Comprehensive nuclear materials 5 01 corrosion and compatibility Comprehensive nuclear materials 5 01 corrosion and compatibility Comprehensive nuclear materials 5 01 corrosion and compatibility Comprehensive nuclear materials 5 01 corrosion and compatibility Comprehensive nuclear materials 5 01 corrosion and compatibility Comprehensive nuclear materials 5 01 corrosion and compatibility Comprehensive nuclear materials 5 01 corrosion and compatibility
Trang 1S Lillard
Los Alamos National Laboratory, Los Alamos, NM, USA
ß 2012 Elsevier Ltd All rights reserved.
5.01.1.4 Reference Electrodes and Their Application to Nuclear Systems 4 5.01.1.5 The Thermodynamics of Corrosion from Room Temperature to the PWR 6
Abbreviations
BWR Boiling water reactor
CNLS Complex nonlinear least squares fitting of
the data
EC Electrical equivalent circuit
EIS Electrochemical impedance
spectroscopy
EPBRE External pressure-balanced reference
electrode
FFTF Fast Flux Test Facility
HIC Hydrogen-induced cracking
HIFER Hi-Flux Isotope Reactor
IG Intergranular
LBE Lead–bismuth eutectic
PWR Pressurized water reactor
SCC Stress corrosion cracking
SS Stainless steel
Symbols
A Surface area
a i Activity of species i
C Capacitance
c i Concentration of species i
CR Corrosion rate
E Potential
EW Equivalent weight
E corr Corrosion potential
f Mass fraction
ƒ Fugacity
F Faraday’s constant
i Current density
i corr Corrosion current density
j i Square root of 1
k Rate constant
L Oxide thickness
M Molecular weight
M M Metal cation
n Number of electrons
N D Donor concentration
Q Reaction quotient
r Rate of reaction
R Gas constant
R p Polarization resistance
R V Solution resistance
S Entropy of transport
t i Transport number of species i
T Temperature
V o Oxygen vacancy
z Charge
Z Impedance
Z 0 Real part of the impedance
Z 00 Imaginary part of the impedance jZj Magnitude of the impedance
b Symmetry factor
b a Anodic Tafel slope
b c Cathodic Tafel slope
d Double layer thickness
1
Trang 2DC0p Change in standard partial molar heat
capacity
DE0 Standard reduction potential
DG Change in Gibbs energy
DG0 Standard Gibbs energy
DS 0
Standard entropy change
« Electronic charge
« 0 Permittivity of space
f Applied potential
h Overpotential
u Phase angle
r Material density
v Frequency
5.01.1 Theory
Mars Fontana identified eight forms of corrosion in
his book Corrosion Engineering1 and it is quite easy
to find examples of almost all of these in nuclear
reactors in both the primary and secondary cooling
water systems For example, galvanic corrosion in
zirconiumstainless steel couples,2,3
crevice corro-sion in tube sheets4 and former baffle bolts,5 and
pitting corrosion in alloy 600 steam generator
tubes.6,7Perhaps the most infamous form of corrosion
observed in nuclear reactors is stress corrosion
crack-ing (SCC), or environmental fracture, as we shall
refer to it here, which has numerous examples in
the literature Environmental fracture includes both
intergranular SCC (IG), such as that which occurs in
austenitic stainless steel, and hydrogen-induced
crack-ing (HIC), frequently observed in nickel base alloys
Failure by one of these mechanisms results from an
interplay between stress, microstructure, and the
envi-ronment (e.g., the electrochemical interface) The goal
of this chapter is not to address each of the corrosion
mechanisms outlined by Fontana individually, that will
be accomplished in the following chapters Rather,
this chapter is meant to provide the reader with the
fundamental electrochemical theory necessary to
critically evaluate the data and discussions in the
corrosion chapters that follow
In this section, we will review the fundamental
theory of the electrochemical interface In the first
three subsections, we review Half Cell Reaction, Cell
Potentials and the Nernst Equation, and Reference
Electrodes in Nuclear Systems In these sections, we
develop the theory necessary to understand the role of
electrochemical potential in environmental fracture
and corrosion mechanisms For example, intergranular stress corrosion cracking (IGSCC) is only observed at potentials more positive than a critical value while HIC
is only observed at potentials more negative than a critical value In the remaining two sections, we review the Thermodynamics from Room Temperature to the pressurized water reactor (PWR) and Kinetics of Dis-solution and Passive Film Formation These sections should help the reader to understand the role of the passive film in the corrosion mechanism and the com-petition that occurs between film formation and metal dissolution rate As the fundamental role of irradiation
in corrosion and environmental fracture mechanism is far from well established, in each section, we incorpo-rate empirical irradiation data as examples and discuss concepts that are more broadly important to nuclear systems
The electrochemical interface is characterized by an electrode (in this case a metal such as a cooling pipe) and an electrolyte (e.g., the cooling water in a reac-tor) While the bulk electrolyte contributes to vari-ables such as solution chemistry and ohmic drop (solution resistance is discussed later in this chapter),
it is the first nanometer of electrolyte that plays the most important role in electrochemistry In this short distance, referred to as the electrochemical double layer, a separation of charge occurs It is this sepa-ration of charge that provides the driving force (potential drop) for corrosion reactions For example,
a 100 mV-applied potential across a typical double layer will result in an electric field on the order of
106V cm2 In the model proposed by Helmholtz,8 the double layer may be thought of as capacitor, with positive charge on the metal electrode and the adsorption of negatively charged cations on the solu-tion side (Figure 1) The capacitance of the double layer is equal to that in its electrical analog e0D/d, wheree0is the permittivity of space, D is the dielec-tric, and d is the thickness of the layer For most electrochemical double layers, C is on the order of
106F cm2 Electrochemical reactions that take place in the double layer are reactions in which a transfer of charge (electrons) occurs There are two different types of cells in which electrochemical reactions may occur9:
Electrolytic cells in which work, in the form of electrical energy, is required to bring about a nonsponta-neous reaction
Trang 3Voltaic cells in which a spontaneous reaction occurs
resulting in work in the form of electrical energy
Electrolytic cells cover a fairly large number of
electrochemical reactions but may generally be
thought of as ‘plating’ or ‘electrolysis’ type reactions
and will not be treated here Corrosion reactions are
voltaic cells and will be the focus of this chapter As in
an electrolytic cell, voltaic cells are characterized by
two separate electrodes, an anode and a cathode In
corrosion, reactions at the anode take the form of
metal dissolution, the formation of a soluble metal
cation:
Because the system cannot store charge, the electrons produced during the anodic reaction must be used This occurs at the cathode where typical reactions may include oxygen reduction:
Acid: O2þ 4Hþþ 4e) 2H2O ½II Base: O2þ 2H2Oþ 4e) 4OH ½III
or hydrogen reduction:
Fromeqns [I] and [III], the general corrosion of an
Fe surface in basic solution may then be written as:
where Fe(OH)2is the corrosion product
An example of what the anodic and cathodic reac-tions on Fe electrode might look like is presented in Figure 2 Though the anodic and cathodic reactions occur at physically separate locations, as shown in this figure, the reactions must be connected via an electrolyte (aqueous solution).Figure 2also suggests that corrosion reactions are controlled by variables such as mass transport (diffusion, convection, migra-tion), concentration, and ohmic drop (resistivity of the electrolyte) These variables will be considered in our discussion of corrosion kinetics
Equation For any chemical reaction the driving force, the Gibbs energy, may be written as10:
where DG0 is the standard Gibbs energy, Q is the reaction quotient equal to the product of the activ-ities (assumed to obey Raoult’s Law for dilute solu-tions and, thus, equal to the concentration) of the products divided by the reactants, R is the gas con-stant, and T is temperature The electrical potential,
E, is related to the Gibbs energy of a cell by the relationship:
where n is the number of electrons participating
in the reaction and F is Faraday’s constant For the reduction of hydrogen on platinum:
Haqþþ eðPtÞ ,1
Fe
Oxide
Figure 2 Diagram of what the anodic and cathodic
reactions may look like on an iron surface depicting the
separation of reactions and ionic conduction.
Excess positive
charge
Excess negative charge Bulk solution
Metal
fmetal
fsolution
Double layer
+ +
+ + +
+ +
+
Figure 1 A diagram depicting the separation of charge at
the electrochemical double layer and the associated
potential drop ( f).
Trang 4The reaction quotient, Q (starting conditions) becomes:
Q ¼½ fH2 1=2
Hþ
where fH2is the fugacity of hydrogen gas Substituting
eqns [2] and [3]intoeqn [1], we find for the reduction
of hydrogen on platinum that:
E¼ DE0þRT
F ln
fH2
½ 1 =2
½Hþ
½4
where F is Faraday’s constant and DE0is the standard
reduction potential for the reaction ineqn [VI]
Equa-tion [10]is commonly referred to as the Nernst
equa-tion and defines the equilibrium reducequa-tion potential of
the half cell and is pH dependent The Nernst equation
is commonly expressed in its generalized form as:
E¼ DE0þRT
Application to Nuclear Systems
In Equation [4], all of the parameters are easily
calculated with one exception, DE0 Therefore, we
define DE0¼ 0 ineqn [4] for a set of specific
para-meters and refer to this cell as the standard hydrogen
electrode (SHE): H2pressure of 1 atm, a pH¼ 0, and
a temperature of 25C This provides a reference
from which we can calculate the standard potentials
for all other reduction reactions usingeqn [5] These
are referred to as standard reduction potentials and a
few examples are provided inTable 1
While the SHE is the accepted standard, from
a practical standpoint, this reference electrode is
difficult to construct and maintain As such,
experi-mentalists have taken advantage of a number of other
reduction reactions to construct reference electrodes
for laboratory use The reaction selected typically
depends on the application One common reference
electrode is the silver–silver chloride electrode
(Ag/AgCl) which is based on the reduction of Agþ
in a solution of potassium chloride:
and the overall reaction being:
The Nernst equation foreqn [XI]is equal to:
E¼ DE0 Ag=AgClþRT
F ln
½aAgCl
½aAg½aCl
¼ DE0
where aCl is the activity of chloride and for which the concentration (mCl) in molal (mol kg1) is fre-quently substituted In the corrosion lab, the refer-ence electrode is constructed by electrochemically depositing an AgCl layer onto a silver wire This wire is then placed in a glass capillary filled with a solution of potassium chloride the concentration of which then defines the cell potential (aClineqn [6]) One end of the capillary is sealed using a porous frit (typically a porous polymer) that acts as a junction between the solution of the reference electrode and the environment of the corrosion experiment While the Ag/AgCl reference electrode construc-tion described above is straightforward for the lab, there are several obstacles that must be overcome before it can be used in a nuclear power plant setting, namely, radiation flux, pressure, and temperature As it turns out, the primary impact of ionizing radiation
on laboratory reference electrodes relates to damage
of the cotton wadding and polymer frits used in their construction and no change in cell potential occurs.11
As such, two approaches based on the Ag/AgCl reference electrode have been used to measure elec-trode potential in nuclear power reactors In the first approach, an internal reference electrode operates in the same high-temperature environment as the reac-tor In this case, one must consider the solubility of
Table 1 Standard reduction potentials for several reac-tions important to the nuclear power industry
Reduction reaction Standard reduction
potential (V)
Cl 2 þ 2e ⇄ 2Cl 1.36
O 2 þ 4H þ þ 4e ⇄ 2H 2 O 1.23
Fe3þþ e ⇄ Fe 2þ 0.77
O 2 þ 2H 2 O þ 4e ⇄ 4OH 0.4 AgCl þ e ⇄ Ag þ Cl 0.22 2Hþþ 2e ⇄ H 2 (NHE) 0.0
Trang 5Ag/Cl complexes that form as a function of
tempera-tureeqn [6].12,13That is, reactions in addition toeqns
[VIII] and [XI]must be considered From a
construc-tion viewpoint, the internal reference electrode
consists of a silver chloride pellet on a platinum foil
(Figure 3).14External potential measurement is made
via contact with a nickel wire which is connected to an
electrometer via a coaxial cable The electrode is
housed in a sapphire tube that is sealed via a porous
sapphire cap In this configuration, there is no internal
electrolyte per se Upon placing the electrode in a
boiling water reactor (BWR), the porous cap allows
the cooling water to penetrate the electrode and the
potential is determined fromeqn [6]and the solubility
of AgCl in high purity water as a function of
temperature.15
In the second approach, an external
pressure-balanced reference electrode is used (EPBRE)
In the EPBRE, the reference electrode is maintained
at room temperature and pressure and the
cor-responding constants are used in eqn [6] The
reference is connected to the high-temperature
envi-ronment via a nonisothermal salt bridge sealed with a
porous zirconia plug (Figure 4).16As a result of this
configuration, the EPBRE is not susceptible to
potential deviations owing to the solubility of AgCl and its complexes as a function of temperature How-ever, the temperature gradient between the reactor and the reference electrode results in a junction potential that must be subtracted fromeqn [6] The corresponding thermal liquid junction potential (ETLJ)17is given by:
ETLJ¼ 1
F
ðT2 T1
tMþSMþ
zMþ þtClSCl
zCl
where t, S, and z represent the transport number, the entropy of transport, and the charge on the cation, respectively The symbol M ineqn [7]represents the metal in the chloride salt, MCl, and is commonly Li, Na,
or K In addition to ETLJ, there is also the isothermal liquid junction potential, EILJ, which arises due to the differences in cation and anion mobilities through the porous frits and the fact that the electrolyte in the external reference (typically KCl) is vastly different from the reactor cooling water in which it is immersed17:
EILJ¼ RT
F
ðT2 T1
ti
zi
SS nut
1/4 OD SS tube
1/4 NPT Compression fitting
Pure water or 0.01 M KCl Glass wick Heat-shrinkable PTFE tube
Restrainer
Rulon adapter
Compression fitting
Ag/AgCl
Rulon sleeve Zirconia plug
Figure 4 A diagram of a pressure-balanced reference electrode is used in BWRs Bottom of figure is sealed into pressure vessel via compression fitting while Ag/AgCl electrode (top) remains at room temperature and pressure Reprinted from Oh, S H.; Bahn, C B.; Hwang, I S.
J Electrochem Soc 2003, 150, E321, with permission from The Electrochemical Society.
AgCl pellet
Sapphire lid Sapphire container Ceramic to metal braze Kovar
TIG weld 304SS
Seal
Coaxial cable
Pt cap
Ni wire
Alumina
insulators
Figure 3 Diagram of an internal Ag/AgCl reference
electrode used in BWRs Top end is inserted into the cooling
loop, while the coaxial cable provides electrical connection.
Reprinted from Indig, M E In 12th International Corrosion
Congress, Corrosion Control for Low-Cost Reliability; NACE
International: Houston, TX, 1993; p 4224, with permission
from NACE International.
Trang 6where the subscript i denotes a species that may be
transported through the zirconia plug and for a nuclear
power reactor may include species ions as Agþ, Cl,
Hþ, OH, Kþ, and B(OH4)
It has been shown that both ETLJand EILJ each
increase by as much as 0.15 V over the temperature
range of 25–350C The result is a decrease in the
measured potential of 0.30 V at 350C While these
junction potentials can be calculated and used to
correct eqn [7], it has been shown that there is
some deviation at higher temperatures (>200C)
and an experimental fitting procedure is the
pre-ferred method for calibration of the reference
electrode
from Room Temperature to the PWR
An atlas of electrochemical equilibria has been
cre-ated by M Pourbaix for metals in aqueous solution at
room temperature.18 This atlas contains potential–
pH diagrams, so-called Pourbaix diagrams, which
define three equilibrium thermodynamic domains
for metals in aqueous solutions: immunity, passivity,
and corrosion Immunity is defined as the state where
the base metal is stable while corrosion is defined
as the formation of soluble metal cations and
passiv-ity the formation of a stable oxide film Pourbaix’s
derivation requires that the values of the standard
chemical potential, m0, for all of the reacting
sub-stances are known for the standard state at the
temperature and pressure of interest For chemical
reactions at room temperature, the equilibrium
con-ditions are defined by the relationship18:
logK ¼
P
nm0
and for electrochemical reactions at room
tempera-ture (Table 1) equilibrium is defined by:
E0¼
P
nm0
where K is the equilibrium constant for the reaction,
m0is in Joules per mole, v is the stoichiometric
coeffi-cient for the species, n is the number of electrons, 5708
is a conversion constant equal to RT/(log10e) where
T is temperature (298.15 K) and R the ideal gas
con-stant (8.314472 J (K mol)1), and 96 485 is Faraday’s
constant in J (mol V)1
As an example of these diagrams, consider the
iron–water system and the solid substances Fe,
Fe3O4, and Fe2O3 Pourbaix18 defined the relevant
equations for this system as:
reactions with two soluble species
Fe2þ¼ Fe3þþ e E0¼ 0:771 þ 0:059log Fe3þ
Fe2þ
½11
solubility of iron and its oxides
Fe¼ Fe2þþ 2e E0¼ 0:440 þ 0:0295logðFe2þÞ ½12
2Fe2þþ 3H2O¼ Fe2O3þ 6Hþþ 2e
E0¼ 0:728 0:177pH 0:059logðFe2 þÞ ½13
Feþ 2H2O¼ HFeO2 þ 3Hþþ 2e
E0¼ 0:493 0:089pH þ 0:0295log½HFeO2 ½14 3HFeO
2þ Hþ¼ Fe3O4þ 2H2Oþ 2e
E0¼ 1:819 þ 0:029pH 0:088log½HFeO2 ½15
reaction of two solid substances 3Feþ 4H2O¼ Fe3O4þ 8Hþþ 8e
2Fe3O4þ H2O¼ 3Fe2O3þ Hþþ 2e
An example of a simplified Pourbaix diagram for Fe
at room temperature based on the reactions in eqns [11]–[17] is presented in Figure 5, where Eq [12] corresponds to figure line 23, [13] to line 28, [14] to line 24, [15] to line 27, [16] to line 13 and [17] to line
17 Note that Eq [11] is the boundary between Fe2þ and Fe3þ and was not drawn in the original figure
In addition to the lines separating the domains for Fe, Pourbaix diagrams will typically include the domains associated with water stability (oxidation and reduc-tion) represented by the dashed lines marked a and b
in Figure 5 Upon inspection of this diagram one would conclude what is know from experience with Fe: that iron is passive in alkaline solutions and at higher applied potentials owing to oxide film forma-tion while at more acidic soluforma-tions Fe is susceptible to corrosion owing to Fe2þ It is worth noting again that these potential–pH domains are defined solely by the thermodynamic stability of the species within them and these diagrams do not consider kinetics which will be addressed later in this chapter This is impor-tant as while a species/reaction may be thermody-namically stable it may be kinetically hindered While the use of Pourbaix diagrams to
Trang 7widespread, these diagrams and the method for
gen-erating them as presented thus far cannot be used at
the higher temperatures associated with nuclear power
reactors This is due to the lack of standard potentials
at elevated temperature as required byeqns [9] and
[10](e.g., the application ofTable 1to higher
temper-ature) In the absence of these high-temperature
ther-modynamic data, Townsend19 used an extrapolation
method introduced by Criss and Coble (the
correspon-dence principle) The method allows for empirical
entropy data of ionic species at 25C to be
extrapo-lated to higher temperatures In this method, the
stan-dard Gibbs free energy is calculated from the
relationship:
DðDG0Þ ¼ DTDS0ð250Þ þ
ðT
25 0
DC0
pðTÞdT
T
ðT
25 0
DC0
where DS0is the standard entropy change and DC0p is
the change in standard partial molar heat capacity The
potential–pH diagram for the Fe–H2O system and the
solid substances Fe, Fe3O4, and Fe2O3at 200C
calcu-lated by Townsend is presented inFigure 6 In
com-parison with the diagram at 25C (Figure 5), the
Fe2O3 and Fe3O4 regions are extended to lower pH
and potentials As a result the area associated with
corrosion at lower solution pH is decreased However,
the most notable change in the diagram is at high pH where the area associated with corrosion owing to the soluble HFeO2has increased dramatically The Criss and Coble method is limited, however, to the 150–200C range and, to extend the Pourbaix to the temperatures of power reactors, Beverskog used a Helgeson–Kirkham–Flowers model to extend the heat capacity data to 300C.20
Thus far, we have described a method for gener-ating electrochemical equilibria diagrams and regions
of passivity, corrosion, and immunity for pure metals from 25 to 300C From an engineering standpoint,
we would like to know this information for structural alloys such as austenitic stainless steels and super nickel alloys At temperatures near 25C, the predom-inant oxide responsible for passivity is Cr2O3and it is sufficient to rely only on the Cr potential–pH diagram for alloys with a high Cr content However, at higher temperatures other oxides form such as Fe(Fe,Cr)2O4, (Cr,Fe)2O3, (Cr,Fe,Ni)3O4, and (Cr,Fe,Ni)2O3, and it is desirable to know the thermodynamic stability of the alloy Beverskog has developed the ternary potential–
pH diagrams for the Fe–Cr–Ni–H2O–H2 system for temperatures up to 300C using heat capacitance data and the revised Helgeson–Kirkham–Flowers model described above.21 However, Fe–Cr–Ni phases lack thermodynamic data and the ternary oxides were,
23
26 17
29 13
27
24
a 28 20
b
Fe
1.5
1.0
0.5
0
-0.5 -1.0 -1.5
pH
15
Figure 6 The potential–pH diagram for the Fe–H 2 O system and the solid substances Fe, Fe 3 O 4 , and Fe 2 O 3 at
200C Most dramatic influence of increased temperature is the presence of a large region of soluble species (corrosion)
at high pH Reprinted from Townsend, H E Corrosion Sci.
1970, 10, 343, with permission from Elsevier.
27
17 26 23
28
a 20
13
Fe3O4
Fe 3+
-2
24
1.5
1.0
0.5
0
-0.5
EH
-1.0
-1.5
pH
15
b
Figure 5 A simplified potential–pH diagram for the
Fe–H 2 O system and the solid substances Fe, Fe 3 O 4 , and
Fe 2 O 3 at 25C based on the reactions in eqns [11]–[17]
Reprinted from Townsend, H E Corrosion Sci 1970,
10, 343, with permission from Elsevier.
Trang 8thus, not considered The diagrams assumed that the
metallic elements in the alloy had unit activity, that is,
equal amounts of iron, chromium, and nickel An
example of the potential–pH diagram for chromium
species in Fe–Cr–Ni at 300C and aqueous species
with a concentration of 106 molal is presented in
Figure 7 Unlike the Fe diagram, where the presence
of soluble HFeO2species increased with temperature
(Figure 6), the diagram for Cr in Fe–Cr–Ni is
domi-nated by passive region where the stable oxides of
Cr2O3, FeCr2O4, and NiCr2O4are formed
Passive Film Formation
The study of dissolution kinetics, corrosion rate,
attempts to answer the question: ‘‘What are the
rela-tionships that govern the flow of current across a
corroding interface and how is this current flow
related to applied potential?’’ Consider the anodic
dissolution of a metal with an activation barrier
equal to Ga1¼ nFE (eqn [2]) If we increase the driving
force (potential) from its equilibrium condition, E0, to
a new value, f, the new barrier is given by the
relationship22:
G2a ¼ G1
whereb is the symmetry factor and reflects the fact
that not all of the increase in potential goes to
decreasing the barrier, that is, not all of the applied potential is dropped across the electrochemical dou-ble layer The rate (ra) of this reaction is expressed in the same, Arrhenius, form as for chemical reactions:
ra ¼ ia=nF ¼ kaco exp DGa
RT
½20 where iais the anodic current density, kais the rate constant, cois the concentration of oxidized species, and DGais the change in free energy for the anodic reaction Substituting Ga Gaineqn [20]for DGain eqn [19], we express the anodic reaction rate as22:
ia¼ nFkacRexp ð1 bÞnF
½21 where (the overpotential) represents a departure from equilibrium and is equal tof E0 We can derive
a similar expression for the cathodic reaction22:
ic ¼ nFkcco exp bnF
RT
½22 where icis the cathodic current density, kcis the rate constant, and co is concentration of oxidized species Combiningeqns [21] and [22]and rearranging them,
we can write an expression for the total current, i:
i¼ io exp ð1 bÞnF
RT
½23 where iois the exchange current density and is equal to nFkccobkabcRb This expression is commonly referred
to as the Butler–Volmer equation
To applyeqn [23]to corrosion reactions, we need
to be able to relate to the corrosion potential, Ecorr, that is, as it stands the Butler–Volmer equation is derived for equilibrium conditions Returning to our definition of the overpotential ¼ f E0, by both subtracting and adding Ecorr from the right side of this definition, inserting the resulting expression back intoeqn [23]and rearranging we find22:
icorr¼ iaexp ð1 bÞF
RT ðEcorr EaÞ
¼ icexp bF
RTðEcorr EcÞ
½25 For small applied potentials around Ecorr, the Stern– Geary approximation ofeqn [24]is used23:
i¼ 2:303icorr baþ bc
babc
where ba and bc are defined as the anodic and
Cr(cr)
-npH 2
1
0
-1
-2
Figure 7 Potential–pH diagram for chromium species in
Fe–Cr–Ni at 300C Concentration of aqueous species is
106molal Reprinted from Beverskog, B.; Puigdomenech, I.
Corrosion 1999, 55, 1077, with permission from NACE
International.
Trang 95.01.2.2) having units of volts and are empirical
fac-tors related to the symmetry factor by the
relation-ships22:
bc ¼ 2:303RT
As it relates to the nuclear power industry,eqn [24]not
only relates the corrosion rate (icorr) to the applied
potential, f, but it can also help us to rationalize
other processes such as the influence of water radiolysis
products on corrosion rate For example, it is generally
observed that ionizing radiation (g, neutron, proton,
etc.) increases Ecorrpotential (Figure 8) and corrosion
rate at Ecorr.24 The flux of ionizing radiation on the
cooling water results in radiolysis, the breaking of
chemical bonds During the course of water radiolysis,
a wide variety of intermediate products are formed,
such as O2, eaq, and the OH radical.25The vast
major-ity of these species have very fast reaction rates so that
the end result is a handful of stable species These
stable products are typically oxidants, such as O2, H2,
and H2O2 That is, these products readily consume
electrons (eqns [II]–[IV]) and increase cathodic
reac-tion rate (Figure 9) Fromeqn [25] we see that an
increase in the cathodic reaction rate, ic, necessarily
results in an increase in corrosion rate, icorr, consistent
with the observation described
While the development of dissolution kinetics is
straightforward, the kinetics associated with passive
film formation and breakdown are less well under-stood yet equally as important to our understanding
of corrosion mechanisms One such example is the case of localized corrosion where the probability for
a pit to transition from a metastable to a stable state is governed by the ability of the active surface to repas-sivate Another example is the initiation of SCC where passive film rupture results in very high dissolution rates and, correspondingly, crack advance rate which is controlled by the activation kinetics described above as the bare metal dissolves.26–28During the propagation stage of SCC, the crack tip must propagate faster than (1) the oxide film can repassivate the surface and (2) the corrosion rate on the unstrained crack sides so that dissolution of the walls does not result in blunt notch
To evaluate the role of repassivation kinetics
in SCC and corrosion mechanisms in general, inves-tigators set about measuring three critical experi-mental parameters, namely film: ductility,29,30 bare surface dissolution rates,31–33and passive film forma-tion rates34,35for various alloys Each of these tech-niques involves the depassivation of a metal electrode using a tensile frame or a nano-indenter (in the case
of ductility studies) or scratching/breaking an elec-trode (bares surface current density and repassivation studies) and measuring the resulting current transient
as a function of time An example of a current tran-sient for SS 304L in chloride solution is presented
in Figure 10 The surface was under potentiostatic control and was bared using a diamond scribe The data were collected using a high-speed oscilloscope
0.4
0.35
0.3
0.25
0.2
0.15
0.1
Time (s)
Beam on at 100 nA
~540 s
Figure 8 Influence of proton irradiation on the E corr of a
SS 304L electrode in dilute sulfuric acid, pH ¼ 1.6 The
increase is caused by the production of water radiolysis
products.
Increasing radiation flux Increasing cathodic reaction rate
Beam = 35 na Beam = 62 na Beam = 100 na
0.1 0 -0.1 -0.2 -0.3 -0.4 -0.5 -0.6 -0.2
-2 )
Potential (V vs SCE)
Figure 9 Influence of proton irradiation on the cathodic reactions on a Au electrode in dilute sulfuric acid, pH ¼ 1.6 The increase is caused by the production of water radiolysis products.
Trang 10The transient is characterized by two separate
pro-cesses, anodic dissolution and repassivation
repre-sented by td and tr in Figure 10 To analyze the
repassivation rates, the period tris typically fit to an
expression and evaluated as a function of solution pH
or electrode potential The most prolific work in this
field is probably on the alloy SS 304L For this alloy, it
has been proposed that the kinetics of film growth are
controlled by ion migration under high electric
field.36–38 The kinetics of high-field film growth
were first proposed by Cabrera and Mott39 to obey
the kinetic relationship:
L
½29
where i is the current density, V is the voltage, L is the
oxide thickness, and A and B are constants
5.01.2 Analytical Methods
In this section, we will review the principle analytical
methods used to probe the electrochemical interface
In the Section 5.01.2.2 Potentiodynamic
Polariza-tion, we discuss linear polarization resistance and
the practical application of corrosion kinetics, eqns
[25] and [26] In that section, we also describe the
salient points of the anodic polarization curve In the
Section 5.01.2.3 Electrochemical Impedance
Spec-troscopy, we introduce an ac method for interrogating
the electrochemical interface This technique is
probably the most versatile experimental method available to scientists and researchers As it relates
to nuclear reactors, this technique has the ability to subtract out the contribution of the solution resis-tance to polarization resisresis-tance measurements which,
if not accounted for in highly resistive cooling water measurements will result in nonconservative corro-sion rates In the final section, we introduce a more seldom used technique, Mott–Schottky analysis While this is by no means a common experimental method, it provides a conduit for the reader to become familiar with defects in the oxide film, their transport and ways to quantify it This has particular interest here as ionizing radiation may promote cor-rosion rates by increasing transport of these defects through the passive oxide film
Regretfully, the scope of this chapter is limited, and we are not able to discuss the step-by-step details
of the experimental methods that are used to make corrosion measurements A comprehensive guide
to experimental methods in corrosion has been published by Kelly et al.40 as well as Marcus and Mansfeld41 while a more broad description of elec-trochemical methods has been published by Bard and Faulkner.42The reader is also encouraged to become familiar with the equipment that is used to make electrochemical measurements and a good introduc-tory chapter on this topic has been presented by Schiller.43 The most important instrument is, no doubt, the potentiostat While this instrument is the cornerstone of corrosion science, it does have its pit-falls including bandwidth limitations and the poten-tial for ground loop circuits when used in conjunction with other equipment such as load frames, autoclaves and cooling loops The latter can be overcome using proper instrumentation such as potentiostat with a floating ground, or isolation amplifiers
To investigate the influence of the neutron flux
on corrosion rates and mechanisms, real-time in-situ corrosion measurements are often made ‘in-reactor’
or at neutron facilities such as Oak Ridge National Lab’s Hi-Flux Isotope Reactor (HIFER) and Argonne National Lab’s Fast Flux Test Facility (FFTF) Alter-nately, neutron damage can be simulated using ion beams As it relates to ion irradiation, this method provides opportunities for studying the interaction
of the components of reactor environments (radia-tion, stress, temperature, aggressive media) that are not possible with in-reactor or neutron irradiation facilities For a full discussion of this topic, see
Beams To summarize these experiments, controlled
td tr
7.0 10 −4
6.0 10 −4
5.0 10 −4
4.0 10 −4
3.0 10 −4
2.0 10 −4
1.0 10 −4
0.0
0 0.002 0.004 0.006
Time (s)
0.008 0.01 0.012
Figure 10 Scratch test current transient from a SS 304L
electrode in 0.1 M NaCl The transient is characterized by a
growth period, t d , and a repassivation period, t r