Similar to (Efc 4) european federation of corrosion guidelines on electrochemical corrosion measurements - prepared by the working party on physico- (20)
(Efc 4) european federation of corrosion guidelines on electrochemical corrosion measurements - prepared by the working party on physico-
1. European Federation of Corrosion
Publications
NUMBER 4
A Working Party Report
Guidelines on Electrochemical
Corrosion Measurements
Published for the European Federation of Corrosion
by TheInstitute of Metals
THE INSTITUTE OF METALS
1990
2. Book Number 497
Published in 1990by The Institute of Metals
1Carlton House Terrace,London SWlY 5DB
and
The Institute of Metals
North American PublicationsCenter
Old Post Road, Brookfield VT 05036
US A
0 1990The Institute of Metals
All rights reserved
British Libray Cataloguing in Publication Data
European Federation of Corrosion.Working Party on Physico-
Chemical Methods of Corrosion Testing
Guidelineson electrochemicalcorrosionmeasurements.
1.Corrosion.Testing.
I. Title 11.Institute of metals 1985- 111.Series
620.11223
Libray -of-Congress -Cataloging-in-Publication-Data
availableon application
I S B N b901462-87-X
Text processingby P i c A Publishing Servicesfrom original
typescripts and illustrations provided by the authors
Printed in Great Britain
3. Contents
European Federation of Corrosion ~ Series Introduction
Foreword
Introduction
Chapter 1 Instrumentation, Performance and Calibration
I. C. Rowlands
1.1Introduction
1.2PerformanceLimitations
1.2.1Impedance Matching
1.2.2Ammeter Loading
1.2.3Amplifier Bias Current
1.2.4Bandwidth Limitations
1.3Error Sources
1.3.1Resistance Noise
1.3.2Triboelectric Noise
1.3.3PiezoelectricEffects
1.3.4Thermoelectric Effects
1.3.5Magnetic Field Loops
1.3.6Earth Loops
1.3.7Earth Potentials
1.3.8CapacitiveTransformer Coupling
1.3.9Long Lead Cable Capacitance
1.4Instrument Calibration
1.5References
Chapter 2 Design of Electrochemical Cells
F. P. IJsseling
2.1 Introduction
2.2 Electrochemical/Electrical Requirements
2.3 Solution Requirements
2.4 Construction Requirements
2.5 References
9
11
12
13
13
13
16
18
18
19
19
19
21
22
25
5
4. Chapter3 Electrode Design
E. Heitz
27
3.1 Introduction
3.2Sizeof Electrodes
3.3 Design with Regard to Various Corrosion Parameters
3.3.1Current and Potential Distribution
3.3.2Mass Transfer
3.3.3Heat Transfer
3.3.4Mechano-chemicalTesting
3.3.5High Temperature/High Pressure Testing.
3.4Problems Concerning ApplicationTechniques
3.5 References
Chapter4 Reference Electrodes
E. Eriksrud and E. Heitz
4.1 Introduction
4.2Choice, Stability, Incompatibility
4.3Checkingof the Reference Electrode,Some Practical Advice
4.4More Specific Requirements
4.5 Some Common Electrodes for Aqueous Systems
4.5.1Mercury-mercurouschloride (calomel)
4.5.2Mercury-mercurous sulphate
4.5.3Mercury-mercuricoxide
4.5.4Silver-silverhalide
4.5.5Reference Electrodes in Non Aqueous Systems
4.6 References
Chapter5 Effectof Specimen Preparation and SurfaceCondition
1. Simpson
5.1 Introduction
5.2 Choiceof Sample
5.3 Samplingand Specimen Preparation
5.4 SurfacePreparation Before Immersion in the Test Solution
5.5 Effect of the Conditions Within the Cell Before Commencementof the Measurement
5.6 Changes Caused by the Measurement Itself
5.7 References
27
27
28
31
32
34
34
34
34
35
35
36
37
37
37
37
38
39
39
39
6
5. Chapter 6 Evaluation and Compensation of Ohmic Drop
L. Clerboisand F. P. IJsseling
6.1 Introduction
6.2Examples
6.2.1 Polarisation Resistance
6.2.2Polarisation Curves in Aqueous Solution
6.2.3Polarisation Curves in Non Aqueous Solution
6.3 Principal Methods
6.3.1 Minimising Ohmic Drop by Cell and Electrode Design
6.3.1.1 Cell Design
6.3.1.2Microelectrodes
6.3.2.1Positive Feedback
6.3.2.2Interrupt Methods
6.3.3.1Measurement of Ohmic Drop
6.3.3.2Direct Calculation
6.3.3.3Mathematical Methods
6.3.2Active Methods
6.3.3Passive Methods
6.4Conclusions
6.5 Acknowledgements
6.6References
Chapter 7 Automatic Measurement Systems
0.For& and 1.Aromaa
7.1 Introduction
7.2Parts of a Measurement System
7.2.1Computers and Peripherals
7.2.2Measuring Devices
7.2.3Data Transfer Busses
7.2.4Software
7.3 Analysis of Measured Data
7.4 Examples of Automatic Systems
Chapter 8 Field Testing
G. Turluer
8.1Objectives
8.2Test Methods
40
40
42
44
52
52
52
55
55
56
58
59
61
61
61
8.2.1 Polarisation Resistance Measurements (R,)
8.2.2Material (or Component) Potential and Process Redox Potential Monitoring
8.2.3Impedance Measurements
8.2.4Potentiodynamic Scanning
8.2.5Galvanic Current Measurements
8.2.6Spatial Potential Scanning
7
6. 8.3 SpecificRequirementsand Precautions
8.3.1General
8.3.2Probe configuration and location
8.3.3 Referenceelectrodes
8.3.3.1 Arbitary electrodes
8.3.3.2 In situ reference electrodes
8.3.3.3External reference electrodes
8.4Interpretation and PossibleLimitations
62
63
8
7. Introduction
At the present time electrochemicaltest methods in corrosionare becomingmore diverse and, as
aconsequence,new techniqueshavebeenadded tothosewellestablished.Eachtechniquehaslimi-
tationsandhasrequirementsforconditionstobesatisfiedforpracticalapplication.Thissubjectwas
the topic of an international workshop held at Ferrara,Italy, 11-13September1985,organised by
theEuropeanFederationof CorrosionWorkingParty on Physico-ChemicalMethods of Corrosion
Testing[l].ThisWorkshopoutlined therange of practicalapplicationof electrochemicalcorrosion
testing aswell as reachingconclusionsconcerningthe use, significanceand limitationsof various
methods. However, this meeting, and others, on electrochemical corrosion measurements have
generallynot provided specificinformationon how experimentsshould be performed[2-41.It was
the Working Party’s intention to fulfil this requirement with the compilation of this booklet. In
undertaking this task, members were confronted with a great variety of techniques to be consid-
ered, although, inevitably, the same experimental principles are involved with the various tech-
niques.Variouspracticalaspectsinconductingexperiments,both inthelaboratoryand inthefield,
have been considered in achieving the objective of this work.
There is a notable absence of specificationsor codes of practice on the use of electrochemical
corrosion testing and it is hoped that these guidelines will fill this gap in corrosion science.
REFERENCES
1.Electrochemical Corrosion Testing. Monograph 101, published by DECHEMA, Frankfurt,
FRG,(ed.E.Heitz, J. C. Rowlands and F. Mansfeld)1986.
2. Electrochemical Techniques for Corrosion. Published by NACE, Houston, USA ( ed. R.
Baboian) 1977.
3. Electrochemical Corrosion Testing. Special Technical Publication727.Published by ASTM
Philadelphia,USA (ed. F. Mansfeld and V. Berucci)1981.
4. ElectrochemicalMethodsinCorrosionResearch.Publishedby Trans.Tech.PublicationsLtd.,
Switzerland (ed.M. Duprat), 1986,8.
12
8. European Federationof Corrosion
Publications
Series Introduction
TheEFC, incorporated in Belgium, wasfounded in 1955with the purpose of promoting European
co-operationin the fields of research into corrosionand corrosionprevention.
Membership is based upon participation by corrosion societies and committees in technical
Working Parties. Member societiesappoint delegates to Working Parties, whose membership is
also expanded by co-optionof other individuals.
The activitiesof the Working Parties cover corrosiontopics associatedwith inhibition, educa-
tion,reinforcementinconcrete,microbialeffects,hotgasesand combustionproducts,environment
sensitivefracture,marineenvironments,surfacescience,physico-chemicalmethodsof measurement,
thenuclearindustry,and computerbased informationsystems.WorkingPartiesonothertopicsare
established as required.
TheWorkingPartiesfunction in various ways, e.g.by preparingreports, organising symposia,
conducting intensivecourses,and producing instructionalmaterial,includingfilms.Theactivities
of the WorkingPartiesareco-ordinated,through a Scienceand TechnologyAdvisory Committee,
by the ScientificSecretary.
The administration of the EFC is handled by three Secretariats:DECHEMA in the Federal
Republicof Germany,theSoci6t6deChimieIndustrielleinFrance,and theInstituteof Metalsinthe
United Kingdom. These three Secretariatsmeet at the Board of Administrators of the EFC. There
isanannual GeneralAssemblyatwhichdelegatesfromallmembersocietiesmeettodetermineand
approve EFCpolicy.News of EFC activities,forthcomingconferences,coursesetc.ispublished in
a range of accredited corrosion and certain other journals throughout Europe. More detailed
descriptionsof activitiesaregiveninan occasionalNewsletterprepared by theScientificSecretary.
Theoutput of the EFC takesvarious forms.Papers on particular topics,forexample, reviewsor
results of experimental work, may be published in scientificand technicaljournals in one or more
countries in Europe. Conferenceproceedings are often published by the organisation responsible
for the conference.
In 1987, the Institute of Metals was appointed as the official EFC publisher. Although the
arrangement isnonexclusive and other routes for publicationare stillavailable,it isexpected that
the Working Parties of the EFC will use the Institute of Metals for publication of reports,
proceedings etc. wherever possible.
A D Mercer
EFC Scientific Secretary
Institute of Metals London, UK
9
9. EFC Secretariatsare located at:
Mr. R. Wood
European Federationof Corrosion
The Institute of Metals
1Carlton House Terrace
LONDON SWI Y 5DB
UK
DrD.Behrens
EuropaischeFoderationKorrosion
DECHEMA
Theodor-Heuss-Allee25
D-6000
FRANKFURT(M)
FRG
M. R. Mas
FederationEuroptkne de la Corrosion
Societe de ChimieIndustrielle
28 Rue Saint-Dominique
FRANCE
F-75007PARIS
10
10. Foreword
Electrochemicalmethodshavebeenused incorrosiontestingeversincethe electrochemicalnature
of corrosionprocesseswas discovered.In the present age electrochemicalmeasurements involve
the use of sophisticated 'black boxes' which are invariablyblamed for any 'pitfalls' which occur.
Hence the European Federation of Corrosion Working Party on Physicochemical Methods of
CorrosionTestingfelt it desirableto remind research workers,studentsand instrument designers
of the more fundamental aspectsof the measurements.
Chaptersin thisbook wereprepared by theWorkingPartymembersnamed in theContentslist.
Copyrightof any particular chapter maybe theproperty of the authorsor their employers,but the
publicationrightshavebeengranted totheInstitute of Metals.TheWorkingPartywishestoexpress
its appreciation to Mr A D Mercer for the final editing of this booklet on behalf of the European
Federationof Corrosion.
Members of the Working Party are as follows:
Belgium L Clerbois
Finland 0Forsen
France P Lacombe
G Turluer
Germany W Fischer
E Heitz
Great Britain J W Oldfield
J C Rowlands
Italy F Mazza
G Rocchini
Netherlands FP IJsseling
Norway E Eriksrud
Spain J M Costa
Switzerland J Simpson
L Clerbois
Chairman EFC Working Party
Physicochemical Methods of Corrosion Testing
11. CHAPTER 1
INSTRUMENTATION,
PERFORMANCEAND
CALIBRATION
J. C. ROWLANDS
ARE Holton Heath, Poole, Dorset, UK
1.I INTRODUCTION
Since most electrochemical measurements relating to corrosion of metals are satisfied with a
sensitivity of 1 pV or 1 pA, modern instrumentation usually employs electronic operational
amplifiers where the noise limits control the range of measurements. The function of the opera-
tional amplifier is to amplify the potential (VJ applied at the input so that it can be displayed on
a low impedance analogue or digital meter (V,) as shown in Fig. 1.1.The output potential of the
operationalamplifierisproportional to thesourcepotentialand isrequired tohave sufficientinput
impedance to avoid polarisation of the potential source.
The basic corrosion instrumentation requirement involves the measurement of potential
difference.Currentsaremeasured asthe potentialacrossaresistor(R,) asshowninFig. 1.2,where
the potential difference is again determined with an operational amplifier. More sophisticated
measurements such aspolarisationcharacteristicsand zeroresistance ammetry involvethe use of
potentiostatswhich again use operationalamplifiersin a differentialmode. The potentiostat is an
instrument for maintaining the potential of an electrodeunder test at a fixed potential compared
witha referencecell,and thebasiccircuitissimilartothat forpotentialmeasurement with theearth
return circuit broken to an auxiliary electrode in the electrochemicalcell. Such a circuit would
maintain the potential of the test electrode at the reference cell potential. This potential may be
variedbyinsertingavariablepotentialsource(V,) intheinput circuitasshowninFig. 1.3.Theactual
cell potential (VJ and the current required to polarise the test electrodeto this potential may be
measured using the basic circuitsshown in Figs. 1.1andl.2 respectively.
A further modificationof thebasicvoltagemeasurement,giveninFig. 1.1,isthezeroresistance
ammeter(ZRA)showninFig.1.4.Themeasurementof currentasthepotentialdropacrossa resistor
(R,) shown in Fig. 1.2 involvesan error due to the value of resistor A. Thismay be overcomewith
theZRAin which thecurrentisdeterminedasthepotentialmeasured acrossa feedbackresistor(R,)
of the high gain operational amplifier.Thus, the current required to maintain electrodes A and B
shown in Fig. 1.4 at the same potential can be determined and displayed on a high impedance
voltmeter for which most commercialdigital multimeters are suitable.
1.2 PERFORMANCE LIMITATIONS
Theperformancecharacteristicsof commercial instrumentation should be supplied by the manu-
facturers.There are, however,measurementlimitationswhichare controlledby theelectrochemi-
cal cellunder investigation.Themajorlimitationsand requirementsarelisted under their separate
headings.
13
13. 1.2.1 Impedance Matching
Themeasurementof apotentialdifferencebetweentwometalsorametalwithrespecttoareference
electrode in an electrolyterequires the determination of a sourcevoltage (V,) which has a source
resistance ($). This source resistance is the sum of the resistivity of the electrolyte and the
polarisation resistanceof the electrodes.In order that the measuring instrument does not draw a
significantcurrent from the source, thus avoiding polarisation,it is necessary that the instrument
ormeterimpedancemustbehighcomparedwiththesourceimpedance.Suchacircuitof thesource
instrument impedance is shown in Fig. 1.5.For this circuit the actual potential measured (Vm)is
related to the source potential (V,) by equation 1. The resultant error if the amplifier input
impedance (R,) is not large is given in equation 2.
Rl
vm=vs-
Rl + RS
VS
Error = vs--vm 100%
Althoughit ismost desirableto haveR, verylargecomparedwith% averyhighvalueof I$may
present problems due to pickup of dc or ac error signals from other instruments and the mains
electricitysupply.A highsourceimpedance,&,isinvariablypresentinconductingelectrochemical
measurements in non aqueous media or when investigatingorganiccoatingson metals. Hence it
may be necessary to compromisethe accuracy of the measurement for the sake of reducing error
signals from other sources by limiting the impedance of the measuring instrument. For most
corrosionmeasurements an error of 1%is acceptable.
1.2.2 Ammeter Loading
The internal resistance of an ammeter (R,) in Fig. 1.2presents a potential drop in the circuit.This
may be overcomeusing the zero resistance ammeter shown in Fig. 1.4.
In many instruments such as potentiostats the value of resistance (R,) is chosen to provide a
limiting current in case the output becomesshorted.
1.2.3Amplifier Bias Current
Electronic amplifiers have a small bias current across the input. In electrochemical corrosion
measurementsit isarequirement that thisbiascurrent isvery small,suchaslessthan 1FAto avoid
polarising the electrodesof the corrosioncell.
1.2.4BandwidthLimitations
When followingpotentialchanges,egelectrochemicalcell capacitancechargingor discharging,ac
impedancemeasurements,orelectrochemicalnoisemeasurements,thebandwidth responseof the
measuring instrument may limit the application.
The frequency response of an instrument is usually specified as the 3 dB point (f3 dB) and is
determined by the R, and C, values for the input of the circuit shown in Fig. 1.5.
1
-- -
2.rrR, C,
‘3 dB (3)
The rise time (t>, i.e. the time to rise from 10%to 90%of the signal amplitude, is given by the
equation:
tr = 2.2 qc, (4)
15
14. Thus, from equations 3 and 4:
t = - 2.2
-r
2T f3dB
(5)
1.3 ERROR SOURCES
1.3.1 Resistance Noise
Resistance or Johnson noise in a resistor is caused by the thermal energy produced due to the
passage of a current. In a metallic conductor the Johnson voltage noise developed in a resistor is
givenby the equation:
E = v4kTRAf (6)
where k = Boltzmann’s Constant, T = temperature K, Af = noise bandwidth in Hz.
In high resistance circuitsthe noise bandwidth is limited by the time constant (t,>of the source
resistance (RJ in parallelwith the input resistance and the input capacitance(q)as shown in Fig.
1.5.The noise bandwidth is then given by the equation:
7 R + R
Asolutiontothisproblemiskeeptheconnectingleadsfromthesourcetotheinstrument asshort
as possible.
1.3.2Triboelectric Noise
Currents are generated by charges created due to friction between an insulator and a conductor.
Suchnoiseisdependent onthelength,degreeof movementand thematerialsof thecable,but may
give rise to large electrostaticcharges which can be minimised using low noise cables.
1.3.3 Piezoelectric Effects
Electrostatic charges are generated when a mechanical stress is applied to some materials, and
hencethecablescoupling thesourceto the instrument should notbe under tensionand not be free
to vibrate.
1.3.4 Thermoelectric Effects
Dueto differentparts of the circuitbeing at different temperatures, or at dissimilarmetal joints as
in thermocouples,an error potential in the millivolt range could be generated. Hence the test cell
and the instruments should be maintained at a uniform temperature and equilibrium reached
beforeundertaking measurements.
1.3.5Magnetic Field Loops
A loopof cablein the circuitbetween the test cell and the measuring instrument may be subjectto
a changein magneticfield whichdevelops a potential.Thispotential (EB)isgivenby the equation:
dB
E , a A-
dt
where dB/dt is the rate of change of magnetic field intensity and A is the area enclosedby the
loop of cable.
Thisis particularly relevant to remote sensingin field tests where potential errors of the order
of millivoltscan be generated.
16
15. 1.3.6Earth Loops
A potentialerror may be developed dueto the useof a commonearth linebetween the test celland
the measuring instrumentation, due to the current carried in the earth line. This error may be
minimised byearthingboth thecelland themeasuringinstrument to thesameearthingpointusing
separate cables.Similarly,the use of any current carryinglead should be avoided when makinga
potential measurement due to the voltage drop in the lead.
1.3.7Earth Potentials
When making measurements in process plant it is not uncommon to have various metallic
conductorsmakingtheconnectiontoearth,resultinginthepresenceof electricalearthsatdifferent
potentials, as shown in Fig. 1.6 where the electrolyte containing the electrodes forms an earth
connection with the plant in which the measurement is being undertaken. It is usually not
acceptableon safetygroundsto removethemainsearth,evenwhen theinstrumentation isearthed
through the test cell and, in such cases, the problem is best avoided using battery operated
instrumentation.
ELECTROLYTE
RESISTANCE
TO EARTH
Fig.l.6: Multiple earth
ELECTRODE
MAINS EARTH
1.3.8Capacitive Transformer Coupling
Although the primary and secondary windings on a mains transformer are considered as an
inductive coupling there is also an intercapacitive effect between the high and low voltage
windings.Thisresultsinanacvoltagetoearthsuperimposedonthesecondaryoutputasillustrated
in Fig. 1.7.This error source is usually only applicablewhen measuring very low potentials and
currentsand may be reduced by transformerearth screens.Theremedy isto usebattery powered
instrumentation.
I I
I1 Fig.l.7: Mains supply earth current
1.3.9Long Lead Cable Capacitance
The use of long leadsbetween the potential sourceand the measuring instrument can result in an
effectivechangeof the output capacitanceof themeasuringinstrument, thus alteringits frequency
response.Typicallythe capacitanceof a twin core cable is of the order of 100pF/m. The effecton
the frequencyresponse can be calculated using equations 3 or 4. The remedy to this problem is to
keepthecablesasshortaspossibleor,wherelongcablesystemsareunavoidable,adriveramplifier
at the source may be required.
17
16. 1.4 INSTRUMENT CALIBRATION
Theinstrumentation describedin this chapterisverydependenton themaker supplyinga reliable
specification.Thereafter it is usually only necessary to check the instrument periodicallyusing a
commercially available standard potential or current source. Any deviation found should be
correctedby followingthe instrument manufacturer's settingup procedure or by returning to the
maker for recalibration.
1.5 REFERENCES
cations. Chapter 13.Published by JohnWiley & SonsInc., 1980.
1.A. J. BARD and L. R. FAULKNER:in ElectrochemicalMethods - Fundamentals and Appli-
2.InstrumentalMethodsinElectrochemistry,SouthamptonElectrochemistryGroup.Published
by EllisHorwood, Chichester, U. K., 1985.
3.J. F. KEITHLEY,J. R. YEAGER and R.J. ERDMAN:inLow LevelMeasurements.Revised3rd
Edition. Published by KeithleyInstruments, Ohio, U. S. A., 1984.
18
17. CHAPTER 2
DESIGN OF ELECTROCHEMICAL
CELLS
F.P. IJSSELING
Corrosion Laboratoryof RoyalNetherlands Naval College, Den Helder, The Netherlands
2.1 INTRODUCTION
The general purpose of an electrochemicalcell is to be able to conduct electrochemical measure-
mentsona metalsample-theworkingortestelectrode-which shouldhaveawell-definedareaand
surface condition. In the cell the working electrode is brought into contact with the corrosive
environment,usually under well defined conditionsi.e.,with respect to environmental composi-
tion, temperature, flow, etc. [l,21. Apart from the working electrode an auxiliary electrode is
required to allow electric current to flow through the cell during polarization; in addition a
reference electrode is needed to be used as a zero point for measuring the potential difference
betweentheworkingelectrodeand thesolution.Mostelectrochemicalcellsarebased on theabove
mentioned three electrodesystem.Only in simplecases,for instancewhen the measurements are
to be made under zero current (e.g.the free corrosion potential) or very low current flow, a two
electrodecell may suffice.Then a referenceelectrodeis used which, under low current flow, can
alsoact as the auxiliaryelectrode.Thecontentsof thischapter areaimed at threeelectrode cellsto
be used on polarizingthe workingelectrode.Inthe followinga list of requirements isgiven, which
are further divided into the following sections:
1.electrochemical/electrical
2. solution
3. construction.
2.2 ELECTROCHEMICAUELECTRICALREQUIREMENTS
The first and foremost requirement is the need to establish a homogeneous electric field at the
workingelectrodesurfaceand consequentlyanevencurrentdistribution.Inthisrespecttherelative
positions of the working-, auxiliary-and reference-electrodesare of paramount importance.
(a).The best way to obtain an even current distribution is to position the auxiliary electrodeas
symmetricallyas possible to the working electrode, the latter preferably having the smaller or at
least the same dimensions (Fig.2.1).A large distance between the electrodes will also promote a
more even current distribution. However, too large a distance may createan unduly high ohmic
resistance in the cell [3,41.
(b).Thereferenceelectrodeshouldbe positioned insucha wayasnot todisturb theevencurrent
distribution by shieldingeffects,at the sametime keepingthe uncompensated ohmicresistancein
the solution at a minimum. The classical solution is to place the reference electrode in a separate
compartment and to use a Luggin capillary,the tip of which is to be placed 1-2outer diameters of
the capillary opening from the working surface (Fig. 2.2a) [5]. In any case, positioning of the
referenceelectrodeshould be possiblein an exactand reproducible way. However, other satisfac-
tory solutionshavebeen found,includingthe sidechannelprobe and the back probe (Fig. 2.2b,c).
The first consistsof a small opening at the side of the working electrodeas a connectionwith the
referenceelectrode,while in the second case the Luggin probe is attached to the back of a narrow
hole drilled through the electrode 16-81.
(c).In all cases the cell should provide a well-defined electricalsystem; in the first place good
electricalconnectionswith the electrodesshould be ensured, which even atlongerexposuretimes
19
18. WE
a.
E Fig. 2.1: Simple three-electrode cell
containingworking electrode(WE),
auxiliary electrode (AE)and refer-
enceelectrode (RE).
rE
b.
I RE
IRE II.
r
c.
Fig. 2.2:Differentmethods of minimizingIRdrop: a. Luggin capillaryb. side channel c.back
channel.
a.
Fig.2.3:Exampleof cellwithseparatecompartmentforAE,connectedviaamembraneorcoarse
sintered glass: a. single separation b. double separation by which an extra compartment is
created in which the possibly contaminated solution can be flushed regularly by means of a
stopcock.
20
19. donot lead to failures.Theproblemismost urgent incaseswhere aggressiveionsarepresent and/
or high temperatures that could eventually lead to oxidationof the contacts.The use of a double
wiring system and detection of possible contact resistance is recommended. The inner electrical
path should alsobe consideredand possibleerrorsresultingfromcloggingof the Luggincapillary
or the formation of poorly conductiveareas in the cell should be avoided as these could give rise
to unduly high electricalresistancesin the power loop or the control loop.
Thepick-up of electricaland electromagneticnoise should be avoided, especiallyin the caseof
the application of electrochemicalpulse methods and sensitivedc measurements involvingsmall
potential perturbations and currents.
The factorswhich should be taken into considerationin such cases,involve:
1.Earthing-inmany laboratoryapplicationsthe workingelectrodeisearthed or kept atvirtual
earth; in a number of commercial potentiostat designs it is possible to disconnect the working
electrodefromtheinstrument zero.Allmeasuringinstrumentsand ancillaryequipment shouldbe
fed from the same line connection. In addition all instrument houses and bodies of ancillary
equipment should be interconnectedand earthed at one point for safetyreasons [9,101.
2. Shortconnections- generallythe control instrument e.g. potentiostat shouldbe located near
the cell to ensure short electricalconnections.If this is not feasiblethe use of a buffer amplifier in
the control loop near the test cell is to be considered.
3. Referenceelectrode-a high resistance which iscommon in many commercial products may
induce noise and even instabilityin the control loop.
The connection with the potentiostat should be shielded; shielded reference electrodes are
commerciallyavailable.However,in the caseof excessivecapacitanceto ground instabilityof the
controlpotentialmay occur,insuchcases theshieldingmaybemaintainedatthereferencepotential
via an operationalamplifier [5].Anotherpossibilityisto provide aplatinum wire tobe used asthe
reference in the control loop, next to the reference electrode proper.
4.Screening-if necessarythecellshouldbeplaced ina Faradaycageor screenedelectricallyand
electromagneticallyby other means as, for instance,by tin plate,copper foil, etc.
(e)Also, in the case of sensitivehigh-performancemeasurements the potentiostat and the cell
design should be matched to each other. Generally the requirements for rapid and accurate
response are: low total cell resistance, low-resistance reference electrode, small-area working
electrode,and low stray capacitances [ll,121.In general it can be said that the faster the response
of thepotentiostat,thegreateritssensitivitytointerference.Soitisnotadvisabletousefastresponse
potentiostats for applications where this feature is not required.
2.3 SOLUTION REQUIREMENTS
The solutionrequirements depend to alargeextenton thepurpose of the measurement. Generally
the solution consists of a solvent, the electrochemically active redox system and possibly
complexingagents,buffers, a supportingelectrolyteto enhance the electricalconductivity,etc. In
thecaseof industrial measurements,e.g.corrosionmonitoring,thesolutionisgenerallythat being
used in the process.
In laboratory experiments special precautions are often required. The main points can be
summarized as follows:
(a). The solution should be free from contaminants as even small concentrations may exert
relativelylargeeffectson theelectrochemicalreactionsby adsorption effects,etc.Soin appropriate
cases freshly double-distilledwater and analytically pure salts should be used.
(b).Moreover, the solution should not be contaminated during the measurements, either by
componentsof the materialsto be used forcellconstructionorby reaction products evolvedat the
current carryingelectrodes.The first possibility should be avoided by proper materials selection.
To eliminate the second one the electrochemicalreactions at the working as well as the auxiliary
electrodeshould be taken into account.For this reason the auxiliaryelectrodeis often placed in a
separatecompartment,thecontentsof whichareelectricallyconnectedwith the test solutionvia a
membrane,glass frit, etc. (Fig.2.3).
21
20. As it isnot feasibleto separate the working electrodefrom the test solutionthe first thing to be
considered is to keep the cell volume large enough to avoid the accumulation of an excess of
corrosion products [131.Monitoringthe corrosionproduct concentrationis advisable and the cell
contents should be refreshed as soon as unduly high concentrations arise. Other possibilities
include the use of special cells which permit continuous or intermittent replacement of the test
solutionand, inthe limit,theuseof once-throughconditions.Apartfromtheauxiliaryelectrodethe
reference electrode is also sometimes placed in a separate compartment, for instance when the
releaseof chlorideionsin the test solutionfroma saturated calomelelectrodecannotbe tolerated.
Even then contaminationby release of chlorideions may become a problem, whichcan be solved
by placinganextracompartmentbetweenthe cellproper and thereferencecompartment(Fig.2.4).
Referenceelectrodeswith intermediate electrolytechambers are commerciallyavailable.
(c).Thetest solutionshould maintain its intended corrosiveproperties during theexperiment,
e.g.O,-content, pH etc.If necessary, the concentrationof the corrodents should be monitored and
kept ata specifiedlevel.Specialcaremustbeexercisedwhenbiologicallyactivesolutionsare used,
in which the bacteriologicalcomponentsinterferewith the corrosionprocess, suchsolutionsoften
being prone to ageing. Seawater is an example of a solution in which significant changes of the
corrosiveproperties may occur on storing and during exposure.
(d).In a number of casesit might be desirableto provide a constant gas-atmosphere in the cell
abovethe solution.Examplesare the removal of oxygenby the addition of nitrogen,hydrogen or
mixtures thereof, or the creation of an oxygen-rich environment by the addition of oxygen. For
many purposes a simple provision to introduce gas in the cell under a slight overpressure,
combined with a device to distribute the gas in the solutionin the form of smallbubbles, suffices
(Fig.2.5).
However,in separate cases,for instancewhen poisonous gasesare involved,a more elaborate
methodology is required.
2.4 CONSTRUCTION REQUIREMENTS
The constructionaldesign of a cell also depends to a large extenton the type of measurements to
be made.Generallythe constructionshould be rugged and allow for easyassemblyand disassem-
bly as required, for instance, for cleaning purposes. Moreover the positioning of the electrodes
should be accomplishedeasily and in a reproducible way.
The followingpoints should be kept in mind:
(a). materials - in many cases glass is satisfactory, although sometimes PVC, PTFE or other
plastics are to be preferred, keeping the contamination problem in mind. Often it is possible to
constructa cell from standard laboratory glasscomponents,although care should be taken in the
useof glassat temperatures aboveabout 6OoCbecauseof the possibility of contaminationby silica
resulting fromdissolution of the glass [141.
(b).temperature - in many cases it will be desirable to study the corrosion process at a well-
defined temperature, necessitatingtemperature control of the cell.The following methods can be
envisaged:
1.directintroductionof heatingand/or coolingelementsintothe test solution(Fig. 2.6a),taking
precautions to avoid current leakage from the element into the solution;
2. externalcirculation of the test solution through a heat exchanger (Fig. 2.6b);
3. positioningof the test cell in a thermostatted bath (Fig. 2.6~);
4. the use of a double walled cell,circulatingthermostatted water through the double wall to
ensure temperature constancy of the inner test solution (Fig.2.6d).
It mustbe remarked that forhigh-temperatureapplicationsthe Luggincapillaryalsoshouldbe
similarly equipped with a double wall for circulationof thermostatted water.
(c).workingpressure-most testsareperformed under atmosphericpressure, obviatingspecial
requirements.However,insomecasesit isessential to testunder increased pressure, so the cellhas
to be adapted to this purpose. Generally autoclaves have been used for this purpose, one of the
problems involved being the gas-tight isolation of the electrodeconnections1151.
(d).solution flow - frequently the flow conditionsat the working electrodesurface are not too
well defined.However, if so required this aspect can also be taken into account, for instance, by
using rotating disc or cylindrical electrodes.
Another possibilityis the use of tubes or plate electrodesin ducts, channels or test loops, e.g.
generally in flow through cells (Fig.2.7 and Chapter 3) [16].
22
21. Fig. 2.4:Electrochemicalcell with
separate compartments for the
reference and the auxiliary elec-
trodes; to prevent contamination
by the solution in the reference
electrode the contents of this
compartment can be flushed out
regularlyby means of a stopcock.
Fig. 2.5: Closed cell with
inlet and outlet for gas.
1
a.
C.
b.
til + dIT0
Fig.2.6:Differentmethods of temperaturecontrol:a.directb. via externalheat-exchangerc.via
temperature bath d. via double walled cell
C.
I RE
d
Fig. 2.7:Different methods for creating solution flow (see
also Chapter 3):a. rotating disk electrodeb. rotating cylin-
der electrodec. pipe system with one test pipe as WE and
two as AE d. pipe system with one test pipe as WE and a
central rod or wire as AE. In cases c and d the RE can be
1
a. b. connected via a hole through the wall of the WE.
23
22. Fig. 2.8:Exampleof electrochemicalcell for SCC testing u n c r potentiostaticcontrc
WE: workingelectrode,consistingof notched rod possibly isolated fromthe solutionexceptat
thenotched position;AE:auxiliaryelectrode,consistingof circularplatinum gauze around the
working electrode; L :Luggin capillary to reference electrodeRE; G :inlet for gas; S :seals.
1 2 3 L 5 6
a. b.
Fig. 2.9:Electrochemicalcell with separated creviced anode and cathode for crevicecorrosion
testing [201.
a. Generalset-up
1.graphite electrode connected to earth for safetyprecautions;
2. heating element;
3. contact thermometer;
4.cathode (in principle the same material as anode material under test);
5. holder for anode, provided with a large number of microcrevices,obtained by pressing a
6. reference electrode.
rubber O-ring on a scratched metal surface;
b. Detail of anode holder:
7. PVC-pressingplug;
8. anode specimen;
9. rubber O-ring;
10.platinum wire making electricalcontact with anode.
24
23. (e).illumination-insomespecialcasesitmightbe desirableto controlthelight conditionsatthe
working electrode, for instance varying between exposure under dark conditions or under
illumination within a given spectralarea.
(f). localcorrosion-alsodependingonthetestinvolved,itmaybenecessarytointroducespecial
conditions, aimed at the creation of local corrosion attack. Well-known examples are stress-
corrosioncrackingand fatiguecorrosioninvolvingtestcellsin whichthe workingelectrodecanbe
exposed under stressed condition (Fig.2.8), and crevice corrosion, in which case the working
electrode may be provided with artificial crevices 117, 181. For some applications, e.g. crevice
corrosionandbimetalliccorrosiontestingthecorrodingmetalsurface(anode)and thecathodehave
been separated in space, being electrically connected externally via an appropriate measuring
instrument, e.g. a zero resistance ammeter (Fig.2.9) 119,201.
In a number of cases the requirements as listed above are conflicting and contradictory,
necessitatinga compromise.However, generally it is not necessary to meet all requirements. The
ultimate requirements to be incorporated in a certain cell design are of course strongly related to
the purpose of the test and the test method and should be selected accordingly. In the official
standardsand guidelines only a fewstarting points canbe found [e.g.211. A number of textbooks
and papers contain usefulbasic information1e.g.1,2,5,6,10,221. In theliteraturea large number
of cells have been described. Apart from the references already given - a more of less random
selectionincluding somegeneral informationin addition is presented in refs. 123-411.
2.5 REFERENCES
1. D.T. Sawyer and J. L. Roberts: in Experimental electrochemistry for chemists, Chapter 3.
Published by JohnWiley & Sons,Inc. 1974.
2. N.D. Greene: in Experimental electrode kinetics. Published by Rensselaer Polytechnic
Institute, Troy,New York, 1965.
3. J.E. Harrar and I. Shain:Anal. Chem., 1966, 38- 1148.
4. W.A. Mueller:Corrosion, 1969,s 473.
5.R. Greef etal.:inAdvancedinstrumentalmethodsinelectrodekinetics,Southampton Electro-
chemistryGroup. Published by Ellis Horwood Ltd., Chichester, U. K., 1985.
6.J.A.vonFraunhoferand C.H.Banks:inPotentiostatand itsapplications. Chapter3. Published
by Buttenvorths, London, 1972.
7. I. Gillet: Bull. Soc.Chim.France, 1962,377,
8. U. Landau, N.L. Weinberg and E. Gileadi:J. Electrochem.Soc., 1988,135,( 2), 396.
9.R. Morrison:in Groundingand ShieldingTechniquesin Instrumentation. Published by John
Wiley & Sons, Inc., 1967.
10.Variouspotentiostat instructionguides and leafletsto be obtained from manufacturers, for
instanceEG and GPrincetonApplied ResearchApplication noteG-2: Groundingand Shieldingin
ElectrochemicalInstrumentation: SomeBasic Considerations.
11, D.D. Macdonald: in Transient techniquesin electrochemistry. Chapter 2.2. Published by
Plenum Press, New York and London, 1977.
12.B.D. Cahan, Z. Nagy and M.A. Genshaw:J. Electrochem.Soc.,1972,119,64.
13.L. Clerbois,E. Heitz,F. P. IJsseling,J. C.Rowlandsand J. P.Simpson:Br.Corros.J., 1985,X!,
(3),107.
14. A. D. Mercer and G. M. Brook:Trib.Cebedeau, 1975,Aug.-Sept., (417/418),299.
25
24. 15.J. Postlethwaiteand R.A. Brierly:Corros.Sci., 1970,'UJ, 885.
16.Yu.V. Pleskovand V.Yu.Filinovskii,(transl.by H.S. Wrobla):in Therotating diskelectrode.
Chapter 9 .Published by Consultants Bureau,New York and London,1976.
17.E.A. Lizlovs:J. Electrochem.Soc., 1970, 117,1335.
18.F.P. IJsseling:Br. Corros J., 1980,
19.I.R. Scholeset al.: in Proc. 6th Eur.Congress on Met.Corrosion. Published by the Societyof
51.
Chemical Industry, London, 1977,161.
20. J.M. Krougman and F.P. IJsseling:in Proc. Intern. Workshop ElectrochemicalCorrosion
Testing, Ferrara; 1985.Monograph vol. 110.Published by DECHEMA, Frankfurt,FRG, 1985,135.
21.A.S.T.M. G5-82:Standard PracticeforStandard ReferenceMethodformakingPotentiostatic
and PotentiodynamicPolarizationMeasurements.Publishedby A. S.T. M., Philadelphia,U. S.A.
22.TechniquesinElectrochemistry,Corrosionand Metal Finishing,a Handbook, part A (ed.A.
T. Kuhn).Publishedby JohnWiley & Sons, Inc., 1987.
23.R.E. Geisert,N.D. Greeneand V.S. Agarwala:Corrosion, 1 9 7 6 , z 407.
24. B. Gerodetti and K.H.Wiedemann:Werkst.und Korr., 1977,a 173.
25.J.R. Scully,H.P.Hack and D.G.Tipton:NACE Intern.CorrosionForum, Boston,1985,paper
no. 214. Published by NACE, Houston, Tx., U. S. A., 1985.
26.T. Hakkarainen: in Proc. 8th European Corrosion Congress, Nice, 1985.Published by Soc.
ChimieIndus., Paris, France 1985,26.
27.P. E. Francis and A.S. Dolphin: Br. Corros.J.,1984,B 181.
28.R. Manner and E. Heitz: Werkst. und Korr., 1978,a559.
29.H. Lajain:Werkst. und Korr., 1972,23- 537.
30.T. Suzuki and Y. Kitamura: Corrosion, 1972,a 1.
31.A. Tamba:Br. Corros.J., 1982,E 29.
32.D.D. Macdonald,B. C.Syrett and S.S.Wing: in NACE Intern. CorrosionForum, Houston
1978,paper no. 25. Published by NACE, Houston, Tx., U. S. A., 1978.
33. F. Hunkeler and H. Bohni: Werkst.u. Korr., 1981,32- 129.
34.J.R. Scully, H.P. Hack and D.G.Tipton:Corrosion, 1986,42- 462.
35.T. Suzuki,M. Yamabe and Y. Kitamura:Corrosion, 1973,a70.
36. R. B. Diegle:Materials Performance, 1982,21, (3),43.
37. E. Bardal, R. Johnsen and P. 0.Gartland: Corrosion, 1984, 628.
38.G.A.Gehring,Jr.andJ. R. Maurer:inNACEIntern.CorrosionConf.,Toronto,Canada, 1981,
paper no. 202; Published by NACE, Houston, Tx., U. S. A.
39.D. A. Jones:Corrosion, 1984,40- 181.
40. 0.Varjonen and T. Hakkarainen: in Proc. 8th Europ. Corrosion Congress, Nice, 1885.
Published by Soc. ChimieIndus., Paris, France, 1985,44.
41. L.L. Shreir:Werkst.und Korr., 1970,2l- 613.
26
25. CHAPTER 3
ELECTRODE DESIGN
E. HEITZ
Dechema Institute, Frankfurt, FRG
3.1 INTRODUCTlON
An electrodemay be defined as a solid electron conductorwhichis in contact with a liquid (solid,
gaseous)ion conductor (electrolyte).At the interface chargetransfer reactionstake place. During
corrosionprocessesthis charge transfer involvesanodic and cathodicpartial reactions of various
kinds which are dependent on many parameters and which have to be taken into account when
designing an electrode.
Theelectrodemaybea metalor any electron-conductingmaterial,forexampleasemi-conductor
which acts as a sourceor sink for electrons.
The electrolyteusually consistsof a solutionof salts,acidsor bases in water or protic solvents,
such asalcohols,carboxylicacids, etc. [ll. Pure solvents,too, can act aselectrolytesif enough con-
ductivityby autodissociationisproduced (water,methanol, ethanol,etc.).Moreover, moltensalts
constituteelectrolyteswith sometimesextremelyhighconductivity.It isimportantto statethat the
electrolyte should be free fromany electronicconductivityotherwiseno electrochemicalreaction
will occur at the electrode/electrolyte interface.
Asdiscussedinapreviouschapterelectrodesareused asworkingelectrodes,counterelectrodes
and reference electrodes.In this chapter emphasis is laid on working and counter electrodes.
Working electrodesare further divided into:
(a)Electrodesfor polarisation measurement in electrochemicalcells
(b)Electrodes for free corrosion experiments.
Designs for working electrodesare diverse.Therefore,in the followingsectionsonly the most
important design principles will be discussed.
Counterelectrodesshould be made froma corrosionresistantmetal, forexamplea noble metal,
and their design should allow for uniform current (potential)distribution and free convectionof
aggressiveagents to the working electrode.
3.2 SIZE OF ELECTRODES
Sizeand geometryareamongthemostimportant aspectstobeconsideredindesigningorchoosing
a particular electrode configuration [21.
In order to study kineticsand mechanisms small plates, foils, spheres,discs, rods or wires are
used.Theypermithighcurrentdensitieswithminimumohmicheatingorcaseswherehigh-current
sourcesare not available.If localisedcorrosion has to be studied a certain minimum area must be
guaranteed. The size has to be chosen so that the corrosion effect occurswith high probability on
the electrodesurface.Thisisespeciallyvalid when measuringthepotentialdependenceof pitting,
stresscorrosion cracking,etc.
27
26. Anotherreasonfor designingelectrodesof a certainsizeisthe combinationof weight losswith
electrochemicalmeasurements.Thus, very smallelectrodesshould not be used when gravimetric
measurements are required.
On the other hand, the use of microelectrodeshas attracted interest in the study of corrosion
effectson a microscopicscale.Microelectrodesthe sizeof a few micrometershavebeen described
[3,41.
Problems associated with size can be assessed by the principles for scalingcorrosion tests. A
report has been published by the European Federation of Corrosion Working Party on
“Physicochemicalmethodsof corrosiontesting-fundamentals and application”[51whichincludes
information on specimen size for testing uniform, bimetallic and pitting corrosion and stress-
assistedenvironmental cracking.
3.3 DESIGN WITH REGARD TO VARIOUS CORROSION PARAMETERS
3.3.1 Current and potentialdistribution
In principle, currentmeasurements onlygiveintegral valuesas current per total electrodesurface
exposed.Onrealelectrodesurfacesthecurrentdistribution ismoreorlessnon-uniform.Thisholds
particularly for cases of non-uniform corrosion attack.
Since measurement of the current distribution is only possible with considerable effort (seg-
mented electrodes),some general rules for minimizing non-uniformity will be presented (for
theoretical background, see refs. [5- 71).
If a working electrode(length L) with a parallelcounter electrode(distanceh), asin Fig. 3.1, is
considered the current distribution is the more uniform the larger the ratio L/h. From this the
followingrecommendationscan be given:
-adesignshouldbechosenwithworkingand counterelectrodesascloseaspossible(takinginto
accountpossible interferenceof anodic and cathodicreaction products)
-the smallestelectrodespossible should be selected(withoutinterferingwith statisticaleffects,
section3.2).
From the concept of the dimensionlessWagner Number [5, 71 the following conclusionsare
derived:
- if possible, low conductivity solutions should be avoided
- polarisationresistancesshould be as high as possible.
Thislast statement means that electrodesproducinghigher currentsduring polarisationshow
a more non-uniformcurrent distribution, sincehigh currents are correlated with low polarisation
resistances.
Therulescited givesomeusefulworking/counter electrodedesignswhichareshowninFig.3.2.
It should also be mentioned that the cylindricalgeometry has also been successfully used to
minimize ohmicdrop in media of low conductivity.
When using plate electrodes it is necessary to consider whether the reverse side should be
insulated. This is not necessary in systems with high conductivity and with sufficient distance
between the electrodeand the cell wall. Insulated plate electrodesproduce problemsas a result of
crevice formation.Thebest solution is, however, to provide the plate electrodewith two counter
electrodessituated on both sides of the plate.
Referenceelectrodesaregenerallyused togetherwith Haber-Luggincapillaries(fordetails, see
Ref. [2]).The design and position of these capillaries pose current and potential distribution
problems. In order to minimize ohmic drop they have to be placed as close as possible to the
electrodesurface.Butif thedistanceistoosmalltheyactasacurrent shieldand non-uniformcurrent
distributionarises.In practice the tip of the Luggin probe should be at a distanceof about 2d from
the working electrode where d is the external diameter of the capillary.
28
27. ..../..#.. ............
Pr
t-
Fig. 3.1:Current distribution on parallel plate electrodesof length L and distance h
9
cylindrical geometry
? Q Q
rotating disc parallel plate
(parallel plane arrangement)
Fig. 3.2:Electrode design with various geometriesto minimize non-uniformity of current
distribution
3.3.2 Mass transfer
Mass transfer influenced corrosion reactionscan only be investigatedon electrodesor specimens
which are exposed under well-defined hydrodynamic conditions.A number of electrodedesigns
used in the past are shown in Fig. 3.3 [8,91.
For basic studies the most suitabletype isthe rotatingdiscwhichiseither placed in the end face
of a rotating cylinder,Fig. 3.2, or isused asa freerotating systemwith anintegral shaft,Fig.3.3 [71.
Further modelsare the rotatingcylinder, freeor co-axial[ll,121,and pipe and channel flow [7,141.
Compared with the situation with the rotating disc channel and pipe flows will generally
becometurbulent atrelativelylow flowratesalthoughtheextenttowhichthishappenswilldepend
on thedimensionsof the cross-section(characteristiclength).Theirpracticaladvantageisthat they
affordbetter accesstothemeasurement sectionthando rotatingmodels,buttheflowisnot soeasily
achieved. The test pieces are arranged and the dimensions chosen so that the test pieces are
electricallyinsulated and inserted flushin the flowpath. Caremust be taken to ensure that thejoin
between the wall of the pipe and the test pieceishydrodynamicallysmooth,thisbeing particularly
necessary with the flowchannel.Another possibilityisto constructthe test piece itself asa rotating
entity. This is the case, for instance,with the rotating disc and the rotating cylinder.
A special arrangement is the ring-discelectrodewhich is currently often used for kinetic and
mechanistic studies [13]. In this case the rotating disc is surrounded by a ring electrode which is
separated by a thin non-conductinggap.
Formeasurementsindisturbedturbulentflowwhichisimportantformanypracticalapplications
the segmentedpipe techniquehasbeensuccessfullyapplied [14,15].Thepipe isdivided into rings
which are arranged in electrical insulation from each other, permitting local mass loss measure-
ments and electrochemicalmeasurements along the pipe axis.
29
28. I
@-
Rotating disc Disc in casing Rotating Coaxial cylinder Flow in Flow in
cylinder a channel a pipe
Fig. 3.3:Electrode arrangements for investigationof flow dependent corrosion [8,91
Counter electrode
Referenrn elnrtrodn- I .
‘R ICooler
Heater
Holder
Thermosensor
PTFE Cover
..-.-. -..-- -.--_.---
--..
PTFE Container
working electrode/
specimen
Thermal insulation
Corrosive medium
Conductivity sensor
Fig.3.4:ElectrodesforCERTexperimentsinaPTFEcellforaggressivecorrosionconditions1211
3.3.3 Heat transfer
Thesimultaneousmeasurementof heattransferand electrochemicalquantities hasbeen described
in the literature [16-181.Tubular and disc-shaped specimensproved to be suitable although diffi-
culties arose in obtaining uniform heat transfer on the surface. For example, it was found that
crevicecorrosioncan occur in the side insulation of disc electrodes.This effect may be attributed
in some cases to the decreased heat flux and therefore higher surface temperature on the outer
circumferenceof the specimen.
30
29. Electrodesof the "cooling finger" typecould alsobe used for controlledheat fluxexperiments.
Specificexperiencewith a rotating coolingfingerhas been made available1191and a combination
of ring-disc with heat transfer measurements has recently been described [201.
3.3.4 Mechano-chemicaltesting
Mechano-chemicaltesting comprisesthe suitable applicationof stress, pressure, bending, thrust
and torsion on specimensin specificmedia either in a time constant or cyclic mode. There are no
specialproblemsinapplyingpotentialsorcurrentstosuchspecimensprovided suitableprovisions
for electricalcontactsand electrodeinsulation are made.
As an example, provisions for electrochemicalmeasurements during CERT(Constant Exten-
sion Rate Testing)are shown in Fig. 3.4 [21,221.
Specialrequirementsforthisset-upare thatforstandardized specimensforCERTthe insulated
specimenholder shouldbe of thesamematerialasthetest materialand there shouldbe lowfriction
sealing of the specimenholder.
Standardized specimensfor CERTareconventionalcylindricalspecimens,notched or smooth,
which are loaded at a slow strain or continuous extension rate 1231.
In order to avoid bimetallic corrosion the specimen holder of the same material has to be
insulated by a polymer coating. More noble materialcan be used if bimetallic corrosionis absent,
but this has to be proved by separate electrochemicalexperiments.
DuringCERTanextremelyslowlymovingspecimenhas tobe sealed.A design asshowninFig.
3.4 may be chosen.
3.3.5 High temperature/high pressuretesting
Testing under high temperature/high pressure conditions is done in normal autoclaves or in
refreshed autoclaves.There are a number of problemsto be overcome,which include sealingand
insulation of the electrodes, the choice of a suitable (pressureand temperature stable) reference
electrodeand possibly control of the parameters discussed above.
Relevant reviews [24, 251 and special papers on technical solutions to problems exist, as, for
example, for high temperature reference electrodes [26,27l, apparatus with exchangeableelec-
trodes formeasurementsup to 330°Cand 300bar [281, CERTexperimentsunder high pressure and
temperature [29]and electrochemicalmeasurements for hydrogen permeation up to 600bar [30,
311.
3.4 PROBLEMS CONCERNING APPLICATION TECHNIQUES
Essential experimental problems to be overcome are [21:
(a)defining the working area
(b)making good electricalcontact
(c) holding the electroderigidly.
If an electrode is exposed in an electrochemicalcell it is necessary to know its working area.
Definedworkingareasmaybeachievedby theuseof coatingsorcompressiongaskets.A systematic
comparative study of a number of systemsrevealed that hot-cured epoxy resin and compression
type gaskets are relatively the best solutions 1321. One of the problems encountered with pas-
sivating systems is crevice corrosion at the threephase boundary electrode-coating-electrolyte.
Due to the low ratio of volume/surface in the crevice small anodic polarisation or even free
corrosionleads to preferentialattackin the crevice.Coatingsand gaskets freeof crevicestherefore
have to be applied. A survey of the literature on attempts to overcome this problem has been
published 121.
Inmakingelectricalcontacttoelectrodestheresistanceof thecontactbetween the specimenand
thecurrentlead shouldbeaslowaspossible.Pressurecontactisoftenused,but problemsmayarise
as a result of passivation or contaminationof the contactif an electrolyte should enter the contact
31
30. area.Spotweldingor solderingareattractivesolutions,but caremustbe takentoavoidpreferential
attack at the contact. It is useful to protect the contact area by PTFE tape or an electroplater's
lacquer[21.
If thecurrent leadisanoblemetal,forexampleplatinum, and if theworkingelectrodeisametal
in the active state, it is possible to exposethe noble metal wire to the electrolyteprovided its area
issmallcomparedwith thatof theworkingelectrodeand itspolarisationresistancelargerthan that
of the working electrode.
Electrodeholders need tobe held rigidlyinplaceand should beeasilyremovable.Ground glass
jointsorscrewcapadapters withcompressionringsareuseful solutionsforsimplelaboratorywork.
Descriptions of techniques for cases of extreme conditions of temperature, pressure, flow and
mechanical load are described elsewherein this volume.
3.5 REFERENCES
M. Fontana and R. Staehle),London, 1974, & 149.
1.E. Heitz:in AdvancesinCorrosionScienceand Technology.Publishedby PlenumPress, (ed.
2. A. T. Kuhn: Techniques in Electrochemistry,Corrosion and Metal Finishing. Published by
JohnWiley & Sons, Inc., 1987.
3. A. M. Bond, M. Fleischmann and M. Robinson: J. Electroanal.Chem. 1984, 168.299.
4. R. T. Atanasoski, H.S. White, W.H. Smyrl:J. Electrochem.Soc. 1986,133,2435.
5. L. Clerbois,E. Heitz, F.P.IJsseling,J.C.Rowlands and J.P.Simpson:Br. Corr.J. 1985,a107.
6.N. Ibl:Current Distribution,in ComprehensiveTreatiseof Electrochemistry(Eds.E. Yeager,
J. 0.M. Bockris, B. Conway,S. Sarangapani)Vol. 6, Plenum Press, 1983,b.
7.E.Heitzand G.Kreysa:PrinciplesofElectrochemicalEngineering;PublishedbyVCH-Verlag,
Weinheim, 1986.
8.U. Lotz and E. Heitz: Werkst.u. Korr. 1983,_34,454.
9. DIN Specification50920 CorrosionTesting in FlowingLiquids: Beuth Verlag, Berlin, 1987.
10.N. Ibl and 0.Dossenbach:ConvectiveMass Transport, see Ref.[61.
11.0.Dossenbach:Ber. Bunsenges.Phys. Chemie 1976, 34.
12. E. Heitz, G. Kreysa and C. Loss:J. Appl. Electrochem. 243 (1979)
13. Instrumental Methods in Electrochemistry;Southampton Electrochemistry Group. Pub-
lished by Ellis Horwood Limited, Chichester,U. K. ,1985,113.
14.U. Lotz, M. Schollmaierand E. Heitz: Werkst. u. Korr. 1985, 36- 163.
15.T. Kohleyand E. Heitz:in The Use of SyntheticEnvironments for CorrosionTesting', STF
970. ASTM, Philadelphia, U. S. A., 1988, 235.
16. Corrosionunder Heat-TransferConditions:MTI Publ.Nr. 17, MaterialsTechnology Insti-
tute, Columbus,Ohio, U. S. A., 1985.
17.Ya.M. Kolotyrkin,V.S. Pakhomov, A.G. Parshin and A.V. Checkhovski:in 'Proceedings of
the 9th International Congresson Metallic Corrosion', NACE, Houston, Tx, U. S. A., 1984,z 1.
18.M. Yasuda, M. Okada and F. Hine: Corrosion -NACE, 1982,38- 256,
19. M. Shirkhanzadeh,V. Ashworth and G.E. Thompson:Electrochim.A. 1988,33- 265.
32
31. 20. ArbeitsgemeinschaftKorrosion e.V., ArbeitsblattWl, Werkst. u. Korr. 1988,B.
21. H.-G. Fellmann,H. Kalfa, U. Schareand E. Heitz:Werkst. u. Korr. 1989,a 34.
22. E. Heitz, R. Henkhaus and A. Rahmel:Korrosionskunde im Experiment, Verlag Chemie,
1983,(Englishedition in preparation).
23. DIN Specification 50922, Untersuchung der Bestindigkeit von Metallen gegen Span-
nungsrisskorrosion, Beuth-Verlag,Berlin.
24. G.Jones,J. Slaterand R.W.Staehle(Eds.):HighTemperature/ High Pressure Electrochem-
istry in Aqueous Solutions;NACE,Intern.CorrosionConferenceSeriesNo.4. NACE, Houston, Tx,
U.S.A., 1973.
25. J.V. Dobson:EMFMeasurementsatElevatedTemperaturesand Pressures;Adv.inCorr.Sci.
and Technol. (Eds.M. Fontana, R. Staehle);Plenum Press, 1980,2,177.
26. P.D. Macdonald: Corrosion, 1978,34- 75.
27. M. Hishida, H. Takabayashi,T. Kawakuboand Y. Yamashina:Corrosion, 1985,4l- 570.
28. M.L. Brown and G.N. Walton:J. Appl. Electrochem.l976,6- 551.
29. H. Hurst, D.A. Appleton, P. Banks and A.S. Raffel:Corros. Sci. 1985,26- 651.
30. D. Festy: Proc. Eurocorr.87,DECHEMA,Frankfurt, 1987,641.
31. G. Schmitt :in ”Wasserstoff und Korrosion” (Ed.D. Kuron), p. 332, Verlag Irene Kuron,
Bonn/FRG (Englishtranslation in preparation).
32. N. D. Greene, W.D. FranceJr. and B.E. Wilde: Corrosion,1965,2l- 275.
33
32. CHAPTER 4
REFERENCE ELECTRODES
E. ERIKSRUD
Veritas Research, Hovik, Norway
AND
E. HEITZ
Dechema Institute, Frankfurt, FRG
4.1 INTRODUCTION
Thebookeditedby Ivesand Janz[llandmorerecently thatbyBard,Parsons,and Jordan[21contain
both theoretical and practicalaspectsrelated to referenceelectrodes.Preparation, applicationand
limitationsof varioustypesof referenceelectrodessuchasthehydrogen electrode,the calomeland
othermercury-mercuroussaltelectrodes,the silver-silverhalideelectrodes,and sulfideand sulfate
electrodesare covered and general referenceto these excellentcriticalreviews is recommended.
Theroleof thereferenceelectrodeinelectrochemicalstudiesistoprovideafixedpotentialwhich
doesnot varyduringthe experiment.Inmost cases,thepotentialof the referenceelectroderelative
to an agreed standard, for exampleto the normal hydrogen electrode,isrequired. In other casesit
isonlynecessaryforthereferenceelectrodeto remain at thesamepotentialduring theexperiment,
for exampleduring a linear polarization resistance or a potential sweep experiment.
4.2 CHOICE, STABILITY, INCOMPATIBILITY
In corrosion testsall the typesof electrodeslisted abovemay be used depending on the corrosive
environment and the aim of the test.
If possiblethe referenceelectrodeshould be designed to be as similar aspossible to the system
underinvestigation.Thisisbothtominimize contaminationof thereferenceelectrode(e.g.chloride
ions diffusing into the sulphate solution of a copper/copper sulphate reference electrode will
changethe potential of the reference electrode)and to avoid misleadingcorrosiontest resultsby
ions from the referenceelectrodeentering the test solution(e.g.chlorideions in a pitting corrosion
test insulphatesolution).Withlargedifferencesinthe ionicconcentrationof thetestand reference
electrodesolution,a ”liquidjunctionpotential”maymakeanessentialcontributiontothepotential
measured between the working and reference electrodes.However, this is usually not a serious
error in corrosiontests.
It shouldbe pointed out, that if aliquid junctionpotential isminimizedby using a commontest
solutionand solutionfor the referenceelectrode,e.g.by dipping a Ag/AgCl wire directlyinto the
test solutioncontaining chlorides,the potentialof the reference electrodemay vary with concen-
trationof the test solution.Forthe Ag/AgCl electrodethechangewillbe roughly60 mV per decade
changein chloride concentration.
4.3 CHECKING OF THE REFERENCE ELECTRODE-SOME PRACTICAL
ADVICE
Itisgood practiceto havereferenceelectrodestobe used onlyforcheckingthereferenceelectrodes
used in the tests. Preferably, three saturated calomel reference electrodes should be kept for this
34
33. purpose. These should be within 2 1-2 mV of each other when checked for example in a chloride
solution.Theseelectrodesshould be stored in adryconditionforperiods of non use,with arubber
sealing ring over the fillinghole and a black rubber cap on the immersion tip during storage.
The working reference electrodeshould be checked both before and after a test. In long term
tests, it should also be checked during the experiment.
When using saturated KC1solutionbridged electrodesit isessential that the solutioncontains
undissolved KCl to ensure that it remains saturated. Air bubbles must not be present inside the
electrode.If KC1crystalsseemto trap air bubblesin the electrodeand thesecannot be removed by
shaking [31,theworking tip shouldbe immersedin warm distilledwater, until mostof thecrystals
are dissolved. The air bubbles can now be removed by shaking the electrode. The inside of the
electrode should never be flushed with anything but saturated KCl solution.
Recently,forenvironmentalreasonsAg/AgCl referenceelectrodesarebecomingmore widely
used as they contain no mercury or mercury salts.
4.4 MORE SPECIFIC REQUIREMENTS
Inpracticethemain requirementof areferenceelectrodeisthat it hasa stablepotentialand that this
is not substantially changed during the experiment. This is the case with the hypothetical,
completelynon-polarizableelectrode, the potential of which is unaffected when electric current
flows across the metal-solution interface. For practical conditions this means that the exchange
current must be largecompared with any net current that it isrequired to pass in use.Ideally ”no”
current flows through the reference electrode (in a three electrode system) if a high impedance
(>lOMn) voltmeter is used.
If temperatures and/or pressuresother than 25 “Cand 1atmospherearetobeused, the stability
of the reference cell with respect to such environments must be ensured.
In linewith requirements for the cell in general, the ohmicresistance of the reference electrode
should be reduced to a minimum in order to make the best use of the sensitivityof the recording
instruments. It may be tempting to use thin capillariesto prevent interdiffusion of the electrode
solutions,but thisis oftenfound toresultinanintolerablelossof sensitivity.Theconflictofinterests
may sometimesbe resolved by insertinga wide-bore tap, opened onlywhen measurements are to
be made,between theelectrodevessels.Unobservedgasbubblestrapped in “solutionbridges”are
sometimesresponsible for high cell resistance.
A lowering of the internal resistance of the electrodeby eliminatingceramic frits and porous
plugs, also has the advantage of a fasterresponse time [4]which is particularly important in the
study of fast electrode reactions using transient techniques.
4.5 SOME COMMON ELECTRODES FOR AQUEOUS SYSTEMS [4]
4.5.1 Mercury-mercurouschloride (calomel)
This is probably the most widely used reference electrode. It is usually made with a saturated
aqueouspotassium chloridesolutionbridge,although 1mol/dm3and 0.1 mol/dm3solutionsalso
are commonly used (see Table 4.1). Calomel electrodes can have very low resistance and good
performance.For this reason they are frequentlyused for checkingother types of electrodes.
4.5.2 Mercury-mercuroussulphate
Thiselectrodecorrespondsto thecalomel,withsulphateinsteadof chloride.Itisusefulforsulphate
solutions,but becomes unstable if the sulphate concentrationfallsbelow 0.1 mol/dm3.
4.5.3 Mercury-mercuricoxide
This electrodeis recommended for use in alkaline solutions.
4.5.4 Silver-silver halide
Theseelectrodesgive very stablepotentialsin halide solutionsprovided the halide concentration
isnot toohigh, when theincreasingsolubilityof the silversaltcausesproblems [2].Theyare easily
prepared in the laboratory, but must be shielded to exclude light and contamination should be
avoided.
35
34. ~~ ~~
Table4.1:Potentialsand fieldsof application of various reference electrodes
Electrode Electrolyte Potential 'Brnpe- Tempe- Field of
system vs SHE rature rature application
coeffi-
cient
mV O C mV/'C
at 25OC range
Hg/Hg, C1 2/C 1- KC1. sat. + 242 0 to 70 -0.65 general
Hg/Ngz Cl 2/C 1- ICC1 1 M + 280 0 to 70 - 0 . 2 4 general
0 to 70 -0.06 generalHglHg2Cl2/C1- KC1 0 . 1 M + 334
Hg/Hg2S04/Sd- K2SO4 sat. + 640 0 to 70 - sulfate con-
4 taining media
Hg/ Hgo/ OH- NaOH 1 M + 98 - - alkaline media
AgIAgClIC 1- KC1 0 . 1 M + 288 0 to 95 -0.5 general
AgIAgC1 I C1- KC1 3 M + 207 80 to 130 -1.00 hot media
HgTI/TlCl KCl 3.5 M - 507 0 to 150 0.1 hot media
InTable4.1 the temperature range of applicability,the temperaturecoefficientand the fieldsof
applicationof various reference electrodesare given.
4.5.5 Reference electrodes in non-aqueous systems
Allreferenceelectrodesareprepared with aqueoussolutions.Theiruse innon-aqueous systemsis
possibleaslongasliquid junctionpotentialsattheaqueous/non-aqueoussolutionphaseboundary
are small. This is the case for primary alcohols such as methanol, ethanol, isopropanol and for
dioxan solutions.The standard electrodepotentials of the calomelelectrodeand the silver/silver
chlorideelectrodehavebeen determined in such solutions.If potentialsin aprotic(nonwater-like)
solventshavetobe measured,difficultiesarise.Moredetailed discussionsaregivenin refs. [SI and
El.
4.6 REFERENCES
Press, 1961.
1.Reference Electrodes, Theory and Practice, Edited by D.J.G. Ives and G.J. Jam, Academic
2. A. Bard, R. Parsonsand J. Jordan,Standard potentials in aqueous solutions. Marcel Dekker,
New York and Basel, 1985.
3. Fixed OffshoreInstallations,Monitoringof cathodicprotectionsystems,TechnicalnoteTNA
705, Det norske Veritas. April 1984.
4. Instrumental Methods in Electrochemistry, Southampton Electrochemistry Group. Pub-
lished by Ellis Horwood, Chichester,U.K.,1985
5. E. Heitz, Corrosionof metals in organic solventsfrom "Advances in Corrosion Scienceand
Technology" (ed. M. Fontana and R. Staehle),Vol.4, Plenum Press, 1974.
6. C. T. Mussini and F. Mazza in "Electrochemical Corrosion Testing", Monograph 101.
PublishedbyDECHEMA,Frankfurt,FRG (ed.E.Heitz,J.C.Rowlandsand F.Mansfeld)1986,pages
67 and 79.
36
35. CHAPTER 5
EFFECTS OF SPECIMEN
PREPARATION AND SURFACE
CONDITION
J. SIMPSON
Sulzer Bros, A. G., Winterthur, Switzerland
5.1 INTRODUCTION
The state ofthe surfaceat the time of measurementcan have a largeinfluenceon the values of the
electrochemical parameters being measured. The surface state is dependant upon the whole
preparation history right up to the moment of measurement.
The stages which must be consideredare:
(a)Choiceof specimenmaterial
(b)Samplingand specimenpreparation
(c) Surfacepreparation beforeimmersionin the test solution
(d)Effect of the conditions within the cellbefore commencementof the measurement
(e)Changes caused by the measurement itself.
From the nature of the subject,it is not possible to treat this topic exhaustively within these
guidelines; this section is intended to increase the reader’s awareness of the many factorsin the
preparation stage which can influence subsequent electrochemicalcorrosionmeasurements.
5.2 CHOICE OF SAMPLE
It is important that the metallurgicalcondition and chemicalcompositionof the sample are well
defined. These parameters can vary considerablydepending on manufacturing process and heat
treatment.
A further problemisthat amaterialisrarelyhomogeneous.Thechoiceof thesourcematerialfor
the specimen will largely depend upon the aims of the experimental work. When studying
corrosionbehaviour for practicalapplications,samplestaken fromcomponentsor materialwith a
similarproduction history are preferable. On the other hand, if material is simply taken fromthe
next availablesourceof materialof nominallythe same specification,it may not be truly represen-
tative of the material as used in practical applications.
5.3 SAMPLING AND SPECIMEN PREPARATION
Metallurgical structure can vary considerably with position and orientation of the sample.
Castingsand weldments areobviousexampleswhere structureand chemicalcompositioncan
vary considerably. Even wrought material, whether drawn, extruded, rolled or forged, is very
rarely homogeneous; these materials will commonly have a texture (i.e. a preferential grain
orientation)and other directionalfeatures such as elongated grains and aligned elongated inclu-
sions. Through section variations must also be considered.
37
36. The surface composition can differ from the bulk after heat treatment: processes such as
decarburization, nitriding, surface oxidation with diffusion of dissolved oxygen into the metal
(oftenobservedin titanium alloys),or depletion of an alloyingelementat the surface,can givethe
surfacevery differentproperties to thebulk material.In suchcasesit must be decided whether the
properties of the modified surfaceor the bulk material are of interest .
It is important to bear such points in mind before sampling and preparing the specimen for
mounting.Specimensamplingprocedures,orientation,surfacepreparationandspecimenmounting
techniquesshould be decided before sectioningthe material rather than later.
To illustrate the above points, one may consider the pitting potential of stainless steels in
chloridesolutions.Thevalue of this potential isstronglydependent upon thedegree of cold work
and the orientationof the sample.A decreasein pitting potential was demonstrated forAISI-316L
stainless steel in physiological saline solutions with increasing degree of cold work.The pitting
potentialwas lowerin the transverse than in the longitudinal direction.Suchanisotropy may also
be present in hot worked and annealed samples due to texture effects [ll.Differencesin electro-
chemical behaviour with cold work and specimen orientation were also noted for austenitic
stainlesssteelsinacid solutions;thecriticalcurrent densityforpassivationin 1MI-$SO,at 25°Cwas
10timesgreater in the transversethan thelongitudinal directionat 30%cold work and the passive
current density in 0.1M HC1 at 25°C differed by up to 1000 times depending upon specimen
orientation [21.
Methodsused to preparethesampleshould not produce anychangesin itscondition.Themost
commonproblemsareheat generationonseparation(e.g.fromhighspeedabrasivecuttingwheels)
and cold work on machining, grinding, shearing or cutting. In general standard metallurgical
specimen preparation techniques are adequate for separation from bulk material. Water cooled
abrasivehigh speedwheelsforcuttingdown to the final sizeare suitablein most cases.For fragile
orextremelyheatsensitivespecimensacoldslowspeeddiamondwheelisoftenthebestalternative.
Specimen size and mounting are discussed in Chapters 2 and 3 as they are highly dependent
upon cell design and the type of measurement envisaged.
5.4 SURFACE PREPARATION BEFORE IMMERSION IN THE TEST
SOLUTION
The specimensurfaceisoftenprepared before measurement to ensure a reproducible and known
surfacecondition.Metallurgicalsurfacepreparation techniquesarecommonlyused, theseinclude:
wet grinding on silicon carbide abrasive papers, polishing with diamond or alumina media,
electropolishing,chemicalpolishing,picklingor etching.
Electrochemical corrosion measurement results can be strongly dependent upon surface
preparation technique.Thecausescan range from a simplesurfacearea effectof different surface
treatments, through secondary effects of the surface preparation technique on the substrate, to
chemicalchangesonthesurfaceduringsurfacepreparationandduringthetimebetweenpreparation
and immersionin the test medium.
Mechanical preparation techniquessuch as grinding can introduce significant cold work into
the surface layers e.g. the pitting resistance of ground austenitic [31 and ferritic stainless steel
surfaces[4] hasbeen showntobe inferiorto that ofelectropolishedsurfaces.Thiswas attributed to
the presence of cold worked surfacelayers from grinding, although chemicalor electrochemical
surface treatments can preferentially remove less resistant phases, e.g.inclusions,which would
otherwise be responsible for an inferior corrosionperformance .
A point often neglected is the handling procedure between preparation of the surface and
immersingthe specimenin the test medium. For instance,oxide film formationon oxide-passive
materialsand or tarnishing layerson copper or iron alloys formed at this stage can influence the
electrochemicalbehaviour considerably; these chemical changes on the surface depend on such
factorsas temperature, humidity and time [5].
38
37. 5.5 EFFECT OF THE CONDITIONSWITHIN THE CELL BEFORE
COMMENCEMENT OF THE MEASUREMENT
The time between placing the specimen in the corrosion cell and the commencement of the
measurement itself must also be counted as part of the surfacepreparation procedure. The value
of the electrochemicalparameter sought may vary considerably if insufficient attention is paid to
this point [6-81.It is preferableto consider this period as part of the measurement itself,reducing
the chances of random variations.
For measurements made at or near the freecorrosionpotential (e.g.polarization resistanceor
some impedance measurements),the most important factors to be considered are:
(i)Control of the medium; e.g. temperature, compositionand agitation. For instance, the rest
potentialof mild steelinaerated saltsolutionsisdependant upon thedegree of agitation [61,in this
systemevenconvectionmaybesignificant,aswould consumptionof dissolvedoxygenbythesteel.
(ii)Time; prepared surfacesare rarely in their stable surfacestate with respect to the medium
when immersed, for instance the zero current potential of a mild steel sample in aerated 0.01 M
NaHCO, at65°Cchangedfrom-400mV(S.H.E.)to+240mVafter 100minutesand 6daysimmersion
respectivelydue to the formation of surfaceoxides [71.
Thesituationbecomesmore complexwhen any in situ surfacepre-conditioningtechniquesare
employed;oftencathodicor anodicpretreatmentsare used.Theseshould be treated with caution.
Cathodictreatments can reduce oxidefilmsand introduce hydrogen into the metal,both ofwhich
caneffectanysubsequentelectrochemicalmeasurement[7,8].Anodictreatment may passivatethe
surfaceor roughen it by anodic dissolution.
5.6 CHANGES CAUSED BY THE MEASUREMENTITSELF
It is important to distinguish between changes in the surface condition caused by the corrosion
process and those forced on the specimen by the measurement technique. The aim may be to
monitor the former, the latter may lead to misleadingreadings or interpretation problems.
Many electrochemicalcorrosiontechniquesare perturbation techniques, i.e. either an external
potential or external current is applied. These can modify the surface, be they anodic causing
dissolution or passivation, or cathodic causing surface oxidereduction or hydrogen production.
The effect is normally one of degree. The measurement of a corrosion potential does not
influence the surfacecondition. Electrochemicalnoise and impedancemeasurements carried out
at thecorrosionpotential alsohave littleeffectasdoesa polarizationresistancemeasurementif the
perturbation is small, although rest potential drift may be a problem if potential control
techniques are used. Techniques involving large potential differences will in general modify a
surface significantly.
It should be borne in mind when repeating an experiment without using a fresh specimenor
fullyrepreparingthe surface,thatthe surfaceconditionisunlikely tobethe sameastheinitialstate,
it willhavebeen modifiedeitherby theexposuretothemedium orby themeasurement technique.
5.7 REFERENCES
1. A. Cigada, B. Mazza, P. Pedeferri and D. Sinigaglia:J. Biomed.Mater.Res.,1977,ll- 503.
2. B. Mazza, P. Pedeferri, D. Sinigaglia,A. Cigada,G. Fumagalli and G. Re: Corros.Sci.,
1979,19,907.
3. R. W. Revie and N. D. Greene: Corros.Sci., 1969,a 763.
4. R. P. Frankenthal:Corros.Sci., 1968,& 491.
5. D. E. Dobb,J. P. Storvickand G. K. Pagenkopf:Corros.Sci., 1986,26- 525.
6. D. M. Brasher:Br. Corros.J., 1967,295.
7. G. K. Glass:Corros. Sci., 1986,&441.
8. C. D. Kim and B. E. Wilde:Corros.Sci., 1970, 735.
39
38. CHAPTER 6
EVALUATION AND
COMPENSATION OF OHMIC DROP
L. CLERBOIS
Solvayet Cie, Brussels, Belgium
AND
F. P.IJSSELING
CorrosionLaboratory of Royal Netherlands Naval College,
Den Helder, The Netherlands
6.1 INTRODUCTION
Electrochemical systems generally contain a working electrode (WE), an auxiliary or counter-
electrode(CE)and a reference electrode(RE).
Duringcurrent flowthe voltagedifferencebetweenWE and CE consistsof two parts :electrode
polarizationand ohmicdrop through the solution (IR).
Thepotentialmeasured betweenWE and RE maybe substantiallyinfluencedby an IR drop, in
particular if high current densitiesareapplied or an electrolyteof low conductivityisused. Thisis
due to the fact that the reference electrodeisconnected with a point in the solutionsomedistance
away from the electrochemicaldouble layer. Thus an ohmic resistance - the so-calleduncompen-
satedresistance(Run>-isincludedin thesolutionbetweenthetipof the referenceelectrodeand the
surfaceof the working electrode(Fig.6.1).As a result an error will be introduced in the measure-
mentof thepotentialdifferenceinsuchawaythat thepotentialdifferencebetweentheworkingand
reference electrodes is not as large in an absolute sense as indicated by the potentiostat or an
auxiliaryvoltmeter.
Theimportanceof knowing the exactvalue of the ohmicdrop or uncompensated resistancein
an electrochemicalsystem has been pointed out by many workers. In studies of the kinetics of
electrodeprocessesby potentiostatictechniques,theohmicpotentialdrop producesadistortion of
the steady state polarization curve which, if uncorrected, will yield erroneous values of the
characteristicparameters (Tafelslope, reaction orders)of the electrodereactions (Fig.6.2).
Also,measurements of the polarization resistance, R might be subjectto considerableerrors
due to the ohmicresistance which may lead to an underestimation of the corrosionrates by up to
several hundred percent.
PI
The effect of uncompensated IR drop on corrosion rate determination using polarization
resistance measurements was discussed in depth by Mansfeld [1-31. He showed that in electro-
chemicalmeasurements of the polarizationresistance theexperimentalvalue R * is the sum of the
true value R and the uncompensated ohmicresistance RU,, which is essentiallythe electrolyte
resistancebut can also contain the resistance of surfacefilms.
P
Po
40
39. T
melectrolyte
WORKING I
ELECTRODE
I
I WE
@
CE e+
RE o X
P
Fig. 6.1 :(a)3-terminalcell schematic;(b)detail showing Runtbetween surfaceof WE and iso-
potential plane of RE; (c) equivalent circuit.
Z = impedance of working electrode
Rmc= uncompensated resistance
Rs = solution resistance
Fig. 6.2 : Effect of ohmic drop on the shape of polarization curves.
The values of parameters are ba= 30 mV, bc= 120mV, IC= 0.1 mA/cm2 [l,21.
Since
R * = Rpo+Run, (1)P
the relative experimental error is
Even if Runcis low, the error can be appreciable if R is also low, e.g., in systems with high
corrosionrates. Thesamemagnitude of error canbe found for systemswith low conductivityand
low corrosion rates. It is, therefore, not a question whether the absolute value Runcis low, but
whether the value Run, / Rpoislow.Similarequationswere derived by other authors, e.g.Rocchini
MI.
PO
41
40. Also in crevice and pitting corrosion, the ohmic potential drop may be responsible for the
stabilityof local attack on passive surfaces.
Thesubject is treated in most textbooks and introductions to electrochemicalexperimentation
(forinstancerefs.[5-14])and two literature reviews have been published 115,161.
6.2 EXAMPLES
Three examplesshowing the importance of ohmic drop correction are given below.
6.2.1 Polarizationresistance
Mansfeldshowed that even in highly conducting solutionsinvolving high corrosionrates, the IR
drop, although small, cannot be disregarded, sincethe polarization resistanceand RmCare of the
sameorder of magnitude. This isin particular true for the corrosionrate determination of carbon
steelin acid solutions [l-31.The dangerousconsequenceof neglectingRunt is that corrosionrates
willbeunderestimated. AnexamplecanbeseeninFig.6.3:withoutcorrectionthecorrosioncurrent
density was very low, whereas weight loss showed a high corrosion rate. After applying a
correctionfortheohmicdropasignificantlyincreasedcorrosioncurrentdensity was found, which
correlatedbetter with the corrosionrate obtained by weight loss.
4
3
2
1
h
Tu
E
4
O O
E -1v
H -2
-3
-4
c o r r e c t e d ii a
-
#' experimental
.................................................. *.................................................
-300 -200 -100 0 100 200 300
Fig. 6.3 :Effect of ohmicdrop on the shape of the polarization curve for the system:low alloy
steel +EDTA 100g/l; pH = 6, T = 100"C,exposed area = 60.5 cm2,flow rate = 1m/s [171
6.2.2 Polarizationcurves in aqueous solution
Mansfeld[1,2]alsopointed outthatthelackof curvatureinexperimentalpolarisationcurvescould
be aresultof a high resistance.Someresultsof theoreticalcalculationshavebeen shown inFig. 6.2.
A practical example is the case of gas bubbles reducing the conductivity of the solution. Gas
evolution from vertical or inclined working electrodesproduces a varying electrolyteresistance,
with the effectbeinggreatestat the top of the electrode.Electrolyteresistancemeasurements made
during gas evolution may thereforehave littlemeaning.
42
41. A cathodic polarization curve for hydrogen evolution on platinum in 2.8 M sulfuric acid is
shown in Fig. 6.4.
1.2
1.0
.8
“E .6
a
.-- . 4
:.2
u
-K
3
u
0
-.150 -300 -,450 0.600 9.750 -.900 -1.050
potential, U, volts
Fig. 6.4:Polarizationcurve during cathodic hydrogen evolution(linearplot without elimina-
tion of ohmicdrop)
Due to a significant IR-drop, this is nearly a straight line. Elimination of IR-drop yields over a
largerange of current density a semi-logarithmicstraight line, with a Tafel slopeof 27mV. Sucha
result which gives also a reasonableTafel slope at high current densities up to 1A/cm2may be
quoted to demonstrate the reliability of the method (Fig.6.5).
. 2
0 . 0
-. 2
-.4
9 -.6
5 -.e
4- -1.0
0) -1.2
--c. -1.4
2 -1.6
5 -1.8
-2.0
-2.2
.-
0
C
L-
/.-
- b = 27 mV
log i = -3.’1
/ .-L
-.225 ,240 7255 -270 ,285 -300
potential (U-IR), volts (SCE)
Fig. 6.5 :Tafel plot of IR-dropcorrected polarizationcurve of Fig. 6.4
6.2.3Polarization curves in non aqueous solution
Depending on the conductivityof the solutionsthe ohmicdrop may vary considerablyin organic
solvents.Whereasin solutions of primary (waterlike)alcoholssuch as methanol, ethanol (EtOH)
and propanol polarization curves with ohmicdrop correctionscan be obtained it is impossibleto
make similar measurements in non water-likesolvents such as long chain alcohols, halogenated
hydrocarbons and other aprotic solvents.
43
42. Figure 6.6showsan examplefor Fe in 0.01N HCl/EtOH with and without IR-dropcompen-
sation.Without IR-drop compensation,a straight line is observed over the entire range of polari-
zation.WhentheIR-dropiscompensated,curvatureisseenandtheshapeof thecurveindicatesthat
ba<bc.Withoutcompensationof theohmicdrop Rp*was 357ohms.Onapplyinga compensation
techniqueRp” was found to be 74 ohms, R,, being 212 ohms.
212
74
The error e = - = 2,86or almost a factor of three.
Runc =: 2120
R: = 7 4 0 1-150
Fig.6.6:ExperimentalpolarizationcurveforFein0.01N HCl/EtOH withand without IR-drop
compensation [21
6.3 PRINCIPAL METHODS
Although ohmic drop cannot be eliminated completelyit can be minimized,while the remaining
effect can be taken into account, either by active methods, e.g. by some means of instrumental
compensationor passivelv,by calculationand subsequent correction.
In practicea combinationof these methods isoftenused i.e.,a good celldesign to minimisethe
ohmic drop, instrumental compensation of the greater part of the remaining error and, finally,
removal of the last part by evaluation, calculation and correction of the experimental data.
6.3.1 Minimizingof ohmic drop by cell and electrode design
Thus, the first step of any remedial action will usually be to minimizethe value of the uncompen-
satedohmicdrop.Thedesignof thecelland theelectrodesare theprincipalmeansof achievingthis
objective.
Both subjects have already been treated in these guidelines and reference should be made to
Chapters 2 and 3.
6.3.2.1Cell design
Generally cell design is dedicated to obtaining a constant current density over the surface of the
working electrodein combinationwith the use of a Luggin capillary for measuring the potential
differencebetween the surfaceof the working electrodeand the solution.As the distancebetween
the tip of the Luggin capillary and the electrode surfacedecreases, so will the ohmic drop error.
However, the distance cannot be made very short without introducing screening effects on the
working electrodesurface.
In practicea capillary tip of outer diameter d may be placed as close as 2d from the electrode
surfacewith negligible shielding error [181.
44
43. Figure6.7showsa schematicdiagram of the equipotentialsin question for the caseof capillary
of diameter d placed 2d from a planar electrode; it can be seen that in this particular case the
potentialthatissampledcorrespondstoanequipotential surfacethatispositioned 5d/3awayfrom
the electrode [191.
E l.ectrode
Desired Potential
7
Measured Potential
Capillary
b
f
d
Fig 6.7 :Schematicdiagram of IR-drop in the electrolytebetween capillary tip and electrode
The IR-drop not only depends on the diameter of the Luggin capillaryand its distance from the
working electrode,but also on the specificconductivity of the solution and the geometry of the
working electrode. In Table 6.1 the IR-drop has been calculated for some simple geometries,
assuming constant specificconductivity [71.
Table 6.1 Theoretical equations for the IR drop for simple geometry types. Values have been
calculated for the probe placed 2d away from the electrodewith 6 = 5d/3; K = 0.02lX'cm-';
ro= 4 x 10-3cm,d = 0.02cm and i = 20mA cm-*[71.
Geornetry Equation V I R /mv
Planar 33
From the equations given in the table it can be seen that the ohmicerror is linearlydependent
on the distance between the RE and WE for a planar electrode.For a cylindricalWE, the error is a
logarithmicfunction of distance, whereas for a sphere of smallradius, the error is lessdependent
on RE position.
Although the applicabilityof spherical electrodesin practical situationsisoften limited, this is
a definiteadvantage, the distancebetween the capillarytip and the electrodebeingnot so critical.
Ahlbergand Parker [20]alsowarned of the dangersof trying to place the Luggintip veryclose
to the WE. Smallvariationsin thepositioningof the tip will make a large differenceto the uncom-
pensated resistance.Theyadvocatetheuse of small-diametersphericalelectrodesand a largeWE-
RE separation.
45
44. In all casesthe ohmic drop is proportional to the current density. Specialcareshould be given
to transients at short times when large currents may flow, possible error sources being current
oscillations,double layer chargingand stray capacitanceto ground.
Apart fromresistance,capacitiveeffectsshouldalsobetakenintoconsideration.Thebestdesign
of a Lugginprobe is one with a narrow capillaryat its tip with thin wallsto prevent shielding,but
with thick wallsin the main body which widen rapidly away from the tip to reduce resistance in
the control loop [211.
A possible solutionto avoid ac interferencesis the use of a specialhigh-frequencyby-pass, for
instancea platinum wire coupled to the normal reference electrodeby a 0.1 pF capacitor.
Any additional resistancein the working electrodeitself, for example due to the formationof
resistive films, will be included in the uncompensated resistance and can only be reduced by
electronic compensation.
In summary :
1. Large capillaries (diameter > 1mm) are useful only at relatively low current densities in
solutionsof high conductivity.
2. Evenwiththe smallestconvenientcapillary(0,2 mmdiameter)placed asclosetotheelectrode
aspossiblewithoutincurringshieldingerrors,theIRcorrectionsgreatlylimitthemaximumcurrent
density for accuratepolarization measurements, especially in solutionsof relativelylow conduc-
tivity.
3. Smallcylinders(wires)or spheres if practicallyfeasiblemay be used advantageously as test
electrodesinpolarizationstudies,sincetheIRcorrectiondecreasesastheelectrodeismadesmaller.
6.3.I.2 Microelectrodes
Thecurrentcanbekept smallby the useof smallelectrodes1221. Theadvantagesof small-diameter
cylindersand sphereshavebeendescribedpreviously.Althoughtherearesomepossibledisadvan-
tages(forinstancethe small surfaceavailablewhen studying pitting),microelectrodesarebecom-
ing popular for use in non-aqueous and aqueous resistivemedia. For instance,Genders,Hedges
and Pletcher [23] described an applicationof microelectrodesto the study of the Li/Li+couple in
ethersolvents. Theyshowed that it ispossibletoobtainhigh-qualitydata for theelectrodeposition
and anodicdissolution of lithium.Theexperimentsrequired only a two-electrodecell and simple
instrumentation because the experimental currents (not current densities) were very small. The
electrodeswere constructed fromCu wires (40and 80 p.m diameter). The electrolytewas always
lithium hexafluoroarsonateLiAsF, in tetrahydrofuran (THF). The resistance of the experimental
cell filled with THF +LiAsF, (0.6mol dm3) measured using an a.c. bridge was 5000R. Only cell
currents below 3 p.A (current density 60 mA cmS2)were analysed quantitatively so that the
maximum IR drop was 15 mV. A test of the absence of significant IR-drops is to repeat the
experiment with an electrode of different area when only the current should change and there
should be no shift in peak potentials or changes in shape of transients.
6.3.2Active methods
A number of active methods are available in which the compensation of the ohmic drop is
necessarily incorporated in the control system of the potentiostat.
One of the methods which has been used is the ac method, based on the application of a high-
frequency signal (e.g. 50 kHz) through the cell, using the ac voltage after amplification and
rectificationas the control signal.The activemethods mostly used are :
a. positive feedback
b. current interrupt.
46
45. 6.3.2.2Positive feedback
Inthepositivefeedbackmethodsavoltagesignalisproduced whichisproportional toIRand which
isadded to the controlinput voltage.TheaimistocompensateforIRautomatically.Dependingon
theprincipleofthepotentiostaticcircuitseveralsolutionshavebeenproposed 1151(forexampleFig.
6.8).
CE
1
negative feedback loop
WE
CE : Counter E l e c t r o d e
RE : Reference E l e c t r o d e
WE : Working E l e c t r o d e
VF : V o l t a g e F o l l o w e r
CF : C u r r e n t F o l l o w e r
C : C o n t r o l A m p l i f i e r
-L
Ro < 7
R, : Measuring R e s i s t o r
R x : Feedback R e s i s t o r
S : C o n t r o l V o l t a g e
Fig.6.8:Basiccircuitforpotentiostat with IRdropcompensationbymeansof positivefeedback
in the control loop
The requirement is that at all frequenciesthe applied positivefeedback signal is smaller than the
negative feedback of the control loop of the potentiostat itself.
Ofteninstabilityproblemswillarisebecausethesystembehavesasapurely capacitiveelectrode
if 100%compensationisobtained[151.Neverthelessthemethodisused frequently,manycommer-
cially available potentiostats being equipped with, or easily adapted to, this form of IR drop
compensation 1241.
The value of the resistance in the potentiostatfeedbackcircuit (Rxin Fig.6.8)maybe set on the
potentiostatin severalways,themostcommonbeingby "trial anderrox".Forexampleonapplying
a square wave signal of small amplitude (e.g.50mV, 50 Hz) to the control input the value of the
resistance is increased gradually until "ringing" (= overshoot + relaxation)is observed with an
oscilloscopeconnected acrossthe WE and RE. The resistance value is then reduced a few percent
until stabilityisrestored and thisvalueisused inthefeedbackcircuit.Analternativemethod isfirst
to measure the resistance, and then to set this value in the feedbackcircuit. The uncompensated
resistanceisassumed constantduringthe subsequentcurrent-voltagemeasurements, leadingto a
constant correction;hence, errors may arise when for somereason the uncompensated resistance
changes,e.g. due to film formation.
It should be realizedthat overcompensationdoesnot alwaysresultin "instability"and sogreat
careisneeded toavoid it.Positivefeedbackin IRcompensated potentiostatshasbeen discussedin
detail by Britz [15] and by Mc Kubre and MacDonald [131.
An analogue techniqueof ohmicdropcompensationhasbeen proposed by Gabrielli et al. 1251,
which avoids some stabilityproblems related with positive feedback.
Many workers subscribeto the view that any compensationisbetter than none and so will use
positivefeedback.
47
46. 6.3.2.2 Interrupt methods
Interrupt methods in particular are becoming popular, being used in conjunction with both
potentiostaticand galvanostatictechniques [lo, 16,26-301. The principle is to measure the rapid
changein potentialof the workingelectrodeas the current is suddenlyswitchedoff (interrupted).
Theohmicdrop is givenby the immediatechangein potential on switching,the potential change
due to polarizationdecayingrelativelyslowlyand certainlymore slowlythan the purely resistive
ohmicdrop componenti.e., because of the very short timeconstant of the latter (Fig.6.9a). Hence
this method is not applicablewhen the double layer capacityis very small.
For determining the ohmicdrop a transient recorder or an oscilloscopeis required. Oneof the
potential problems associated with this method is the possible measurement of artefacts, i.e. the
decay characteristics of the measuring circuit overlapping the desired potential decay of the
working electrode. Other difficultiesare associated with the need for reliable triggering and fast
data acquisition.Themethod ismostlyappliedby interrupting thecurrentperiodically,necessitat-
ing fast current switchingof possibly large currents.
f- lnterruptlon- tlme
A unc = A U / A I
I
topulse + t i m e
L
C
m
m
E
n
m Runc = lim Z K J Id
Real part 2'
R unc
Fig. 6.9 Differenttechniques for the measurement of the ohmic potential drop (schematic)
(a)interrupt technique (b)pulse technique (c) high frequencyac technique [311
Someof the moreexpensivecommerciallyavailablepotentiostats are, or canbe,equipped with
these formsof ohmicdropcompensationtechnique.Inprinciple, themethod canbe applied when
the values of Run, and the measuring resistance changeduring the measurements.
Some special designs have been published, in which the potentiostat, the interrupter and the
dataacquisitionapparatus havebeen combinedinasophisticatedcomputer controlledinstrument
[28-311.Anexampleof this isthe measuring equipment describedrecentlyby Heitzet al.(Fig.6.10)
[XI.
48
47. F'I COMPUTER - - - - - -I
I 1
1 I
1 I
1 I
I
I
I
I I
I I
TRANSIENT
RECORDER
POTENTIO-
I
1
INTER- POTENTIO-
RUPTER STAT
'
-
I
1
I- - - - - - - - - - -
Fig. 6.10 :Flow diagram of the computer-controlledmeasuring equipment used to eliminate
ohmic potential drop [321
RE WE
The method is based on a setup which has been succesfullyused to eliminateohmic potential
dropat gas-evolvingelectrodesat highcurrent densities. With modem equipmentforelimination
of ohmic drop as previously described it is possible to make rapid correctionsin solutions with
conductivitiesbelow 1k s cm-'.
CE I -
6.3.3Passive methods
This heading coversall methods in which the ohmicdrop is measured or calculated,followed
by mathematicalcorrectionafterwards when theexperimentaldata havebeen collected [15,16,25,
331.
6.3.3.1 Measurement of ohmic drop
The measurement of the ohmic drop can be performed by a number of methods, the most well-
known being :
a.Positivefeedbackmethod asdescribedabove,determiningtheresistanceat whichthe poten-
tiostat oscillatesby manual adjustment of the feedbackresistance.
b.Current interruptmethod asdescribedaboveoralternatively,themeasurementof theinstan-
taneous potential change when the current is switched on (pulsemethods :Fig. 6.9b).
c. Alternating current methods, including the use of bridges and four-point methods [MI.
d. An extensionof method c isthe applicationof impedancetechniques,either the full Nyquist
diagram being determined or the impedance at one or more discrete frequencies(Fig.6.9c) [331.
e.Extrapolationtechnique,inwhichthedistancebetweenthetipof the Luggincapillaryand the
surfaceof the workingelectrodeis varied;the potentialdifferencebetween the working electrode
and the reference is measured as a functionof the distance and extrapolated to zero distance [351.
f. An extensionof method e is the multiple potential method, in which the potential of the WE
is measured with respect to severalRE positioned at fixed, known distances 1331.
An exampleof the application of impedance techniques(method d)is shown in Fig. 6.11.
49
48. 1.2
-8
-cv
.6
n
N
- . 4
-2
0
Fig.6.11 :Determinationof solutionresistance for differentsystemsconsistingof low alloyed
ferritic steelsin contact with aqueous solutions [36]:
-
- - _ _ _ _ _ _ _ -eo-- -*- -
1 -
-
_ c _ c _ - c - - a U l a-- *- -m- -
-
-
-
UAA.& -A- - -A-
=- _ _ -- - - - -
-
I l l I 1 I t 1 1 I 1
(a)100g/1 EDTA,T = 100"C,pH = 6 (0)
(b)100g/1 EDTA, T = 100"C,pH = 7 (B)
(c) 150g/1 HC1+ 1g/1 inhibitor, T = 76 "C (A)
Theevaluationof ohmicdropwithalternatingcurrentathighfrequenciesisbasedontheelectric
equivalent of the metal-solutioninterfaceof Fig. 6.12, where Run,represents the electric solution
resistancebetween working and referenceelectrodes.
Fig. 6.12:Equivalentcircuit of metal-solutioninterface
The impedance,Z, between the points A and D, at angular frequency o,is given by
Z = Runc+Rt/Q - j o CR2,/Q, (3)
50