E.C. Bucharsky^{#}, E.B. Castro and S.G. Real
Instituto de Investigaciones Fisicoqu�micas Te�ricas y Aplicadas (INIFTA),
Suc. 4, C.C.16 (1900) La Plata, Argentina.Email:
The protective performance of zinc rich paints(ZRP)/topcoating systems has been investigated using electrochemical impedance spectroscopy (EIS) and corrosion potential measurement, at different immersion times. ZRP formulations included zinc pigment content close to 78% and different organic binders, such as alkyd resins and epoxypolyamide, an alkyd type resin was used as a topcoating. Experimental impedance diagrams were described in terms of transfer function analysis using nonlinear fit routines. Changes in the dielectric properties of the topcoating, monitored by EIS, were used to evaluate the water uptake.
Key words: organic coatings, electrochemical impedance, cathodic protection.
Coatings or linings with organic and inorganic compounds are a widespread method to provide corrosion protection in order to improve the durability of engineering structures. Transport properties of water and corrosion species are very important factors in the coating formulation design.
In the development of coatings it is therefore necessary to optimize the permeability of the coating for water and corrosive species. Water, becomes the major cause of swelling, loss of adhesion, deterioration of the mechanical properties and the start of the corrosion process. Modern electrochemical techniques, such as electrochemical impedance spectroscopy, has been used for the investigation of the protective coating properties on metals [15]. In the present work, changes of coating properties have been monitored employing electrochemical impedance spectroscopy to evaluate the water transport and protective behaviour in final coating system containing an alkydtypetop and either sprayed zinc or zinc rich paint (ZRP) based on different organic binders such as epoxy and alkyd. For this purpose, a short description of the theory related to the capacitance method of evaluating water uptake is presented.
Water is transported through the coating by two different and simultaneous mechanisms: convection through pores and imperfections and diffusion through the polymer matrix.
The convection of the electrolyte through the coating pores is responsible for the initiation of the corrosion process at relatively short times, (t � 1h), allowing the electrical contact between Zn particles in the ZRP and with the steel substrate.
Water diffusion through the topcoating layer, is a much slower process, resulting in an increase of the permittivity of the polymer with immersion time. Accordingly, higher values of the film capacitance, C, may be determined from impedance measurements recorded at increasing immersion times. C being related to the polymer permittivity, e_{p}, by:
Where e_{o} is the permittivity of vacuum, A is the area of the polymer film and l is the thickness.
In this way impedance measurements may be used as a method to monitor the diffusion of water in the coating.
The empirical equation of Brasher and Kinsbury [6], accounts for the relationship between the coating capacitance and the volume fraction of water in the coating layer, as already described by de Witt et al [7]:
Where f is the volume fraction of water in the coating, C(t) is the capacitance
at time t and C(0) is the capacitance at t = 0, 80 is the relative permittivity
of water.
Parameters related to water diffusion in the paint film, such as the diffusion coefficient D, the solubility S and the permeation coefficient P, may be derived from impedance data recorded at different immersion times. For this purpose the transport equation of water in the coating must be solved with the adequate boundary conditions.
As a first approximation water diffusion may be described by Fick�s laws. The solution of Fick�s second law , for a free film taking up water from the environment, was given by Crank and Park [8,9], considering as boundary conditions:
 For t � 0,����� The concentration of water in the coating is uniform (or zero).
 For t > 0, ���� The surface concentrations of water at x = 0 and x = l are equal and constant. The existence of water , is assumed at the ZRP/topcoat interface due to the fast convective transport through the pores, as indicated by EIS data described below.
The flux of water, at x=l/2 ( half thickness of free film) is J_{w }= 0.
Taking into account eq.(2), the following expression may be derived[7], valid at long immersion times:
The simulation of experimental C(t) values, in terms of eq. (3) allows the determination of the diffusion coefficient, D, if the polymer layer exhibits ideal Fickean behavior even at long immersion times.
This procedure requires good estimated values of the initial coating capacitance, C_{0}, and of the capacitance at infinite time, C_{�}.
The solution of Fick�s laws indicates a linear dependence of C(t) vs t^{1/2}, for t � 0 [11], so C_{0 }may be calculated from the extrapolation of this plot at t = 0.
C_{�} may be determined from capacitance values at very long immersion times, provided that no swelling of the coating takes place, ie C(t) is constant for t � �.
SAE 1020 (UNS G10200) steel plates 20 x 8 x 0.2 cm were used as metallic substrate. Metal surfaces were initially sandblasted to AS 2 1/2−3 degree (SIS Standard 05 59 00/ 67), degreased with vapor toluene, and finally coated with different ZRP and topcoating alkyd resins; in all the cases panels were prepared in duplicate and stored seven days for curing at 20�2 �C before beginning the tests. The primer compositions and dry film thickness are assembled in Table 1. ZRP formulations included spherical zinc as unique pigment, whose physical properties are given in Table 2. For the sake of comparison, some steel samples were covered with sprayed Zn (sample S3 in Table 1).
Table 1.� ZRP and Top Coat Compositions
ZRP Paint 
Composition 

Thickness 
S1 
Alkyd Resin 
21.5% 
60 
S2 
Epoxipolyamide 
22.2% 
60 
S3 
Zinc Sprayed 
100% 
75 
Top Coating 
Alkyd Resin 
45.4% 
30 
Table 2. Zinc Pigment Physical Characteristics
Zinc Pigment 

Form 
Powder (Spherical Particles) 
Color 
Gray 
Odor 
Odorless 
Density (g cm^{3}) 
7.1 
Apparent density (g cm^{3}) 
2.4 
Oil Absorption (g/100 g) 
13 
Purity (%) 
99.02 
The ZRP/Topcoat dry film thickness was measured with an electromagnetic gauge employing bare sanded plates and standards of known thickness as reference. Potentials were measured and referred to in the text against a saturated calomel electrode (SCE). Electrochemical impedance Spectroscopy (EIS) measurements in the 3 mHz � f � 65 kHz frequency range were performed in the potentiostatic mode at the corresponding corrosion potential attained after different exposure times in artificial sea water using a frequency response analyzer and an electrochemical interface (Solartron, FRA 1250 and EI 1186, respectively) integrated with a PC system. The exposed geometrical area of samples was 16 cm^{2}. For impedance measurements, an activated Pt probe was coupled to the SCE through a 10 �F capacitor to reduce phase shift errors at high frequencies. Artificial sea water was prepared according to the ASTM Standard D 114190. Detailed descriptions of the experimental setup and data processing have been described elsewhere [1214].
The standarized procedures ASTM B 11785 (Salt Spray Chamber) and ASTM D 454185 (PullOff Adhesion) were also performed on the painted steel samples for comparing their results with the electrochemical ones. After these tests, the painted panels were assessed with the ASTM Standards D 61068 and D 71487 in order to evaluate the degree of rusting and of blistering, respectively, in an attempt to correlate visual observations and electrochemical data.
The dependence of the corrosion potential E_{corr} on immersion time (Fig. 1) in artificial sea water illustrates changes in the protective mechanisms supplied to the steel substrates by the studied final coating. At the beginning of the exposure the oscillations of the open circuit potential indicate that E_{corr} remains undefined due to an initial barrier protective mechanism.
With increasing immersion time, (t > 1h) E_{corr} attains a potential that lies in the range of the corrosion potential of pure Zn electrodes in sea water [15]. This fact indicating that water has reached the ZRP layer, giving rise to the electric contact among the Zn particles and with the steel substrate. In this way the cathodic protection mechanism period is established, at relatively short immersion times.
Figure 1.������� E_{corr} dependence on exposure time in artificial seawater for the tested samples
The fast initial transport of water into the ZRP layer must occur by convection through the pores and imperfections of the topcoating polymer. Diffusion of water through the polymer matrix is a much slower process, producing changes in the topcoating capacitance, C, at longer times, as described above.
After prolonged exposure time in the electrolyte, t > 120 days, E_{corr} attains values corresponding to the corrosion potential range of steel in sea water (0.65V), indicating that the cathodic protection is no longer active.
The comparison of E_{corr }vs t plots, of ZRP (of the same Zn content) without topcoating, previously published[1], with the results depicted in fig.1, shows a longer period of cathodic protection for the samples with topcoating. This fact can be interpreted considering that the topcoat acts as a barrier for the transport of both water and oxygen.
Impedance spectra at different immersion times for sample S1, S2 and S3 are depicted in Figure 2. Impedance data display similar features as those reported for ZRP without the topcoating [1,10], indicating the presence of water in the ZRP layer at short immersion times, as described above. Nevertheless significant diferences are evident: Higher impedance values are determined at comparable immersion times, and a continuous decrease of the modulus lZl is observed, in contradiction with EIS data for ZRP without Topcoat. For t > 30 days relatively constant values of lZl are attained. These experimental facts indicate a continuous increment of the exposed active Zn area, due to a much longer wetting period, in comparison with the data of ZRP without Topcoat. The whole corrosion process of active Zn is also hindered by the smallest rate in the oxygen diffusion process.
Figure 2������� Nyquist plots for samples S1, S2 and S3 at different immersion time.
An increase of the whole system impedance is observed after t > 100 days, as the formation of Zn corrosion products reduce the active area, at long exposure times.
The set of impedance diagrams contains valuable information related to the characteristic coating parameters as well as to the kinetics of corrosion process. Impedance spectra were analyzed using a non linear fit routine according to the following transfer function:
The transfer function described by eqs. (4) and (5) corresponds to the dynamic behaviour of the equivalent circuit:
Zi corresponds to:
Where Rs is the electrolyte resistence, C is the toppaint film capacitance and R_{p }the pore resistance in the topcoat. Z_{i }corresponds to the impedance related to the whole corrosion process taking place in the ZRP layer[3], R_{C }is related to the charge transfer resistance of the oxygen reduction reaction and R_{A }�accounts for the charge transfer resistance of zinc dissolution reaction. C_{dl }is the capacitance associated with the Zn/electrolyte interface. A finite diffusion impedance, Zd, was considered in order to account for the transport process involved in the cathodic partial reaction:
Zd=R_{D0 }(jS)^{1/2 }tanh(jS)^{�}, where
d and D_{o} being the diffusion length and diffusion coefficient of oxygen.
Figure 3���� Experimental and simulated Bode and Nyquist plots, for the samples S1, S2 and S3 at different exposure time.
Figure 3 shows a fairly good agreement between experimental and fitted results obtained according to the transfer function described by equations (4) and (5).
The C values obtained from the fitting procedures were used to determine the diffusion coefficient of water through the coating according to the theoretical analysis presented above. Equation (3) is valid when the transport of water is properly described in terms of Fick�s laws. If this is the case, two limiting conditions must be observed: a linear dependence of C vs t^{1/2} as t � 0, and a plateau in the C vs t curve as t � �, indicating the value of C_{�}.
The plot C _{�}vs t ^{1/2} corresponding to the system S_{2}, is depicted in Figure 4.
Figure 4.���� Dependence of
C on� t^{1/2 }��for sample S2.
�From the extrapolation to t = 0, a value of C_{0 }= 3x10^{−}^{9} F cm^{−}^{2}, may be obtained. A value of C_{� }= 4x10^{−}^{7} F cm^{−}^{2}, corresponding to the experimental capacitance after 28 days of immersion, was considered as the capacitance at infinite time.
Figure 5 depicts experimental and theoretical [log(C(t)/C_{0})/log(C_{�}/C_{0})] vs t plots
Figure 5.����� Experimental and theoretical [log(C(t)/Co) / log (C�/Co)] vs t plot, corresponding to sample S2. The theoretical curve was calculated in terms of Eq.3.
The theoretical curve was calculated in terms of eq.(3), taking into account seven terms of the series and D = 2x10^{−}^{11} cm^{2}s^{−}^{1} for the diffusion coefficient of water in the polymer layer. Good agreement between experimental and calculated data may be observed indicating that the assumption of Fickean behaviour can be considered as a good approximation for the diffusion of water through the polymers presented in this work.
The dependence of fit parameters on immersion time is assembled in Figure 6 for the different ZRP formulations.
In the absence of Topcoat[1], the continuous decrease of the capacitance C_{dl} as well as the increase of the resistance R_{A} and R_{C }can be assigned to a progressive decrease of the electrochemically active area due to the accumulation of� corrosion products in the ZRP.
In the present case the existence of a topcoat introduces modifications in this behaviour: An increase in the Zn active area, with immersion time, is deduced by the decrease of the parameters R_{A} and R_{C} and the increse of C_{dl }. Higher values of R_{DO} are also related to higher values of the diffusion length for oxygen transport, d, due to the presence of the toplayer.
Figure 6.��� Dependence of fitting parameters C,� C_{dl} , R_{A}� and� R_{c} on exposure time for samples S1, S2 and S3.
Dynamic system analysis employing small signal perturbation allows the determination of specific parameters of the system, which characterize the protective performance of the final paint scheme with increasing immersion time in sea water.
The dependence of E_{corr}, and of the fitting parameters with immersion time indicates that the topcoating acts as a barrier, hindering both water and oxygen transport. As a consequence the wetting period of the ZRP layer is much longer, and the oxygen diffusion transport is slower. In this way Zn particles provide active cathodic protection for longer exposure times than ZRP without topcoating.
EIS has proved to be a powerful tool allowing to monitor changes in the coating properties, with immersion time as well as providing information related to both water transport and corrosion process taking place at the metal/coating interface.
This research project was financially supported by Consejo de Investigaciones Ci�ntificas y T�cnicas (CONICET), Comisi�n de Investigaciones Ci�ntificas de la Provincia de Buenos Aires (CIC), and Fundaci�n Antorchas. Part of the equipment used in this work was provided by the DAAD and the Alexander von HumboldtStifung.
[1]������� E.C. Bucharsky, S.G. Real and J.R. Vilche, Corrosion Reviews, 14, No12, (1996)
[2]������� C.G. Munger, Corrosion Prevention by Protective Coatings, NACE, 2nd ed. Houston, Texas,(1986).
[3]������� C.A. Gervasi, A.R. Di Sarli, E. Calvacanti, O. Ferraz, E.C. Bucharsky, S.G. Real and J.R. Vilche, Corros Sci., 36,1963,(1994).
[4]������� C.A. Gervasi, R. Armas, A.R. Di Sarli, E.C. Bucharsky, S.G. Real and J.R. Vilche, Materials Science Forum, Vols.192194, 357, (1995).
[5]������� R.D. Armstrong, B.W. Johnson and J.D. Wrigth, Electrochim. Acta 36, 1915 (1991).
[6]������� D. M. Brasher and A. H. Kinsbury , J. Appl. Chem. 4, 62 (1954).
[7]������� E. P. M. Van Westing, G. M. Ferrari and J. H. W. de Witt, Corrosion Sci., 36, 957 (1994).
[8]������� J. Crank, The Mathematics of Diffusion, Oxford University Press, London (1967).
[9]������� J. Crank and G. S. Park, Diffusion in Polymers, Academic Press, London (1968).
[10]���� C. M. Abreu, M. Izquierdo, M. Keddam, X. R. Novoa and H. Takenouti, Electrochim. Acta, 41, 2405 (1996).
[11]���� F. Belluci and L. Nicodemo, Corrosion, 49, 235 (1993).
[12]���� R.A. Armas, C.A. Gervasi, A.R. Di Sarli, S.G. Real and J.R.Vilche, Corrosion 48, 379 (1992).
[13]���� S.G. Real, A.C. El�as,C.A. Gervasi, A.R. Di Sarli and J.R. Vilche, Electrochim.Acta 38, 2029 (1993).
[14]���� E.B. Castro,S.G. Real, R.H. Milocco and J.R. Vilche, Elctrochim.Acta 36, 117 (1991).
[15]����� M. Pourbaix , Atlas
d�Equilibres Electrochemiques, GanthierVillars, Paris, p. 312 (1963)