Corrosion Science 48 (2006) 35653583

www.elsevier.com/locate/corsci

Determination of susceptibility to intergranular

corrosion and electrochemical reactivation
behaviour of AISI 316L type stainless steel
G.H. Aydogdu, M.K. Aydinol

Metallurgical and Materials Engineering Department, Middle East Technical University, Ankara 06531, Turkey
Received 11 April 2005; accepted 9 January 2006
Available online 15 March 2006

Abstract
In this study, double loop electrochemical potentiokinetic reactivation (DLEPR) test was applied to
determine the degree of sensitization in 316L type stainless steel, where obtained results were correlated
with revealed microstructures after oxalic acid test and weight loss measurements of Streicher and
Huey acid tests. Best agreement was provided with test parameters which are 1 M H2SO4 and
0.005 M KSCN at 0.833 mV/s scan rate at 30 C. Specimens were classied structurally as absence
of chromium carbides step, no single grain completely surrounded by carbides dual and one or more
grain completely surrounded by carbides ditch, in the as-etched structure, if the Ir:Ia (100) ratios
were obtained to be between 0 and 0.2, 0.2 and 5.0 and 5.0 and higher, respectively. It was also found
that at high KSCN concentrations, reactivation current prole skewed to higher potentials where this
was attributed the formation of metastable pits, during the anodic scan of the test procedure.
 2006 Elsevier Ltd. All rights reserved.
Keywords: Intergranular corrosion; DLEPR test; AISI 316L austenitic stainless steel

1. Introduction
The best known example of metallurgical eect on corrosion is intergranular corrosion
which is mostly observed on the use of austenitic stainless steels. Austenitic stainless steels
*

Corresponding author. Tel.: + 90 312 210 25 23; fax: +90 312 210 12 67.
E-mail address: kadri@metu.edu.tr (M.K. Aydinol).

0010-938X/$ - see front matter  2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.corsci.2006.01.003

3566

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

(containing 18% chromium8% nickel) are widely used in steam generating plants as piping and superheating tube materials due to their good mechanical properties and corrosion
resistance at elevated temperatures. However, when austenitic stainless steels have undergone a treatment like welding in the temperature range between 500 and 800 C, there is a
breaking corrosion resistance intergranularly as a result of segregation of chromium carbides. High concentration of chromium in M23C6 particles decreases locally the chromium
content in the region that is adjacent to these chromium rich precipitates. Since chromium
diuses much more slowly than carbon, there is not enough time for chromium to diuse
to the carbide from all over the grains, so, in the region that is near grain boundaries, chromium content lowers to below 13%, which is a critical value for required stainless corrosion behaviour. However, at prolonged treatments, chromium diusion from the bulk of
the grain increases the concentration above the critical limit and heals the boundaries [1].
In addition to M23C6 carbide, the exposure of AISI 316L type austenitic stainless steel
to elevated temperatures for long periods of time can result in formation of various other
phases (sigma, chi, laves phases) [26]. The formation of the intermetallic phases, which is
delayed due to the slower diusion of substitutional elements required for their nucleation
and growth, results in a depletion of chromium and molybdenum in austenite matrix.
Sigma phase with formula FeCr, which is more generally expanded as (FeNi)x(CrMo)y,
is a severe problem due to its detrimental eect on the mechanical properties and localized
corrosion resistance [7]. It nucleates mainly on the grain boundaries and is found in 316L
type stainless steels approximately in 100 h at 800 C [4].
In austenitic stainless steels, corrosion resistance is provided by a very thin surface lm,
known as passive lm that is an invisible lm of oxide, formed by the metal reacting with
the ambient environment. Normally these lms are free of pores, but their stability may be
weakened locally. It therefore has dierent properties in areas where the steel surface is
altered due to grain boundary precipitates. This heterogeneous microstructure is very dangerous since it weakens steel without much change in the outward appearance.
Intergranular attack is accelerated by potential dierences between grain and grain
boundaries, that is, attack is determined by availability of anodic sites at grain boundaries.
Therefore, making it anodic passivates the specimen. At that time, the chromium depleted
alloy sets up passiveactive cell of appreciable potential dierence, the grains (exhibit passive behaviour) constituting large cathodic areas relative to small anode areas at grain
boundaries (exhibit active behaviour). During decreasing the potential, the protective passive lm over chromium-depleted areas is more easily dissolved than that over undepleted
(non-sensitized) surfaces. The electrochemical potentiokinetic reactivation (EPR) test is
based on the assumption that only sensitized grain boundaries become active, while grain
bodies are unsensitized.
To provide a rapid, quantitative and non-destructive test method, lead many researchers to develop electrochemical potentiokinetic reactivation tests. The history and review of
EPR method were presented by Cihal and workers [8,9]. Detection of sensitization of
stainless steel started with potentiostatic polarization for etching of grain boundaries by
Cihal and Prazak in 1956. Introduction of reactivation from transpassive or passive state
with double loop EPR technique was presented by Cihal et al. [10]. Clarke et al. on the
bases of this rapport found out rstly the single loop EPR test to quantify the sensitization
[11]. This technique was also developed by Novak et al. [12]. On the other hand, double
loop technique was presented by Desestret et al. [13], Knyazheva et al. [14], Charbonier
[15], Umemura et al. [16,17] and Borella and Mignona [18] between the years 1971 and

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

3567

1980. This technique for especially detecting sensitization of 304 stainless steels was developed by Majidi and Streicher [19].
In the double loop test, specimen is rst polarized anodically through the active region
then the reactivation scan in the reverse direction is carried out. When it is polarized anodically at a given rate from the corrosion potential to a potential in the passive area, this
polarization leads to the formation of a passive layer on the whole surface. Then when
scanning direction is reversed and the potential is decreased at the same rate to the corrosion potential, it leads to the breakdown of the passive lm on chromium depleted areas.
As can be seen in Fig. 1, two loops are generated, an anodic loop and a reactivation loop.
A ratio of maximum current generated in the double loop test (Ir:Ia) is used as a measure
for the degree of sensitization.
In the literature, there are many studies which deal with verication of EPR test
method, comparison of double loop and single loop test methods and improving or applying test methods for dierent type of materials susceptibility to intergranular corrosion.
Cihal presented a study [20], which was a continuation of an earlier study dealing with
electrochemical determination of sensitivity to intercrystalline corrosion of stainless steel,
based on reactivation from the passive state. He showed that test method was veried on
austenitic chrome nickel steels with increased carbon content. The ratio of charge during
reactivation gave optimum quantitative criterion of tendency of steels to intercrystalline
corrosion. Discrepancies between the standard test (ASTM A-262 Practice E) and electrochemical reactivation test method was shown by Novak et al. [12]. They observed that the
reactivation method detects both continuous and local chromium depleted region in the
steel structure, however acid test exposes only continuous depletion zones leading to intergranular corrosion.
The original work by Majidi and Streicher [19], which proposes the double loop
method, compares the results of this new method and the single loop and acid test and
concludes that; the agreement between measurement made with double loop and single
loop EPR test was good and gave a quantitative measure of sensitization. It is also
concluded that the reproducibility of the double loop test is excellent when optimum
conditions are maintained. The optimum conditions were determined by examining

Fig. 1. Diagram for the procedures of double loop EPR test method.

3568

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

parameters such as surface nish, scan rate, temperature and potassium thiocyanate concentration which is used as an activator. They have determined the optimum DLEPR test
conditions to be electrolyte of composition 0.5 M sulphuric acid and 0.01 M potassium
thiocyanate and a scan rate of 1.667 mV/s. Some specics are such that there was an
increase not only in intergranular corrosion but also in general corrosion when lower scan
rates are used. Another result was the increased Ir:Ia ratio as the amount of potassium
thiocyanate concentration increases because there is again general and intergranular corrosion at the entire surface of specimen due to the activator property of potassium thiocyanate. However, above 0.03 M potassium thiocyanate there was no further increase in
Ir:Ia ratio.
The area under the large loop generated in the curve of potential versus current is proportional to the electric charge, Q that depends on surface area and grain size and is normalized by total grain boundary area. This normalization is called Pa that can be selected
as a tolerable level of sensitization for a given application. Majidi and Streicher also studied the eects of some parameters on Pa values in the single loop method for 304 and 304L
type stainless steels. They proposed that Pa value increases when decreasing scan rate,
increasing roughness from 1 to 23 lm and temperature of test solution [21]. Similarly
for 304 type stainless steels, degree of sensitization was evaluated with single loop method
by Jargelius et al. [22]. They argued that EPR test results are strongly dependent on the
testing temperature. Increasing temperature increases Pa value.
In the literature, there were also some studies on the applicability of DLEPR method on
high nickel alloys like Inconel 600, since this alloy also suers from intergranular corrosion. Inuence of some test parameters on EPR response were investigated in order to
assess the optimum conditions which were determined to be 0.1 M sulphuric acid
and 0.001 M potassium thiocyanate for sensitivity of EPR test method by Maday and
Mignone [23]. They discovered that at too low sulphuric acid concentration chromium
depleted regions were not detected, while too high acid concentration caused other types
of attack. On the other hand, the optimal modied DLEPR test condition for alloy 600
was obtained in 0.01 M sulphuric acid and 10 ppm potassium thiocyanate at 25 C and
at 0.5 mV/s scan rate by Ahn et al. [24]. They observed that standard test conditions cause
pitting and general corrosion in addition to intergranular corrosion. On the other hand,
the eect of potassium thiocyanate addition and its concentration on the reactivation
behaviour at single loop EPR test method of Alloy 600 in sulphuric acid solution were
investigated by Tsai and Wu [25]. They discovered that at high potassium thiocyanate concentrations, passivation is enhanced. Tsai, Wu and Cheng also proposed that for sensitized
Alloy 600, three anodic peaks appear in the reactivation loops. While higher anodic potential correspond to pitting corrosion and matrix corrosion at lower potential. The middle
potential range of +60 to 10 mV SCE was associated with grain boundary corrosion [26].
Similarly, the sensitization to intergranular corrosion of AISI 316 type stainless steel
was evaluated quantitatively both by microscopy and by electrochemical tests. The conformity of EPR test methods (single and double loop) and Strauss test on 18Cr12Ni2.5Mo
austenitic stainless steel was examined by Zahumensky and Tuleja [27]. An excellent agreement was observed between the results of these test methods. They also found that annealing at 650 C for 100 h led to the highest sensitization among experimental states. In
another study by Matula et al. [28], the degree of sensitization of AISI 316L type stainless
steel to intergranular corrosion was determined by means of electrolytic etching in oxalic
acid and EPR test method followed by metallographic inspection. Also the kinetics of

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

3569

precipitation of second phases were studied by means of quantitative metallography and

rst M23C6 carbides at grain boundary were detected. Chromium depletion were quantitatively evaluated by analytical electron microscope. They concluded that chromium
depleted zones increases with aging time.
In the present work, optimum test parameters of DLEPR method were investigated for
evaluation of susceptibility to intergranular corrosion in 316L type steels. These parameters are scan rate, potassium thiocyanate and sulphuric acid concentration in the test solution. Nitric acid and ferric sulphatesulphuric acid tests were also applied to specimens
which were classied as step, dual and ditch structure by means of oxalic acid tests. Results
of DLEPR and acid tests were correlated and nally reproducibility of convenient test
parameters of DLEPR test method was controlled.
2. Experimental procedure
The chemical analysis of 316L type stainless steel used in this study was given in Table 1.
Dierent heat treatment procedures were applied in order to simulate dierent degrees of
sensitization, details were given in Table 2.
All specimens were dipped into a solution of 100 g of oxalic acid crystals dissolved in
900 ml of deionized water. The specimens were made anode in the stainless steel beaker,
which was made the cathode. The specimens were etched at 1 A/cm2 for 90 s according
to ASTM A-262 Practice A. After applying the oxalic acid test, etched surfaces were rinsed
with deionized water and alcohol and then dried. The microstructural characterization was
made by scanning electron microscopy (SEM) using JEOL JSM-6400 Electron Microscope, and the etched structures are classied as step (absence of chromium carbides), dual
(no single grain completely surrounded by carbides), ditch (one or more grain completely
surrounded by carbides).
Nitric acid and ferric sulphatesulphuric acid tests were also conducted on all heat-treated samples. After heat treatments, surface was ground by 120 grit emery paper to remove
oxide scale which should be done with care. In ferric sulphatesulphuric acid test ASTM
A-262 Practice B, 236 ml of sulphuric acid is added slowly to 400 ml deionized water in
order to prevent boiling by heat evolution so that the concentration of the solution is
maintained. Then, 25 g ferric sulphate is added to sulphuric acid solution. The specimens
were not immersed with the cradle in the Erlenmeyer ask, unless the ferric sulphate was
completely dissolved in the solution. During boiling period of 120 h, the colour of solution
has been controlled and when it changed to dark green, ferric sulphate inhibitor was
added. In nitric acid test ASTM A-262 Practice C, a fresh 65% nitric acid was boiled
and specimens were kept at this condition for ve 48 h periods.
For acid tests, specimens were weighed with 0.00001 g sensitivity analytical balance
before and after these experiments. The corrosion rate was calculated as the loss in weight
as inch per month (ipm) according to ASTM A-262 as follows:

Table 1
Chemical composition of 316L type stainless steel (wt.%)
C

Cr

Mo

Ni

Si

Mn

Fe

0.021

16.82

2.44

11.5

0.406

1.50

0.0338

0.0478

66.19

3570

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

Table 2
Specimen nomenclature and applied heat treatments
Name of the specimen

Heat treatment time and temperature

NS
S-160
S-233
S-285
S-336
S-406

At 1050 C 2.4 ks (40 min) + water quench

NS + at 650 C 576 ks (160 h) + water quench
NS + at 650 C 838.8 ks (233 h) + water quench
NS + at 650 C 1026 ks (285 h) + water quench
NS + at 650 C 1209.6 ks (336 h) + water quench
NS + at 650 C 1461.6 ks (406 h) + water quench

Table 3
Variation of DLEPR test parameters and their representative experiment codes
Experiment code

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

ipm

278  W
Atd

Test parameters
H2SO4 (M)

KSCN (M)

Scan rate (mV/s)

0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1
1
1
1
1
1
1
1
1
1
1.5
1.5
1.5

0.005
0.005
0.005
0.005
0.01
0.01
0.01
0.01
0.02
0.02
0.005
0.005
0.005
0.005
0.01
0.01
0.01
0.01
0.02
0.02
0.005
0.01
0.02

0.277
0.833
1.667
2.5
0.277
0.833
1.667
2.5
0.833
1.667
0.277
0.833
1.667
2.5
0.277
0.833
1.667
2.5
0.833
1.667
1.667
1.667
1.667

where t is the time of exposure in hours, A is the total surface area in cm2, W is the weight
loss in grams and d is the density, where for chromiumnickelmolybdenum stainless steels
it is taken as 8 g/cm3.
In the DLEPR test, rst a hole of 2.5 mm diameter was drilled on one side of the 20 mm
long cylindrical specimens before the heat treatment procedure. Then the regarding heat
treatment was applied to specimens after which a 3 mm diameter thread was opened, so
that the contact between the specimen and current transfer rod is clear. Finally, all surface
of specimen was ground by 400 up to 1200 grit emery paper. The ner nish is used for this
test to enhance the quality of photomicrographs, and also specimen was polished 1 lm

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

3571

alumina paste for shining appearance. During cutting, grinding and polishing operations
work piece was cooled with water to minimize temperature increase. The DLEPR test
solution is prepared freshly under a ventilated hood with stirring and used not more than
ve times due to possible breakdown of solution purity. The solution temperature is held
constant at 30 1 C with the use of water bath where its temperature is controlled by
thermostat control heater. The variety of test parameters can be seen in Table 3.
The electrochemical polarization cell design in this study is due to ASTM G-5 standard.
In this design, the cylindrical working electrode is centrally located between the two counterelectrodes which are placed at the sides of the cell for better current distribution and
made of materials that are inert to test solution even under strong anodic polarization.
The specimen, two counterelectrodes and a calomel reference electrode, which is positioned in a salt bridge, are connected to Solartron 1480 Multi Channel potentiostat.
The potential of the working electrode is measured through the luggin probe, which is exibly mounted to the cell and probe tip was placed near the specimen surface to minimize
IR-drop. The multistat is controlled by Corrware software, which enables the test variables to be set and the results to be implemented. The procedure is composed of two steps.
Firstly, the specimen was subjected to open circuit conditions for 300 s so that Ecorr develops. Then voltage is scanned anodically from Ecorr to +0.3 V versus SCE with the regarding scan rate, after which it is reversed back to Ecorr.
3. Results and discussion
The resulting microstructures after the oxalic acid test were given in Fig. 2, along with
their classication according to ASTM A-262. As can be seen, the NS specimen exhibits
the step structure, whereas S-160, S-233 and S-285 exhibit the dual structure. Although
the number of completely encircled grains in S-406 is more than in S-336, both are classied as the ditch structure.
Results of nitric acid test (Practice C) and ferric sulphatesulphuric acid (Practice B)
test methods were given in Fig. 3. It is seen that both methods gave similar results, where
corrosion rate initially increases with aging time, afterwards however, it is slowed down
and even a slight decrease is seen. This is believed to be due to chromium re-enrichment
of the grain boundaries because of the availability of time for chromium to diuse from
the grain to the boundary.
In any of the combinations of the test parameters, imperceptible reactivation behaviour
was obtained for the non-sensitized specimen, which clearly depicts the state of the structure. However all of the sensitized specimens, of dierent degree, showed a clearly recognizable reactivation behaviour, as it is seen from the polarization curves given in Fig. 4.
In general, for all specimens, potassium thiocyanate is more eective than sulphuric
acid to increase the passivation potential and current almost irrespective of the scan rate
used. Moreover, at the same test conditions, all specimens gave very similar activation
behaviour, which is desired, so that Ia can be used as a reference state for the reactivation
behaviour. The reactivation current itself, however, showed a quite complex behaviour
depending on the concentrations of potassium thiocyanate and sulphuric acid, and the
scan rate. Therefore in order to understand the eect of these parameters, univariate
analysis of variance was performed on the Ir:Ia (100) values to obtain a general linear
model (GLM). The GLM univariate procedure, which was implemented in many popular
statistical analysis software, provides regression analysis and analysis of variance for one

3572

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

Fig. 2. SEM micrographs of specimens after the oxalic acid etch. (a) NS step, (b) S-160 dual, (c) S-233 dual,
(d) S-285 dual, (e) S-336 ditch and (f) S-406 ditch.

dependent variable by one or more factors and/or variables. Using this general linear
model procedure, one can test null hypothesis about the eects of other variables on the
means of various groupings of a single dependent variable. Therefore one can investigate
interactions between factors as well as the eects of individual factors, some of which may
be random. In short, univariate analysis provides error estimates for each factor, and its
eect on the dependent variable, see Fig. 5, where the analysis for S-233 and S-406 specimens were given. Furthermore the test of null hypothesis gives the importance of the fac-

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

3573

Fig. 3. Weight loss acid test results according to ASTM A-262, ipm for all specimens.

Fig. 4. Polarization curves for the NS, S-285 and S-406 specimens. Test conditions were 0.005 M KSCN + 1 M
H2SO4 and 1.667 mV/s scan rate.

tors eect. If the signicance of the test is lower than 0.05 then it should be understood
that the eect of the factor is very signicant. In Table 4, the signicance of the GLM analysis were given for the dependent variables Ir:Ia for all of the specimens. In Fig. 5(a), for
example, after GLM analysis we have obtained quite horizontal graphs, considering the
range of error bars on the data points, for both specimens S-233 and S-406. This means
that, going from 0.5 M to 1 M and then to 1.5 M of sulphuric acid concentration did
not aect the Ir:Ia (100) ratio signicantly, where we have obtained 0.413 and 0.568 as
GLM signicance values for S-233 and S-406 specimens respectively, as given in Table
4. These numbers are very much larger than 0.05, meaning that the eect of sulphuric acid

3574

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

(a)

(b)

(c)
Fig. 5. Results of GLM analysis. Factors variables are (a) H2SO4 molarity, (b) KSCN molarity and (c) scan rate
(mV/s), and the dependent variable is Ir:Ia (100). The eect of these factor variables, including the error bars, on
Ir:Ia (100) was given for S-233 and S-406 specimens.

is insignicant. In Fig. 5(b), however, a clear dependence of Ir:Ia (100) ratio on potassium
thiocyanate concentration was seen, as it is also seen in Table 4 from the GLM signicance
values, which are 0.009 and 0.001 for S-233 and S-406 specimens respectively.
The analysis of Table 4, Fig. 5 and such gures of all of the specimens for the dependent
variable Ir:Ia, resulted in the following conclusions to be made. The sulphuric acid concentration has a weak eect on Ir:Ia, regardless of the state of the specimen (dual or ditch),
and randomly either increases or decreases the ratio. The potassium thiocyanate concen-

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

3575

Table 4
GLM signicance values for the dependent variable Ir:Ia (100) ratio for all specimens
Name of specimen

S-160
S-233
S-285
S-336
S-406

Factor variables
H2SO4

KSCN

Scan rate

0.873
0.413
0.035
0.149
0.568

0.024
0.009
0.096
0.000
0.001

0.220
0.003
0.001
0.019
0.010

Signicance value less than 0.05 for any factor variable means that Ir:Ia (100) ratio depends strongly on that
factor variable in that specimen state.

tration, however, has a strong eect on the dependent variable, such that increased potassium thiocyanate always decreases the ratio. Scan rate also aects the ratio considerably,
such that increasing scan rate generally decreases the Ir:Ia ratio. If the GLM analysis were
made to see the eects of factor variables on Ia and Ir separately, following observations
were made. Other factors being constant and regardless of the state of the specimen, the
increased concentration of sulphuric acid causes an increase on both Ia and Ir, as can
be understood from its weak eect on the ratio of Ir:Ia. Moreover, going from dual to ditch
structure, it was observed that the absolute values of Ia did not change considerably for the
respective concentrations of sulphuric acid, whereas, there was a slight increase in the
absolute values of Ir.
The eect of scan rate on Ia, other factors being constant, was quite low for all states of
the specimen, but a slight decrease can be noticed. Its eect on Ir, on the other hand, is
very pronounced and as scan rate increases, Ir drops considerably. Similarly, the absolute
values of Ia did not show much dependence on the state of the specimen, but for Ir, there
was again a slight increase as going from dual to ditch structure. In addition, it was
observed that, the reactivation curve expanded to active potentials with lower scan rates.
This can be the sign of an increase in general corrosion rather than intergranular corrosion. The low Ir values at high scan rates, is most probably because of the insucient time,
where the passive lm breakdown cannot occur eectively during reactivation scan. Therefore, Ia being almost an invariant and strong dependence of Ir on scan rate, it is very probable to come to wrong conclusions about the state of the steel. In Fig. 6, the polarization
curves of S-233 and S-406 were given, in which the dual structure was appeared to be
exposed to more corrosion attack although its grain boundaries are more resistant to intergranular corrosion than the ditch structure.
The eect of potassium thiocyanate concentration especially on Ir was found to be
somehow dierent from the other factors. Its eect on Ia, others being constant, was such
that Ia increases considerably as potassium thiocyanate molarity increases. This increase
was observed for all specimens and the absolute values at respective potassium thiocyanate
concentrations were similar. Its eect on Ir, however, was dierent. It was observed that at
high concentrations of potassium thiocyanate Ir drops. More important than the drop
itself was the change in the reactivation prole. In Fig. 7, reactivation proles for the dual
and ditch structures were given depending on the potassium thiocyanate concentration. It
can be seen that as potassium thiocyanate increases there is a drop in the Ir, but also the
prole became skewed to higher potentials.

3576

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

Fig. 6. Polarization behaviour of S-233 and S-406 specimens under fast and slow voltage scanning conditions
respectively. Note the lower Ir value for the ditch structure compared to dual. Test solution concentration was
0.5 M H2SO4 and 0.01 M KSCN.

This is very prominent especially in the ditch structure at higher scan rates. The reason
for this eect may be explained by the observation of a similar eect that was made in Inconel 600 alloy [25,26]. In that study, the reactivation curve having two distinguishable
peaks were deconvoluted to several reactivation curves. Wu et al. [25,26] arrived to the
conclusion, by comparing the microstructure of the alloy that showed the two peaks
and the one that not, that the peaks were due to dierent types of corrosion occurring
in the alloy. The deconvoluted curve appearing at higher potentials were attributed to
the pitting type of corrosion occurred in the alloy.
In this regard, the skewed reactivation prole we obtained, can be because of the combined behaviour of two corrosion processes taking place simultaneously, where the one
taking place at higher potentials dominating over the other one. Considering the conclusion of Wu et al. [25,26], we have investigated the microstructure of the S-406 specimen,
after it has been exposed to DLEPR test with dierent potassium thiocyanate concentrations. The micrographs are given in Fig. 8. It can be seen that, there is denitely a dierent
activity taking place at the surface of the specimen, which is not rather the intergranular
corrosion. However this activity could not clearly be attributed to pitting type of corrosion
too. As can be seen the number of stable pits observed in low and high potassium thiocyanate solutions is not so dierent, but an increased metastable pit formation in the latter.
The formation of such metastable pits can be either because of the microstructural features
of the alloy, the repassivation mechanism or the eect of remaining surface state after the
anodic scan, where the last two seems to be more likely. In order to understand this phenomenon, the microstructure of S-406 specimen just after the anodic scan (without reactivation) was examined. The micrographs were given in Fig. 9 for the 0.005 M and 0.02 M
potassium thiocyanate solutions. A similar microstructure was obtained compared to the
complete scan cycle picture, Fig. 8. Therefore it is understood that the metastable pit formation is because of the simultaneous occurrence of passivation and lm breaking, and it

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

3577

Fig. 7. Reactivation proles of specimens (a) S-233 and (b) S-406 during the DLEPR test with dierent KSCN
concentrations at 1 M H2SO4and 1.667 mV/s scan rate.

is enhanced in solutions containing higher amounts of potassium thiocyanate which eases

lm breaking. Such a remaining state, then causes the observed change in the reactivation
prole, and a drop in the value of Ir.
In order to determine what combination of DLEPR test parameters would give the best
prediction, we correlate the results of the DLEPR and weight loss acid tests. The results
were given in Table 5. The correlation coecients close to one indicates that the two
results (DLEPR and acid tests) are correlated to each other and the signicance which
was given in brackets gives how strong is the correlation, e.g. signicance value less than
0.05 means that there is a very strong correlation.
We have arbitrarily chosen the lower limit of correlation coecient (for the current
ratio) to be 0.9 that is to be satised for all acid test, or 0.95 for one test and 0.85 for
the other acid test. The DLEPR test parameters that yielded good correlation between
the acid tests according to the above criteria were given with the experiment codes 8, 9,
10, 12, 15 and 19. The Ir:Ia (100) ratios of the above-mentioned experiments and the

3578

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

Fig. 8. SEM micrographs of S-406 specimen after DLEPR test in 1 M H2SO4 solution at 1.667 mV/s scan rate
with (a, b) 0.005 M KSCN and (c, d) 0.02 M KSCN.

Fig. 9. SEM micrographs of S-406 specimen just after the anodic scan in 1 M H2SO4 solution at 1.667 mV/s scan
rate with (a) 0.005 M KSCN and (b) 0.02 M KSCN.

ipm of Practice B and Practice C for all specimen types were given in Fig. 10. Experiments
8 and 15 have been carried out with scan rates 2.5 and 0.277 mV/s, respectively. As we discussed before, very high or very low scan rates may be deceptive for the determination of
the state of the specimen and it is wise not use these scan rates along with any other test

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

3579

Table 5
Correlation and in brackets the signicance values between Ir:Ia (100) of DLEPR test and ipm of weight loss
acid tests for Practice B and Practice C of ASTM A-262
Experiment
code

Practice B

Practice C

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

0.867 (0.025)
0.862 (0.027)
0.637 (0.174)
0.649 (0.163)
0.874 (0.023)
0.851 (0.032)
0.698 (0.123)
0.91 (0.012)
0.944 (0.005)
0.989 (0)
0.928 (0.008)
0.98 (0.001)
0.667 (0.148)
0.983 (0)
0.881 (0.02)
0.898 (0.015)
0.892 (0.017)
0.703 (0.119)
0.98 (0.001)
0.938 (0.006)
0.871 (0.024)
0.814 (0.049)
0.841 (0.036)

0.946 (0.004)
0.738 (0.094)
0.722 (0.105)
0.749 (0.087)
0.802 (0.055)
0.87 (0.024)
0.597 (0.211)
0.912 (0.011)
0.908 (0.012)
0.869 (0.025)
0.876 (0.022)
0.871 (0.024)
0.706 (0.117)
0.837 (0.037)
0.954 (0.003)
0.865 (0.026)
0.897 (0.015)
0.875 (0.023)
0.951 (0.004)
0.851 (0.032)
0.825 (0.043)
0.903 (0.014)
0.968 (0.002)

Correlation value close to one means that two results are correlated to each other.

parameter even if they yield good correlation. Moreover the Ir:Ia (100) ratios of the
experiment 8, regarding the specimen state, are close to each other, so that the resolution
of this particular experiment is low.
In the experiments 9, 10 and 19, solutions containing 0.02 M potassium thiocyanate
were used. In this condition, we should keep in mind that, during reactivation scan, not
only intergranular corrosion but also pitting type of corrosion may take place. This is seen
more obviously in the ditch structure rather than in the dual structure, so sensitization
degree of the ditch structure should appear to be higher, where this behaviour is not inconvenient for our purposes. However, as can be seen from Fig. 10, for experiments 9 and 10,
due to their low sulphuric acid content, their resolution again seemed to be low.
The nal experiment in the list was given with code 12, which does not indicate any negative concerns mentioned before and also it predicts the results of the Streicher acid test
with exceptional good agreement and resolution. The polarization curves of specimens
tested with parameters as given in the experiment 12, which are 1 M sulphuric acid and
0.005 M potassium thiocyanate and 0.833 mV/s scan rate, were given in Fig. 11. As can
be seen, there is a smooth transition as the state of the structure goes from step to dual
and to ditch.
Finally to check the reproducibility of the test results, the S-233 (dual) and S-406 (ditch)
specimens were tested successively 10 more times under the conditions of the experiment
12. The mean, standard deviation, standard error and 95% condence limits for potential

3580

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

Fig. 10. Ir:Ia (100) ratios for the experiment codes 8, 9, 10, 12, 15 and 19, and corrosion rates according to
ASTM A-262 (a) Practice B and (b) Practice C.

and current values were given in Table 6. It is found that, the passivation and depassivation potentials can be precisely obtained. Similarly the activation current can be
reproduced within a slight error margin. However for the reactivation currents there is
some variation, where its magnitude increases for the ditch structure. Nevertheless it is
believed that, the results were reproduced within an acceptable error margin, with standard errors on the reactivation current as 0.12 and 0.36 for the dual and ditch structures,
respectively.
In order to classify the state of the structure according to the obtained result in the
DLEPR test, we analyzed and correlated the results of all tests made and postulate that
(under experiment 12 conditions), specimens giving Ir:Ia (100) ratio higher than 5.0 have
the ditch structure. The upper limit for the dual structure is therefore 5.0. When the results
of all experiments for the NS and S-160 specimens were examined, the highest value
obtained for NS is in the order of 0.2 and lowest value obtained for S-160 is in the order
of 0.5. The values for the experiment 12 conditions are 0.036 and 0.782 respectively. Therefore, in order to set a clear limit from step to dual, an additional heat treatment was

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

3581

Fig. 11. Polarization curves of all specimens tested under experiment 12 conditions (1 M H2SO4 and 0.005 M
KSCN solution at 0.833 mV/s scan rate).

Table 6
Reproducibility of the DLEPR results for the S-233 and S-406 specimens with experiment 12 conditions
Specimen

Mean

Standard
deviation

Standard
error

95% condence
range, lower:upper

S-233
Ea (V SCE)
Ia (A/m2)
Er (V SCE)
Ir (A/m2)
Ir:Ia (100)

0.146
144.5
0.172
3.5
2.39

0.009
4.0
0.007
0.6
0.40

0.003
1.2
0.002
0.2
0.12

0.153:0.140
141.7:147.2
0.177:0.168
3.0:3.9
2.12:2.65

S-406
Ea (V SCE)
Ia (A/m2)
Er (V SCE)
Ir (A/m2)
Ir:Ia (100)

0.154
160.8
0.142
9.1
5.67

0.007
6.7
0.023
2.2
1.19

0.002
2.0
0.007
0.6
0.36

0.158:0.149
156.3:165.3
0.158:0.127
7.7:10.6
4.87:6.47

Statistical analysis was made over 11 samples.

applied so as to yield the step structure. After the NS treatment the specimen was heat
treated at 650 C for 183.6 ks. The microstructure of this specimen was characterized as
step in the oxalic acid test, see Fig. 12, and the obtained ipm values of weight loss acid
tests were 0.00416 and 0.02952 for Practice B and C, respectively. The corresponding values for the NS specimen were 0.0015 and 0.00166 and for the S-160 specimen they were
0.01244 and 0.15747. The Ir:Ia (100) result of DLEPR test for this new specimen, under
experiment 12 conditions, was 0.3979 0.065, where the statistical analysis was made over
six experiments. Then, to be on the safe side, an approximate value of 0.2 was assumed to
be the upper limit for the step structure.

3582

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

Fig. 12. SEM micrograph of the specimen, after the oxalic acid etch, heat treated at 650 C for 183.6 ks (51 h),
revealing the step structure.

4. Conclusions
In this study, the eect of scan rate and composition on the anodic polarization and the
reactivation behaviour of AISI 316L stainless steel was investigated, from which a criteria
can be obtained for the determination of susceptibility to intergranular corrosion. This criteria, Ir:Ia ratio, is the basis of the DLEPR method. The arbitrary choice of these test
parameters might be misleading. Therefore in order to devise a procedure for the correct
prediction of the degree of susceptibility to intergranular corrosion in this steel, DLEPR
test parameters were systematically varied and correlated with the results of the weight loss
acid tests and with the analysis of the microstructure, where nally, the following conclusions were drawn.
In general, Ir and Ia values increase similarly with the increase of sulphuric acid content,
thus constituting a weak functional dependence for the Ir:Ia ratio.
The activation parameters weakly depends on the scan rate, but there is a strong eect
on the reactivation parameters. At high scan rates, during reactivation, time is not sucient to breakdown the passive lm, whereas there is plenty of time at low scan rates,
which causes general corrosion to take place too. Therefore for low scan rates Ir increases,
so does the Ir:Ia ratio. Therefore the improper choice of the scan rate can yield wrong
results to be used for the prediction of susceptibility.
There is an increase in anodic current with the increased potassium thiocyanate content
in the test solution, but for the reactivation current a more complex behaviour is seen.
Ir always decreased for all specimen types when potassium thiocyanate is at 0.02 M concentration compared to lower concentrations. Moreover, reactivation current prole
changes with potassium thiocyanate in such a way that it becomes skewed to higher potentials, where it is very obviously seen in the ditch structure. It is believed that this behaviour
is due to some surface activity taking place over the remaining microstructure of the anodic scan, resembling metastable pits.
DLEPR test presents quantitative results. In the evaluation of sensitization in 316L
type steel, in terms of Ir:Ia the best agreement with the weight loss acid tests were obtained

G.H. Aydo
gdu, M.K. Aydinol / Corrosion Science 48 (2006) 35653583

3583

with the following test parameters, 1 M sulphuric acid and 0.005 M potassium thiocyanate
solution at 0.833 mV/s scan rate and with 30 C solution temperature. Increasing the
potassium thiocyanate concentration, generally, still correlates well with the acid test
results, but the resolution decreases slightly.
Using the proposed test parameters, 316L type steels are classied as ditch, dual and
step when they yield Ir:Ia (100) higher than 5.0, between 5.0 and 0.2, and less than 0.2,
respectively in the DLEPR test. No tolerance was allowed for step to dual limit due to
safety reasons but for the dual to ditch limit, there can be a 1.0 deviation.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]

[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]

H. Sahlaoui, K. Makhlouf, H. Sidhom, J. Philibert, Mater. Sci. Eng. A 372 (2004) 98108.
T. Sourmail, Mater. Sci. Technol. 17 (2001) 114.
E.L. Hall, C.L. Briant, Metall. Trans. 15A (1984) 793811.
B. Weiss, R. Stickler, Metall. Trans. 3 (1972) 851866.
J.K.L. Lai, Mater. Sci. Eng. 58 (1983) 195209.
Y. Minami, H. Kimura, Y. Ihara, Mater. Sci. Technol. 2 (1986) 795806.
M. Schwind, J. Kallqvist, J.O. Nilsson, J. Agren, H.O. Andren, Acta Mater. 48 (2000) 24732481.
V. Cihal, R. Stefec, Electrochim. Acta 46 (2001) 38673877.
V. Cihal, T. Shoji, V. Kain, Y. Watanabe, R. Stefec, EPR A Comprehensive Review 2004, FRRI
Publication, Graduate School of Engineering, Tohoku University, Sendai, Japan.
V. Cihal, A Desestret, M. Froment, G.H. Wagner, Etude de nouveaux tests potentiocinetiques de corrosion
intergranulaire des aciers inoxydables, Etude CETIM Centre de Researches de Firminy, C.A.F.L., Rapport
no. 958, December 1969.
W.L. Clarke, W.M. Romero, J.C. Danko, Corrosion77, National Association of Corrosion Engineers,
Houston, Texas, 1977, print no: 180.
P. Novak, R. Stefec, F. Franz, Corrosion 31 (1975) 344347.
A. Desestret, P. Guiraldenq, M. Froment, Extended Abstracts, 23rd Meeting of I.S.E., Stockholm, 1972,
p. 87.
V.M. Knyazheva, Zashch. Metall. 4 (1972) 420.
J.C. Charbonier, IRSID COS Report, 1974, 74/58.
F. Umemura, M. Akashi, T. Kawamoto, B. Gijutsu, Corr. Eng. Jpn. 29 (1980) 163.
F. Umemura, T. Kawamoto, B. Gijutsu, Corr. Eng. Jpn. 28 (1979) 24.
A. Borella, A. Mignone, Br. Corr. J. 17 (1982) 176.
A.P. Majidi, M.A. Streicher, Corrosion 40 (11) (1984) 584593.
V. Cihal, Corr. Sci. 20 (1980) 737744.
A.P. Majidi, M.A. Streicher, Corrosion-NACE 40 (8) (1984) 393408.
R.F.A. Jargelius, S. Hertzman, E. Symniotis, H. Hanninen, P. Aaltonen, Corrosion 47 (1991) 429435.
M.F. Maday, A. Mignone, M. Vittori, Corr. Sci. 28 (9) (1988) 887900.
M.K. Ahn, H.J. Kwon, J.H. Lee, Corrosion 51 (6) (1995) 441449.
T.F. Wu, W.T. Tsai, Corr. Sci. 45 (2003) 267280.
T.F. Wu, T.P. Cheng, W.T. Tsai, J. Nucl. Mater. 295 (2001) 233243.
P. Zahumensky, S. Tuleja, J. Orszagova, J. Janovec, Corr. Sci. 41 (1999) 13051322.
M. Matula, L. Hyspecka, M. Svoboda, V. Vodarek, C. Dagbert, J. Galland, Z. Stonawska, L. Tuma, Mater.
Charact. 46 (2001) 203210.