyPMo reduction was performed and ( ) with addition of 10 3 M Na3PMo12O40.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3227
pyrrole:anion = 1:4, and the second the maximal volume fraction of cores in the monolayer is estimated to be 60%. However, the SEM investigation of the Ppy coating cross-section revealed that at some points the coating is thicker than a monolayer. A precise quantification is difficult to make, as this does not show up in the top view SEM images. In any case all particles seem to be active. In the UV–vis spectrum recorded for this electrolyte, an adsorption peak at approximately 210 nm and small adsorption peaks at 310 and 330 nm, corresponding to the [PMo12O40]3 anion adoption bands [45], can be found. The UV–vis spectrum obtained from the electrolyte in which PpyMo reduction has been performed shows a broad adsorption shoulder at 220–230 nm and a small peak around 260 nm, which can be assigned to MoO24 [46,47]. This indicates that the MoO24 was released from the polymer matrix. These results indicate that both [PMo12O40]3 and MoO24 can be released from Ppy coating by electrochemical reduction. When the PpyPMo or PpyMo sample has been just stored in the electrolyte (i.e. without electrochemical reduction) anions were not present in the solution. 3.4. SKP experiments
200
300
100
200 100
0
E / mV vs. SHE
E / mV vs. SHE
Fig. 10 shows the delamination behaviour of the different Ppy coatings. As can be seen, at the upper right side of each diagram, the potential of the intact Ppy coating
-100 -200 -300
0 -100 -200 -300 -400
-400 -500
(a)
-500 0
-600
200 400 600 800 1000 1200 1400 1600 1800
0
200 400 600 800 1000 1200 1400 1600 1800 2000
(b)
d / µm
d / µm
200 100
E / mV vs. SHE
0 -100 -200 -300 -400 -500
0
200 400 600 800 1000 1200 1400 1600 1800
(c)
d / µm
PpyS2 O28
Fig. 10. SKP delamination profiles: (a) -reference sample, (b) PpyMo and (c) PpyPMo. Intervals between profiles are in case of all samples the same and equal, 20 min.
3228
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
remains between 0.1 and 0.2 V vs. SHE. The border between defect and coating is at d = 0. Profiles moving towards the right side of the diagram represent the propagation of the delamination front, from the defect into the intact polymer coating on the right side (see also Fig. 5). Intervals between the profiles (lines) are in case of all samples the same and equal, 20 min. The reference sample as well as the Ppy coating doped with the MoO24 delaminates fast, i.e. hundreds of micrometers in a few minutes. No passivation of the defect could be observed. In the case of Ppy doped with [PMo12O40]3 (Fig. 9c) a different behaviour is observed. The first profile does not differ from the one measured in the case of the other samples, but already the second one shows a significant increase of the potential, DE = +0.17 V, in the delaminated area. Consequently, the potential difference between the delaminated area and the intact polymer coating, which is the driving force for a further delamination [23], is getting smaller. This induces a significant decrease in the delamination speed. The potential in the already delaminated area changes further and becomes even more positive with time. The change is in the case of the presented sample 0.28 V, and usually being between 0.15 V and 0.3 V. This observation shows a good agreement with a separate measurement of the open circuit potential at the defect, by using a micro-reference electrode [48], where also a significant anodic potential shift equal 0.17 V has been observed.
4. Discussion In the following section the results concerning delamination experiments on Ppy coatings will be discussed. 4.1. PpyPMo XPS analysis performed at the surface of the iron sample shows the presence of molybdate at the defect as well as at coating peeled off the delaminated area. The amount of molybdate varies in the range between 1% and 3% (atomic per cent) for individual experiments. Since from one released [PMo12O40]3 anion only 1 mole of phosphate (V) arises, compared to 12 moles of molybdate, it was usually impossible to identify it with XPS. Only in the case of two samples less then 1% of phosphate have been found at the metal surface. In the delaminated Ppy film K+ can be found, thus indicating that parallel to the anion release reaction, to some extent, cation incorporation takes place. It has been already reported [49,50] that during the reduction process of Ppy, not only anion release takes place but also cation incorporation is usually possible for the charge compensation. This parallel reaction is of course not desired, since it decreases the amount of released anions and thus the achievable inhibition effect. The XPS analysis shows that the Mo 3d peak found at the interface appears within the typical binding energy range for MoVI compounds [51,52]. The peak is very slender (Fig. 11) indicating the presence of only one molybdate oxidation state. Also the molybdate found in the defect is clearly MoVI.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3229
2
x 10 24 22 20
CPS
18 16 14 12 10 8 240
238
236
234
232
230
228
226
Binding Energy (eV) Fig. 11. XPS spectra of molybdenum (3d3/2, 3d5/2) found under PpyPMo polymer coating delaminated in O2 atmosphere.
According to observations reported in the literature [53–55], the role of the MoO24 as corrosion inhibitor is, next to other effects, strengthening of the native iron oxide layer by incorporating and healing its defects. Hence, it is possible that the molybdate released at the metal/polymer interface is, to some extent, already incorporated into the iron oxide. According to some authors [27,28,56] OH ions being strong nucleophiles can react with Ppy. As a consequence of this interaction, Ppy can be overoxidised and can release anions from the polymer chain (Fig. 12). Knowing that during the delamination experiments Ppy is exposed to OH attack, and the observed anion release might be due to the mentioned attack and not to the electrochemical reduction of the Ppy, additional experiments were performed. The PpyPMo coating was delaminated, more precisely said reduced, in controlled N2 atmosphere. Also in this case XPS analysis showed the presence of molybdate at the iron surface. This result shows that the [PMo12O40]3 anions are released by a truly electrochemical reduction and not due to interaction with OH ions. During the reduction experiments in N2 atmosphere, due to a lack of OH , no significant decomposition of the [PMo12O40]3 anion should take place. Indeed, a H AN +
N
OH-
N H
N
O
Fig. 12. Nucleophilic attack of OH on Ppy.
3230
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
Table 1 Binding energies (eV) of measured Mo 3d3/2 and Mo 3d5/2 peaks Description of measured Mo 3d peak
Molybdate oxidation state
Binding energy (eV) Mo 3d3/2
Binding energy (eV) Mo 3d5/2
Molybdate at the metal surface/delamination experiments in O2 atmosphere Molybdate in the defect area/delamination experiments in O2 atmosphere Molybdate in the Ppy[PMo12O40] film Molybdate in the Na3[PMo12O40] powder Molybdate at the metal surface/reduction experiments in N2 atmosphere
+6
234.7
231.5
+6
234.5
231.4
+6 +6 +6
235.8 235.1 235.0
232.7 231.9 231.9
significant shift of the Mo 3d peak, in comparison to the position of peak observed in the case of samples delaminated in O2 atmosphere, is observed (Table 1). This indicates that a different molybdate form exists at the metal/polymer interface. To assign molybdate spectra properly, reference XPS spectra of non-delaminated PpyPMo coatings on iron and Na3PMo12O40 powder have been measured. The comparison of the Mo 3d peak positions indicates, that in the case of coating reduction in N2 atmosphere [PMo12O40]3 anions are released and can be identified at the iron surface. According to these observations the following sequence of processes can be proposed: due to the potential decrease at the interface, caused by the delamination, i.e. more precisely by the related ohmic contact with the corroding defect, the [PMo12O40]3 anion is firstly released from the thusly reduced Ppy chain; subsequently this anion reacts with the OH ions available at the iron/polymer interface, yielding mainly MoO24 , which is responsible for the observed defect passivation. Although the ability of these coatings to passivate even quite large defects and to stop delamination is obvious, it is also obvious that it takes some time to release enough anions. Clearly, for technical applications the coatings need to be significantly improved. One approach would be to prepare thicker coatings. However, preliminary experiments suggest that thicker coatings cannot be prepared with such a high content of Ppy particles, and for lower contents the conductivity and electrochemical activity of the coatings decrease significantly. These questions are topics of further research. 4.2. PpyMo Also, one would assume that the direct use of the smaller MoO24 anion should result in a much better performance. There is no need for an additional decomposition reaction, and also the release and transport should be easier. This means that the competition with the parallelly occurring cation incorporation, which decreases the amount of releasable anions, should be favourably affected. But as already mentioned, the coatings containing Ppy doped with MoO24 show no inhibition of the delamination process. XPS analysis shows that after the delamination experiment no MoO24 anion is present at the iron/polymer interface or in the
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
3231
defect, meaning that no measurable anion release takes place. Two hypotheses have been considered to explain this observation. The first hypothesis refers to the possibility that, contrary to the just discussed, cation incorporation is predominant in the PpyMo system, which would explain the remaining of MoO24 anions in the polymer matrix. However, XPS analysis shows that only small amounts of K+ are incorporated inside of the delaminated film. The second hypothesis considers the influence of OH on the release mechanism. A significant difference between the PpyMo and the PpyPMo release system is the anion decomposition reaction [34] that takes place after the [PMo12O40]3 anion is set free from PpyPMo. In this reaction OH is being consumed, which buffers the usually high pH present at the iron/polymer interface during the delamination process to lower values. The hypothesis proposed here is that, if the [PMo12O40]3 anion decomposition reaction provides this kind of buffering effect, in the case of PpyMo coating the polypyrrole is exposed to a much higher pH existing at the delaminating interface. That may cause a faster and more efficient Ppy overoxidation process. It is possible that depending on conditions and the used Ppy dopants, as a consequence of the overoxidation process the dopants will stay trapped in the non-conducting, overoxidised Ppy matrix [26]. To analyze if in the case of Ppy doped with MoO24 anions this undesired OH influence, resulting in anion immobilization in the coating, may be the reason for its inability to release the anions under ambient conditions, reduction experiments of the PpyMo coating in controlled N2 atmosphere were performed. The results of the corresponding XPS analysis of these samples clearly show that under these conditions the MoO24 anion is indeed released from the polymer coating. Analysis of the changes in the C 1s and O 1s detail spectra measured on the thusly reduced coating show a significantly lower amount of C@O and O–C@O groups, compared to the same coating delaminated in oxygen. This observation strongly supports the hypothesis that the PpyMo coating is overoxidised during the delamination in air, with involvement of OH , which results in the MoO24 anions being trapped in the Ppy matrix. One might think that for a further verification of this hypothesis a comparison of changes of the C 1s and O 1s spectra between the PpyPMo and PpyMo, both delaminated in oxygen atmosphere, would be informative. Although the [PMo12O40]3 decomposition reaction obviously provides the mentioned buffering effect, that lasts long enough for successful anions release, the buffering effect certainly decreases with time. But while the amount of anions responsible for the buffering effect is limited, this is not the case for the amount of OH that may be produced at the iron/polymer interface (unless the inhibition of the defect occurs much faster than is the case now). Hence, the XPS analysis of the C 1s and O 1s spectra on delaminated PpyPMo coatings also shows a presence of some amounts of C@O and O–C@O groups.
5. Conclusions The presented SKP results show that in the case of the Ppy doped with the [PMo12O40]3 anion, a significant inhibition of the delamination is provided by on
3232
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233
demand functioning release. This result is supported by the observation of a partial passivation of the defect, where due to MoO24 anion accumulation the iron dissolution is inhibited. Additionally presented results confirm a truly electrochemical release of [PMo12O40]3 anions, not influenced by high pH at the metal/polymer interface. The second studied coating, doped with the MoO24 anions, showed no inhibition effect on the delamination process. According to the proposed mechanism this results from Ppy overoxidation, due to the highly alkaline pH at the buried metal/polymer interface. In the case of the PpyPMo coating, the [PMo12O40]3 anion decomposition reaction plays an essential role in the successful delamination inhibition. This reaction provides a significant buffering effect, preventing too high pH values at the metal/polymer interface. In the case of the Ppy doped with MoO24 coating this buffering effect cannot appear. Resuming, a Ppy release system, unaffected by leaching and pH changes, that safely stores the corrosion inhibitor ions can be used as an intelligent corrosion protecting coating. However, the influence of conditions during the cathodic delamination process on the release behaviour, especially the effect of small easy incorporable cations and the high pH, needs further investigations.
Acknowledgement The financial support from the Bundesministerium fu¨r Buildung und Forschung is gratefully acknowledged.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]
J.O. Iroh, W.C. Su, Synthetic Met. 114 (2000) 225. N.V. Krstajic, B.N. Grgur, S.M. Jovanovic, M.V. Vojnovic, Electrochim. Acta 42 (1997) 1685. W.K. Lu, R.L. Elsenbaumer, B. Wessling, Synthetic Met. 71 (1995) 2163. A.A. Pud, G.S. Shapoval, P. Kamarchik, N.A. Ogurtsov, V.F. Gromovaya, I.E. Myronyuk, Y.V. Kontsur, Synthetic Met. 107 (1999) 111. K. Idla, J. Reut, A. Opik, Synthetic Met. 102 (1999) 1392. G.M. Spinks, A.J. Dominis, G.G. Wallace, D.E. Tallman, J. Solid State Electron. 6 (2002) 85. D.W. Deberry, J. Electrochem. Soc. 132 (1985) 1022. V.T. Truong, P.K. Lai, B.T. Moore, R.F. Muscat, M.S. Russo, Synthetic Met. 110 (2000) 7. B. Wessling, Adv. Mater. 6 (1994) 226. N. Ahmad, A.G. MacDiarmid, Synthetic Met. 78 (1996) 103. T.D. Nguyen, M. Keddam, H. Takenouti, Electrochem. Solid St. 6 (2003) B25. T. Schauer, A. Joos, L. Dulog, C.D. Eisenbach, Prog. Org. Coat. 33 (1998) 20. P.J. Kinlen, Y. Ding, B.C. Silverman, Corrosion 58 (2002) 490. J.D. Stenger-Smith, Prog. Polym. Sci. 23 (1998) 57. R.J. Holness, G. Williams, D.A. Worsley, H.N. McMurray, J. Electrochem. Soc. 152 (2005) B73. J.N. Barisci, T.W. Lewis, G.M. Spinks, C.O. Too, G.G. Wallace, J. Intel. Mater. Syst. Str. 9 (1998) 723. M. Kendig, M. Hon, L. Warren, Prog. Org. Coat. 47 (2003) 183. P.J. Kinlen, D.C. Silverman, C.R. Jeffreys, Synthetic Met. 85 (1997) 1327. P.J. Kinlen, V. Menon, Y.W. Ding, J. Electrochem. Soc. 146 (1999) 3690.
G. Paliwoda-Porebska et al. / Corrosion Science 47 (2005) 3216–3233 [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56]
3233
D.E. Tallman, G. Spinks, A. Dominis, G.G. Wallace, J. Solid State Electron. 6 (2002) 73. A. Cook, A. Gabriel, D. Siew, N. Laycock, Curr. Appl. Phys. 4 (2004) 133. J.E. Pereira da Silva, S.I. Cordoba de Torresi, R.M. Torresi, Corros. Sci. 47 (2005) 811. M. Rohwerder, M. Stratmann, MRS Bull. 24 (1999) 43. W. Furbeth, M. Stratmann, Prog. Org. Coat. 39 (2000) 23. A. Leng, H. Streckel, M. Stratmann, Corros. Sci. 41 (1999) 579. F. Palmisano, C. Malitesta, D. Centonze, P.O. Zambonin, Anal. Chem. 67 (1995) 2207. Y.F. Li, R.Y. Qian, Electrochim. Acta 45 (2000) 1727. H.L. Ge, G.J. Qi, E.T. Kang, K.G. Neoh, Polymer 35 (1994) 504. C. Malitesta, I. Losito, L. Sabbatini, P.G. Zambonin, J. Electron Spectrosc. 76 (1995) 629. M.S. Vukasovich, J.P.G. Farr, Polyhedron 5 (1986) 551. V.A. Kumari, K. Sreevalsan, S.M.A. Shibli, Corros. Prevent. Contr. 48 (2001) 83. E.A. Lizlovs, J. Electrochem. Soc. 114 (1967) 1015. A. Weisstuch, C.E. Schell, Corrosion 28 (1972) 299. G.A. Tsigdinos, Topics Curr. Chem. 76 (1978) 1. D.M. Basher, J.E. Rhoades-Brown, Br. Corros. J. 8 (1973) 50. M. Stratmann, H. Streckel, Corros. Sci. 30 (1990) 681. M. Stratmann, H. Streckel, Corros. Sci. 30 (1990) 697. A. Leng, H. Streckel, M. Stratmann, Corros. Sci. 41 (1999) 547. Y. Lu, Polypyrrole-Containing Composite Particles: Preparation, Characterization and Application, TU/Dresden, Dresden, 2005. M. Rohwerder, E. Hornung, M. Stratmann, Electrochim. Acta 48 (2003) 1235. R. Hausbrand, M. Stratmann, M. Rohwerder, Steel Res. Int. 74 (2003) 453. M. Stratmann, R. Feser, A. Leng, Electrochim. Acta 39 (1994) 1207. M. Stratmann, H. Streckel, K.T. Kim, S. Crockett, Corros. Sci. 30 (1990) 715. G.A. Snook, G.Z. Chen, D.J. Fray, M. Hughes, M. Shaffer, J. Electroanal. Chem. 568 (2004) 135. L.F. Zhong, Y.M. Zhang, Y. Tang, Y. Bai, Polyhedron 22 (2003) 2647. B.E. Erickson, G.R. Helz, Geochim. Cosmochim. Acta 64 (2000) 1149. T. Ozeki, H. Adachi, S. Ikeda, J. Chem. Soc. Jpn. 69 (1996) 619. A.W. Hassel, K. Fushimi, M. Seo, Electrochem. Commun. 1 (1999) 180. J.Y. Lim, W.K. Paik, I.H. Yeo, Synthetic Met. 69 (1995) 451. S.J. Dong, G.H. Lian, J. Electroanal. Chem. 291 (1990) 23. W.E. Swartz, D.M. Hercules, Anal. Chem. 43 (1971) 1774. K.S. Kim, W.E. Baitinger, J.W. Amy, N. Winograd, J. Electron. Spectrosc. Relat. Phenom. 5 (1974) 351. M.A. Stranick, Corrosion 40 (1984) 296. A.K. Bairamow, S. Zakipour, C. Leygraf, Corros. Sci. 25 (1985) 69. A.K. Bairamov, Br. Corros. J. 27 (1992) 128. F. Beck, P. Braun, M. Oberst, Ber. Bunsen. Phys. Chem. 91 (1987) 967.