1991 NASA/ASEE SUMMER FACULTY FELLOWSHIP PROGRAM
JOHN F. KENNEDY SPACE CENTER
UNIVERSITY OF CENTRAL FLORIDA
DEVELOPMENT OF AN ACCELERATED TEST METHOD FOR THE DETERMINATION
OF SUSCEPTIBILITY TO ATMOSPHERIC CORROSION
UNIVERSITY AND DEPARTMENT:
John R. Ambrose, Ph.D., P.E.
University of Florida
Department of Materials Science
Materials Science Laboratory
Failure Analysis and
Coleman J. Bryan
July 26, 1991
University of Central Florida
NASA-NGT-60002 Supplement: 6
I would like to express my appreciation for being selected to
participate in the 1991 NASA/ASEE Summer Faculty Fellowship
program, to Ray Hostler [University of Central Florida] and Mark
Beymer [NASA] for their capable administration Q f the program,
and to my NASA colleagues Cole Bryan and Rupert Lee for their
assistance and encouragement. Special thanks must go to Jim Jones
for his assistance in expediting the various chemical analyses
which 1 requested and to Helein Hitchcock, Hae Soo Kim, Stan
Young and Bill Carman for their capable technical assistance* And
to Lee Underhill, without whose assistance in finding the various
apparatus with which to set up my experiments i would never have
even begun this project, I offer a nomination for the "Golden
And most of all, my gratitude to the Materials Testing
Laboratory staff - fo Steve Mcpanel , whose progress from a former
student to a capable failure analyst makes this teacher proud, to
Scott Murray for his patience and cooperation in allowing this
curious child of 51 years to tag along, seeing things again for
the first time , and most especially to Peter Marciniak, whose
instruction and encouragement allowed me to find satisfaction in
doing things undone for some 25 years . May Scott always find one
last jelly doughnut each morning and Peter his film for "just one
...and to KSC and its surrounding Merritt Island Wildlife
Refuge for being what and where they are* 1 can sincerely say
that no summer has meant more to me or will mean more as time
goes by . Even though endless summers must always end, thanks for
the memories ...
The theoretical rationale is presented for utilization of a
repetitive cyclic current reversal vol tammetric technique for
characterization of localized corrosion processes, including
atmospheric corrosion. Applicability of this proposed
experimental protocol is applied to characterization of
susceptibility to crevice and pitting corrosion, atmospheric
corrosion and stress corrosion cracking. Criteria upon which
relative susceptibility is based have been determined and tested
using two iron-based alloys commonly in use at NASA/KSC - A36, a
low carbon steel and 4130, a low alloy steel.
Practicality of the procedure has been demonstrated by
measuring changes in anodic polarization behavior during high
frequency current reversal cycles of 25 cycles per second with
1 mA/cm* current density amplitude in solutions containing C'l’*.
The results demonstrated that, due to excessive polarization
which affects conductivity of barrier corrosion product layers,
A36 was less resistant to atmospheric corrosion than its 4130
counterpart - behavior which has also been demonstrated during
Based on an analysis of factors which are known to contribute
to the overall corrosion process, gal vanostatic electrochemical
procedures show greater promise for creating an environment
during accelerated testing which more closely simulates natural
environments than do conventional potentiostatic methods.
Similarly, since both anodic and cathodic reactions occur
simultaneously on the surfaces of freely corroding metals, it is
necessary to periodically reverse the direction of current flow.
On the basis of these modelling simulation requirements, the
Repetitive Current Reversal Voltammetry technique has been
Using a galvanostat to supply current, and a storage
oscilloscope to measure potential transients as a function of
time, a methodology and associated predictive criteria have been
generated using two representative alloys found in common use at
NASA\KSC - A36 steel and 4130 low alloy steel. By trial and
error, the optimum combination of impressed current amplitude and
current reversal cycle frequency has been found which will
differentiate environmental degradation behavior for these two
metals - +/- 2 milliamperes current amplitude with a 25 cycles
per second current reversal frequency.
Results to date suggest that susceptibility to various
localized corrosion forms of attack may be associated with "over-
polarization", that is, too high an electrical resistivity for
insoluble corrosion product barrier layers which form at the
metal/environment interface. In the presence of aggressive ions
such as the chloride ion, these high potential drops or fields
across the barrier layers "draws M these aggressive anions into
the film, causes its decrease in resistivity and eventually leads
to the inability of passivity being maintained.
This report summarizes many of the insights, opinions and
perspectives of the author which may prove useful to others as
they contemplate developing experimental procedures for the study
of environmental degradation of materials.
TABLE of CONTENTS
1.1 EXPERIMENTAL BASIS FOR EXPERIMENTAL PROCEDURE
1.1.1 CRITERIA FOR CORROSION SUSCEPTIBILITY
1.1.2 EXAMPLES OF CRITERIA APPLICATION
1.1. 2.1 ATMOSPHERIC CORROSION
1.1. 2. 2 CREVICE CORROSION
1.2 PROPOSED EXPERIMENTAL PROTOCOL
2.0 EXPERIMENTAL - MATERIALS AND EQUIPMENT
3.0 EXPERIMENTAL RESULTS AND DISCUSSION
3.1 LOW FREQUENCY CURRENT REVERSAL CYCLING
3.2 HIGH FREQUENCY CURRENT REVERSAL CYCLING
3.2.1 POLARIZATION IN STOCK SOLUTION
3.2.2 EFFECT OF CHLORIDE ADDITIONS
5.0 RECOMMENDATIONS FOR FUTURE WORK
LIST OF ILLUSTRATIONS
Schematic showing physical sense of current .
Schematic representation of the Separation of Siicide
and cathode in an electrochemical cell utilizing
potentiostatic control .
Schematic representation of limiting cases in
polarization behavior - metal dissolution v .
ihSolUble corrosion product .
Schematic representation of how increased solute
concentrations can be uSed to simulate evaporation of
condensed moisture 1 ayeis .
Schematic representation of how a crevice leads to
1 ocal ized breakdown of passivity .
Photograph of machined samples assembled into the bi-
specimeh electrode configuration.
Photograph of electrochemical cell used.
Schematic of experimental procedure With reSpect to
the RcRv control variables .
Conventional Polarization Plot [ I v . V] showing the
effects of oxygen depletion on corrosion kinetics .
RCRV plot for A36 alloy in air-saturated stock.
Polarization kinetics fof A36 alloy in the stock
solution; experimental parameters shown oh graph .
Replot of Figure 11 , plotting -Lh(voits) v . time .
A comparison of polarization kinetics for the A36 and
4130 alloy steels.
A comparison of polarization kinetics for A36 and
4i30 Steels With chloride ions present .
A comparison of polarization behavior for A36 with,
and without, chloride ions.
Polarization behavior of 4130 with and without
Effect of 5 minutes exposure to chloride ions on the
polarization behavior of A36 and 4130 alloys.
The vast majority of electrochemical test methods used in
corrosion science experimentation derive from potentiostatic
polarization techniques - i.e. a measurement of current required
to maintain the potential of a metal at some reference value. As
the name implies, the electronic instrumentation fixes potential
by regulation of current. As a consequence of this functionality,
corrosion rates [current] vary with time as the externally
maintained potential remains fixed [potentiostatic] or is varied
[potentiodynamic sweep] during the course of the experiment.
A number of "accelerated" corrosion test procedures are based
upon potentiostatic polarization procedures - linear current
versus applied potential relationships or "polarization curves",
low amplitude linear polarization for determination of
"polarization resistance", low amplitude cyclic polarization or
"AC impedance" and linear cyclic voltammetry or "LCV" to name but
a few. Although a great deal of information has been obtained
through such experimentation, some of it even useful, attempts to
correlate long term corrosion behavior with the results of such
procedures have often been fraught with inconsistencies,
irreproducibi 1 ities and difficulties with interpretation. The
problem, as I see it, is that the basis for these techniques -
potential or polarization control - is not a suitable or accurate
model for the way metals behave in "real life" situations. The
true basis is, in fact, just the opposite.
When a metallic material is exposed to a corrosive
environment, there develops, rather quickly, an electrochemical
potential across the metal /envi ronment interface which serves to
drive those chemical reactions which contribute to that
particular electrochemical potential. Corrosion processes consist
of one or more anodic or oxidation reactions which, in a physical
sense, carry current out from the surface of the metal and into
the environment. Anodic current is balanced by a cathodic current
of equal amplitude which flows into the metal from the
environment, carried by one or more reduction reactions. The
respective current flows produce, due to a electrical resistance
which opposes the flow of current [polarization resistance], a
shift in potential or polarization equal to the product of the
current flow and the resistance through which that current flows.
The resulting so-called "mixed" or corrosion potential of the
material is positioned between the reversible potentials for the
various electrochemical processes, the displacement being
dependent upon the polarization resistance for the various
contributing electrochemical reactions. Mathematical models of
electrochemical kinetics [e.g. Butler-Volmer equation] have been
used to characterize the relationship between current and applied
potential, resulting in what have come to be known as "Evans
Diagrams". A schematic representation of current flow across the
metal /environment interface is shown in Figure 1.
uniform anodic current density
-J(A): M 0 =M +Z +z«-
uniform cathodic current density
-J(C): 0 2 +2H^)=40H"
insoluble corrosion product
uniform corrosion current density
- electrical resistance
Figure 1 - Schematic showing physical sense of current
direction - anodic current leaving metal and cathodic current
The point to be made is simply this: Since the corrosion
potential of the material is determined by polarization,
specifically polarizibility of the material as influenced by the
contributing electrochemical reactions, then it is the flow of
current which controls potential, not the other way around.
Fixing or controlling the potential of the material will cause
current flow to conform to the degree of polarization - if there
is any change in resistance to current flow which occurs as a
natural consequence of the chemical reactions taking place during
service exposure, then current will change accordingly. One such
occurrence, the evolution of insoluble reaction products which
block the flow of current across the interface, is also shown in
Figure 1. Under "natural", free corroding conditions, it would
seem that development of insoluble corrosion products with
whatever inherent intrinsic electrical and ionic transport
resistivity they possess would cause the corrosion potential to
shift in one direction or another with respect to a film-free
metal/electrolyte interface. Since many metal oxides and
hydroxides [iron based, e.g.] provide greater resistance to the
anodic current flow [ion transport control], the corrosion
potential under natural conditions tends to shift toward more
positive values. When experimental simulation of corrosion
processes are under potentiostatic control, the potential, of
course, is fixed. There are two consequences of this control.
One, the corrosion current would not be expected to be the same
as under natural control conditions, resulting in lack of
correspondence with long term corrosion rates. Secondly, since
the potential of the metal determines what electrochemical
reactions can occur, differences in potential may result in
altogether different reaction products being generated. Since it
is the reactions products which determine the nature of insoluble
reaction products, and it is the insoluble reaction products
which affect both degree of polarization and amplitude of the
corrosion current, the end result is most likely a continual
divergence of experimental corrosion rates from the real ones.
A second negative characteristic associated with
potentiostatic control of electrochemical processes derives from
the configuration of the system used to control the process being
studied. Under anodic polarization - when the potential of the
material is driven to higher [more positive] values than its
corrosion or steady-state value - the current which produces this
polarization leaves the material [working electrode], flows
through the electrolyte and enters a remote auxiliary [counter]
electrode. In a freely corroding metal, on the other hand,
current leaves and
enters the same exposed
surface, albeit not
necessarily at the same
point on that surface.
What this means is that
cathodic and anodic
reactions do not occur
concurrently on the
same metal surface -
more significantly, the
reaction products of
processes are not
produced at a common
surface. Since most
products which form in
contain metal cations,
hydroxide ions and
anode from cathode in the electrochemical cell changes the
localized chemistry associated with the natural or free corrosion
process [Figure 2].
anions [SO, , Cl"
r*r\ "2 _ i. _ T - 1
1 . 1 THEORETICAL BASIS for PROPOSED EXPERIMENTAL
Measurement of degree of polarization which result from
impressed current is riot a new technique, having been used for
some time as an analytical chemistry technique -
chronopotentiometry , Stripping voltammetry, etc. It has not been
used to any great extent in corrosion science applications. It
would seem, however , to be most appropriate in measuring the
behavior of a material in response to flow of current across the
metal /environment interface . By impressing a constant
[gal vanostatic] current between an inert electrode [platinum
counter] and the material being characterized [working
electrode] , the potential change or polarization can be measured
as a function both of time and of the amplitude of the impressed
current . In order to simul ate "natural " conditions , the current
direction should be regularly reversed in order to develop
concentrations of both kinds of reaction products at the metal
interface - anodic and cathodic . Thus evolves the name of the
technique - "CYCLIC CURRENT REVERSAL VOLTAMMETRY" . The
experimental variable to be measured will be the rate of change
in material potential , or "polarization rate".
1.1.1 CRITERIA for CORROSION SUSCEPTIBILITY
Polarization in response to current flow can be of three
types , individually or in combination :
ii a potential drop across an ohmic resistance . This
polarization is characterised by a V-IR response , and is
Virtually time independent - i.e. instantaneous
polarization with application of current . Capacitance or
interfacial charging processes are included in this
category as is the voltage drop across the electrolyte
between the working and counter electrodes .
ii . polarization due to the resistance to charge transfer
acrOSs the electrified interface - i.e. so-called
H Tafel" overvoltage . This kind of polarization is
characterized by a 1 ogarithmic dependence upon current
flow - the "Taf el Equation":
where n ; = degree of polarization produced by current I
and B- = charge transfer resistance,
iii. the potential drop across an insoluble reaction product
or film which forms at the metal/electrolyte interface.
The -degree of polarization is a function of the
resistivity of the reaction product, the polarization
rate is a function of the nucl eation/growth kinetics of
the deposition process. It is this polarization process
with which we will be most interested.
A schematic representation of the polarization extremes -
polarization resistance during active metal dissolution versus IR
resistance across an insoluble corrosion product - is shown in
Under conditions of
reversal , any change in
either degree or in
rate of polarization
signifies changes in
the one of the three
above. Of the three,
only the third should
provide any significant
contribution. Thus, by
changes, we should be
able to establish
criteria for evaluating
stability of a
particular material in
a given environment.
films grow on bare or
air-formed film covered
metal substrates, there should be a regular increase in degree of
polarization with each consecutive anodic cycle. Furthermore, the
degree of polarization should progressively decrease as well, if
the protective film is becoming more and more protective. Any
change in this trend will be interpreted as an indication of
development of instability in the system - a loss in ability of
the system to resist the corrosive actions of the environment.
limiting cases in polarization
behavior - metal dissolution v,
insoluble corrosion product.
X.1.2 EXAMPLES of HOW CRITERIA. ARE APPLIED
NASA/KSC is interested in predictive capability with respect
to a large number of environmental degradation of materials
problems. Specifically, accelerated test methods for predicting
long term resistance of coated and uncoated metals to atmospheric
corrosion are needed for rapid, early screening of candidate
materials. Susceptibility to crevice corrosion, pitting and
stress corrosion cracking - all localized forms of corrosion
attack •* is another area to which this approach can be applied in
developing accelerated test methodology.
X. 1.2.X ATMOSPHER I C CORROSION
Atmospheric corrosion can occur in two forms - at elevated
temperatures by direct reaction between material and corrosive
gasses, and through electrochemical means by interaction between
material and condensed aqueous liquid films. KSC is considered
one of the most, if not the most, severe locations for
atmospheric corrosion in the world. This is due to its proximity
to the ocean and to the corrosive nature of combustion products
from Space Shuttle solid fuel rocket boosters [SRBs] .
During atmospheric corrosion under condensed moisture films,
the actual corrosion process naturally takes place only when the
film is present upon the surface - long term exposure testing as,
for instance, is performed at NASA-KSC's seaside test site,
provides an integrated measure of atmospheric corrosion behavior
under those conditions where condensation is present. Ocean
spray, rain, condensation of dew, etc. provide the natural
environment, The process of Condensation and evaporation actually
increases the severity of the attack over what it would be under
constant exposure conditions - and, as such, provides a clue as
to an appropriate acceleration procedure.
As Figure 4 demonstrates, although initial, as-condensed
electrolyte concentrations may fall below a critical level to
initiate corrosive attack [<5xlO' J M Cl’ 1 for carbon Steels, e . g . ] ,
a critical level can be achieved during evaporation. During test
site exposures, this condition is only reached during some
fraction of the total exposure times. Rainwater and dew
condensation tend to return the process back to square One.
It is therefore proposed that one possible acceleration test
would be to increase chloride [or any Other aggressive species]
by aliquot additions to a stock solution while the metal
undergoes repetitive current reversal voltammetry. The time or
number of cycles to a point where polarization behavior is
affected would be a criteria for comparison of material
susceptibi 1 ity .
Assume some critical [minimum]
concentration of chemical species X
causes resistivity of film to decrease.
If C(X) In environment Is 0.5C(*),
no problem, but..
if C (X) increased, or
if solvent Is evaporated,
we have a problem -
film begins to thin
Figure 4 - Schematic representation of how increased solute
concentrations can be used to simulate evaporation of
condensed moisture layers.
1.1. 2. 2 CREVICE CORROSION
Metals with the capacity to passivate themselves with barrier
surface films [aluminum, stainless steels, nickel and titanium
alloys] require a finite hydroxide ion concentration [from the
cathodic reduction of oxygen] to retain their passivity.
Occluding any area of a passive metal surface - with a crevice,
for example, restricts oxygen transport to the creviced area, and
leads to breakdown of passivity and active corrosion [Figure 5].
Degree of susceptibility of a material to this form of localized
corrosion might also be rapidly characterized by an accelerated
test combining Repetitive Current Reversal Voltammetry [RCRV]
with oxygen depletion from the electrolyte.
It is proposed that an accelerated method for determination of
susceptibility to crevice corrosion would involve the effect of
oxygen depletion on RCRV while deaerating the electrolyte.
1. Uniform "general" corrosion
- "passive" film on surface
t h mm
2. Add crevice -
Impede replenishment of
cathodically reducible species -
decrease cathodic current
Inside the crevice,
some decrease In anodic current
3. With lower [OH ], passive film 1
under crevice dissolves - WBBS&Bk
anodic current Inside crevice Increases -
anodic current outside crevice decreases.
Figure 5 - Schematic representation of how a crevice leads to
localized breakdown of passivity.
1 . 2 PROPOSED EXPERIMENTAL PROTOCOL.
a. Reversal frequency and current amplitude experimental
b. Determination of the effect of oxygen depletion on RCRV
c. Determination of the effect of increased chloride ion
concentration on RCRV kinetics.
d. Comparison of RCRV kinetics for two different materials for
the purposes of distinguishing behavior - procedural
e. Determination of effects of other atmospheric contaminants
on RCRV kinetics.
f. Development of a "standard" test medium composition and
methodology for characterization of atmospheric corrosion
MATERIALS and EQUIPMENT
Feasibility studies were performed using two relatively common
ferrous-based alloys in use at NASA/KSC - A36 [basically a carbon
steel] and 4130 [a low alloy steel]. Their compositions are given
in Table I .
Based upon these
compositions, the A36
alloy would be expected
to show the lesser
degree of resistance
Specimens of the two
materials were machined
assembled into the
shown in Figure 6. This
selected for a> number of reasons - immersion of the tip end of
the assemoly beneath the electrolyte precludes the need for
isolation coating and their inherent tendency of providing
crevices for crevice corrosion attack; multiple material
Table I - Elemental Compositions of Iron
Alloys Selected for this Study.
A36 : 0.29 C(max), 0.8-1.20 Mn,
4130: 0.28-0.33 C, 0.40-0.60 Mn,
0.15-0.30 Si, 0.8-1.10 Cr,
Figure 6 - Photograph of machined samples
assembled into the bi-specimen electrode
experimental runs can be made consecutively in the same
electrolyte without concern for changing, cleaning and
repositioning electrodes between experiments.
The electrochemical cell used in these studies was designed
and constructed prior to my arrival at KSC - it is shown in
Figure 7. It features, in addition to conventional items
V i r ;
Figure 7 - Photograph of Electrochemical Cell used
in these studies.
[reference electrode well/luggin capillary, isolated counter
electrode well, working electrode access and gas dispersion tube
input], two separate vessels for addition of test solutions.
Electrochemical measurements were obtained using a Tacussel
BiPad potentiostat/gal vanostat coupled with a signal generator.
The waveform selected for these experiments was a square wave
function in which the input variables were current amplitude and
cycle frequency [Figure 8].
Current amplitudes used here were as high as +/- 2
milliamperes with reversal frequencies from 10 milliseconds to 10
seconds. Polarization kinetics were measured using a Tektronics
storage oscilloscope. Data was obtained by analysis of
photographs taken of the retained screen image.
The standard or stock solution selected was 0 . 1M KH^PO^
adjusted to pH=7 with 0 . 1M NaOH. The basis for this selection was
that phosphate solutions represent in innocuous environment
insofar as steels are concerned. 0 . 1M NaCl was added when effect
of aggressive ion concentration was to be evaluated. Unless
otherwise noted, all solutions used were air-saturated and
Prior to running any experiment, specimens were mechanically
polished using 100 grit silicon carbide paper, rinsed with
distilled water, then dried with absolute ethanol rinse and a
blast of warm air before assembling the electrode and placing the
assembly in the
Solution was added from
one of the reservoirs
to a level just above
the bottom of the
Rest or corrosion
recorded once they had
reached some steady
Appfed Current, mAnps
Figure 8 - Schematic of Experimental
Procedure with respect to the RCRV
3.0 EXPERIMENTAL RESULTS and DISCUSSION
In Figure 9, for
those who derive
appears a pair of
for A36 steel in air-
saturated and deaerated
[with N 2 ] KHjPOi
solution [pH 7j. The
only feature of
interest is the graphic
demonstration of the
ability of the A36
alloy to passivate in
this solution, whether
0 2 is present or not.
Figure 9 - Conventional Polarization
Plot [I v. V] showing the Effects of
Oxygen Depletion on Corrosion Kinetics
of A36 Alloy.
Notice that the
requires a critical anodic current of better than 500 mA to
achieve passivation at potentials around -600 mV v SCE. Also note
that the critical current requirements are higher in the
When the same experiment is performed using the low frequency
RCRV method, with a +/- 10 second frequency and a +/- 1.25 mA
current amplitude, passivation is produced by the 5th cycle with
one four hundredth [1/400] of the anodic current required for the
potentiostatic polarization method [Figure 10].
These results, as contained within Figures 9 and 10, serve to
endorse my argument criticizing the reliance on potentiostatic
polarization procedures for prediction of corrosion behavior.
Their use, particularly in determining , from Tafel slope
measurements, corrosion current densities are misleading to say
RCRV measurements of the 4130 alloy in the stock solution
using low frequency current reversals were similar to the results
displayed in Figure 10, only the number of cycles to produce
passivity were noticeably less - by the 2nd cycle. However,
experimental results for both alloys in this stock solution were
extremely irreproducabl e . The problem lies with the
electrochemical history of the process. Length of time at open
circuit was of prime consideration - after passivation was
achieved with either alloy, it became impossible to distinguish
behavior between the two alloys. Specimen pretreatment prior to
the experiments was also an influencing factor - surface
preparation and finish would affect the results.
Addition of solutions containing chloride ion resulted in
inability of both alloys to achieve passivation - the film
resistance appeared, however, slightly greater for the 4130 alloy
than for the A36. These results also were difficult to reproduce.
The point is, unless a definitive criteria can be developed which
is independent of operator procedure, the technique will never be
applicable as a standard test method.
One problem with the electrochemical cell was that, with its
present design, aliquots of solutions containing the aggressive
species could not be rapidly mixed with the stock solution. This
suggests a design change such that second solution aliquots be
added through the gas dispersion valve circuit - in this way,
solutions can be rapidly mixed during RCRV sequences.
3.1 HIGH FREQUENCY CURRENT REVERSAL CYCLING
Increasing the cycle frequency increases the stability of the
system - at least with respect to reproducibility in measurement.
In Figure 11 is plotted polarization kinetics for a single
typical cycle when a
±10 mA amplitude
current is impressed
for 10 mS cycles . A
number of relevant
observations are worth
noting - When forward
and reverse current
amplitudes are equal
[current balanced], the
median potential read
from a slow response
voltmeter is remarkably
constant. So constant,
in fact, this procedure
might be considered as
a standard test method
for measurement of
Changes in current
amplitude results in a
shift in median
potential to higher or lower values, respectively.
When current amplitudes are unbalanced, the median potential
shifts in the direction of the current imbalance. This procedure
should also be studied further in that the rate of polarization
A36 alloy in the stock solution;
experimental parameters shown on graph.
shift appears to reflect the degree of control of the
contributing anodic and cathodic electrochemical kinetics on the
overall corrosion process.
Replotting the data of Figure H using the negative natural
logarithm of the cell voltage, produces the graph shown in Figure
12. Only the anodic polarization portion of the overall cycle is
shown. Although the plot is not a linear function of the plotting
variables over the whole range, it fits the later portion of the
cycle quite well . Neither an exponential or parabolic plot would
do as well. Without drawing any conclusions as to the mechanistic
implications to be derived from this, experimental results will
be presented in this format, since it is easier to see variations
when comparing different alloys and when chloride ion solutions
are introduced into the stock solution.
3.2.1 POLARIZATION BEHAVIOR IN O.IM KHjPO,^
- In (Cell Voltage) [v. SCE] stock solution
Elapsed Time, msec
Figure 13 - A comparison of polarization kinetics for the A36 and
4130 alloy steels in the stock solution [0.1 KHjPO^ , pH 7].
Current amplitude was ±2 mA, cycle reverse frequency was 120 mS.
In Figure 13 are compared polarization kinetics for our two
reference materials, the A36 carbon steel and the 4130 low alloy
steel. The operational variables were selected by trial and error
to produce the optimum conditions for producing the greatest
difference in behavior between the two - 20 mS current reversal
frequency with a 2 mA current amplitude. What is interesting to
note is that the A36 alloy evidences a larger degree of
polarization than does the 4130 steel. Remember, now, the 4130 is
more resistant to atmospheric corrosion than the A36.
Furthermore, the IR drop [film electrical resistance] is
substantially greater for the A36 material. Intuitively, you
might think that this is contradictory in terms of the known
corrosion resistance of the low alloy steel. It must be
emphasized that corrosion resistance lies not necessarily with
electrical resistivity, but with passivity - with the resistance
to ion transport through the barrier layer . The proof of the
pudding, so to speak, Will lie with ability of a barrier layer to
retain its ionic resistivity in the presence of aggressive or
corrosive chemical species. Whatever, we can say at this point
that polarization behavior of the A36 alloy is distinguished by a
larger electrical resistance which, in the presence of impressed
anodic currents , results in larger [more positive] degrees of
polarization - larger electrical fields or potential drops across
the barrier corrosion product layer.
- lri(C&lj Voftage) [V.
Stock solution +0. 00626 M Cl
41 iC polarization
j i. -A "
. * A
. * .. & '
A comparison of polarization kinetics for A36 and
in the stdck solution to ,whiOh has been sufficient
to Bfitii the overall Cl -1 concentration to 6.25x10'
lire 14 -
Immediately upon addition bf chloride ions^ here producing a
concentration in the stock solution of 6 . 25xlO’ 3 H; just above the
minimum concentration required to produce pitting in low carbon
steels , there are changes in polarization behavior between the
two metals. These results are shown in Figure 14. There is little
apparent difference between polarization kinetics for the two
materials . However , there are differences in polarization degree .
in order to better visualize what the differences are , the
data contained in Figures 13 and 14 are compared by replotting
and comparing the polarization behavior of each alloy with, and
without, chloride ions being present.
Notice that there is
little change in the
polarization, but a
noticeable decrease in
the IR drop across the
film - the film has
[Figure 15]. Similarly,
when the behavior of
the 4130 alloy is
compared in stock and
solution, there is also
a decrease in IR drop
across the corrosion
product, but the
relative amount of the
decrease is far less
[Figure 16] .
There is little other
than speculation as to
the interpretation of
this behavior. The fact
that the degree of
polarization is so much
greater for the A36
alloy indicates that
there is a greater
driving force for
migration of negatively
charged anionic species
into the film toward
interface. This, by
itself, does not
necessarily result in
conductance - unless,
the chloride ion
somehow leads to the
within the corrosion
product. Perhaps, just perhaps, chloride ions within the
corrosion product lead to an increased concentration of negative
charge carriers, increasing the n-type extrinsic semiconductivity
production of electronic charge carriers
behavior for A36 with, and without,
chloride ions being present [from
Figures 13 and 14].
of the barrier layer.
Whatever the explanation, within a few minutes, the IR drop
decrease trend continues, until, after 5 minutes, the A36 has
completely lost its capacity for producing a electrically
resistive barrier layer completely [Figure 17].
- fn(Cell Voltage) [v. SCE] 5 minutes after Cl addition
| . ..
2 4 6 8 10 12 14 16 18
Elapsed Time, msec
Figure 17 After 5 minutes of RCRV in the presence of chloride
ion, A36 has lost its capacity for production of electrically
resistive corrosion products while 4130 is hardly affected at
Notice that the 4130 is hardly affected at all.
Bear in mind that this behavior has been measured without the
occurrence of noticeable corrosion attack on the A36 alloy - only
the tendency has been measured.
a. Gal vanostatic measurements represent a more natural
approach to the modelling and development of accelerated
test methods than do conventional potentiostatic methods.
b. Periodic or cyclic current reversal tends to simulate
concurrent anodic and cathodic reactions on a freely
corroding metal surface better than do potentiostatic
methods where anodic and cathodic processes occur on
separate electrode surfaces.
c. Development of highly resistive barrier layers which result
in large degrees of polarization during anodic current
cycles appear to be directly associated with susceptibility
to atmospheric corrosion.
d. Repetitive Current Reversal Voltammetry [RCRV] has been
demonstrated to be a highly sensitive means with which to
differentiate relative susceptibility to localized
Experimentation must J?e continued with attention focussed
on variations in polarization behavior as a function of
solution composition. This would require modification to
the existing electrochemical cell used in these studies .
Development of a standard test solution and procedure to be
applied to a larger list of candidate materials. Emphasis
most be placed on establishing experimental methodology
wbi ch is not limited by the experience or skill of the
operator - ’’idiot proof".
For scientific purposes , parallel studies of mechanistic
implications of polarization kinetics should be
accomplished . This will require "up-grading" pf the
equipment used - a digital oscilloscope with computer
interfacing would speed up the data acquisition and
interpretation process .
The relationship between polarization behavior and both
current amplitude and frequency needs to be established in
more detai 1 than has been done for these feasibility
studies . It is possible that the results of these
experiments may be integrated with experimentation using AC
impedance techniques thus permitting more meaningful
interpretation of alternating current methods than has been
achieved to date .