Electrochemical impedance spectroscopy

for the detection of stress corrosion cracks

in aqueous corrosion systems at ambient
and high temperature
Rik-Wouter Bosch
SCK-CEN, Belgian Nuclear Research Centre, Boeretang 200, B-2400 Mol, Belgium
Received 24 September 2003; accepted 13 May 2004
Available online 20 July 2004
Abstract
Electrochemical impedance spectroscopy (EIS) has been used as a tool to detect stress
corrosion cracking in a stainless steel sample exposed to an aqueous environment at ambient
and high temperature. A model has been developed to describe the impedance of a cracked
surface. This model could distinguish between a at electrode surface and an electrode surface
with cracks. To test this technique, three case studies were considered: (1) Slow Strain Rate
Tests (SSRT) on sensitised Type 304 SS specimen in a 5 N H
2
SO
4
+0.1 M NaCl solution at
room temperature, (2) constant load tests on Type 304 SS in a boiling (110 C) acidied
sodium chloride solution, and (3) Slow Strain Rate Tests (SSRT) on Type 304 SS in an oxygen
containing solution of 0.01 M Na
2
SO
4
at 300 C.
EIS measurements were always performed simultaneously on two identical samples: one
stressed with a SSRT or a constant load test and one free of stress. Kramers Kronig Trans-
formations have been used to validate the experimental data obtained with the EIS mea-
surements. It was shown that the phase shift between the two samples could be related to the
stress corrosion cracking process. Analysis of the fracture surfaces conrmed that stress
corrosion cracks were formed in all three cases.
2004 Elsevier Ltd. All rights reserved.
Keywords: Stress corrosion cracking; EIS; Stainless steel; Corrosion monitoring
E-mail address: rbosch@sckcen.be (R.-W. Bosch).
0010-938X/$ - see front matter 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.corsci.2004.05.018
www.elsevier.com/locate/corsci
Corrosion Science 47 (2005) 125143
1. Introduction
Impedance measurements have only been used occasionally for the investigation
of stress corrosion cracking phenomena. Oltra et al. [1] have used a transmission line,
similar to the cylindrical pore model developed for the potential distribution in
porous electrodes of de Levie [2], to model the impedance of a stress corrosion crack.
Petit et al. [3] have used impedance measurements during Slow Strain Rate Tests
(SSRT) of Type 304 SS in boiling 44% MgCl
2
solution. They measured the phase at a
particular frequency during the tests and try to correlate crack initiation to changes
in this phase. We have also used this approach in previous work where it was shown
that an evolution in the phase of the impedance at certain frequencies could be linked
to the occurrence of stress corrosion cracking [4].
In this paper we will summarise the results of stress corrosion cracking detection
with impedance measurements of three dierent stress corrosion cracking systems.
First however we will use a mathematical model of the crack impedance to inves-
tigate the relationship between the crack length, the perturbation frequency and the
outer surface area of the metal specimen that holds the (possible) stress corrosion
crack. For this model the transmission line model is used and extended to model the
Nomenclature
C
crack
double layer capacity at the crack tip, C
crack
C
dl
pr
2
(F)
C
dl
double layer capacity (F/mm
2
)
C
surface
double layer capacity of the outer surface of the working electrode (F)
C
wall
double layer capacity of the crack wall per unit length, C
wall
C
dl
2pr
(F/mm)
e potential along the crack (V)
E
L
potential at the crack tip (V)
i current along the crack (A)
L crack length (mm)
r crack radius (mm)
R
crack
polarisation resistance at the crack tip R
crack

R
passive
pr
2
(X)
R
ohm
ohmic resistance (X)
R
passive
polarisation resistance of passive stainless steel (Xmm
2
)
R
sol
solution resistance of the solution in the crack per unit length (X/mm)
R
surface
polarisation resistance of the outer surface of the working electrode (X)
R
wall
polarisation resistance of the crack wall per unit length, R
wall

R
passive
2pr
(Xmm)
z distance along the crack (mm)
Z
crack
impedance of the total crack (X)
Z
cracktip
impedance of the crack tip (X)
Z
wall
impedance of the crack wall (Xmm)
j conductivity of the electrolyte solution (X
1
mm
1
)
126 R.-W. Bosch / Corrosion Science 47 (2005) 125143
crack impedance. This resulted in a crack model that incorporates: the geometry of
the crack, an active or passive crack tip and the inuence of the ratio between the
electrode surface area and the crack surface area [4].
The three experimental cases yielding stress corrosion cracks that have been
studied are:
1. SSRT on sensitised Type 304 SS specimens in a 5 N H
2
SO
4
+0.1 M NaCl solution
at room temperature [5].
2. Constant load testing on Type 304 SS in a boiling acidied sodium chloride solu-
tion [6].
3. SSRT testing in an autoclave on Type 304 SS in an oxygen containing solution of
0.01 M Na
2
SO
4
at 300 C [7].
The rst case study results in intergranular stress corrosion cracking at room
temperature [5]. This material/solution combination was chosen, for we wanted to
avoid the possible inuence of an electric heater (electric noise) on the impedance
measurements in a rst experiment. The second case study is based on the ASTM
standard G 123 Standard Test Method for Evaluating Stress-Corrosion Cracking of
Stainless Alloys with Dierent Nickel Content in Boiling Acidied Sodium Chloride
Solution [6]. Due to the accelerated nature of a SSRT experiment, the measurement
time of the impedance measurements was limited to frequencies ranging from 1000
to 1 Hz. However, it was expected that more information regarding the occurrence of
SCC could be retrieved from impedance measurements at lower frequencies. The
duration of an experiment using a constant load test is longer than with a SSRT test.
Hence, the impedance measurements can go to lower frequencies and the results
could be dierent. The last case study is the stress corrosion cracking of (sensitised)
Type 304 SS in an oxygen containing solution of 0.01 M Na
2
SO
4
at 300 C [7]. This
test should show that electrochemical impedance measurements also work in auto-
clave systems.
An experimental set-up was developed which allowed testing of two identical
samples: one under stress (by means of a SSRT or constant load test) and one free of
stress. During all the experiments the impedances of both samples were determined
quasi-simultaneously. To make sure that the EIS data were of good quality, Kramers
Kronig Transformations (KKTs) were used to validate the experimental data.
2. Model of the crack impedance
The mathematical derivation and description of the crack model has been given
elsewhere [4], but will be briey summarised here. The model of the crack impedance
can be described by an equivalent circuit as shown in Fig. 1.
In this equivalent circuit R
surface
is the polarisation resistance of the outer surface
of the working electrode, C
surface
the capacity of the electrochemical double layer of
the outer surface of the working electrode and R
ohm
the ohmic resistance between the
R.-W. Bosch / Corrosion Science 47 (2005) 125143 127
reference electrode and the outer surface of the working electrode. The crack
impedance Z
crack
is given by Eq. (1).
Z
crack
R
sol

Z
wall
p

R
sol
Z
wall
p
Z
cracktip
coth L

R
sol
Z
wall
q
R
sol

Z
wall
p
Z
cracktip
coth L

R
sol
Z
wall
q

R
sol
p
1
In this modelling attempt, the cracks are represented by cylindrical holes. The
solution resistance, R
sol
, of this cylinder is then,
R
sol

1
jpr
2
2
where r the crack radius and j the conductivity of the electrolyte solution.
The crack tip impedance consists of a charge transfer resistance at the crack tip
R
crack
and the crack tip capacity C
crack
, due to the electrochemical double layer. Then
the crack tip impedance is,
Z
cracktip

R
crack
1 jxC
crack
R
crack
3
with,
R
crack

R
passive
pr
2
and C
crack
C
dl
pr
2
4
The impedance of the crack wall Z
wall
follows from an analogue treatment,
Fig. 1. Equivalent circuit for the crack impedance based on a transmission line.
128 R.-W. Bosch / Corrosion Science 47 (2005) 125143
Z
wall

R
wall
1 jxC
wall
R
wall
5
with,
R
wall

R
passive
2pr
and C
wall
C
dl
2pr 6
All units of the parameters mentioned in the above equations are summarised in the
nomenclature. The crack impedance Z
crack
is part of the total equivalent circuit as
shown in Fig. 2. The total impedance Z
tot
is then given by,
Z
tot

R
surface
Z
crack
R
surface
Z
crack
jxC
dl
R
surface
Z
crack
7
If there is more than one crack, for instance the number of cracks is equal to m, then
the total crack impedance consists of m crack impedances in parallel as demon-
strated in Fig. 2.
The total crack impedance Z
crack;total
is then,
Z
crack;total

1
P
m
i1
1
Z
crack;i
8
assuming that all the crack impedances are equal, the total crack impedance
Z
crack;total
becomes,
Z
crack;total

Z
crack
m
9
Fig. 2. Equivalent circuit for m cracks.
R.-W. Bosch / Corrosion Science 47 (2005) 125143 129
3. Simulations with the crack model
The just developed crack model has been used to investigate two phenomena that
are of interest when applying EIS to stress corrosion cracking. First, the inuence of
the total surface area of an electrode on the impedance response for a xed crack
surface area was investigated. Secondly, the optimum test or perturbation frequency
for the impedance measurement for a given crack length was determined.
The parameters used are summarised in Table 1.
The contribution of the crack impedance to the total impedance of an electrode
could be so small, that it might not be noticed. This constitutes a problem especially
for small stress corrosion cracks on a large electrode surface area. Therefore some
simulations were performed with the crack model shown in Fig. 1. The impedance
was calculated for a small crack of xed dimensions on an electrode surface of
varying diameter.
The phase shift ( phase
stressed
phase
nonstressed
, see Section 4.4) of the impedance
versus the frequency for various electrode diameters is shown in Fig. 3.
Fig. 3 shows that the inuence of the crack impedance on the phase shift is
decreasing with increasing surface area (or sample diameter D) of the electrode.
However the inuence of a small crack on the phase shift is still noticed even when
the surface area of the electrode is 225 times larger than the surface area of the cross
section of the crack i.e.
3
2
0:2
2

9
0:04
225.
We would also like to nd the optimum test or perturbation frequency for the
impedance measurements i.e. the frequency where the phase shift due to cracks has
the largest value. In order to nd this optimum the relation between the frequency
and the phase shift for various crack length has been calculated. The results are
shown in Fig. 4.
The theoretically calculated curves show that there is an optimum frequency for
the detection of a crack for each crack length. For small cracks the optimum fre-
quency is bigger than for large cracks. That means that for a growing crack, the
optimum perturbation or test frequency is continuously changing. It is therefore
necessary to use a range of perturbation frequencies. For instance if we only test at
0.1 Hz we will not nd cracks smaller than 1 mm in depth. It is also clear that a large
crack generates a much larger response than a small crack.
Table 1
Parameters for the crack model
Parameter Value
Crack radius r 0.1 mm
Double layer capacity C
dl
30 10
8
F/mm
2
Polarisation resistance passive stainless steel R
passive
10
8
Xmm
2
Relative conductivity j 0.1799 10
3
S mm
1
Number of cracks m 1
Ohmic resistance R
ohm
100 X
Outer surface area diameter D 3 mm
Crack length L 1 mm
130 R.-W. Bosch / Corrosion Science 47 (2005) 125143
4. Experimental
4.1. Experimental set-up for constant load testing
An experimental set-up was developed, which used two identical samples. One of
the samples was subjected to stress corrosion cracking by means of a constant load
device, while the other was free of stress. Fig. 5 shows a schematic of the constant
load test set-up.
Fig. 3. Phase shift versus frequency for various sample diameters, data from Table 1.
Fig. 4. Dependence of the phase shifts on the test frequency for various crack lengths.
R.-W. Bosch / Corrosion Science 47 (2005) 125143 131
The load was applied by means of a spring that was compressed with a nut. The
frame of the constant load unit was used as counter electrode. All parts were isolated
from each other using Teon tape. Also the tensile specimens were covered with
Teon tape, accept for the gage length. The stainless steel 304 tensile specimens were
small size specimens according to ASTM A370-95 with gage length of 25 mm and
diameter of 6.25 mm. The samples were used in the as received condition. The
sample under stress was subjected to an initial stress of 300 N/mm
2
.
The experiments were performed in boiling acidied sodium chloride solution
according to ASTM standard G 123-94 [6]. This solution contains 25% sodium
Fig. 5. Experimental set-up to perform constant load testing.
132 R.-W. Bosch / Corrosion Science 47 (2005) 125143
chloride (by mass) with a pH of 1.5, which was obtained by adding phosphoric acid.
We used a test temperature of 105 C, which was very close to the boiling point of
the solution (106110 C).
4.2. Experimental set-up for impedance measurements in an autoclave
An experimental set-up was developed which allows us to test two identical
samples: one under stress (by means of a Slow Strain Rate Test (SSRT)) and one free
of stress. A schematic of the experimental set-up is shown in Fig. 6.
The experiments were performed on Type 304 SS samples in the solution annealed
condition and sensitised at 621 C during 24 h. The samples were tested in an oxygen
containing solution of 0.01 M Na
2
SO
4
at 300 C, which should result in intergran-
ular stress corrosion cracking (IGSCC) [7]. Before heating up the autoclave, the
autoclave was pressurised with 10 bar of air to ensure that the test solution contains
a large amount of oxygen. The experiments were performed under static conditions,
i.e. no recirculation of the test solution was performed. One sample was subjected to
a constant elongation of 1 10
6
s
1
. The other sample was free of stress.
4.3. Multiplexed EIS measurements
Every hour, Electrochemical impedance spectroscopy (EIS) measurements were
performed quasi-simultaneously at both samples. The EIS measurements were per-
formed with one potentiostat and one frequency response analyser connected to
a multiplexer. The impedance measurements were performed from 1000 to 1 Hz
for the SSRT experiments and from 1000 Hz to 0.1 Hz for the constant load
Fig. 6. Schematic of the experimental set-up: autoclave, tensile specimens, reference electrodes (Pt-wire)
and counter electrodes (CE).
R.-W. Bosch / Corrosion Science 47 (2005) 125143 133
experiments. For both cases amplitude of 10 mV RMS was used. One measurement
(containing all frequencies) took about 3 min for the rst case and approximately 20
min for the second case. Fig. 7 shows the set-up with the frequency response analyser
(FRA), the potentiostat and the electrochemical multiplexer.
4.4. Data interpretation
The phase shift is measured between a sample without a crack and a sample with a
crack. In this way we hope to lter out the phase shifts due to an evolution in time of
the corroding surface. This is illustrated in Fig. 8.
In order to compare the phase shifts at dierent frequencies in one graph we
sometimes need to do a translation. This is explained in Figs. 9 and 10.
Fig. 9 shows that the phase of the stressed sample starts at 23 and that the phase
of the non-stressed sample starts at approximately 30. It can be clearly seen that
there is a dierent evolution of the phase for the stressed and stress free samples. Fig.
10 shows the phase shift.
Fig. 7. Experimental set-up for the multiplexed impedance measurements showing the potentiostat, the
frequency response analyser and the electrochemical multiplexer.
Fig. 8. Phase shift is determined between sample with and without crack.
134 R.-W. Bosch / Corrosion Science 47 (2005) 125143
Fig. 10 shows that the evolution of the phase shift starts at 6. To be able to plot
the phase shift at dierent frequencies in one plot, the graph is translated to zero.
5. Results and discussion
5.1. Slow Strain Rate Tests on Type 304 SS in a 5 N H
2
SO
4
+0.1 M NaCl solution
The results in this solution have been published previously [4]. We have inter-
preted these experimental data according to the previous paragraph resulting in
Fig. 11.
Fig. 9. Evolution of the phase during a stress corrosion cracking test, data from Fig. 12.
Fig. 10. Evolution of the phase shift during a stress corrosion cracking test, data from Fig. 12.
R.-W. Bosch / Corrosion Science 47 (2005) 125143 135
Impedance measurements and corrosion potential measurements were performed
at time intervals of half an hour during the Slow Strain Rate Tests.
The phases at 1000 Hz were almost equal for the tests with and without stress and
are not plotted. At 100 Hz a clear dierence between the phases of the tests with and
without stress can be notice from approximately seven hours onwards. At 1 and 10
Hz this dierence is even clearer, also starting from approximately seven hours
onwards. For application of EIS to the monitoring of stress corrosion cracking
under these circumstances a frequency between 100 and 1 Hz seems the most
appropriate choice.
5.2. Constant load testing on Type 304 SS in a boiling acidied sodium chloride
solution
Impedance measurements were performed during approximately 80 h. At the
beginning of the test both samples already showed a dierence in impedance
behaviour. Therefore phase dierences were calculated for dierent frequencies.
These dierences in phase are related to the phase shift at the beginning of the test,
when it is sure that there are no cracks in the sample surfaces. Fig. 12 shows the
evolution of the phase shift during the constant load experiments for dierent fre-
quencies.
The best result i.e. largest measured phase shift, is obtained for a frequency of 1
Hz. For higher and lower frequencies the dierences are less pronounce, which is in
agreement with the simulations in Figs. 3 and 4. With an increase in phase shift we
assume that there is an increasing number and length of cracks in the stressed
sample.
Fig. 11. Phase shift evolution during a SSRT test with sensitised Type 304 SS in a 5 N H
2
SO
4
+0.1 M
NaCl solution.
136 R.-W. Bosch / Corrosion Science 47 (2005) 125143
Fig. 13 shows a photograph of both samples after the experiments. A clear dif-
ference between the stressed and non-stressed sample can be observed.
5.3. Slow Strain Rate Tests on Type 304 SS in an oxygen containing solution of 0.01 M
Na
2
SO
4
at 300 C
Tests were carried out with stainless steel 304 in the sensitised and solution an-
nealed condition. Proper treatment of the raw data improves the possibility to make
Fig. 12. Evolution of phase during the constant load test for frequencies of 100, 10, 1 and 0.1 Hz on Type
304 SS in a boiling acidied sodium chloride solution.
Fig. 13. Photograph of the two samples after testing under constant load on Type 304 SS in a boiling
acidied sodium chloride solution, the upper sample was free of stress, the lower sample was subjected to a
constant load.
R.-W. Bosch / Corrosion Science 47 (2005) 125143 137
a link between the phase and the stress corrosion cracking process. Fig. 14 shows the
evolution of the phase shifts for frequencies of 1 and 10 Hz during a SSRT test.
When the SSRT test starts, the phase shifts start to increase suggesting the formation
of stress corrosion cracks.
Fig. 15 shows SEM pictures of the fracture surface at two magnications. It can
be clearly seen that the specimen was broken due to intergranular stress corrosion
cracking.
Another test has been performed with Type 304 SS in the annealed condition. Fig.
16 shows the evolution of the phase shifts for frequencies of 1 and 10 Hz during a
SSRT test. When the SSRT test starts, the phase shifts start to increase suggesting
the formation of stress corrosion cracks. Again the 1 Hz response is more pro-
nounced than the 10 Hz response.
Fig. 14. Crack initiation and propagation is related to the evolution in phase shifts during a SSRT test of
sensitised Type 304 SS in 0.01 M Na
2
SO
4
at 300 C.
Fig. 15. Intergranular stress corrosion cracking (IGSCC) of sensitised Type 304 SS in 0.01 M Na
2
SO
4
at
300 C.
138 R.-W. Bosch / Corrosion Science 47 (2005) 125143
Fig. 17 shows SEM pictures of the fracture surface at two magnications. It can
be clearly seen that the specimen was broken due to transgranular stress corrosion
cracking.
6. Data validation with the Kramers Kronig Transformation
The Kramers Kronig Transformations (KKT) are mathematical relationships
between the real and the imaginary parts of the impedance that must be obeyed by
valid impedance or admittance data. That means that when Z
imag
is known for all
frequencies, it is possible to calculate Z
real
at all frequencies and vice versa [813]. The
Fig. 16. Crack initiation and propagation is related to the evolution in phase shift during a SSRT test of
solution annealed Type 304 SS in 0.01 M Na
2
SO
4
at 300 C.
Fig. 17. Transgranular stress corrosion cracking (TGSCC) of the solution annealed specimen of Type 304
SS in 0.01 M Na
2
SO
4
at 300 C.
R.-W. Bosch / Corrosion Science 47 (2005) 125143 139
derivation of the Kramers Kronig Transformation (KKT) is based on four general
conditions of the system [8,9]:
1. Causality. The response of the system is due only to the perturbation applied and
does not contain signicant components from spurious sources.
2. Linearity. The perturbation and response of the system are linearly related i.e. the
impedance is independent of the amplitude of the perturbation signal.
3. Stability. The system must be stable in the sense that it returns to its original state
after the perturbation is removed.
4. The impedance must be nite-valued at x !0 and at x !1 and must be con-
tinuous and nite-valued function at all intermediate frequencies.
It has been shown that when a corroding system obeys the just mentioned four
criteria the impedance data will converse correctly [10,11]. However the inverse is not
always true. It is still possible to have a correct KKT when impedance data are non-
linear [12]. In the case of our impedance measurements we are mainly concerned
about the stability of the system and for this case the KKT is an excellent tool for
data validation. For instance, the stability of the stress corrosion cracking system can
be distorted by a change from a passive to an active crack tip during an EIS mea-
surement. Secondly a growing crack will also change the system. If such changes are
small and occur slowly, the stress corrosion system will at least be stable during an
EIS measurement. To verify these KKT transforms were carried out with experi-
mental data for all three case studies at a time that it was sure that stress corrosion
cracks were present. With the equivalent circuit [14] software the imaginary part of
the impedance of a set-up of EIS data is calculated with the experimental values of
Fig. 18. Kramers Kronig Transformation of impedance data obtained during SSRT testing with sensitised
Type 304 SS specimens in a 5 N H
2
SO
4
+0.1 M NaCl solution at room temperature after 60 h of testing.
140 R.-W. Bosch / Corrosion Science 47 (2005) 125143
the real part of the impedance. Then the KKT data are plotted together with the
experimental data. If both graphs are similar, it is assumed that the experimental
data are of good quality i.e. the stress corrosion cracking system has not remarkably
changed during the time of measurement.
Fig. 18 shows a Bode diagram with both the KKT and experimental data ob-
tained during SSRT testing with sensitised stainless steel 304 specimens in a 5 N
H
2
SO
4
+0.1 M NaCl solution at room temperature after 60 h of testing.
Fig. 18 shows that there is excellent agreement between the experimental and
simulated data. There is a small deviation at high frequencies, which is behind the
Fig. 19. Kramers Kronig Transformation of impedance data obtained during constant load testing with
stainless steel 304 in a boiling acidied sodium chloride solution after 70 h of testing.
Fig. 20. Kramers Kronig Transformation of impedance data obtained during SSRT testing in an auto-
clave with Type 304 SS in an oxygen containing solution of 0.01 M Na
2
SO
4
at 300 C, after 25 h of testing.
R.-W. Bosch / Corrosion Science 47 (2005) 125143 141
range of interest. Fig. 19 shows a Bode diagram with both the KKT and experi-
mental data obtained during constant load testing with Type 304 SS in a boiling
acidied sodium chloride solution.
Here again, there is general agreement between both the experimental data and
simulated data, although the phase shows signicant uctuations. This can also
be seen in Fig. 19. These uctuations are related to back ground noise as they
occur at all frequencies. Concluding, the general stability of the system has not
changed during the measurement. Fig. 20 shows a Bode diagram with both
the KKT and experimental data obtained during SSRT testing in an autoclave
with Type 304 SS in an oxygen containing solution of 0.01 M Na
2
SO
4
at
300 C.
Fig. 20 shows that there is excellent agreement between the experimental and
simulated data.
7. Conclusions
A crack model was developed that described the impedance of a stress corrosion
crack. This crack model showed that a phase shift of the impedance at particular
frequencies is related to the length of a stress corrosion crack. Theoretical calcula-
tions showed that there is an optimum test frequency for each crack length. The
longer the crack, the lower was the optimum test frequency.
Three experimental case studies were considered to generate stress corrosion
cracks:
SSRT testing on sensitised Type 304 SS specimens in a 5 N H
2
SO
4
+0.1 M NaCl
solution at room temperature.
Constant load testing on Type 304 SS in a boiling acidied sodium chloride solu-
tion.
SSRT testing in an autoclave on Type 304 SS in an oxygen containing solution of
0.01 M Na
2
SO
4
at 300 C.
The EIS measurements were always performed (quasi-) simultaneously on two
similar tensile specimens; one subjected to stress corrosion cracking by means of a
SSRT test or constant load test and one free of stress. Kramers Kronig Transfor-
mations showed that the impedance data were of good quality. It was shown that the
evolution of the phase shift of the impedance at various frequencies could be related
to the stress corrosion cracking of one of the samples. Hence, it was shown that
Electrochemical impedance spectroscopy could be used to detect stress corrosion
cracks in various corrosive environments and under various loading conditions.
Depending on the system investigated, the phase shifts could be small. In these cases,
the sensitivity of the measurement technique should be improved to have more
discriminating results.
142 R.-W. Bosch / Corrosion Science 47 (2005) 125143
References
[1] R. Oltra, M. Keddam, Corrosion Science 28 (1) (1988) 1.
[2] R. de Levie, Advances in Electrochemistry and Electrochemical Engineering 6 (1967) 329.
[3] M.C. Petit, M. Cid, M. Puiggali, Z. Amor, Corrosion Science 31 (1990) 491.
[4] R.W. Bosch, F. Moons, J.H. Zheng, W.F. Bogaerts, Corrosion 57 (2001) 532.
[5] G. Cragnolino, D.D. Macdonald, Corrosion 38 (8) (1982) 406.
[6] Standard Test Method for Evaluating Stress-Corrosion Cracking of Stainless Alloys with Dierent
Nickel Content in Boiling Acidied Sodium Chloride Solution, ASTM Standard G 123-94, Annual
Book of ASTM Standards, vol. 03.02, 1995.
[7] M.E. Indig, A.R. McIlree, Corrosion 35 (1979) 288.
[8] J.R. Macdonald, Impedance Spectroscopy, Emphasizing Solid Materials and Systems, John Wiley &
Sons, 1987, p. 274.
[9] R. Cottis, S. Turgoose, Electrochemical impedance and noise, in: B.C. Syrett (Ed.), Corrosion Testing
Made Easy, NACE International, 1999, p. 135.
[10] D.D. Macdonald, M. Urquidi-Macdonald, Journal of Electrochemical Society 132 (10) (1985) 2316.
[11] M. Urquidi-Macdonald, S. Real, D.D. Macdonald, Journal of Electrochemical Society 133 (10)
(1985) 2019.
[12] M. Urquidi-Macdonald, S. Real, D.D. Macdonald, Electrochimica Acta 35 (10) (1990) 1559.
[13] R.L. Van Meirhaeghe, E.G. Dutoit, F. Cardon, W.P. Gomes, Electrochimica Acta 20 (1975) 995.
[14] B.A. Bouwkamp, EQUIVALENT CIRCUIT, computer assisted electrochemical ac-immittance data
analysis system, University of Twente, 1995.
R.-W. Bosch / Corrosion Science 47 (2005) 125143 143