CORROSION SCIENCE

Evaluation of Corrosive Degradation

in Coated Steel Using Alternating Current
Impedance Spectroscopy

D.F. Wei,* I. Chatterjee,** and D.A. Jones***

ABSTRACT INTRODUCTION

Alternating current (AC) impedance spectroscopy was Electrochemical impedance spectroscopy (EIS) has
applied to measure the degradation of steel coatings been used widely and successfully to study the
previously exposed to corrosive solutions such as tap behavior of coated metals immersed in electrolyte
water and 3% sodium chloride (NaCl). The method used solutions.1-6 Simpson, et al., utilized an atmospheric
was simple and quick and did not utilize electrochemical
electrochemical monitor that used EIS to detect paint
(potentiostatic) control. Previously immersed specimens of
three types of coated steel were sandwiched between degradation in the atmosphere.7 EIS can determine a
metallic foil electrodes and connected to the impedance number of fundamental parameters related to the
measurement instrumentation. Typical Nyquist and Bode electrochemical kinetics of coating degradation from
plots were obtained showing progressive losses in coating water and ion absorption.8 Conventional EIS methods
resistance with exposure time. The advantage of the use a frequency response analyzer, an electrochemi-
method was that immersion in the electrolyte solution was cal interface (potentiostat), and a computer controller.
not necessary during measurements for atmospherically In the present work, a simplified method referred
exposed coatings, as required in electrochemical to as alternating current (AC) impedance spectros-
impedance spectroscopy (EIS). As a result, weathered and
copy was used to evaluate the degradation of coated
otherwise degraded coatings in previous atmospheric
exposures could be evaluated without the additional effects steel that previously had been subjected to a corro-
present during electrolyte immersion in EIS. The method sive environment. In this method, test samples of
successfully discriminated between low- and intermediate- coated steel were sandwiched between copper (Cu)
resistance alkyd and high-resistance epoxy coatings. foil electrodes and connected directly to a dynamic
signal analyzer (DSA) controlled by a computer
KEY WORDS: alkyd coatings, alternating current,
workstation. No potentiostat was necessary.
atmospheric corrosion, coatings, degradation, epoxy
coatings, impedance, resistance, sodium chloride, steels Three types of coatings were evaluated. The first
was a high-resistance, epoxy-based coating. The
second was an intermediate-resistance, low-solids,
Submitted for publication January 1994; in revised form, July 1994. alkyd-based coating. The third was a low-resistance,
* Department of Electrical Engineering, University of Nevada, Reno, NV,
89557. Present address: Cermetek, Sunnyvale, CA. high-solids, alkyd-based coating. Nyquist and Bode
** Department of Electrical Engineering, University of Nevada, Reno, NV, plots obtained for these coated samples under
89557.
*** Department of Chemical/Metallurgical Engineering, ENGR 170,
various immersion conditions were analyzed to obtain
University of Nevada, Reno, NV, 89557. equivalent circuit representations of the electrical
0010-9312/95/000025/$5.00+$0.50/0
CORROSION–Vol. 51, No. 2 NACE International 97
CORROSION SCIENCE

known. A schematic of the clamp-on Cu electrodes is

shown in Figure 1. The area of the electrodes was
1 cm2. Three coated samples of the same cold-rolled
steel were used in this study. The first sample was
coated with a primer of high-solids alkyd, the second
with a primer of low-solids alkyd, and the third with an
epoxy primer. The sample panels (10 cm by 30 cm
by 0.8 mm) were coated only on one side with these
single-phase primers to a thickness of 0.03 mm. Test
specimens (2.5 cm by 2.5 cm) were cut from each
of the three sample panels, cleaned and degreased
with ethanol, and stored in a desiccator at room
temperature for later use. The specimen area was
5.25 times or 84% larger than the electrode area,
thus avoiding fringing effects.
To accelerate degradation, specimens of the
FIGURE 1. Pressure clamp and metallic electrodes.
three types of coated samples were immersed fully,
one set in tap water and a second set in 3% sodium
chloride (NaCl) solution. Individual specimens were
removed from the tap water and 3% NaCl after 3, 5,
10, 15, and 25 days. Specimens of the epoxy-coated
sample also were immersed fully in tap water and
3% NaCl for 90 days when it was observed that there
was no degradation in this type of coated steel after
25 days of immersion.
AC impedance measurements were conducted
initially on specimens of the as-received samples.
The immersed specimens then were removed at the
appropriate time, carefully wiped dry, and
immediately subjected to the AC impedance
measurements. Each specimen was sandwiched
carefully between the clamp-on Cu foil electrodes
such that one of the electrodes made contact with
the coating surface and the other electrode was in
FIGURE 2. Electrical connections used to measure impedance contact with the uncoated steel surface. The
for a coated sample. electrical contact was confirmed each time by
ensuring that the measured impedance values were
independent of the clamping pressure.
The electrodes were connected to the DSA as
behavior of the coated steel exposed to a corrosive shown in Figure 2. The DSA was controlled
environment. It was shown that the simplified automatically by the computer using software written
measurement setup easily could track the degrada- in HP-BASIC language. The DSA was interfaced to
tion of coated steel subjected to corrosive exposures the workstation using a HP interface bus cable. The
for varying periods of time. As expected, the study DSA input AC signal (Vin) was 100 mV. The voltage
showed the epoxy-based coating had better across the test specimen (Vout) was measured
resistance to corrosion than the alkyd coatings. automatically by the DSA over the frequency range
0.1 Hz to 100 kHz. The current (i) through the test
EXPERIMENTAL PROCEDURE specimen was calculated from this voltage and the
accurately known value of the standard resistor (Rst)
The instrumentation consists of a Hewlett using Equation (1). The impedance (Z) of the test
Packard (HP) Model 3562A† DSA,9 a HP9000/Series specimen then was calculated using Equation (2),
300† computer controller, a pair of clamp-on Cu foil and results were plotted by the software in the form
electrodes, and a standard resistor whose value was of Nyquist and Bode plots.

Vin – Vout
i= (1)
†
Trade name. Rst

98 CORROSION–FEBRUARY 1995
 CORROSION SCIENCE

Vout
Z= (2)
i

To test the capability of the instrumentation to

measure accurately impedances of the order of the
input impedance of the DSA, a dummy test circuit
was set up. It consisted of a 1-MΩ resistance in
parallel with a 100-pF capacitance, this parallel FIGURE 3. Equivalent circuit for a coated metal system.
combination being in series with a 2-MΩ resistor. It
was verified that the impedance of this test circuit
measured by the DSA was in very good agreement
with that obtained from circuit theory. To verify that
there was no alternative current path through the
rubber pads and metallic clamp, a measurement of
impedance was made with paper layers of thickness
approximately equal to that of the coated steel
sample. No impedance was measured, since the
rubber pads, electrical tape holding the rubber pads,
and the paper presented an impedance beyond the
range of the instrument. Thus, no alternate current
path was present through the metallic clamps.

RESULTS AND DISCUSSION

The coated steel system may be described in
terms of the equivalent circuit shown in Figure 3,10 FIGURE 4. Nyquist plots for the specimen coated with epoxy
where Cc and Rc represent the capacitance and primer immersed in 3% NaCl (short-term immersion).
resistance of the coating, respectively; Zw represents
the Warburg impedance; and Rg is a combination of
the resistance of the solution that penetrated the
coating to the steel surface and the contact
resistance of the foil electrodes. The value of Rg
generally depends upon the interface characteristics
between a specimen and electrodes and the
properties of the electrodes. It is believed that
changes in Rc are the best measures of the
degradation of coated metal systems.
The measured impedance data were plotted in
Nyquist (imaginary impedance vs real impedance)
and Bode (log impedance magnitude vs log
frequency) formats. The value of Rc was determined
from the Nyquist plot, where Rc was equal to the
diameter of the semicircle. The real axis intercept of
the Nyquist plot at the high-frequency end yielded Rg,
the equivalent solution resistance. Cc was determined FIGURE 5. Bode magnitude plots for the specimen coated with
from the slope of the straight line obtained when the epoxy primer immersed in 3% NaCl (short-term immersion).
high-frequency imaginary admittance was plotted vs
angular frequency (imaginary admittance =
capacitance x angular frequency). Zw = K/␻1/2, where
K was the Warburg coefficient, which could be immersion time of 25 days in tap water or 3% NaCl.
obtained using a Randles plot in which the real part The Nyquist and Bode magnitude plots for the
of impedance was plotted against ␻–1/2 (where ␻ is specimen coated with epoxy primer when immersed
angular frequency). These results were computed in 3% NaCl are shown in Figures 4 and 5. The
automatically by the computer workstation. corresponding plots for an identical specimen
The coating impedance for the specimen coated immersed in tap water were almost indistinguishable
with epoxy primer did not change even after an from Figures 4 and 5. Only results for the as-received

CORROSION–Vol. 51, No. 2 99

CORROSION SCIENCE

FIGURE 6. Nyquist plots for the specimen coated with a primer FIGURE 7. Bode magnitude plots for the specimen coated with
of high-solids alkyd immersed in tap water. a primer of high-solids alkyd immersed in tap water.

specimen and those obtained after 25 days of

immersion are shown, since results for 3-, 5-, 10-,
and 15-day immersions were almost identical to
those for the as-received specimen.
It was evident from visual observation that the
coating surface underwent no deterioration. There
was obvious degradation in the specimen coated with
the primer of low-solids alkyd when it was immersed
in 3% NaCl but no degradation in an identical
specimen immersed in tap water for 25 days. The
Nyquist and Bode magnitude plots for the specimens
coated with a primer of high-solids alkyd are shown in
Figures 6 through 9. There was significant degrada-
tion in the specimen coated with a primer of
FIGURE 8. Nyquist plots for the specimen coated with a primer high-solids alkyd for immersion in tap water as well
of high-solids alkyd immersed in 3% NaCl (the impedance for as in 3% NaCl. Degradation commenced when the
the 15-day immersion is too small to be seen). specimen was immersed in tap water for 5 days and
in 3% NaCl for 3 days. The Nyquist and Bode
magnitude plots for the specimens coated with the
primer of low-solids alkyd are shown in Figures 10
through 13. Degradation occurred only after immer-
sion in 3% NaCl for > 10 days. A summary of the
experimental results is presented in Table 1. A large
decrease in impedance was measured whenever a
number of large blisters were present on the coating
surface. Only in the case of the low-solids alkyd
coating immersed in tap water for 25 days and in 3%
NaCl for 5 days were a few small blisters formed. In
both these cases, there was no substantial change in
impedance.
Circuit simulation calculations were performed on
the specimen coated with the primer of high-solids
alkyd. Figure 14 shows the measured Nyquist plot
and that calculated from the simulated circuit for the
as-received specimen. Figures 15 and 16 show the
FIGURE 9. Bode magnitude plots for the specimen coated with Nyquist plots for this same specimen after 20 days of
a primer of high-solids alkyd immersed in 3% NaCl.
immersion in tap water and 10 days of immersion in

100 CORROSION–FEBRUARY 1995

 CORROSION SCIENCE

FIGURE 10. Nyquist plots for the specimen coated with a FIGURE 11. Bode magnitude plots for the specimen coated
primer of low-solids alkyd immersed in tap water. with a primer of low-solids alkyd immersed in tap water.

FIGURE 12. Nyquist plots for the specimen coated with a FIGURE 13. Bode magnitude plots for the specimen coated
primer of low-solids alkyd immersed in 3% NaCl. with a primer of low-solids alkyd immersed in 3% NaCl.

TABLE 1
Rc and Visual Observations of the Coating Surface(A)
Immersion Time Rc Coating
Sample Solution (days) (kΩ) Surface

High-solids alkyd 3% NaCl 3 610 Many large blisters

High-solids alkyd Tap water 5 478 Many large blisters
(B)
High-solids alkyd 3% NaCl 25 Peeling
Epoxy 3% NaCl 25 1,000 No change
Epoxy Tap water 25 1,000 No change
Epoxy 3% NaCl 90 72.5 Many large blisters
Epoxy Tap water 90 1,000 Many large blisters
Low-solids alkyd Tap water 25 1,000 Few small blisters
Low-solids alkyd 3% NaCl 5 905 Few small blisters
Low-solids alkyd 3% NaCl 15 345 Many large blisters

(A)
Rc = 1 MΩ for all the as-received samples.
(B)
Coating film peeled off the steel surface. No measurements could be made.

CORROSION–Vol. 51, No. 2 101

CORROSION SCIENCE

FIGURE 15. Simulation calculation for a specimen coated with

FIGURE 14. Simulation calculation for the specimen coated a primer of high-solids alkyd after a 20-day immersion in tap
with a primer of high-solids alkyd, as-received. water.

TABLE 2
Rg Values for the Specimen
Coated with the Primer of High-Solids Alkyd
Immersion Time Rg (kΩ) Rg (kΩ)
(days) in 3% NaCl in Tap Water

5 9.3 6.1
10 7.8 5.4
(A)
20 2.3

(A)
Coating film peeled off the steel surface. No measurements could
be made.

FIGURE 16. Simulation calculation for the specimen coated

with a primer of high-solids alkyd after a 10-day immersion in
3% NaCl. dilute solutions and, hence, Rg will become smaller,
as seen in Table 2.
A specimen coated with epoxy primer was
immersed in tap water and an identical specimen
immersed in 3% NaCl for 3 months. Figures 17
3% NaCl solution, respectively. The linear section through 20 are the measured Nyquist and Bode
at low frequencies in Figures 15 and 16 represented magnitude plots for the specimens immersed in tap
the diffusion impedance (i.e., Zw), which indicated water and 3% NaCl, respectively. There was obvious
that the rate of the electrochemical reaction at the degradation in the specimen immersed in 3% NaCl,
metal-coating interface was limited by diffusion as evidenced by the change in impedance. However,
through the coating. K was obtained using the no change was observed in the case of immersion
Randles plot in which the real part of the impedance in tap water. In the latter case, blistering was present
was plotted vs ␻–1/2. The equivalent electrical circuit on the coating surface, although the coating film did
models and K values are shown in Figures 14 not peel off the steel surface. Circuit simulation
through 16. calculations were performed for the specimen
Table 2 shows Rg values for the specimen exposed to 3% NaCl. A comparison of the measured
coated with the primer of high-solids alkyd when it impedance with that calculated from the simulated
was immersed in tap water and in 3% NaCl for electrical circuit representation is shown in Figure
different periods of time. It has been shown that the 21.The linear section at low frequencies was
more dilute the solution, the greater is the water indicative of a diffusion-controlled process (Zw). The
uptake by the coating.11 This means that there are straight section made an angle of 45° with respect to
more charged ions on the coating surface for more the axes. The equivalent simulated electrical circuit

102 CORROSION–FEBRUARY 1995

 CORROSION SCIENCE

FIGURE 17. Nyquist plots for the specimen coated with epoxy FIGURE 18. Bode magnitude plots for the specimen coated
primer immersed in tap water (long-term immersion). with epoxy primer immersed in tap water (long-term immersion).

FIGURE 19. Nyquist plots for the specimen coated with epoxy FIGURE 20. Bode magnitude plots for the specimen coated
primer immersed in 3% NaCl (long-term immersion). with epoxy primer immersed in 3% NaCl (long-term immersion).

also is shown in Figure 21. Good agreement was

obtained between the measured and calculated
results.

CONCLUSIONS

❖ The AC impedance spectroscopy method

described in this paper offers a simple, convenient,
and fast measurement approach for evaluation of
degradation in coated metals. The method uses a
DSA interfaced with a computer workstation and
does not require electrochemical controls.
❖ The method measured successfully the
degradation in three samples of cold-rolled steel
coated with epoxy and alkyd single-phase primers
and exposed to full immersion in tap water and NaCl FIGURE 21. Simulation calculation for the specimen coated
with epoxy primer after a 3-month immersion in 3% NaCl.
solutions. Simulated equivalent electrical circuit

CORROSION–Vol. 51, No. 2 103

CORROSION SCIENCE

models gave impedance results that agreed closely REFERENCES

with the measured data.
1. S.J. Downey, O.F. Devereux, Corrosion 45 (1989): p. 675.
❖ Further work using this method would involve 2. J.R. Scully, J. Electrochem. Soc. 136 (1989): p. 979.
measurement of degradation of more complex 3. H. Leidheiser Jr., J. Coatings Technol. 63 (1991): p. 21.
4. E. Frechette, C. Compere, E. Ghali, Corros. Sci. 33 (1992): p. 1,067.
coatings in simulated and actual atmospheric service 5. S.A. McCluney, S.N. Popova, B.N. Popov, R.E. White, J. Electrochem.
exposure conditions. Soc. 139 (1992): p. 1,556.
6. P.R. Roberge, E. Halliop, V.S. Sastri, Corrosion 48 (1992): p. 447.
7. T.C. Simpson, P.J. Moran, W.C. Moshier, G.D. Davis, B.A. Shaw, C.O.
ACKNOWLEDGMENTS Arah, K.L. Zankel, J. Electrochem. Soc. 36 (1989): p. 2,761.
8. J.R. Macdonald, ed., Impedance Spectroscopy Emphasizing Solid
Materials and Systems (New York, NY: John Wiley & Sons, 1987).
The authors acknowledge grants from the U.S. 9. Hewlett Packard 3562A Dynamic Signal Analyzer Operating Manual,
Department of Energy (through the Center for Hewlett-Packard, Everett, Washington, 1985.
10. “Evaluation of Organic Coatings by Electrochemical Impedance
Infrastructure Studies, University of Nevada, Reno, Measurements”, Application Note: AC-2, Electrochemical Instruments
Nevada) and from the Nuclear Waste Projects Office, Group, EG&G Princeton Applied Research, Princeton, New Jersey,
1984.
Yucca Mountain, Nevada. 11. A. Amirudin, D. Thierry, Brit. Corros. J. 26 (1991): p. 195.

104 CORROSION–FEBRUARY 1995