Content uploaded by Fritz Hunkeler
Author content
All content in this area was uploaded by Fritz Hunkeler on Aug 05, 2017
Content may be subject to copyright.
Metastable Pitting of Stainless Steel
G. S. FRANKEN L. STOCKERT,^ F. HUNKELER.^ H. BOEHNI"
Abstract
Current transients resulting from metastable pitting events on stainless steel at potentials
below the pitting potential are studied. Metastable pits are covered during growth and exhihit
a constant current density. Rupture of the cover during metastable growth leads to repass'va-
tion of the pit. If a salt film precipitates on the pit surface before the cover ruptures, pit growth
is stabilized. The kinetic factors hindering the replacement of a passive film by a salt fflm are
discussed.
Introduction
Many investigators nave oüserved that passive film break-
down and repair events can occur on stainless steels (SSs) at
Potentials below the potentiat required for stable pitting {Ep).1-6
These events result in current transients such äs those shown
in Figure 1. The potential region where current transients were
observed has been called the repassivation region3 or the
region of unstable pitting* It will be shown below that such
current transients may be observed during a potentiostatic
lest at Potentials far below Ep and also above Ep during the in-
".ubation period before the onset of stable pitting. Further-
more, when stable pits first Start growing, they behave in the
same manner äs the pits responsible for the current tran-
sients. It will become clear that a more appropriate term for
this type of pitting is metastable pitting.
It has recently been suggested ttiat a salt film is more
stabte than the passive film on SS in chloride-containing solu-
tions at potentials above a value E,, (-190 mV SCE for AISI
302 SS).7 However, stable breakdown of the passive film does
not occur until the much higher value of Ep because the
replacement of the passive film by a salt film is a klnetically-
hindered process. From this viewpoint, it Is possible to explain
several characteristics of the pittlng process such äs the varia-
tion in the reported values of pitting and repassivation poten-
tials.
Through the analysis of metastable pitting, it is possible
to better understand same of the kinetic factors hindering the
replacement of the passive film on SS by a salt film at poten-
*Submitted for publication May 1986.
'I.B.M. Corp., T. J. Watson Research Center, P.O. Box 218,
Yorktown tieights. New York 10598.
"Institute of Materials Chemistry and Corrosion, Swiss
Federal Institute of Technology, CH-8093 Zürich, Switzer-
fand.
TIME
FIGURE 1 - Metastable pitting current transients
observed on 302 SS in 0.1 M NaCI at 420 mV SCE.
tials above Eg,. The stability criteria for pitting may also be ac-
curately determined.
Experimental Procedure
AISI 302 SS wire with a diameter of 0.5 mm and composi-
tion shown in Table 1 was tested. The wire loop electrode con-
figuration described previously was used to eliminate crevice
corrosion.8'9 In this work, the wire was not bent to form a com-
Vol. 43, No. 7, July 1987
0010-9312/87/000237/$3.00/0
© 1987, National Association of Corrosion Engineers
429
TABLE1
Composition of
AISI 302 SS in
wt% (balance
iron)
.^:.,
^fe
Cr
Ni
Mn
Si
Mo
c
Co
s
Ti
18.1
9.2
1.3
0.55
0.26
0.076
< 0.02
0.005
< 0.002
plete loop. Instead, the middle section of a lengtti of wire was
simply dipped into the solution, resulting in an exposure area
of - 1 cm2. The wire specimens were first prepared by manual-
ly polishing with 400 grlt paper. They were then cleaned se-
quentially in ultrasonic baths of distilled water, pure gasoline,
and ethanol. After each bath, they were rinsed with distilled
water and dried with hat air.
The solutions were prepared from distilled water and
reagent grade chemicals and purged with nitrogen before and
during testing. All experiments were performed at room tem-
perature (RT).
After 10 min at the rest potential, the specimens were
polarized relative to an SCE at a congtant potential below Ep.
The current, measured äs the potential drop across a shunt
resistor, was monitored by a digital oscilloscope at 50 Hz. The
data collection System has been previously reported.10
To further study satt film precipitation under potentio-
static condjtions, experiments were also performed with ar-
tificial pit electrodes.11'12 AISI 302 SS wire of diameter 0.3 mm
and pure iron wire of diameter 0.25 mm were imbedded in
epoxy and attached to the bottom of a cetl facing upward. The
specimens were first dissolved to a depth of at least twice the
diameter at a potential of 800 mV SCE in 0.1 M HCI, 1.0 M HCI,
1.5 M Nid;, or 3 M NiClz. The potential was then switched off
for 10 min and the pit was flushed with bulk electrolyte. Salt
film precipitation was studied by jumping the potential to a
given value between 200 and 1000 mV SCE and observing the
current. A1 the point of precipitation, the current decreased
rapidly. Between measurements, the specimen was atlowed to
dissolve for 10 min under diffusion control and then held at
open circuit for 10 min with frequent flushing of the pit to
dissotve the salt film. Experiments were also performed with
artificial pit electrodes using the galvanostatic technique in
0.1 M HCI since no investigations in dilute solutions have been
reported.11'12
Results
Metastable pitting transients exhibited by SSs in chloride-
containing solutions have a characteristic shape, äs displayed
in Figure 1. The current increases abovethe background pas-
sive current äs the pit nucleates and begins to grow. After a
short growth period, the metastable pits repassivate and the
current quickly decreases to the level of the original passive
current. During the growth stage, the current increases ap-
proximately with the square of time. The current frequentty in-
creases at a faster rate just before the decay. The lifetimes of
metastable pits have been observed in this work to be äs lang
FIGURE 2 - SEM micrographs of metastable pits on
302 SS after testing in 0.1 M NaCI at 340 mV SCE (a)
after lest but before ultrasonic cleaning; (b) after
uttrasonic cleaning.
äs 15 s, but typically less than 5 s. The transient heights varied
from 4 /iA to äs much äs 20 /xA.
The metastable pits associated with the current tran-
sients are readily observable by scanning electron microscopy
(SEM) after a run (Figure 2(a)]. They are covered by a layer that
is a remnant of the passive film. Puncturing ordestroying such
covers early in the growth of pits in SSs have been found to
result in repassivation oi the pits.13-14 In this work, every pit
observed had an opening in the cover [see Figure 2(a)] or was
directly adjacent to another pit whose cover had an opening.
This suggests that the creation of openings leads to the
repassivation of these metastabte pits and (hat during the
growth stage they are covered by an intact layer. The covers
are easily removed by cleaning in an ultrasonic water bath to
reveal the insides of the pits. They are hemispherical in shape
[Figure 2(b)], indicating that the current density (CD) is con-
stant aver the whole pit surface. Metastable pits observed by
SEM had diameters from less than 1 ^m to more than 10 ^m.
The pit volume, area, and thus CD äs a function of time
may be calculated by integrating the current, using a conver-
sion factor of c, = 3.24 x 10-5 cm3/C and assuming
hemispherical shapes.15 The CD for a hemispherical pit with
dissolution current increasing proportional to t2 should be
constant aver time and the radius should grow linearly with
time.16 CDs calculated from the current transients are indeed
found to be rather constant aver the tifetime of the metastable
pits, which may be äs lang äs 15 s. An example is shown in
Figure 3. Higher CDs are always determined at the very begin-
ning of growth äs a result of error in the calculation when the
current and the area are both very small. The CD typically in-
creases just before the repassivation, which is connected to
430
CORROSION-NACE
1.08 ^
'< 1.0^ h
UJ
(E
a:
=?
0
0.96h
0.92l-
rsi
'§
<
>-
h-
ül
z
UJ
a
ÜJ
Q:
CE
TIME
FIGURE 3 - (a) Metastable pit current transient ob-
served in 0.1 M NaCI at 340 mV SCE; (b) CD äs a function
of time calcutated from the current transient.
the faster rate of current increase usually observed in this time
period. The average CD during the growth stage, V, deter-
mined after elimination of both the initial region of higher CD
and the final decay, is dependent upon the potential and also
the bulk chloride concentration of the solution (Figure 4). Also
shown in Figure 4 are the CDs measured at the point of max-
imum current, i*, which are slightly higher than the average
CDs. The vatues shown in Figure 4 are averages of values
determined from many transients. The Standard deviations
ranged from 20 to 40%.
Pretreatment of a specimen by oxidizing in air at 250 C for
8 h has an effect on metastable pitting. The nucleation rate is
strongty reduced and the average CD at a given potential is
higher (Figure 4). The CD at the point of peak current is, how-
ever, relativety unaffected. Ep decreases from 440 mV SCE for
specimens prepared in the Standard fashion and tested in 0.1
M NaCI to 420 mV SCE for specimens thermally preoxldized at
250 C for 8 h. Electropolishing before thermal oxidation results
in a much larger change in Ep. Oxidizing at 250 C for 2 h after
electropolishing for 4 min in a solution containing 60% phos-
phoric acid, 20% sulfuric acid, and 20% water at 10 V reduces
Ep to 250 mV SCE.
Prepolarization in solution at Potentials below Ep has the
opposite effect äs it decreases the susceptibility 6f SSs to
stable pitting. By increasing the potential stepwise from
Potentials below Ep, specimens are able to withstand an ex-
tended period at 530 mV SCE in 0.1 M NaCI (90 mV above Ep)
without the onset of stable pitting. Note that Ep is.defined here
äs the potential at which stable pitting occurs on specimens
prepared in a given fashion without prepolarization at lower
Potentials.
"e 3
u
i 2
UJ
a
Ut
ai
cc
3
0
6 M LiCI.ODOIM NaOH
01 M NaCI
OLI M NaCI. proxidized
J.
J.
-200 0 200 ^00
POTENTIAL (mV SCE)
FIGURE 4 - CD of metastable pits äs a function of
Potential, chloride concentration, and preparation
method.
Prepolarization also promotes the growth of just one pit
when stable pitting does begin äs has been reported by
Hisamatsu, et ai.3 The growth of a single stable pit may in this
fashion be examined, an example of which is shown in Figure
5 for 0.1 M NaCI. During the initial stage of growth, the dissotu-
tion current of a stable pit increases approximately with t2 and
the calculated CD is approximately constant. At same point
the current decreases, but not back to the background level.
Rather, it continues to increase in an erratic fashion approx-
imately with t1<2, which has been observed frequently for
stable pit growth.13-17-18 It is clear that, until the first current
drop, stable pits behave like metastable pits.
Discussion
Four questions may be asked regarding the fundamental
nature of metastable pits: What is the mechanism of initia-
tion? What is the rate controlling step during the growth
stage? What is the rate controlling step during the decay
stage? What are the stability criteria? (Why do metastable pits
stop growing while stable pits continue to grow?) These ques-
tions will be presently addressed in turn.
Initiation
Analysls of metastable pitting has not, äs of yet, provided
conclusive evidence to support any of the suggested theories
of pit initiation.19 However, some observations are of interest.
Since stable pits are metastable when they first Start growing,
it is likely that the mechanism of pit inltiation is the same for
stable and metastable pits. Therefore, pits can initiate at
Potentials far below the Potential required for stable pit
growth. This indlcates (hat pit initiation is not the critical step
in the development of a pit.
It has already been reported that the nucleation rate of
metastabte pits decreases exponentially with time at a given
Potential.9-10 The rate also depends on potential,9 bulk chloride
concentration,20 pH,9 alloy composition,20 and surface pre-
treatment. Initiation occurs at specific susceptible sites an
the specimen surface. The exact nature of the passive film and
the weak Spots in the film are obviously dependent on many
factors.
Vol. 43, No. 7, July 1987
431
-\ 2s ^.
I l t
TIME
FIGURE 5 - (a) Stahle pit observed in 0.1 M NaCI at 500
mV SCE; (b) initial portion of stable pit growth.
Growth
It is possible to achieve same understanding of the rate
controlling step during metastable growth using the observa-
tions presented above. Dissolution is not under mass trans-
port control since the CD is potential dependent and constant
with time. The ohmic potential drop within the pit is also not
important äs this too should result in a decreasing CD. It has
been shown previously that ohmic potential drops within small
pits are insignificant.21
Metastable pits grow on SS with covers by undercutting
the passive film. These covers must be sufficiently porous to
allow flow of electrolyte into the pit and metal ions out of the
pit. Pores in the cover are unobservable by SEM and thus must
have diameters less than ~ 50 ^m. Large resistances and
therefore ohmic potential drops may develop äs a result of the
constriction of current flow by the pores. The resistance of a
porous pit cover may be calculated by assuming that the pore
radii, r?, pore lengths, lp, and density of pores, pp, are uniform
(Figure 6). The resistance resulting from a single pore, Rg, is
composed of two parts: the resistance through the pore and
the resistance outside the pore. Assuming a cylindrical pore
and using Newman's formula for the resistance of the primary .
current distribution of a small active site,22
FIGURE 6 - Schematic dlagram of pit with porous
cover.
RS=
XpTrri
4Xpfp
(1)
where Xj, and Xp are the conductivities of the bulk and pore
electrolytes. The total resistance of several pores acting äs
resistors in parallel, R(, becomes
R.= -^=
pp^
(
1
XpTfg
4Xbfp
)
(2)
where n? is the number of pores in a cover and FI, is the hole or
pit radius. According to Equation (2), the resistance varies with
the inverse square of the pit radius. Since the current in-
creases with the square of the radius, the ohmic potential
drop, IR, is a constant during the growth of a covered pit. This
could explain why the CD is constant.
Pretreating a specimen by oxidizing in air at 250 C for 8 h
alters the average CD of metastable pits. This indicates that
the exact structure of the pit cover is important in determining
the resistance and thus controlling growth. Similarly, the
curvature of the current-potential curves (Figure 4) may result
from an influence of applied potential on the passive film
structure. The effect of bulk chloride concentration, however,
is caused largely by a change in the electrotyte conductivity.
Using reasonable estimates for the pore shape and distribu-
tion, the resistance of the cover äs calculated from Equation
(2) could account for all or only part of the total measured
resistance. The control of dissolution may be mixed in nature
with Charge transfer also playing a rote.
Decay
It has been suggested above that a metastable pit stops
growing and the current decays when an opening appears in
the pit cover. The covers are very thin slnce they are remnants
of the pre-existing passive film. The large dissolution current
within the pit could result in an osmotic pressure on the cover,
which spans the pit mouth. Internal stresses present within
the passive film would further enhance rupture. Immediately
after the cover ruptures to form an opening, the ohmic resist-
ance decreases and the CD increases to a value i * (see Figures
3 and 4). Repassivation of the active pit surface quickly follows
äs the pit and bulk electrolytes mix.
Figure 7 shows ari~expanded view of the current decay
region of a typical metastable pit transient. It is clear that the
shape of the decay is different than the exponentiat form
typically found for a repassivation process. Furthermore, the
time for the current decay is much larger than that found by
other investigators for repassivation events.23-24 Marshall and
Burstein have reported equations to describe the repassiva-
tion of SSs in nonchloride-containing solutions.23 A curve
representing the current decay predicted by their equation for
an area equivalent to the pit size is also shown in Figure 7.
Although the chloride ions will have same effect, the current
measured during the decay of metastable pits is much larger
432
CORROSION-NACE
0.6-
<
^0.^
LÜ
Q:
cc
=3
u
0.2^
A metastable pit
ASDU53
Eqn8'/N^A
,I=AJO'3f1'1
10.02 s
TIME
FIGURE 7 - Current decay portion of a transient ob-
served in 0.1 M NaCI at 420 mV SCE after subtracting the
passive current The solid line is Equation (8) (itted to
the data points with the following fitted parameters:
ii = 3.72 A/cm2, ^ = 1.55/im, and k = 22.2 /tirts. The
dashed line is the repassivation equation reported by
Marshall and Burstein23 for 304 SS in a nonchloride-
containing solution o( pH 7.9 at 400 mV SCE.
than the predicted repassivation current. By assuming that
two dimensional spreading of the passive film along the pit
surface is the slow step and that the repassivation current is
small in comparison to the dissolution current of the remain-
ing unfilmed area, such large currents can be explained.
A mode! developed by Avrami25'27 has been used to study
the growth of oxide films on metals.28-29 This mode) considers
monolayer oxide patches to be growing with a CD at the patch
edges of i; on a surface dissolving with CD i,. A patch radius
will grow according to
r =
i,6t
QOX
(3)
where Qox is the Charge per unit area needed to form the oxide
and ö is the film thickness. The fractional coverage of filmed
area is
0, = Nirt, -)2t2 =-Kt2
(4)
where N is the number of patches per unit area. To account for
overlapping areas at later stages of growth, Equation (4) must
be modified to25-29
0,= 1-exp(-Kt2) (5)
The dissolution current of the unfilmed metal area is
l = iiAod - ö,)
(6)
where Ag is the original active surface area. The decreasing
rate of film spreading at langer times predicted by Equation (5)
is not observed in the current decay of metastable pits and
Equation (6) can not be easily fitted to the data if Equation (5)
is used to describe 6,. On the other hand, Equation (6) repre-
openlng
7
FIGURE 8 - Schematic diagram of a repassivating
metastable pit
^ ^0
=1.
30
^.
£1
^1
LU Ö
a ö
20
10
o stable pits, first decay
A metastable pits
200 1.00
POTENTIAL ( mV SCE )
FIGURE 9 - Decay rate constant, k, äs a function of
Potential for metastable pits in 0.1 M NaCI. Also shown
is the value o( k (itted to the first current decay of two
stable pits grown at 500 mV SCE. The bars indicate
Standard deviation.
sentsthedataquitewellusingEquation(4)withN = 1-Thisin-
dicates (hat only one nucleus of oxide is formed, and one
patch spreads aver the whole surface. When an opening forms
in a pit cover, it is likely that the diluted electrolyte first
reaches the pit bottom at same point under the opening to
nucleate the passive film. As the electrolyte under the remain-
ing edges of the cover dilutes, the film spreads aver the whole
pit surface.
Equation (4) may be modified to account for the
hemispherical geometry of metastable pits (Figure 8). Here, it
is assumed that the film nucleates at one point at the bottom
of the pit and grows at a constant rate, forming a bowl with
angle a increasing according 1o
kt
a=-r.T (7)
where k is a constant. Solving for 0, and substituting into
Equation (6), the following is obtained for the dissoiution cur-
rent of the unfilmed area in a hemispherical pit
l = ii27T^ cos
kt
1-h
(8)
Equation (8) fits well to the data, äs is shown in Figure 7, for
the decay portion of one transient. The kinetic parameter k is
Potential dependent for metastable p.its in 0.1 M NaCI (Figure
9). The film spreads faster at higher Potentials, indicating a
Vol. 43, No. 7, July 1987
433
[CI-] (mol/1)
FIGURE 10 - Factor \^T äs a function of bulk chloride
concentration_For real and artificial pits tested poten-
tiostatically, i Vr is plotted.
larger driving force for passivation. The fitted values of i, and
r», were atways dose to the values of i* and r,,, determined by
Integration of the current transients.
Pit Stability Criteria
It has been suggested that pit stability is related to the
precipitation of a salt film on the pit surface.19 The kinetics of
salt film precipitation may be represented by Sand's equation
for the case of constant CD, i, and one dimensional
diffusion,'1.12.30
iVr=
nFVrDAC
= B"
(9)
where r is the time for precipitation, D is metat ion diffusivity,
and AC is the differerice between the supersaturation concen-
tration and the bulk concentration.15 Experimental results with
artificial pit electrodes qf pure iron have verified Sand's equa-
tion and shown that i-s/r depends^nly on bulk chlpride con-
centration.11-12 Using average CD, T~, the value of V^/T for ar-
tificial pit electrodes held at constant potential is the same äs
i-s/r measured under constänt current, äs long äs the CO does
not vary greatly during the period before precipitation (Figure
10). For artificial pits in the dilute solution of 0.1 M HCI, the CD
varies considerably because the ohmic resistance within the
pit decreases äs the pit electrolyte becomes more concen-
trated. Only in this case are the values of 1~^ far from the
values determined galvanostatically. The values determined
for iron and SS electrodes are similar in all solutions.
It has been shown above that the CD during metastable
pitting is approximately constant with time. Therefore, Sand's
equation is applicable. The factor T-^/tp-, where tp is the pit
lifetime, is plotted in Figure 11 for a number of metastabte pit
transients observed in 0.1 M NaCI at Potentials from - 100 to
+530 mV SCE. The transients observed above Ep were ob-
tained by increasing the potential stepwise from below En.
Also ptotted in Figure 1 1 are four stable pits analyzed for V\/\
|(mV SCE)|-100to *<.20|^0|üOl500l53Ö
1-V
r^
E
u
\
<M
<n
<
12^
10.7|
Q-\
ka
^.-1
symbol
^L* l a l A
560
A a
a ^
+
A *
&
.0»-
a
- i7T
0 /. 8 12 16 20
FINAL RADIUS (um)
FIGURE 11 - rVtp" for metastable pits and stable plts
(filfed Symbols) in 0.1 M NaCI at a ränge o( potentlals.
at the point just below the first current decrease. The values of
i^/tp for metastable pits vary aver a wide ränge, indicating
that salt film precipitation is not responsible for the current
decay of metastable pits. The scatter of values for the stable
pits, however, is quite small, and a critical value of 1-Vr =
10.7 As1'2 cm-2 may be found to separate the stable f rom the
metastable pits.
The value of i^/r for artificial pit electrodes in a solutfon
containing 0.1 MCI- is ~4As1/2cm-2.11.12Strehblowhasdis-
cussed how this product should be larger for real pits because
of hemispherical diffusion away from very smal! pits.19 Sand's
equation has been jnodified to allow for hemispherical diffu-
sion resulting in19'31
air r
AC = --(1 - -=-)
nFD ' VDtT ' (10)
where a =3 is a factor derived by Vetter and Strehblow21 to ac-
count for the concave geometry inside a hemispherical pit.
Other equations describing hemispherical diffusion under
constant potential give the same results äs Equation (10) for
metastable pits.31-34 When the pit current increases with t2,' ttie
radius increases tinearty with time according to
r = Ciit
(11)
where c, was giyen above. Using Equation (11) and knowing
the values of 1^7 given by Equation(9) for one dimensional dif-
fusion, So, Equation (10) can be rearranged to give
«V,)3_^:,^,^,o
C1 ' ' ac^ (12)
Equation (12) can be solved for iVr using appropriate values of
Bg and D at different bulk chloride concentrations (see Figure
10).15 Also shown in Figure 10 are experimental results for ar-
tificial pit electrodes measured potentiostatically and
galvanostatically, äs well äs real pits in 0.1 M Cl- and6 M Cl-
solutions. The critical values determined for real pits are
somewhat less than those predicted by Equation (12) because
of the effect of the pit covers, which act äs diffusion barriers
and thus decrease the time required for precipitation.
The stability of pits in SS may be clearly understood using
Figure 12, where average CD is plotted against final pit radius
434
CORROSION-NACE
for the same pits shown previously in Figure 11. The dashed
lines are lines of constant time. It is ' clear that most
metastable pits have lifetimes less than 5 s. In this figure, the
boundary separating stable pits and metastable pits is of the
form i x r = constant. Equation (11) can be rearranged to
show that i x r is also constant when i^/r is a constant. This
boundary is, therefore, equivalent to the boundary shown in
Figure 11.
Pits grow initially with increasing radius at a constant CD
(a horizontal tine in Figure 12). If a pit grows lang enough at a
givenCDtocrosstheboundarywhereiVT = 10.7As1/2cm-2,a
salt film will precipitate on the pit surface and it will be stable.
On the other hand, if the cover ruptures before the pit crosses
the boundary, the pit will repassivate. Vetter and Strehblow
have discussed the necessity of a large ohmic potential drop
to stabilize the active dissolution occurring in a pit directly
next to a passive area.21 As lang äs an intact cover is present,
the ohmic resistance of the pores in the cover äs described by
Equation (2) will stabilize the dissolution. This dissolution is
only metastable since, if the cover ruptures, the ohmic poten-
tial drop disappears and the pit repassivates äs the pit electro-
tyte dilutes. If the cover ruptures after a salt film has precipi-
tated on the pit surface, the lost potential drop is compen-
sated by an increase in the salt film thickness and thus an in-
crease in the ohmic potential drop across the salt film.35 As a
result, the pit cover is no langer critical to pit stability.
The properties of the pit cover are thus very important in
attaining stability. A strong, resistant cover facilitates pit
stability, while a weak or stressed cover hinders it. Specimen
pretreatment affects the achievement of pit stability through
its influence on the properties of the passive film. Mechanical
abrasion followed by cleaning, a treatment that is perhaps
similar to those used in the field, is a beneficial pretreatment.
The passive film is stressed because of the deformation of the
metal surface, and the pit covers easily rupture. Oxidizing at
250 C builds a more stable film that enhances the precipitation
of a salt film by being resistant to rupture, even .though it
decreases the nucleation rate of pits. As a result, preoxidation
reduces Ep. Electropolishing before oxidation apparently
removes the damaged surface layer and allows a much more
stable oxide film to form. Isaacs and Kissel observed the iden-
tical effects of pretreatment on the lifetimes of pits growing in
SSs by watching the pits with a scanning reference
electrode.36 They found that the pit lifetime after thermal ox-
idation of the specimen increased slightly over the case where
the specimen was only abraded. Electropolishing before ox-
idation increased the pit lifetime by a factor of almost 500.
Precipitation of a salt film will cause a decrease in the
current äs the concentration at the surface decreases from ttie
supersaturation to the saturation value. The first current drop
during stable pit growth can be approximately fitted to Equa-
tion (8). However, the kinetic constant k is much less than what
is expected for passive film spreading at the same potential
(see Figure 9). This current drop tikely resutts from the salt film
precipitation, which marks the transition to stabte pitting. The
erratic current behavior after that point may result from
naturat convection in the pit. In the 6 M LiCI + 0.001 M NaOH
solution, the current drop is not äs pronounced and the current
increases afterward in a more uniform fashion.
It is interesting to note that the CDs calculated during the
metastabte stage of the stable pits observed in this work were
lower than those of many metastable pits that preceded them
(Figure 12). Larger CDs could enhance rupture of the covers by
causing a larger osmotic pressure inside the pit. Therefore, the
major influence of applied potential is probably the effect it
has on the properties of the passive film rather than its effect
on pit CD.
The replacement of a passive film on SS by a salt film can
be seen to occur in two Steps, both of which are kinetically-
hindered. First, the passive film must break down to initiate a
pit. This breakdown can only occur at susceptible sites in the
passive film since the film is, in general, very protective. Film
(mVSCE)|-100to^20
0 4 8 12 16 20
FINAL RADIUS ( |im)
FIGURE 12 - Average CD vs final radius for metastable
pits and stable pits (filled Symbols). Dashed tines in-
dicate constant time.
breakdown, äs evidenced by metastable pit transients, has not
been observed at potentials lower than -210 mV SCE, which
is approximately equal to ES(. Therefore, pitting can only ini-
tiate at Potentials above Es». Once a pit initiates and Starts
growing, achievement of the conditions necessary for satt film
precipitation and thus stable growth is also difficult. The com-
bination of the CD and the mechanical properties of the
passive film at the specific site must be such (hat the pit can
remain covered for the time required for the protection of a salt
film.
Conclusions
1. Metastable pitting current transients, resulting from
passive film breakdown and repair events, can occur at poten-
tials äs low äs -210 mV SCE.
2. The CD is approximately constant during the lifetime
of a metastable pit. Growth is controlled largely by the ohmic
resistance of the porous pit cover.
3. When their covers rupture to form openings,
metastable pits repasslvate. Two dimensional spreading of
the passive film is the slow step in the repassivation process.
4. If the pit cover remains intact for a period lang enough
to allow for salt film precipitation on the pit surface, pit growth
is stabilized.
5. A strong passive layer facilitates pit stability while a
weak or stressed layer hinders it.
Acknowledgment
The authors acknowledge the Schweizerischer National-
fonds zur Foerderung der wissenschaftlichen Forschung for
supporting this research through its program, Rohstoff- und
Materialprobleme. »
References
1. W. Schwenk, Corrosion, Vol. 20, No. 4, p. 129t, 1964.
2. P. Forchhammer, H.-J. Engel), Werk. und Korr., Vol. 20, No.
1, p. 1, 1969.
3. Y. Hisamatsu, T. Yoshii, Y. Matsumura, Localized Corro-
sion, NACE-3, R. W. Stachle, B. F. Brown, J. Krüger, A.
Vol. 43, No. 7, Juty 1987
435
Agrawal, £ds.. National Association of Corrosion Engi-
neers, Houston, Texas, Vol. 3, p. 427,1974.
4. N. Sato, J. Electrochem. Soc., Vol. 129, p. 255, 1982.
5. D. E. Williams, M. Fleischmann, J. Stewart, T. Brooks,
Electrochemical Methods in Corrosion Research,
M. Duprat, Ed., Materials Science Forum, Vol. 8, p. 151,
1986.
6. G. Daufin, J. Pagetti, J. P. Labbe, F. Michel, Corrosion,
Vol. 41, No. 9, p. 533, 1985.
7. F. Hunkeler, G. S. Frankel, H. Boehni, Corrosion, Vol. 43,
No. 3, p. 189, 1987.
8. L. Stockert, F. Hunkeler, H. Boehni, Corrosion, Vol. 41,
No. 11, p. 676, 1985.
9. L. Stochert, H. Boehni, "Proc. 8th European Corr. Cong.,
Centre Francais de la Corrosion, Paris, Vol. 2, p. 22, 1985.
10. R. Holliger, H. Boehni, Proc. Symp. Computer Aided Ac-
quisition and Analytical Corros. Data, Electrochem. Soc.,
Pennington, New Jersey, Vol. 85-3, p. 200,1984.
11. H.-H. Strehblow, J. Wenners, Electrochem. Acta, Vol. 22, p.
421, 1977.
12. H. C. Kuo, D. Landolt, Corros. Sei., Vol. 16, p.915,1976.
13. l. L Rosenfetd, l. S. Danilov, Corros. Sei., Vol. 7, p. 129,
1967.
14. H. S. Isaacs, Localized Corrosion, NACE-3, R. W. Staehle,
B. F. Brown, J. Krüger, A. Agrawal, Eds., National Asso-
ciation of Corrosion Engineers, Houston, Texas, Vol. 3,
P.158,1974.
15. F. Hunkeler, H. Boehni, "Mass Transport Controlted Pit
Growth on Stalntess Steels and Nickel," Int. Conf. on Cor-
ros. Chem. within Pits. Crevices and Cracks, National
Physics Laboratory, Middlesex, England, 1984.
16. H.-H. Strehblow, J. Wenners, Zeit. fuer Physik. Chemie
Neue Folge, Vol. 98, p. 199,1975.
17. F. Hunkeler, H. Boehni, Passivity of Metals and Semi-
conductors, M. Froment, Ed., Elsevier, Amsterdam, Hol-
land, p. 655, 1983.
18. R. C. Newman, E. M. Franz, Corrosion, Vol. 40, No. 7, p.
325,1984.
19. H.-H. Strehblow, Proc. 9th Int. Cong. on Metallic Corros-,
National Research Council, Ottowa, Canada, Vol. 2, p. 99,
1984.
20. L Stockert, hl. Boehni, submitted to die Umschau.
21. K. J. Vetter, H.-H. Strehblow, Localized Corrosion, NACE-3,
R. W. Staehle, B. F. Brown, J. Krüger, A. Agrawal, Eds., Na-
tional Association of Corrosion Engineers, Houston,
Texas, Vol. 3, p. 240, 1974.
22. J. Newman, J. Electrochem. Soc., Vol. 113, p. 501, 1966.
23. P. l. Marshall, G. T. Burstein, Corros. Sei., Vol. 23,p.1219,
1983.
24. C. GabrieDi, F. Huet, M. Keddam, R. Oltra, J. C. Colson,
Proc. Int. Symp., M. Pourbaix, R. P. Frankenthal, J. Krüger,
Eds., Electrochem. Soc., Pennington, New Jersey, Vol.
84-9, p. 522, 1984.
25. M. Avrami, J. Chem. Phys, Vol. 7, p. 1103, 1939.
26. M. Avrami, J. Chem. Phys, Vol. 8, p. 212, 1940.
27. M. Avrami, J. Chem. Phys, Vol. 9, p. 177, 1941.
28. H.-J. Raetzer-Sheibe, H. Buhl, Werk. und Korr., Vol. 30, p.
846,1979.
29. H.-J. Raetzer-Scheibe, Proc. 8tti Int. Cong. on Metaltic
Corros., Deutsche Gesellschaft fuer Chemisches Ap-
paratewesen. Chemische Technik und Biotechnotogie
e.V., Frankfurt, West Germany, Vol. 1, p. 212,1981.
30. R. Alkire, D. Emsberger, T. R. Beck, J. Electrochem. Soc.,
Vol. 125, No. 9, p. 1382, 1978.
31. A. Ruis, S. Polo, J. Llopis, Anales de Fisica y Quimica,
Series B, Vol. 45, p. 1029, 1949.
32. A. J. Bard, Anal. Chem., Vol. 33, No. 1, p. 11,1961.
33. P. J. Lingane, Anal. Chem., Vol. 36, No. 9, p.1723, 1964.
34. C. R. Ito, S. Asakaura, K. Nobe, J. Electrochem. Soc., Vol.
119, No. 6, p. 698, 1972.
35. F. Hunkeler, A. Krolikowski, H. Boehni, submitted to Elec-
trochim. Acta.
36. H. S. Isaacs, G. Kissel, J. Electrochem. Soc-, Vol. 119,
No.12,p.1628, 1972.
436
CORROSION-NACE