Self-healing protective coatings of polyvinyl butyral/polypyrrole-carbon black composite on carbon steel

Abstract

Self-healing polyvinyl butyral (PVB) based organic coating formulations were prepared by incorporating polypyrrole-carbon black (PPyCB) composite as an inhibiting pigment. The redox properties and diffusion barrier nature of PPy imparted self-healing to the PVB/PPyCB composite coatings in aggressive environments. PPy induced the formation of a stable passive layer on the metal surface through the interaction of released dopant, from the organic sulfonic acid doped PPy with metal iron oxide. SEM images and Raman spectroscopy confirmed the formation of a protective passive layer on the metal surface. Furthermore, reduced PPy hindered the diffusion of water and oxygen through the coating. The addition of more conducting particles like graphene further enhanced the protective nature of the PVB/PPyCB composite coatings. This work demonstrates a possible application of conducting particles in enhancing the protective nature of organic coatings used widely in industry simply as a barrier coating.

---

Introduction

Intrinsically conducting polymers (ICPs) have attracted much research interest in the area of protective coatings because of their properties such as self-healing and environmental friendliness, which make them promising candidates to replace the coatings based on hexavalent chromium.^1–3 Several possible corrosion inhibition mechanisms have been reported for ICPs, though there still exists some ambiguity concerning the exact protection pathway. The most cited mechanism is the surface ennobling and anodic passivation of steel induced by the inherent redox capability of conducting polymers.^4–6 Another proposed mechanism is the so-called self-healing, where ICPs doped with suitable anionic inhibitors act as a reservoir of corrosion inhibitors. On initiation of corrosion, oxidation of the substrate changes the oxidation states of these conducting polymers which forces the doping agent to be released at the affected area.^7–13 These smart coatings are able to prevent corrosion even in the scratched areas where bare steel comes in direct contact with the aggressive environment.^7,10

Among the different ICPs, polypyrrole (PPy) is promising in terms of its high conductivity, stability and ease of synthesis.^14,15 However, some inherent characteristics of PPy, such as its poor mechanical properties, porosity and poor solubility in common solvents, have limited its application in coatings.⁴ Also, it is difficult to cast PPy films due to its poor solubility in solvents resulting from the strong intermolecular attraction between PPy chains. Direct deposition either by chemical or electrochemical methods was commonly used for the fabrication of PPy based protective coatings.^8,9,11,12 Recently, a new strategy for corrosion protection by using conducting polymers as additive components or inhibiting pigments in composite coatings has been developed, by which the undesirable properties of both ICPs and organic polymers can be minimized.^{4,7,10,13–15} A number of literature studies have reported the use of polyaniline (PANI) with other additives as inhibitive agents in commercial coatings.^16–22 However, a limited number of studies have been reported describing the use of PPy as inhibitive pigment.^4,7,23,24 Yan et al. used oxyanion doped PPy deposited on aluminium flakes as an inhibitive pigment in epoxy primer for aluminium.⁴ PPy functioned both as a reservoir of corrosion inhibitor and an oxygen scavenger, thus resulting in the self-healing of the artificially created defects on the epoxy coatings.

Corrosion protection by organic coatings is brought about by the barrier effect of the coating as well as internal sacrificial electrode formation, which protects the underlying metal from further corrosion.¹⁷ The barrier properties of the coating can also be further enhanced by using suitable reinforcing fillers.^18,25 The addition of filler is generally reported to improve the adhesion strength between the polymer and metal as well as to increase the diffusion paths of water and oxygen molecules through the coating, thus enhancing the corrosion resistance of the coating.^17,26 Graphene, a two dimensional sheet of sp² carbon atoms, was recently reported as an ideal filler for corrosion inhibition in reinforced polymer coatings because of its barrier nature towards corrosion promoting species.^26–35 Recently, Chaudhry et al. discussed the anti-corrosion behavior of graphene incorporated into a self-cross linked polyvinyl butyral composite.²⁷ Graphene platelets increased the diffusion path of oxygen and ionic species through the coating, which improved corrosion resistance of the coating. Similar mechanisms were reported for graphene/epoxy,^28,29 polyurethane/graphene³⁰ and poly(sodium styrene sulfonate)/graphene³¹ composites as protective coatings. The use of polymer composites consisting of graphene and conducting polymers presents a new approach for making corrosion resistant coatings. The electrical conductivity of graphene imparts corrosion protection by enhancing the passive oxide layer formation at metal polymer interface, in addition to its diffusion barrier nature. Sreevatsa et al. investigated the potential of graphene as an ionic barrier for steel in conjunction with a polypyrrole ‘topcoat’ to form a p–n junction.³² Chang et al. reported the application of polyaniline/graphene composites (PAGCs) for corrosion protection of steel.³³

In the present study, fabrication of stable self-healing organic coatings were reported by incorporating polypyrrole-carbon black (PPyCB) composite pigment into a polyvinyl butyral (PVB) matrix. The effect on the corrosion protection of PVB/PPyCB composite coatings by the addition of conducting particles like graphene was also analyzed. The protective nature of these coatings was studied by immersion test and electrochemical methods.

Experimental

Materials

Polypyrrole (doped, conductivity 30 S cm⁻¹ (bulk), extent of labeling: 20 wt% loading, composite with carbon black, melting point > 300 °C), polyvinyl butyral with trade name Butvar B-98 (molecular weight 40 000–70 000 and specific gravity of 1.1 at 23 °C), methanol and hydrochloric acid were purchased from Sigma-Aldrich. Graphene nanoplatelets (with a trade name of N002-PDR) were purchased from Angstron materials, USA, and were used as received.

Preparation of PVB/PPyCB formulations

The required amount of PVB was dissolved in 20 ml methanol by stirring for 6 h. To avoid the evaporation of methanol, coating formulations were prepared in round bottom flask fixed with a water condenser. The methanol level on the flask was marked after the addition of PVB and was maintained during the stirring and sonication processes. After dissolving PVB completely in methanol, the PPyCB composite pigment was added slowly with magnetic stirring. The total amount of PVB and PPyCB composite was fixed at 2 g. The mixture was subjected to continuous stirring for 12 h, followed by 8 h sonication to obtain a well dispersed PVB/PPyCB composite formulation. Three different formulations were prepared by using 5%, 10% and 20% PPyCB composite in the PVB matrix (denoted as PVB/PPyCB5, PVB/PPyCB10 and PVB/PPyCB20). Table 1 shows the amounts of different components in the PVB/PPyCB composite formulations. In order to incorporate graphene in the PVB/PPyCB composite, 5 wt% of graphene was added into the PVB/PPyCB20 formulation and sonicated for 12 h followed by stirring for another 6 h.

Table 1 Amount of different components used in the PVB/PPyCB composite formulations

Sample	PVB (g)	PPyCB (g)	Total (g)
PVB	2.00	—	2.00
PVB/PPyCB5	1.90	0.10	2.00
PVB/PPyCB10	1.80	0.20	2.00
PVB/PPyCB20	1.60	0.40	2.00

---

Application of composite formulations on carbon steel

Carbon steel of grade RST37-2 DIN 17100-80 was purchased from the Qatar Steel Industries Factory, Qatar. Its composition in weight% was C (0.125), Mn (0.519), Si (0.016), P (0.014), S (0.005), Al (0.034) and Fe (99.287). Carbon steel coupons of 5 cm × 2 cm × 2 cm were first pickled with hydrochloric acid for 2 h in order to remove the oxide layer from the surface. After acid treatment, the coupons were polished with sandpaper (60, 150 and 180 grits) followed by water and acetone rinsing. Finally, the coupons were sonicated in acetone for 10 min and dried in oven at 90 °C for 1 h. The prepared PVB/PPyCB formulations were applied on the metal coupons by dip coating using QPI-128 dip coater from Qualtech Product Industry Co., Ltd. The metal coupons were dipped in the solution with an immersion rate of 100 mm s⁻¹ for 30 s, and then withdrawn from the solution with the same rate as immersion. This process was repeated 3 times to obtain a suitable coating thickness. After the coating, the metal coupons were cured in the oven at 190 °C for 1 h to obtain cross-linking between PVB, PPyCB and metal surface, so as to improve adhesion. After curing, the metal coupons were cooled down to room temperature. The thickness of each coating was measured using PosiTector 6000 coating thickness gages from DeFelsko Corporation, New York. Adhesion strength of the coatings was measured by the Cross Cut Tape Test, according to the ASTM standard test method D3359-09.

Electrochemical measurements

Open circuit potential (OCP), electrochemical impedance spectroscopy (EIS) and polarization measurements were carried out at room temperature in a three-electrode corrosion cell consisting of a saturated calomel reference electrode (SCE), a platinum counter electrode and composite coated steel coupons as the working electrode. The electrolyte solution used for the electrochemical analysis was 4 wt% NaCl solution. The exposed area of working electrode was 1 cm². All measurements were performed on an electrochemical analyzer (supplied by BioLogic, France) using a software EC-Lab V10.39. Ultra-low current cables connected to the potentiostat was used for the accurate measurement of the current. This option includes current ranges from 100 nA down to 100 pA with additional gains extending the current ranges to 10 pA and 1 pA. The resolution on the lowest range was 76 aA. Open circuit potential was measured continuously for 48 h without disturbing the experimental set up and impedance measurements were incorporated in-between the OCP cycles at regular intervals. Impedance measurements were performed as a function of open circuit potential by applying a sinusoidal potential in frequencies from 10⁴ Hz to 10⁻² Hz, with an amplitude of 15 mV. The same software was used to simulate the impedance behavior of the samples. Polarization measurements were performed by polarizing the working electrode from an initial potential of −250 mV up to a final potential of +250 mV as a function of open circuit potential. The values of the corrosion current density (I_corr), corrosion potential (E_corr), and cathodic (β_c) Tafel constant were extracted from the polarization curves by Tafel extrapolation of the cathodic branch.^36–38 A scan rate of 0.1667 mV s⁻¹ was used for the polarization sweep. To study the reproducibility of the measurement, each set of experiments was repeated three times on newly coated samples.

Immersion test

The edges of the coated coupons were sealed using a Nippon epoxy primer. After 24 h drying at room temperature, the surface of the samples was scratched in an X shape using a surgical blade fixed into a cutter (with an ergonomic handle/holder) included with an adhesive test kit from GARDCO Paul N. Gardner Co., Inc. USA. This special design of holder gave comfortable and precision control, and helped to make identical scratches on all the samples. Scratched coupons immersed in 4 wt% NaCl solution were followed up daily using an optical microscope to monitor the self-healing effect of the coatings.

Characterization techniques

The microstructure of the passive layer formed on the scratches was analyzed using a scanning electron microscope (SEM) (FEI Quanta, FEG250, USA) at an accelerating voltage of 5 kV. The chemical structure of the passive layer formed on the scratched areas was also analyzed using Raman spectroscopy. Raman spectra were recorded using a LabRAM HR spectrometer (Horiba Jobin Yvon). Laser light from a He/Ne source with a wavelength of 633 nm was used for excitation. A long working distance objective with magnification 50× was used to collect the scattered light as well as to focus the laser beam on the sample surface. A digital microscope, KH-7700 (Hirox Co. Ltd, USA), was used for the optical images of the coatings. Tensile testing was carried out using an Instron 2519-107 universal testing machine (Instron Corp.), according to the ASTM standard test method D6382V. Testing was done at room temperature using plain rubber grips and the average of three replicas is reported here.

Results and discussion

Morphology of the PVB/PPyCB composite coatings

To generate stable organic coatings of PVB/PPyCB composites on the carbon steel substrate, curing of coatings at different temperatures was carried out at the conventionally used ambient conditions²⁵ for generating the PVB coatings that did not provide corrosion protection on the incorporation of PPyCB pigments. These coatings were non-uniform and started to peel out while immersed in corrosive media. Curing for 1 h at 190 °C was observed to be optimal for attaining stable coatings of 20 ± 5 μm thickness. The morphology of the composite coatings was compared through optical microscopy in Fig. 1. The coatings exhibited uniform dispersion of the PPyCB pigment in the PVB matrix. The PVB/PPyCB5 coatings were smooth with no cracks on the surface (Fig. 1a). However, as the amount of PPyCB was increased, the coatings became more prone to the appearance of surface cracks (Fig. 1b and c). This might have occurred due to the increased amount of carbon black particles in the composite coatings which hindered a continuous film formation by reducing the contact between PVB and PPy.

Fig. 1 Texture of PVB/PPyCB composite coatings (a) PVB/PPyCB5, (b) PVB/PPyCB10, and (c) PVB/PPyCB20.

The mechanical properties of PVB/PPyCB composites studied in this work revealed that all the composite films are brittle materials with high Young’s moduli (E) and low elongations at break (ε_σmax). The Young’s modulus of the PVB films modified with PPyCB increased with the concentration of conducting pigments. The elongation at break and tensile strength decreased in the PVB/PPyCB composites, though these mechanical parameters were independent of the PPyCB concentration which showed an irregular behavior at high concentrations (Table 2). However, a strong adhesion of PVB and PPyCB pigments on the metal surface occurred during the fabrication of the PVB/PPyCB coatings at high temperature. Fig. 2 show the images of cross cut tape test conducted on PVB/PPyCB20 coatings before and after 1 d immersion in 4 wt% sodium chloride solution. The edges of the cut in Fig. 2a were smooth and the squares of the lattices were not detached, indicating the strong adhesion of the composite matrix on the metal surface. During the immersion, even though the corrosion products were formed mainly at the cross cut regions, most of the coating areas were not detached while the tape test (Fig. 2b), which showed the stability of the PVB/PPyCB20 coating, indicated that strong binding on the metal surface prevented the ingress of water through the edges of the scratches.

Table 2 Tensile properties of the pure PVB and PVB/PPyCB composites

Sample	Young’s modulus (E) MPa	Tensile strength (σmax) MPa	Elongation at break (εσmax) %
PVB	1181 ± 190	32.0 ± 8	2.77 ± 0.1
PVB/PPyCB5	1838 ± 220	29.9 ± 10	2.23 ± 0.4
PVB/PPyCB10	2134 ± 188	23.5 ± 9	1.65 ± 0.2
PVB/PPyCB20	2170 ± 325	25.1 ± 7	1.70 ± 0.2

---

Fig. 2 Cross cut tape test on PVB/PPyCB20 composite coating (a) before and (b) after 2 d immersion in 4 wt% NaCl solution.

Electrochemical measurements

The corrosion resistance of the PVB/PPyCB composite coatings was studied by electrochemical measurements as both metals and conducting polymers lead to redox reactions when in contact with electrolyte. OCP and polarization measurements were conducted on the PVB and PVB/PPyCB composite coatings by exposing a 1 cm² area to 4 wt% NaCl solution. Evolution of OCP with time of immersion is shown in Fig. 3. The initial OCP values of all coatings were observed to be more positive than that of bare steel (−0.6 V vs. SCE),⁸ thus confirming the barrier nature of the coatings. The OCP of the PVB coating exhibited a passive potential of steel at 0.05 V vs. SCE. After 30 h of immersion, it decreased sharply to a potential (−0.66 V vs. SCE) close to the corrosion potential of bare steel and remained in the same range till the completion of measurement. Thus, protection of the metal surface by the PVB coating persisted over a short time period and as corrosion started on metal surface by the ingress of electrolyte through the defects in the coating, OCP decreased rapidly to the active dissolution potential of iron. Since PVB is not an inhibitive polymer, the corrosion initiated on the metal surface continued to propagate and no further corrosion protection was offered by the PVB coating. After a slight decrease of OCP from −0.12 V vs. SCE to −0.22 V vs. SCE, 5% the PPyCB coating exhibited a shift in the potential to −0.15 V vs. SCE after 1 h immersion, which was retained till the end of the 2 d analysis. A similar trend was observed for the 10% PPyCB coating. After the initial decrease from −0.26 V vs. SCE to −0.30 V vs. SCE, OCP required 2 h to reach plateau behavior at −0.19 V vs. SCE. Unlike the PVB coating, the tendency of PPyCB coatings to shift the OCP to potentials nobler than bare steel increased with enhancing PPyCB content in the composite coating. The initial passive potential of the PVB/PPyCB20 composite coating at −0.03 V vs. SCE decreased to a pseudo-plateau at −0.20 V vs. SCE and was retained for 10 h. At this plateau stage, some oscillations of the potential could be observed, which were attributed to the anion exchange of the PPy coating with the environment during equilibration in NaCl solution.¹² Such oscillations were also observed for the 5% and 10% PPyCB coatings, however, the time taken for the stabilization was lower than for the 20% PPyCB coating. After the stabilization, the potential for 20% PPyCB coating was decreased to the dissolution potential of iron at −0.30 V vs. SCE. Two hours after the start of corrosion, OCP increased slowly to a plateau value of −0.04 V vs. SCE, which was close to the initial potential, thus indicating that metal subjected to corrosion was shifted to a more noble side. The positive shift of OCP basically shows the passive state of underlying metal because of the good corrosion protection ability of the surface film.³⁹ Therefore, the potential jump of 260 mV was associated with re-passivation of the defects on the coatings which might have resulted due to the deposition of protective layers on the metal surface. Kowalski et al. have also reported a similar behavior of self-healing on bi-layered PPy coatings doped with hetero-polyanions, which resulted due to the synergistic effect of PPy redox reactions and catalytic reaction of dopant anions with metal ions.¹⁰ The first potential plateau at around −0.20 V vs. SCE of PVB/PPyCB20 coatings was associated with the extended passive state of the iron, which was noted by Hien et al.¹³ and Nguyen et al.⁹ for analysis of PPy coated iron substrates. According to these reports, this potential also results in the decreased ingress of chloride anion in the coating. However, the first plateau, extended for a short duration, moved to a second potential plateau at −0.30 V vs. SCE. At this potential, active dissolution of iron with Cl⁻ attack and subsequent pit formation took place, which drove the polypyrrole into an un-doped state. The released dopant formed stable iron complexes inside the defect, and thus provided a blocking effect for further iron dissolution. The return of OCP towards the initial potential (−0.04 V vs. SCE) suggested that the deposited iron complex and the PPy polymer effectively blocked the diffusion of corrosive ions to the passive oxide layer formed beneath the deposit. It should be noted here that, before the recovery of noble potential, the steel remained in the active potential region in which iron dissolved and the duration of potential shift to the noble region was also high. Such a shift of OCP towards the noble potential of steel was also reported by Kowalski et al., where a passive oxide layer was stabilized by the iron molybdate deposition at the defects.¹⁰

Fig. 3 OCP vs. time plots of PVB and PVB/PPyCB composite coatings in 4 wt% NaCl solution.

Corrosion protection of PPyCB pigments in the composite coatings was studied further by measuring the impedance at the open circuit potential. 5% and 10% PPyCB composites showed low frequency impedance in the same order of PVB at 10⁹ with a slightly decreased magnitude. However, Bode plots of the 20% PPyCB coating measured after 5 min immersion showed a higher low frequency impedance than the pure PVB coating (Fig. 4a). Corrosion protection of organic coatings are mainly associated with their barrier properties, which prevent oxygen and moisture from reaching the metal surface.¹⁷ Such barrier coatings generally show a time constant at high frequency regions in the EIS plots.⁴⁰ Therefore, the plateau in the high frequency region in the Bode phase angle plot of the PVB coating could be attributed to its barrier effect, which remained close to 90° over half of the measurement frequencies (10¹ to 10⁴, Fig. 4b). This plateau nature of the Bode phase angle plot was not affected by the addition of PPyCB pigment into the PVB matrix, even at a high concentration of 20% PPyCB composite. It suggests that the PVB/PPyCB composite coatings had a barrier nature, which was sufficient to protect the underlying metal from corrosion. Such a barrier nature would have resulted from good dispersion and better interaction of PPyCB pigments with the metal and PVB matrix. The low frequency phase angle at 45° of PVB/PPyCB20 composite coating in comparison to nearly zero (6.5°) of pure PVB coating shows the enhanced corrosion protection of the coatings at high content of PPyCB pigments (Fig. 4b). Furthermore, 2 d immersion in 4 wt% NaCl solution decreased the corrosion resistance of pure PVB coating. It was observed in the Bode plot by the decrease in Z modulus at the low frequency region. Changes were more obvious in the Nyquist plot, where the diameter of the semicircle decreased significantly after 2 d immersion (Fig. 5a). However, in the case of PPyCB composite coatings, low frequency Z modulus in the Bode plot increased after 2 d immersion (Fig. 4a). Impedance in the Nyquist plot also increased significantly, especially at high concentrations of PPyCB composites (Fig. 5a and b). Significant changes in the low frequency phase angle were also obvious in the Bode phase angle plot of PVB/PPyCB20 coating which shifted to 69° after 2 d exposure in sodium chloride solution (Fig. 4b). These changes in the EIS plots indicated that PVB/PPyCB composite coatings imparted protection that occurred through the synergistic effect of redox reactions and electrical conductivity of the PPyCB pigments.

Fig. 4 EIS plots of PVB and PVB/PPyCB composite coatings in 4 wt% NaCl solution: (a) Bode plot and (b) Bode phase angle plot.

Fig. 5 Nyquist plots of (a) PVB, PVB/PPyCB5, PVB/PPyCB10 composite coatings and (b) PVB/PPyCB20 composite coating after 5 min and 2 d immersion in 4 wt% NaCl solution.

To study the behavior of the PPyCB coatings in a corrosive environment, Nyquist plots were fitted with an equivalent circuit shown in Fig. 6 and for which the obtained best fitting parameters are given in Table 3. The symbol R_s represents the solution resistance of the bulk electrolyte between the reference electrode and working electrode, R_c is the coating resistance, R_ct is the charge transfer resistance, Q_c and Q_dl are the constant phase elements (CPE) used instead of the pure capacitances C_c and C_dl respectively.⁴¹ PVB exhibited R_c values typical for a barrier coating which decreased with immersion time. On the other hand, the continuous immersion increased the R_c values of the PPyCB coatings. Coating resistance (R_c) is related to the resistance of the electrolyte in pores, cracks and pits, and hence indicative of the barrier properties of the coating.^42,43 Therefore the observed increase in R_c after 2 d immersion suggested that direct ingress of electrolyte through the coating is diminished. It can be further explained based on the constant phase element (CPE). According to the previous references, constant phase element permits the simulation of phenomenon that deviates from a pure capacitive behavior.^19,20,41 Therefore, the changes in CPE can be explained by evaluating the expression in terms of capacitance:

C = Y₀(ω_max)ⁿ⁻¹

where Y₀ is the magnitude of the CPE, ω_max is the frequency at which the imaginary impedance reaches a maximum for the respective time constant, and n is the exponential term of the CPE which can vary between 1 for a pure capacitor and 0 for a pure resistor.^19,20 If n is equal to 1, the CPE behaves as a pure capacitor; when n is equal to 0, it represents a resistor and when n is equal to −1, it represents an inductor. Since the constant n given in the Table 2 is close to 1, CPE would be very similar to a pure capacitor, and hence, the constant Y₀ follows the same trend as the capacitance.⁴⁴ The capacitance of the coating (C_c) is proportional to its dielectric constant and can be, therefore, attributed to the amount of water absorbed by the coating.⁴³ The Q_c constant Y₀ of the PVB coating increased significantly after 2 d immersion in 4 wt% NaCl solution indicating the continuous ingress of electrolyte into the coating (Table 3). At the same time, PPyCB coatings showed a decrease in Y₀ values. Cascales et al. also analyzed the polypyrrole/water interface using a molecular dynamics simulation study and stated that reduced polypyrrole was hydrophobic and oxidized polypyrrole was hydrophilic in nature. Due to the high hydrophobicity of the reduced polypyrrole, water molecules were repelled from the core of the polymer matrix and so did not penetrate into the polymer matrix.⁴⁵ In the case of PPyCB composite coatings, the reduction of PPy by the diffusion of counter ions (un-doping) increased the hydrophobicity of the coating. As a consequence of this hydrophobicity, the water permeability of the coating decreased and thus the total impedance of the system as well as its barrier nature increased. It was observed in the OCP plots in Fig. 3 as retaining the barrier nature after the healing of defects on PVB/PPyCB20 coatings. In the PVB/PPyCB composite coatings, anodic dissolution of iron by the attack of corrosive species initiated un-doping of polypyrrole, resulting in the formation of passive film at the defected areas which prevented corrosion. The diminished corrosion process of the metal surface by the passive layer deposition are obvious in the R_ct values. R_ct, corresponding to the resistance to charge transfer processes on the metal surface, increased significantly for 20% PPyCB coatings with immersion time, whereas it decreased for pure PVB coating (Table 3). During the process of the un-doping of the PPy, the oxidation charge of PPy changed to the reduced form. Ions from the electrolyte (Cl⁻ and OH⁻) balanced the charge within the polymer.^4,9,46,47 At prolonged exposure time, the reduction of the polymer became the main cathodic reaction and the percentage of reduced polymer increased the hydrophobicity of the coating and thus decreased the water permeability.¹⁴ The reduced PPy had an ability to capture dissolved oxygen in the coating, which was reported by Yan et al.⁴ This oxygen-scavenger effect of PPy pigments decreased the O₂ permeability of the coating and, subsequently the corrosion rate of the substrate. A schematic representation of corrosion protection provided by PVB/PPyCB composite coatings is shown in Fig. 7. As per the scheme, the redox properties of PPy provided an anodic protection to the steel by inducing a passive film formation at the defect area, which in turn increased the amount of reduced PPy near the passivated areas. This passive film along with the hydrophobic nature of the reduced PPy limited the charge transfer reaction rate by acting as a diffusion barrier.

Fig. 6 Equivalent circuit used for EIS modelling.

Table 3 EIS parameters of PVB and PVB/PPyCB composite coatings at different time of immersion in 4 wt% NaCl solution

Coatings	10 minutes							2 days
	Rs^a, (Ω cm^2)	Rc^b, (Ω cm^2)	Qc		Rct^e, (Ω cm^2)	Qdl		Rs^h, (Ω cm^2)	Rc^i, (Ω cm^2)	Qc		Rct^l, (Ω cm^2)	Qdl
			n^c	Yo^d (Ω^−1 cm^2)		n^f	Yo^g (Ω^−1 cm^2)			n^j	Yo^k (Ω^−1 cm^2)		n^m	Yo^n (Ω^−1 cm^2)
a 5% probable error.b 9% probable error.c 1% probable error.d 8% probable error.e 12% probable error.f 2% probable error.g 7.8% probable error.h 3% probable error.i 10% probable error.j 1% probable error.k 10% probable error.l 11% probable error.m 1.5% probable error.n 5.6% probable error.
PVB	1447	9.27 × 10^9	0.991	59.45 × 10^−12	17.45 × 10^9	0.989	52.80 × 10^−12	1440	7.76 × 10^5	0.977	0.116 × 10^−9	1.20 × 10^9	0.505	0.554 × 10^−9
PVB/PPyCB5	705	1.73 × 10^9	0.985	0.109 × 10^−9	7.045 × 10^9	0.802	0.156 × 10^−9	2368	2.96 × 10^9	0.999	0.080 × 10^−9	8.60 × 10^9	0.832	0.110 × 10^−9
PVB/PPyCB10	1716	1.31 × 10^9	0.978	0.101 × 10^−9	3.154 × 10^9	0.799	0.211 × 10^−9	3775	8.67 × 10^9	0.980	0.056 × 10^−9	18.9 × 10^9	0.825	0.168 × 10^−9
PVB/PPyCB20	1016	0.19 × 10^12	0.977	97.04 × 10^−12	23.2 × 10^9	0.818	4.37 × 10^−9	3686	0.46 × 10^12	0.999	82.25 × 10^−12	0.22 × 10^12	0.981	1.71 × 10^−9

---

Fig. 7 Schematic representation of corrosion protection imparted by PPy pigments in PVB/PPyCB composite coatings.

Fig. 8 shows the cathodic and anodic polarization curves recorded for the PVB and PVB/PPyCB composite coatings after 1 d of exposure in 4 wt% NaCl solution. It is obvious in the plots that anodic polarization curves deviated from the linear Tafel behavior. This deviation would have occurred by the formation of passive film and pitting. The existence of passivation in conjunction with a dissolution reaction, due to pitting, does not result in a well-defined experimental anodic Tafel region.^38,45 Due to the absence of linearity in the anodic branch, accurate evaluation of the anodic Tafel slope by Tafel extrapolation of the anodic branch was not possible. Therefore, Tafel extrapolation of the cathodic branch of the polarization curve to the corrosion potential (E_corr) was used for the determination of I_corr.³⁶ Thus the obtained electrochemical parameters, corrosion potential (E_corr), corrosion currents densities (I_corr) and cathodic Tafel slope (β_c) are tabulated in Table 4 as a function of PPyCB concentration in the composite coating. 5% and 10% PPyCB coatings showed E_corr in a more negative side than pure PVB coating after 1 d of immersion. However, an anodic shift of E_corr occurred for 20% PPyCB coating, thus confirming the protection provided by PPyCB at high concentrations. The penetration of electrolyte down to the metal surface through the pores or cracks present on the coating initiated the corrosion processes. With time, PPy in the coating released dopant anions by redox reactions and suppressed the corrosion by passivating the defected areas on the metal surface, which shifted the corrosion potential of the metal to anodic side. The suppression of the anodic dissolution process by the deposited passive layers was attributed to the deviation of anodic polarization curves from the linear Tafel behavior. The deposited passive layer along with the reduced PPy hindered the diffusion of oxygen and water down to the metal surface which diminished the cathodic reduction reactions, as observed from the high value to cathodic Tafel slope (β_c) of 20% PPyCB after 1 d exposure (Table 4). Therefore, it can be suggested that PPy imparted anodic and cathodic protection to the underlying metal substrate. However, the corrosion current densities (I_corr) of the 20% PPyCB coating were higher than those of the pure PVB coating. The enhancement in I_corr might be caused by the contribution from PPy redox reactions in the PVB matrix, and it was observed to be increasing with increasing PPy content in the coatings. A similar increase in the corrosion current density of an aluminium substrate with a PPy coating has also been reported by Liu et al.¹¹ Though the E_corr of 5% and 10% PPyCB coatings were more negative than the PVB coating, the I_corr of the 5% PPyCB coating was lower than PVB and nearly the same value as the PVB was observed for the 10% PPyCB coating, indicting the protection provided by these coatings as well. The increased I_corr of the 10% PPyCB coating in comparison to the 5% PPyCB coating would have resulted from the redox reactions of PPy. Though 5% and 10% PPyCB coatings also provided protection, this effect was not sustainable for longer periods of time due to the lesser amount of PPy in the PPyCB composite coatings. For comparison, the corrosion resistance behavior of other nanocomposite coatings of PPy and polyaniline (PANI) reported previously are given in Table 5.^17,48–50 The E_corr of these coatings was recorded more in the cathodic side and I_corr was higher than those noted for PVB/PPyCB composite coatings in Table 3. It clearly shows the enhanced resistance of PVB coating to corrosion processes of the underlying metal, because of the inhibition behavior of incorporated PPyCB pigments.

Fig. 8 Polarization curves of PVB and PVB/PPyCB composite coatings after 1 d immersion in 4 wt% NaCl solution.

Table 4 List of electrochemical parameters obtained from polarization curves of PVB and PVB/PPyCB composite coatings after 1 d immersion in 4 wt% NaCl solution

Coatings	Ecorr (V vs. SCE)	Icorr (mA cm^−2)	βc (mV per decade)
PVB	−0.246	3.60 × 10^−7	135
PVB/PPyCB5	−0.265	0.67 × 10^−7	137
PVB/PPyCB10	−0.348	3.88 × 10^−7	132
PVB/PPyCB20	−0.160	2.00 × 10^−6	747

---

Table 5 Comparison of E_corr and I_corr of other nanocomposite coatings of polypyrrole and polyaniline reported in ref. 17 and 48–50 ^a

Material	Coating thickness	Corrosive media	Ecorr (V)	Icorr (mA cm^−2)
a PVB: polyvinyl butyral, PANI: polyaniline, MMT: montmorillonite clay, PEA: poly(o-ethoxy aniline), PPy: polypyrrole.
PVB/PANI	17	3.5 wt% NaCl	−0.30	—
MMT/PEA	20	5.0 wt% NaCl	−0.48	2.69 × 10^−4
Na-MMT/PPy	—	3.5 wt% NaCl	−0.43	0.2827
Epoxy/PPy-flyash	—	3.5 wt% NaCl	−0.50	1.40 × 10^−4

---

Immersion test

In order to further analyze the protection of the PVB/PPyCB coatings and the nature of protective film formed beneath the coatings, defects were created intentionally on the coating surfaces, followed by immersion in 4 wt% NaCl solution. The images of the PPyCB composite coatings in Fig. 9 were taken when the scratches became invisible. Optical images of the scratches on the coatings’ surface indicated that the corrosion was stopped in composite coatings, specifically at high PPyCB pigment concentrations. Healing of the scratches on PPyCB coatings was identified by looking for the black and red coloured products expected to form at the corroded areas, similar to the optical image of bare steel in Fig. 9b. Optical images of the scratches on the PVB coating indicated complete corrosion with black corrosion products during the immersion in a 4 wt% NaCl solution. The PVB/PPyCB5 coating also exhibited a similar result as pure PVB coating, except for certain areas of the scratch which were covered by a passive film, instead of red and black corrosion products. The effect of passive film formation was more prominent in the PVB/PPyCB10 and PVB/PPyCB20 composite coatings. In addition, the time taken for the passive film formation was related to the PPyCB content in the coatings. For the PVB/PPyCB10 coating, passive film formation took about 3 d to cover most of the scratch, whereas the scratch was protected within 2 d for the PVB/PPyCB20 coating. Significant healing occurred for the PVB/PPyCB20 coating where the self-healed scratches were difficult to identify from the rest of the areas (Fig. 9j). In addition, no red and black corrosion products were observed on the healed scratches of PVB/PPyCB10 and PVB/PPyCB20 coatings (Fig. 9h and j), which were observed on the scratches of PVB and PVB/PPyCB5 coatings (Fig. 9d and f), indicating the enhanced protection of PVB/PPyCB composite coating at high percentages of PPy pigment. In order to analyze the nature of passive film, scratched areas were subjected to SEM and Raman spectroscopic measurements. Fig. 10a and b show SEM images of scratches on the PVB surface and PVB/PPyCB20 composite coating respectively after 2 d immersion in 4 wt% NaCl solution. Corrosion products distributed on the metal surface as agglomerated particles were obvious in the scratch on the PVB coating (Fig. 10a). Raman spectra of these particles exhibited characteristic bands of Fe₃O₄ oxides at 405 cm⁻¹, 655 cm⁻¹ and 1308 cm⁻¹ and characteristic Raman shifts of γ-Fe₂O₃ at 223 cm⁻¹ and 291 cm⁻¹ (Fig. 10c).^5,51 The scratch on the PVB/PPyCB20 composite coating was observed to be completely covered by a thick layered material (Fig. 10b). The Raman spectrum of this layer exhibited sharp signals in the region of γ-Fe₂O₃ oxides and less intense signals in the Fe₃O₄ oxides region (Fig. 10d). In addition to these iron oxide signals, two broad Raman shifts were apparent at 1288 cm⁻¹ and 1103 cm⁻¹, attributed to the asymmetric and symmetric axial deformation of O=S=O groups of sulfonic acid.¹¹ Therefore, it can be assumed that the anodic dissolution of iron activated organic sulfonic acid doped PPy to release the dopant at the oxidizing areas, which created a stable complex with metal irons and stabilized the passive oxide layers on the metal surface, thus preventing the underlying metal from further corrosion. Thus, from electrochemical and microscopic analysis, it should be concluded that doped PPy in the PVB/PPyCB composite coatings provided protection by acting as an inhibitive additive. In an earlier study, Armelin et al. generated PPyCB composite coatings of ∼140 μm thickness with epoxy matrix on carbon steel and PPy concentration was optimized to 1 wt% for the maximum corrosion protection.¹⁵ Ruhi et al. also fabricated a three component system of chitosan–polypyrrole–SiO₂ composite coatings with epoxy polymer and observed an adequate corrosion protection at a maximum of 2 wt% of chitosan–polypyrrole–SiO₂ pigment in epoxy matrix.²³ However, in the case of PVB/PPyCB composite coatings, a maximum of 20 wt% was incorporated in the PVB matrix at a much lower coating thickness of 20 ± 5 μm and a significant protection was observed at 48 h of immersion in 4 wt% NaCl solution.

Fig. 9 Optical images of scratches during immersion in 4 wt% NaCl solution. Bare steel (a) before and (b) after 2 d; PVB (c) before and (d) after 7 d; PVB/PPyCB5 (e) before and (f) after 4 d; PVB/PPyCB10 (g) before and (h) after 3 d; PVB/PPyCB20 (i) before and (j) after 3 d.

Fig. 10 SEM images of scratches on (a) PVB coating and (b) PVB/PPyCB20 composite coating after 2 d immersion in 4 wt% NaCl solution. Raman spectra of marked areas on the scratches for (c) PVB coating and (d) PVB/PPyCB20 composite coating.

Protective performance of graphene incorporated PVB/PPyCB composite coatings

Fig. 11a shows the changes in OCP of PVB/PPyCB20/Gr5 coating as a function of immersion time in 4 wt% NaCl solution. The OCP behavior of the graphene incorporated coating was very similar to the PVB/PPyCB20 composite coating. The OCP shifted from −0.12 V vs. SCE to −0.26 V vs. SCE during the initial 5 h immersion, followed by an increase to a value of −0.16 V vs. SCE at the end of the measurement. Such an OCP pattern suggested that corrosion inhibition took place through the healing of defects on the PVB/PPyCB20/Gr5 coating. EIS was performed to study the electrochemical responses of PVB/PPyCB20/Gr5 coating on steel substrate. The Bode plot measured during the initial time of immersion, shown in Fig. 11b, exhibited two regions of distinct electrochemical responses. The first time constant at high frequencies (10² Hz to 10⁴ Hz) was governed by the coating response. At low frequency, the kinetics of a charge-transfer process governed the response as manifested by the second time constant which indicated the electron transfer reaction between metal and conducting polymer. Such an electrochemical reaction (metal oxidation) at the metal–PVB/PPyCB20/Gr5 coating interface resulted in a low frequency impedance in the Bode plot (Fig. 11b). These results indicated that PVB/PPyCB20/Gr5 coatings were more prone to defects due to weak interaction of Gr with PVB matrix and metal. However, prolonged immersion in 4 wt% NaCl solution increased impedance especially in the low frequency region by an order of 10². Since low frequency impedance is directly related to the protective efficiency of the coating,⁵² it can be suggested that conducting particles, PPyCB pigment and graphene platelets, induced an inhibition to corrosion processes on the metal surface. These changes are more obvious in Bode phase angle plot shown in Fig. 11b. The first time constant attributed to the coating response changed to a resistive plateau at 90° which indicated that the coating barrier effect enhanced with immersion time. The coating barrier is related to the resistance of the electrolyte in pores, cracks and defects on the coating.⁴² Therefore, the behavior observed in the PVB/PPyCB20/Gr5 coating can be considered to be caused by the passive layer deposition on the coating defects, which prevented the direct ingress of electrolyte down to the metal surface. It was more obvious in the second time constant, attributed to the charge transfer reactions on the metal/polymer interface, which completely vanished after 2 d immersion. In addition, the low frequency phase angle increased to 45°. These changes suggested that corrosion attack was effectively stopped by the passive layer deposition on the metal surface. Changes in the coating and on the metal surface were studied by fitting the EIS plots with an equivalent circuit shown in Fig. 6 and the fitting data are displayed in Table 6. R_ct, corresponding to resistance to charge transfer process on the metal surface, increased significantly with immersion time which indicated the diminished corrosion process on the metal surface by the deposition of passive layer. The deposition of protective layers could be confirmed further from the deposited layer capacitance (C_dl):

C_dl = εε_o(At⁻¹)

where ε and ε_o represent dielectric constant of deposition and the permittivity of free space (8.9 × 10⁻¹⁴ F cm⁻¹) respectively, A is the area of the exposed metal surface and t is the deposited layer thickness.⁵³ As Q_dl was used instead of pure capacitance, the decrease in Y₀ constant of Q_dl with immersion time indicated an increase in the thickness of deposited layers. The Y₀ values also indicate the area of metal surface exposed for the corrosion process,^42,43 which decreased with immersion time. Such deposition of layers on the metal surface can be attributed to an increased n₂ value. The exponential term of the constant phase element n also measures the surface inhomogeneity; the lower its value, the higher the surface roughening of the metal/alloy and vice versa.⁵⁴ Therefore, the increased R_ct, n₁ and decreased Y₀ constant of Q_dl (Table 6) confirmed the formation of passive layers at the defect areas. As per the scheme shown in Fig. 7, corrosion protection of the coating was controlled by polypyrrole’s ability to capture and transport electrons from the metal surface. Since both conducting polymer and graphene platelets possess conjugated π systems, an electronic interaction via π–π stacking can be expected to occur.^34,35 It was observed in the PVB/PPyCB20/Gr5 coating that graphene promoted the electron transfer for the PPy redox processes and thus enhanced the protective passive layer formation on metal surface. Furthermore, it has been reported that graphene is an excellent barrier to oxygen and water diffusion.^28–35 The diminished water permeability of PVB/PPyCB20 coating with the addition of graphene resulted a decreased Y₀ value of Q_c in 2 d immersion. It was attributed further to an increase of coating resistance (R_c) (Table 6) as well. Therefore, it can be suggested from these results that the reinforced effect of electrical conductivity along with the reduced oxygen and water permeability by graphene incorporation enhanced the protective behavior of PVB/PPyCB20 coatings.

Fig. 11 Electrochemical analysis conducted on graphene incorporated PVB/PPyCB20 composite coating during immersion in 4 wt% NaCl solution: (a) OCP vs. time of immersion, (b) Bode plots measured after 5 min and 2 d immersion, solid lines show the fit results.

Table 6 EIS parameters of PVB/PPyCB20/Gr5 composite coatings at different time of immersion in 4 wt% NaCl solution

Time	Rs^a, (Ω cm^2)	Rc^b, (Ω cm^2)	Qc		Rct^e, (Ω cm^2)	Qdl
			n1^c	Y0^d (Ω^−1 cm^2)		n2^f	Y0^g (Ω^−1 cm^2)
a 3.6% probable error.b 8% probable error.c 1.5% probable error.d 14% probable error.e 10% probable error.f 1.2% probable error.g 9% probable error.
10 min	760	2.201 × 10^6	1	3.680 × 10^−10	10.01 × 10^6	0.756	22.76 × 10^−9
2 d	1860	6.915 × 10^9	1	40.56 × 10^−12	6.50 × 10^9	0.999	0.139 × 10^−9

---

Polarization measurements on the PVB/PPyCB20/Gr5 coating were conducted after 1 h and 1 d immersion in 4 wt% NaCl solution (Fig. 12). E_corr values obtained from the polarization curves were −0.154 V vs. SCE and −0.156 V vs. SCE respectively for 1 h and 1 d exposed samples (Table 7). It indicated that the healing of defects by the deposited passive layers helped the coating to maintain its protective performance. It is also seen that the cathodic branches of the polarization curves displayed a typical Tafel behavior whereas anodic polarization curves were deviated from the linear Tafel behavior over the complete applied potential range. Therefore, Tafel constants (I_corr and β_c) given in Table 7 were calculated from the extrapolation of the cathodic branch of the polarization curves. The curvature of the anodic branch was attributed to the deposition of passive layers on the metal surface. The protection offered by the deposited passive layer on the metal surface could be confirmed from the increase of β_c value after 1 d exposure in sodium chloride solution. The corrosion protection performance of PPyCB composite coatings resulted due to synergistic effect of electrical conductivity and redox processes of PPy. The presence of graphene particle in the composite improved the electrical conductivity of the coating, which in turn enhanced the redox processes of PPy as well. This probably attributed to the unexpected increase of I_corr of PVB/PPyCB20/Gr5 coating observed after 1 d immersion in sodium chloride solution. In addition uniform distribution of graphene platelets enhanced the diffusion paths of water and ions through the coating, thus effectively prevented its availability on the metal surface for the corrosion processes.

Fig. 12 Polarization curves of graphene incorporated PVB/PPyCB20 composite coating after 1 h and 1 d immersion in 4 wt% NaCl solution.

Table 7 List of electrochemical parameters obtained from polarization curves of PVB/PPyCB20/Gr5 composite coating after 1 h and 1 d immersion in 4 wt% NaCl solution

Time	Ecorr (V vs. SCE)	Icorr (mA cm^−2)	βc (mV per decade)
1 h	−0.154	3.91 × 10^−7	144
1 d	−0.156	4.38 × 10^−7	151

---

Conclusions

PPyCB pigment, up to a concentration of 20 wt%, was successfully incorporated into PVB coating formulation without affecting the barrier effect and adhesion properties of organic coatings. PPy imparted a self-healing character to PVB/PPyCB coatings that shifted the OCP of the metal from a dissolution potential to a noble potential on immersion in aggressive corrosive media. It was also reflected in the polarization curves of the PVB/PPyCB20 composite coatings with an anodic shift of E_corr. A synergistic effect of the redox properties and diffusion barrier nature of PPy resulted in the corrosion resistance performance of the PVB/PPyCB coatings. SEM and Raman spectra of the surface of intentionally made scratches on the coatings exhibited formation of passive layers due to the interaction between released dopant from organic sulfonic acid doped PPy and iron oxides. The deposited passive layer along with the reduced PPy hindered the diffusion of oxygen and water down to the metal surface, which diminished the corrosion processes. This was attributed to an increased R_ct, significant for 20% PPyCB composite coatings, with immersion time. Incorporation of conducting graphene particles in the PVB/PPyCB20 composite coatings enhanced the process of passive layer deposition on the metal surface, which hindered electron transfer between metal and polymer along with preventing water and gas from reaching the metal surface, thus preventing metal from further corrosion. It can be concluded from these results that the use of conducting composite particles such as PPyCB as corrosion inhibiting pigments imparts a high degree of self-healing protective nature to the organic coatings without sacrificing the inherent coating properties.

References

1. Z. Tian, H. Yu, L. Wang, M. Saleem, F. Ren, P. Ren, Y. Chen, R. Sun, Y. Sun and L. Huang, RSC Adv., 2014, 4, 28195.
2. K. Qi, Y. Qiu, Z. Chen and X. Guo, Corros. Sci., 2012, 60, 50.
3. C. Xing, Z. Zhang, L. Yu, L. Zhang and G. A. Bowmaker, RSC Adv., 2014, 4, 32718.
4. M. Yan, C. A. Vetter and V. J. Gelling, Corros. Sci., 2013, 70, 37.
5. B. Wessling, Adv. Mater., 1994, 6, 226.
6. W. K. Lu, R. L. Elsenbaumer and B. Wessling, Synth. Met., 1995, 71, 2163.
7. G. Paliwoda-Porebska, M. Stratman, M. Rohwerder, K. Potje-Kamloth, Y. Lu, A. Z. Pich and H. J. Adler, Corros. Sci., 2005, 47, 3216.
8. J. O. Iroh and W. Su, Electrochim. Acta, 2000, 46, 15.
9. T. Nguyen, T. Le, B. Garcia, C. Deslouis and X. Q. Le, Electrochim. Acta, 2001, 46, 4259.
10. D. Kowalski, M. Ueda and T. Ohtsuka, J. Mater. Chem., 2010, 20, 7630.
11. A. S. Liu, M. C. Bezerra and L. Y. Cho, Mater. Res. Bull., 2009, 12, 503.
12. A. C. Balaska, I. A. Kartsonakis, G. Kordas, A. M. Cabral and P. J. Morais, Prog. Org. Coat., 2011, 71, 181.
13. N. T. L. Hien, B. Garcia, A. Pailleret and C. Deslouis, Electrochim. Acta, 2005, 50, 1747.
14. N. V. Krstajic, B. N. Grgur, S. M. Jovanovic and M. V. Vojnovic, Electrochim. Acta, 1997, 42, 1685.
15. E. Armelin, R. Pla, F. Liesa, X. Ramis, J. I. Iribarren and C. Aleman, Corros. Sci., 2008, 50, 721.
16. W. J. Bae, K. H. Kim and W. H. Jo, Macromolecules, 2005, 38, 1044.
17. S. Radhakrishnan, S. R. Siju, D. Mahanta, S. Patil and G. Madras, Electrochim. Acta, 2009, 54, 1249.
18. M. R. Mahmoudian, Y. Alias, W. J. Basirun and R. Yousefi, Metall. Mater. Trans. A, 2013, 44, 3353.
19. M. kraljic, Z. Mandic and L. Duic, Corros. Sci., 2003, 45, 181.
20. J. E. Pereira da Siva, S. I. Cordoba de Torresi and R. M. Torresi, Prog. Org. Coat., 2007, 58, 33.
21. S. Radhakrishnan, N. Sonawane and C. R. Siju, Prog. Org. Coat., 2009, 64, 383.
22. Y. Shao, H. Huang, T. Zhang, G. Meng and F. Wang, Corros. Sci., 2009, 51, 2906.
23. G. Ruhi, O. P. Modi and S. K. Dhawan, Synth. Met., 2015, 200, 24.
24. M. Selvaraj, S. Palraj, K. Maruthan, G. Rajagopal and G. Venkatachari, J. Appl. Polym. Sci., 2010, 116, 1524.
25. Y. Tong, S. Bohm and M. Song, Austin Journal of Nanomedicine & Nanotechnology, 2013, 1, 1.
26. K. Singh, A. Ohlan and S. K. Dhawan, in Nanocomposites - New Trends and Developments, ed. F. Ebrahimi, InTech, Hampshire, UK, 2012, ch. 3, p. 37.
27. A. U. Chaudhry, V. Mittal and B. Mishra, RSC Adv., 2015, 5, 80365.
28. K. C. Chang, M. H. Hsu, H. I. Lu, M. C. Lai, P. J. Liu, C. H. Hsu, W. F. Ji, T. L. Chuang, Y. Wei, J. M. Yeh and W. R. Liu, Carbon, 2014, 66, 144.
29. P. A. Okafor, J. S. Beemat and J. O. Iroh, Prog. Org. Coat., 2015, 88, 237.
30. Y. Li, Z. Yang, H. Qiu, Y. Dai, Q. Zheng, J. Li and J. Yang, J. Mater. Chem., 2014, 2, 14139.
31. S. Mayavan, T. Siva and S. Sathiyanarayanan, RSC Adv., 2013, 3, 24868.
32. S. Sreevatsa, A. Banerjee and G. Haim, ECS Trans., 2009, 19, 259.
33. C. H. Chang, T. C. Huang, C. W. Peng, T. C. Yeh, H. I. Lu, W. I. Hung, C. J. Weng, T. I. Yang and J. M. Yeh, Carbon, 2012, 50, 5044.
34. N. T. Kirkland, T. Schiller, N. Medhekar and N. Birbilis, Corros. Sci., 2012, 56, 1.
35. M. Merisalu, T. Kahro, J. Kozlova, A. Niilisk, A. Nikolajev, M. Marandi, A. Floren, H. Alles and V. Sammelselg, Synth. Met., 2015, 200, 16.
36. Z. Shi, M. Liu and A. Atrens, Corros. Sci., 2010, 52, 579.
37. E. M. Cafferty, Corros. Sci., 2005, 47, 3202.
38. M. A. Amin, S. S. Abd El Rehim and H. T. M. Abdel-Fatah, Corros. Sci., 2010, 52, 140.
39. S. L. A. Maranhao, I. C. Guedes, F. J. Anaissi, H. E. Toma and I. V. Aoki, Electrochim. Acta, 2006, 52, 519.
40. J. Carneiro, J. Tedim, S. C. M. Fernandes, C. S. R. Freire, A. J. D. Silvestre, A. Gandini, M. G. S. Ferreira and M. L. Zheludkevich, Prog. Org. Coat., 2012, 75, 8.
41. R. Akid, M. Gobara and H. Wang, Electrochim. Acta, 2011, 56, 2483.
42. M. L. Zheludkevich, R. Serra, M. F. Montemor, K. A. Yasakau, I. M. M. Salvado and M. G. S. Ferreira, Electrochim. Acta, 2005, 51, 208.
43. A. Latnikova, D. Grigoriev, M. Schenderlein, H. Mohwald and D. Shchukin, Soft Matter, 2012, 8, 10837.
44. T. F. Conceicao, N. Scharnagl, C. Blawert, W. Dietzel and K. U. Kainer, Corros. Sci., 2010, 52, 2066.
45. J. J. L. Cascales, A. J. Fernandez and T. F. Otero, J. Phys. Chem. B, 2003, 107, 9339.
46. K. L. Levine, in Polymer Nanocomposite Coatings, ed. V. Mittal, CRC Press, Boca Raton, London, New York, 2013, ch. 11, p. 277.
47. J. Reut, A. Opik and K. Idla, Synth. Met., 1999, 102, 1392.
48. J. M. Yeh, C. L. Chen, Y. C. Chen, C. Y. Ma, K. R. Lee, Y. Wei and S. Li, Polymer, 2002, 43, 2729.
49. A. Olad, A. Rashidzadeh and M. Amini, Adv. Polym. Technol., 2013, 32(3), 21337.
50. G. Ruhi, H. Bhandari and S. K. Dhawan, Am. J. Polym. Sci., 2015, 5(1A), 18–27.
51. W. K. Lu, R. L. Elsenbaumer and B. Wessling, Synth. Met., 1995, 71, 2163.
52. T. Sugama and M. Cook, Prog. Org. Coat., 2000, 38, 79.
53. D. A. López, S. N. Simison and S. R. De Sánchez, Electrochim. Acta, 2003, 48(7), 845.
54. S. Chongdar, G. Gunasekaran and P. Kumar, Electrochim. Acta, 2005, 50(24), 4655.

---