Public I 0 hour pe, response. inlud~ng the time for rviewing .ntr tOns, searching existing data sucsgather U i U . J k ecollection of information. Send Comments rega'rig this burden estimate or any other aspect ofthis co A D A 3 7 Washington Heladquarters services, Directorate for Information Operations and Reports. 1215 JeffersonDavis inagement and Budget. Paperwork Reduction Project (070.4-0188). Washington. DC 20503.

1. Agency use Only (Leave blank). 2. Report Date. 3.Report Type and Dates Covered.

4. Title and Siubtitle. 5. -unding Numbers.

KTechnical Review of ElectrochemiceL Techniques Appltied to Hicrobiologicalky Program Element No. 61153Inf Luenced Corrosion

______________________________________________________________Project No. 03102

6. Author(s).

Fbrian Manlsfetl and Brenda Little TakN.310

Accession No.DM094463

7. Performing Organization Name(s) and Address(es). 8. Performing OrganizationNaval. Oceanographic and Atmospheric Research Laboratory* Report Number.Ocan Science Directorate

Stenntis Space Center, MS 39529-5004 JA 333:021:90

9. SponsoringlMonitoring Agency Name(s) and Address(es). 10. Sponsoring/MonItoring Agency

Naval Oceanographic and Atmospheric Research Laboratoryiseot ubrOcean Science Directorate JA 333:021:90

Stennis Space Center, HS 392-5004

11. Supplementary Notes.

CR ~aaFoI,.erLy Naval Ocean Research and Development Activity

12a. DistvibutionlAvailability Statement. .1 Istifrnb1Jtin Code.

Approved for ptiiLic release; distribution is untimited.G

13. Abstract (Maximum 200 words).

A critical. review of the Literature concerned with the application of clectrochemical techniques in the study of

microbiologically influenced corrosion (MI1C) is presented. The electrochemicaf techniques covered in this review

include measurements of the corrosion potential., the redox potential., the polarization resistance, the electro-

chemical impedmnce, electrochemical. noise, and polarization curves including pitting scans. For each experimental

technique some discussion concerning experimental procedures, advantages and disadvantages of the technique for the

'~study of MIC is presented. Applications range from studies of the corrosion of steel pipes in the presence of sulfate

reducing bacteria to investigations of the formation of blofilm end calcareous deposits on stainless steels in sea-

water and the destruction of concrete pipes in sewers by microorganism producing very Low pHt soLutions. For each

experimental. technique at Least one examipe for typical, experimental data Is presented.

14. Subject Terms. I. Number of Pages.

CU) Corrosion; (U) Siodeterioration; (U) Electrochemistry 2

_____________11:6.Price Code.

17. Security Classification 18. Security Classification 19. Security Clasuification 20. Limitation of Ahetract.of Report. of This Page. of Abstract.I

Unclassified Unclassified Uniclassified j AtS 50-l20-50Standard Form 296 (Rev. 2-ag)

P'95C-D" by ANSI SIC Z39-18298- '02

C(,rrosion Scence. Vol. 32. No. 3, pp. 247-272. 1991 001 -938X/91 $3.(0) + ()(KPrinted in Great Britain Pergamon Press pli

A TECHNICAL REVIEW OF ELECTROCHEMICALTECHNIQUES APPLIED TO MICROBIOLOGICALLY

INFLUENCED CORROSION

FLORIAN MANSFELD and BRENDA LITTLE*

Corrosion and Environmental Effects Laboratory, Department of Materials Science, University ofSouthern California. Los Angeles. CA 9(X)89-0241, U.S.A.

*Naval Oceanographic and Atmospheric Research Laboratory. Stennis Space Center, MS 39529-5004,

U.S.A.

Abstract-A critical review of the literature concerned with the application of electrochemical techniquesin Ihe ot'dv of microbiolooicdly influcnc:" corro-i.n (MIC) is presented. The clcctroc-cmica1 technique-covered in this review include measurements of the corrosion potential, the redox potential, thepolarization resistance. the electrochemical impedance, electrochemical noise, and polarization curvesincluding pitting scans. For each experimental technique some discussion concerning experimentalprocedures. advantages and disadvantages of the technique for the study of MIC is presented. Appli-cations range from studies of the corrosion of steel pipes in the presence of sulfate reducing bacteria toinvestigations of the formation of biofilms and calcareous deposits on stainless steels in seawater and thedestruction of concrete pipes in sewers by microorganisms producing very low pH solutions. For eachexperimental technique at least one example for typical experimental data is presented.

INTRODUCTION

MICROBIOLOGICALLY influenced corrosion (MIC) has received increased attention bycorrosion scientists and engineers in recent years. MIC is due to the presence ofmicroorganisms on a metal surface which leads to changes in the rates and sometimesalso the types of the electrochemical reactions which are involved in the corrosionprocesses. It is therefore not surprising that many attempts have been made to useelectrochemical techniques to study the details of MIC and to determine itsmechanisms. Applications have ranged from MIC of many structural materials inseawater to soils, aircraft fuels and sewers.

The evidence provided so far suggests that the presence of microorganisms onmetal surfaces often results in highly localized changes of the concentration of someof the electrolyte constituents, pH and oxygen levels. t ' 2 For passive metals theselocalizcd changes can lead to localized corrosion in the form of pitting or crevicecorrosion. As with other corrosion systems, difficulties exist in the interpretation ofelectrochemical data obtained for cases of localized corrosion. The prevailingapproach to estimate the tendency towards localized corrosion based on pittingpotentials rather than to determine rates of localized corrosion often preventsadvances in this area.

Electrochemical techniques have been shown to be very useful for mechanisticstudies in laboratory investigations and for monitoring purposes in field studies. Aswith all studies of corrosion phenomena more detailed and reliable information canbe obtained when a number of different electrochemical techniques are combined.

Manuscript received 8 January I1990; in amended form 1 March 1990.

247

91 o (61

248 F. MANSFELD and B. LJrnLE

In the special case of MIC it is of course necessary to use microbiological in additionto surface analytical techniques such as Auger electron spectroscopy (AES), X-rayphotoelectron spectroscopy (XPS) and scanning electron microscopy (SEM) withenergy dispersive X-ray analysis (EDAX). These considerations lead to the con-clusion that real advances in the understanding of the mechanisms of MIC can onlybe expected from the work of teams which consist of experts in the varioustechniques discussed above.

Dexter et al.3 have recently presented their conclusions concerning the use andlimitations of electrochemical techniques for investigating MIC. Their reviewfocused mainly on the background of different electrochemical techniques andpresented the authors' understanding of the advantages and possible problems withthe techniques which are routinely used in corrosion studies. The aim of the presentreview is to focus on the information obtained so far with the use of electrochemicaltechniques in investigations of MIC and to use these results as a basis for a discussionof the type of information which can be obtained with each of these techniques. Theprcct iev;ew is b,sed on a computerized literature search which produced over 100papers and the results of the authors' on-going research in the area of MIC. Thefollowing discussion is based mainly on papers published after 1970 when moresystematic studies of MIC were reported.

RESULTS OBTAINED WITH ELECTROCHEMICAL TECHNIQUES IN THE

STUDY OF MIC

In the following the background of the electrochemical techniques which areconsidered suitable for the study of MIC will be summarized briefly and someimportant experimental aspects will be discussed for each technique. This will befollowed by a summary of the experimental results obtained so far with eachtechnique for different metals and alloys in various environments. A criticalassessment of the information obtained will be given if this is possible from ananalysis of the data presented by the authors.

The corrosion potentialMeasurement of the corrosion potential Ec,,,r is the easiest electrochemical test,

but it provides the least amount of mechanistic information. The measurement ofE,,,rr requires a stable reference electrode, a high impedance voltmeter and in mostcases a suitable recording device. The possibility of problems with the properfunctioning of the reference electrode due to the formation of a biofim hasapparently not received much attention.

The main problem with the use of E,,,r measurements is the overinterpretation ofsuch data. This will be illustrated for the case of stainless steels in seawater asdiscussed below. Based on the information obtained with Ecor alone it is not possibleto determine whether corrosion rates have increased or decreased. It is thereforeadvisable to determine not only Ec,,,. but also the polarization resistance Rp at thesame time. The changes of both parameters can then be interpreted in terms ofchanges in the rates of the anodic and/or the cathodic partial reactions whichdetermine the corrosion rate.

Early measurements of Ec,rr which were carried out for steel in the presence ofsulfate reducing bacteria (SRB) suffer from the lack of additional information andare therefore difficult to interpret. Diffrent authors have explained the obscrved

Electrochemical techniques applied to MIC 249

300

200-3C3

0- -2

15-200 /1 Sandvik 5R60 4 254 SMO-30 J2 SAF 2205 5 M'onit00

3 Sanicro 28 6 984 LN4 00 ' I ' ' ' . I .0 4 8 12 16 20 24 28 32 36

EXPOSURE TIME (days)

FIG. 1. Open-circuit potential vs time for six stainless steels exposed to flowing naturalseawater."

changes of Eorr in the negative direction by a reduction of the rate of cathodicreaction 4 and by an increase of the rate of the anodic reactions. 5 These are the twosimplest possibilities which can be expected to lead to the observed result. However.without additional data no valid conclusions concerning the effect of SRBs onelectrochemical corrosion reactions can be drawn. Only when the exact mechanismof MIC has been established for a given system will it be possible to use Ec,,rr data formonitoring purposcs such as the detection of an increase of uniform corrosion ratesor the initiation of localized corrosion due to the presence of bacteria.

Numerous reports have appeared in which the time dependence of Ecorr has beenmonitored for stainless steels in natural seawater. Most published data show a rapidennoblement of Ecor. during the first days of exposure. 6 Figure 1 shows as arepresentative example potential-time curves reported by Johnsen and Bardal8 forseveral stainless steels in flowing (0.5 m s-) natural seawater. Ecor, changed from-200 to -250 mV(SCE) at the beginning of the test to -50 mV for two of the alloysand to +200 mV for the remaining four alloys after 28 days. Johnsen and Bardai 8

offered a different interpretation than Mollica 6 for the cause of the observedennoblement of Ec,,r, but agreed that this effect was caused by a change of tilecathodic properties of the stainless steels as a result of microbiological activity on thesurface.

Dexter and Gao12 have recently reported Ec, data and polarization curves forSS 316 exposed to natural seawater. The authors concluded that the observed scatterof the Err data in seawater was due to the presence of microorganisms. However ina subsequent discussion of these data' 4 it became apparent that the very negativeEc,,rr values for some, but not all, of the samples were due to crevice corrosion undera lacquer and should therefore be discarded. Based on the results obtained withcathodic polarization curves Dexter and Gao 2 concluded that the increase of Ecorrfor the remaining samples of SS 316 was due to an increase of the rate of the cathodicreduction of oxygen at a given potential. In an analysis ot the results ot tle variousauthors who have reported an ennoblement of Ecor for stainless steels duringexposure to natural seawater, one has to consider that these results do not demon-strate directly an icica.,cd .usccptibil:ty LU iiulAn,. L., f.. the sustained noblepotentials show that the stainless steel surface retained its normal corrosion resist-

.. ... ...

250 F. MANSFELD and B. LITrTLE

ance in seawater for the time of the experiment used by different authors. If localizedcorrosion would have occurred, E,.rr would have dropped to the active potentialstypical of pitted stainless steels such as the samples of Dexter and Gao12 whichsuffered from crevice corrosion.

As far as the mechanism responsible for the ennoblement of E,, is concerned, itis not possible from the experimental data for Ecorr or the polarization curves todecide whether the increase of Ecorr is due to thermodynamic effects, kinetic effectsor both. The reversible potential of the oxygen reduction reaction can increase bothdue to a local increase of the oxygen concentration or a decrease of pH. Thecomplexity of this situation is illustrated by the fact that naturally occurringmicroorganisms within a marine biofilm can decreasc or increase the local oxygenconcentration and pH. Little et al.2 15 have reported some of the first measurementsof the interfacial chemistry and have demonstrated its impact on the development ofEc,,rr with exposure time. An increase of the exchange current density for the oxygenreduction reaction would also lead to an increase of Ecorr. Since Ecorr is a mixedpotential, one has to consider that its value can also change due to changes in thepassive current density (c.d.) of the stainless steel. At constant kinetics of the oxygenreduction reaction, Ecor, would become more noble if the passive c.d. woulddecrease with exposure time. This discussion shows that a simple measurement ofEcor, cannot provide enough information for mechanistic interpretations.

The groups of Little and Mansfeld have shown recently 2' 16 that details of theexposure conditions can have pronounced effccts on the constituents of the biofilm,intei racial chemistry and the resulting Ecorr. In exposure to flowing natural seawaterit was found for the stainless steels SS 304, SS 316 and SS AL6X and for Ti grade 2that Ec,,rr did not change very much over a period of four months (Fig. 2).16 Thisresult which is quite different from the results of other authors discussed above(" 3

was considered to be due to the exposure under reduced light conditions and theresulting very low population density of photosynthetic, oxygen-producing diatomsin the biofilm. The surface properties of these samples have been further character-ized by electrochemical impedance spectroscopy (EIS) as discussed below and bySEM with EDAX. 16 No evidence of localized corrosion was detected on any of thesamples exposed to Pacific Ocean water for four months. Laboratory tests for thesame materials in sterile media such as NaCI and artificial seawater produced anennoblement of Ec,,rr in a few days. 17

The discussion presented above has focused on measurements of Eor forstainless steels in natural seawater. Additional Ec,,r data have been reported for mildsteel in the presence of SRB which is an important topic due to the problems withlocalized corrosion of buried steel pipes.18- 20 Kasahara and KajiyamaX havepresented an interesting application of the combination of electrochemical tech-niques such as the recording of Ecorr, linear polarization and impedance data formonitoring purposes. Ec, data have also been used in the evaluation of thecorrosion problems caused by fungal attack of Al alloys used as tanks for aircraftfuel. 2'

The redo. potentialThe reduction-oxidation (redox) or solution potential which is uvia!ty meas~ired!,, In iijert electrode such as Pt is an indicatoi ot the oxidation power of the

electrolyte. It could be used as an indicator of the corrosivity of the electrolyte if

Electrochemical techniques applied to MIC 251

suitable calibration were provided. In fact, measurements of the redox potentialhave been used to monitor changes in the corrosive properties of solutions as a resultof the bacterial metabolism. However, it has to be realized that a solution of a givencorrosivity, i.e. redox potential, can produce varying degrees of corrosion fordifferent metals such as mild steel, stainless steel or Al exposed to the same solution.Statements such as 'another potential useful for predicting whether or not a givenmetal will suffer an increase in corrosion due to microorganisms is the reduction-oxidation (or redox) potential of the envircnment "1 3 are therefore to be consideredwith great caution. A useful application of the redox potential would be in acombination with measurements of Ecorr and Rp for monitoring purposes. For theassessment of potential problems with MIC in soils the combination of measure-ments of the redox potential, soil conductivity and water content seems to haveprovided useful information in the past.2 -

It has been suggested recently to use a combination of a prepassivated Ptelectrode and the material of interest to follow the development of the biofilm anddetermine its effects on the corrosion behavior of structural materials.2 3 The timedependence of Ec,,rr of several stainless steels and Ti has been compared to theopen-circuit potential of the prepassivated Pt electrode. Figure 3 shows the timedependence of Ec,,,rr for SS 304L in a solution with and without A6F bacteria, while

80

40 -

0

-40

-80MV

-120

T

-160

*S30400* S31600

-200 - xN08366

-240-TTi

-280 _ _ _ _ _ _ _ I_ _ _I

0 20 40 60 so 100 120DAYS

Fi(;. 2. Open-circuit potential vs time for three stainless steels and Ti grade 2 exposedflowing natural seawater. 16

252 F. MANSFELD and B. Lirrt.E

W 00

~Medium

00

0.)

0 24 48 72TIME (hours)

FiG. 3. Open-circuit potential vs time for SS 304 in sterile medium (curve 1) and inmedium with bacteria (curve 2).23

Fig. 4 shows the corresponding data for Pt. 23 The authors concluded that the datashown in Fig. 3 and 4 reflect the processes of bacterial growth, changes of the localenvironment induced by the metabolic activities of the bacteria and the resultingcorrosion effects of the metal. The role of the Pt electrode was to indicate changes inthe local oxygen concentration which were assumed to be the same as those on thestainless steel under investigation. Since the potential of Pt changes in the positivedirection when the oxygen concentration is increased or the pH is decreased, a pHindicator was added to the solution to determine independently the pH changes. 23

The dependence of the reversible potential of the oxygen electrode (E",) on pH orthe partial pressure of oxygen (Po.) at 25'C is given by:

(dE0,)/(d pH)0 , = -59 mV (la)

(dE0,)/(d ogpo,)pH = + 15 mV. (1b)

~Medium

024 48 72TIME (hours)

Fir- 4. Open-circuit potential vs time for platinum in sterile medium (curve 1) and inmedium with bacteria (curve 2) .23

L 'U

'" - I .m m m ,mmm k mmmm l hmm m-

Electrochemical techniques applied to MIC 253

An increase of the local oxygen concentration at the electrode surface is thereforeexpected to produce only a small shift of the potential of a Pt clectrode which is usedas a redox electrode, especially since only parts of the surface aie usually coveredwith microorganisms and experience increases of po,.

The polarization resistance technique

The main advantage of the polarization resistance technique is the possibility to

continuously monitor the instantaneous corrosion rate of a metal exposed to acorrosive environment. 24 Therefore this technique is very suitable for the detectionof changes in corrosion rates due to the presence of bacteria, inhibitors, sunlight,biocides. etc. As with the measurement of Ecrr, a persistent problem with the use ofRP data is the tendency to overinterpret the experimental data as exact corrosionrates.

A thorough review of the use of the polarization resistance technique for themeasurement of corrosion currents has been given earlier. 24 Rp is defined as:

RP = (aE/ai)i = 0. (2)

The value Rp should therefore be measured as the slope of a potential E-currentdensity i curve at E,,r, where i = 0. The c.d. icor is calculated from Rp by:

ic,,rr =B/Rp, (3a)

where

B = b;,b,2.303(ba + b.). (3b)

The exact calculation of t .... for a given time during an experiment requirestherefore the simultaneous measurements of RP and the Tafel slopes b, and b, forthis time. Computer programs such as CORFIT,2 5 POLCURR 6 and POLFIT2 7

have been developed by Mansfeld and co-workers for the determination of precisevalues of ic,,,r according to equation (3). In the analysis of experimental R, data, ithas to be recognized that these measurements contain a contribution from theuncompensated solution resistance Ru. 24

R = Rp + R u . (4)

For electrolytes of low conductivity or systems with very high corrosion rates (lowR.) significant errors in the calculation of corrosion rates can occur if a correction forR, is not applied. The corrosion rate will be underestimated in these cases.Additional problems can arise from the effects of the sweep rate which is used todetermine R. according to equation (2). If the sweep rate is too high, the experimen-tal value of RP will be too low and the calculated corrosion rate will be too high. Carehas to be taken in the interpretation of Rp data for corrosion systems in whichlocalized corrosion is the main corrosion mechanism. For these systems, the

experimental Rp data should be used only as a qualitative indication that rapidcorrosion is occurring. Large fluctuations of the R data with time are often observedfor systems undergoing pitting or crevice corrosion.

A simplification of the polarization resistance technique is the linear polarizationtechnique- 4

21 in which it is assumed that the relationship between E and i is linear ina narrow range around U Usually only two (E.i) points are measured and it is

254 F. MANSFELD and B. LurrtE

often assumed that B has a constant value of about 20 mV. This approach is mainlyused in field tests and forms the basis of commercial corrosion rate monitors. For theapplications related to MIC, the combination of measurements of Eorr, the redoxpotential and RP would be very useful for monitoring purposes, however, very fewauthors have reported the results of such investigations. hp can also be determined asthe d.c. limit of the electrochemical impedance. This approach will be discussedbelow in the section dealing with electrochemical impedance spectroscopy (EIS).

Applications of the polarization resistance technique have been reported by Kinget al. 19 in a study of the corrosion behavior of iron pipes in environments containingSRB. Other techniques used in this study were potential/time monitoring, zeroresistance ammetry, electrochemical noise and a.c. impedance. The authors con-cluded that 'linear polarization resistance monitoring (LPRM) was the most usefultechnique'. In a similar study, Kasahara and Kajiyamal8 used R, measurements withcompensation of the ohmic drop and reported results for active and inactive SRBconditions. Nivens et al.29 calculated the corrosiop c.d.i,.,,. from experimental R.data and Tafel slopes for SS 304 exposed to a seawater medium containing the non-sulfate-reducing bacterium Vibrio natriegens.

Mansfeld et al.30 have used the linear polarization technique to determine R formild steel sensors embedded in concrete. The concrete samples were exposed to asewer environment in Los Angeles County for about nine months. One sample wasperiodically flushed with sewage in an attempt to remove the acidic environmentproduced by the biofilm, the other sample was used as a control. A special datalogging system wa. used to collect RP data at 10 min intervals simultaneously for thetwo corrosion sensors and two pH electrodes which were placed on the concretesurfaces. Figure 5 shows the cumulative corrosion loss 1 INT, where INT wasobtained by integration of the 1/Re-time curve as:

INT = dt/Rp. (5)

A qualitative measure of the corrosion rate can be obtained from the slope of thecurves in Fig. 5. The integrated corrosion loss INT is given in Fig. 5 units of ohm -Due to the presence of the uncompensated ohmic resistance aiid tre lack of valuesfor the Tafel slopes (equation 3), the data in Fig. 5 should be viewed only asindicative of significant changes in corrosion rates. The corrosion loss remained verylow during the first two months and then showed a large increase for both samples.This increase occurred at about the same time that the surface pH reached very lowvalues of I or even less. The total corrosion loss as determined from the integrated R.data was less for the control than for the flushed sample. These results wereconfirmed by the EIS data shown below which were obtained for the two systems atthe end of the field test.

The dual-cell techniqueThe dual-cell or split-cell technique developed by Little et al. for the study of

MIC 1.31 allows continuous monitoring of the changes in the corrosion rate of a metaldue to the presence of a biofilm. In this technique two identical electrochemical cellsare biologically separated by a semipermeable membrane. The two working elec-trodes are connected to a zero resistance ammeter (ZRA) or a potentiostat set to anapplied potential of 0 mV. Bacteria are then added to one of the two cells and the

lectroch,- nic. l techniques applied to MIC 255

pH =2.0 /

70 -E

V4q 35

0

I I I I I0 60 120 180 240 300DAYS

150(b)pH = 4.0 -, -

100

V. 50

0 60 120 DAYS 180 240 300

ho S. 5 Intcgirated corrosion loss 1 INT for mild steel sensors embedded in concrete.30

resulting galvanic current is recorded on a strip chart recorder to provide acontinuous record of the time dependence of the corrosion process. The sign and themagnitude of the galvanic current can be used to determine details of the corrosiveaction of the bacteria. It has to be recognized that the galvanic current is a measure ofthe increase of the corrosion current of the anode due to the coupling to the cathodeand does not allow calculation of the corrosion rate of either electrode in a simplemanner. In each experiment with the dual-cell technique it has to be establishedthat current flow does not occur between supposedly identical electrodes. Additionalmechanistic information could be obtained if the potential of the galvanic couplewere also to he recorded as a function of time.

Dauns e, ,1.33 have used the concept termed a 'biological battery' to study thecorrosion mechanism of K55 steel in the presence of hydrogenase-containing(Ilase ) and non-hydrogenase (Hase ) bacteria. Figure 6 shows the galvanic c.d.between a sterile cell and an inoculated cell in the presence of SRB Hase (curve A)and in the presence of SRB Hase + (curve B). The authors concluded based on theseresults and others obtained in the absence and presence of sulfide that the oxidationof cathodic hydrogen by the bacteria was the dominant corrosion mechanism for the

256 F. MANSM-i) and B. LDi-rrE

Hase organisms. They suggested that FeS played an important role in the corrosionreaction with Hase - organisms.

Little et al.3 have used the split-cell technique to simulate the oxygen-deficient oreven oxygen-free conditions which occur inder biofilms when oxygen is taken up bythe respiring microorganisms as rapidly as it diffuses to the metal surface. Whenoxygen was bubbled though both cells with Ni 201 as working electrodes, the galvaniccurrent remained at zero. Similarly, no current flow was observed when nitrogen waspassed through one cell at 23°C (Fig. 7). However, when the temperature %kas raisedto 60'C under the same conditions the electrode in the deacrated compartmentbecame the anode due to complete removal of oxygen (Fig. 7). Other applications ofthe dual-cell techniques can be found in Refs 34-36.

Ele-trochemical impedance spectros opyEIS is a relatively new technique in corrosion research which has found many

successful applications. A recent summary of the background of FIS and appli-cations in many artas of electrochemistrv can be found in the proceedings of the FirstInternational Symposium on EIS which was held int 1989. 37 Mansfeld 1,s discussedsome basic concepts concerni"g the recording and anlalysis of impedancc data 3S 31)

and with Lorenz has compared the results obtained with a.c. and d.c. techniques foriron in the presence of corrosion inhibitors .

In the EIS technique the impedance data are recorded as a function of thefrequency of the applied signal at a fixed working point (E,i) of a polarization curve.In corrosion studies this working point often is the corrosion potential (E = E,_11i = 0). Usually a very large frequency range has to be investigated to obtain thecomplete impedance spectrum. In most corrosion studies this frequency rangee,tends from 65 kHz. which is the upper limit of a commonly used frequencyresponse analyser (FRA), to 10 mHz. The impedance data are usually determinedwith a three-electrode system, although it is also possible to use a two-electrodesysteim in which both electrodes are of the same material. A potentiostat is used toapply the potential at which the data ate to be collected. The FRA is programmed to

E

-A HASEInoculatiron IS HASE I

-'- A

I III12 24 36 48

TIME (hours)

Fi. 6. Time dependence of the galvanic current in the presence of SRB Flase (curve A)and SRB Hase ' (curve B).

Electrochemical techniqies applied to MIC 257i

50

6000- j - - -

E All

I-uJ I

ZD .23*C0 ....

-100 25 50

TIME (hours)FIGi. 7. Time dependence of the galvanic current resulting from ahioticully created

differential aeration.31

apply a series of sine waves of a constant amplitude, which needs to be small enoughto remain in the linear potential region. and varying frequency. The impedance dataare determined by the FRA at each frequency and stored in its memory. Since a verylarge number of data has to be collected, displayed and analysed it is essential to useadequate software for these purposes.

The type of information obtained with EIS differs from that determined with theother techniques described in this review since the corrosion system is analysed at afixed potential (or current). The properties of the system at this potential can then bedetermined through the analysis of the frequency dependence of the impedance.One of the advantages of EIS is the fact that only very small signals arc applied. Inmany cases. EIS data have shown that existing models for corrosion systems based onresults of studies with d.c. techni,iies were too simplistic. Only when all details of animpedance spectrum can be explained in the entire frequency range can one besatisfied that the model employed is indeed correct. The inhomogeneous surfacemodel proposed byiJuettneretal.4 1 is a good example for the development of a modelwhich explains the coupling of charge transfer and mass transport reactions in asupposedly simple system such as the corrosion of iron in neutral aerated media. Avery important point in the use of the EIS technique is the development of softwarefor the analysis of the experimental data. Such software needs to be tailored to theparticular system to be studied as will be shown for the analysis of data obtained forcalcareous deposits on stainless steels.

Several reports have already been published in which EIS has been used for thestudy of MIC. The evaluation of some of these reports is often made difficult by theunfortunate tendency to display the impedance data using linear coordinates incomplex plane plots. As pointed out elsewhere, 42 this approach neglects the fact thatthe impedance often changes over many orders of magnitude and that the frequencydependence of the impedance cannot be recognized in complex plane plots. Thepreferred fornat for the display of EIS data is therefore the Bode-plot in which logA. and the phase angle ar, plotted vs the frequency Y) of the applied signal with IZ

being the modulus of the impedance. Often ihe EIS-data have been only used to

258 F. MANSFELD and B. Llrir.

determine the polarization resistance RP which is defined as the d.c. limit of the realpart of the impedance:

RP =im {Re(Z)} - R u, (6)

where Ru is the solution resistance. In this case all the other important informationcontained in the impedance spectrum is ignored.

Several applications of EIS are contained in the Proceedings of the InternationalConference on Biologically Induced Corrosion. These papers deal with the role ofSRB in the corrosion of buried pipes 18 ' and reinforced concrete.4 7 The analysis ofthese data is qualitative and no models for the impedance behavior in thesecomplicated systems have been presented. Dowling et al. have studied the effects ofMIC on stainless steel weldments in artificial seawater using EIS, small amplitudecyclic voltametry and ESCA.43 From the frequency dependence of the impedancedata it was concluded that two relaxations were associated with the as-weldedinoculated surface, while only one time constant was detected in the Bode-plots forthe as-welded sterile surface or the polished surfaces. 43 The authors speculated thatthe occurrence of a second time constant was due to the development of pits.Dowling et al.44 have also used EIS to study the corrosion behavior of carbon steelsaffected by bacteria. Attempts were made to determine R. from the EIS data.

The formation of the biofilm and calcareous deposits on three stainless steels andon Ti grade 2 during exposure to natural seawater has been followed using EIS andsurface analysis with SEM/EDAX by the groups of Mansfeld and Little."6 Sampleswere either exposed at E.)rr or polarized at -850 mV(SCE) for certain time periodsand then removed for analysis. Figure 8 shows a typical example for the type of dataobtained for the unpolarized samples. The impedance is dominated by the capacitivecomponent. No changes in the shape of the spectra were detected as the biofilmformed with increasing exposure time. Figure 9 shows the time dependence of theelectrode capacitance (C11) for the four materials. The small decrease of thecapacitance with time could be due to the formation of the biofilm which was shownby observation with SEM to be of a patchy nature. 16

Much more pronounced changes in the impedance spectra were observed for thepolarized samples. Figure 10 gives spectra at three different exposure times for

6 [- - __- - 906 2 SS304, unpolarized 9

1. 5days5t", 2_1 2.43 days t75-- F" ,' 3 " 3. 81 days i-

1.~3 8- 1

-9 -1 0 1 2 3 4 5

LOG f (H)

Fi(i. 8. Bode-plots for unpolarized SS 304 exposed to flowing natural seawater for

different time periods..

Electrochemical techniques applied to MIC 259

100

I. SS304a AL8X

80- TiL

Ezt60a

40

20 "I0 20 40 60 80 100 120 140

TIME (days)

FIG. q. Time dependence of the electrode capacitance Cad for unpolarized stainless steelsand Ti grade 2 exposed to flowing natural seawater. lb

90SS304, polarized

1. 5 days75 -2. 13 days

75-

2 1 da

-

60 3- 3

c_,45

< 230

15

5

E4E

0

-2 -1 0 1 2 3 4 5LOG f (Hz)

FiG. 10. Bode-plots for SS 304. polarized to -850 mV(SCE). as a function of exposuretime to flowing natural seawater.

260 F. MANSFELD and B. LrrLE

90Ti, polarized

75 2. 13 days~3. 43 days

603

ZD45

15-

5 -

E4 -

E

3 3

22

-2 -1 0 1 2 3 4 5

LOG f (Hz)

Fi(;. 1I. Bode-plots for Ti grade 2. polarized to -850 mV(SCE). as a function of exposuretime to flowing natural seawater.

SS 304. Similar data were obtained for SS 316 and AL6X. At very short exposuretimes one time constant was observed, similar to the data for the unpolarizedsamples. As the calcareous deposit was formed, a second time constant appeared dueto the presence of this porous deposit. SEM analysis showed that the area which wasnot covered by the calcareous deposit was partially covered by bacteria. When thecalcareous deposit covered the entire surface and its porosity became quite small, thetime constant at the higher frequencies disappeared again. The data for Ti grade 2after 43 days were different from those for the stainless steels insofar as two timeconstants were still apparent (Fig. 11). This behavior was explained by the differentmorphology of the calcareous deposit on Ti using the model for the impedance of aporous polymer film on metals. 45 A full analysis of the data with the COATFITsoftware is given elsewhere. 16 It is remarkable that the predictions concerning thedifferences in the structure of the deposits for the stainless steels and Ti are fullyconfirmed by the SEM data as shown in Fig 12

Mansfeld and Shih have shown recently that EIS can be used to detect theinitiation and growth of pits on Al alloys as well as on Al-based metal matrixcomposites.46 This possibility would be very useful for the study of MIC. where manycorrosion reactions are considered to occur because of the localized action of

,:. - .; . . •.

FIG. 12. Surface of (a) SS 304 and (b) Ti grade 2 after polarization to -850 mV(SCE) inflowing natural se -water for 43 days. 16

261

,"vF

Electrochemical techniques applied to MIC 263

7 90

1315

LOG f (Hz)Fin. 13. Bode-plots for SS 304 exposed to natural seawater (curves I and 2) and to

pasteurized, filtered seawater. 17

microorganisms. In the study of stainless steels and Ti exposed to seawater, no signsof localized corrosion were detected visually or by EIS over the entire test period ofmore than four months. However, in laboratory tests conducted in seawater, largechanges in the shape of the impedance spectra were observed when a faulty 0-ringwas used which created crevice corrosion of SS 304. Curve 1 in Fig. 13 shows the typeof spectra which were recorded in the presence of crevice corrosion. 7 Two timeconstants dominate the spectrum in this case as can be seen by the occurrence of apeak in the phase angle at frequencies between 100 and 1000 Hz. Rp has decreased toabout 3 × 104 ohm cm 2. The surface pretreatment for the sample in curve 1 consisted

of polishing and degreasing. When the surface was passivated by immersion inconcentrated HNO 3 . crevice corrosion did not occur and only one time constant wasIobserved (curve 2). Some deviations from capacitive behavior were also detected fora sample which was exposed to sterilized seawater by the changes of the impedancebetween 100 and 1000 Hz indicating the initiation of localized corrosion. Care has tobe taken not to interpret these impedance data as being due to the formation of aporous biofilm. 17

Impedance measurements have confirmed the conclusions discussed above forthe corrosion monitoring effort in a corrosion chamber fed by sewer gases (Fig. 5).3°

After the field tests had been terminated the samples were returned to the laboratoryand E1S-data were collected after immersion in tapwater for 48 h. The impedancedata were determined for the two-electrode system at E pn= 0 inV. As shown in Fig.14. the impedance at the d.c. limit and therefore also R,, were lower for the mild steelsensors which were embedded in the concrete samples flushed periodically withsewage. This result is in agreement with the Rpdata collected independently on thebasis of the linear polarization resistance technique (Fig. 5). 3o

Electrochemical noise analysisIn electrochemical noise measurements the fluctuations of the potential, often

Ec ,,, ort the current are determined as a function of time or the experimentalconditions. One of the advantages of this technique is that no external signal needs to

2) ,M ,A.

, : ,;R',,. =. .

Mn,

264 F. MANSFELD and B. LITTLE

be applied for the collection of the noise data. In some cases the potential of a testelectrode is measured versus a stable reference electrode whereas in others the noisedata are recorded for two electrodes of the same material which are exposed underidentical conditions. In principle, a statistical analysis of the noise data can provideinformation concerning the nature of the corrosion processes and the magnitude ofthe corrosion rate. However, most analyses of electrochemical noise measurementshave been qualitative and used to determine the occurrence of localized corrosion.The frequency and the amplitude of the fluctuations are sometimes used to estimatethe value of the corrosion rate. The results reported so far for the application ofelectrochemical noise measurements suggest that this technique at its present state ofdevelopment is more suited for monitoring purposes than for mechanistic studies.

A very useful approach for corrosion monitoring is that used by the researchgroups at the University of Manchester Institute of Science and Technology(UMIST), who combined impedance and electrochemical noise measurements in aseries of experiments which were designed to simulate conditions in concrete oilstorage tanks. 47 The effects of SRB and other communities of microorganisms on thecorrosion of concrete and the steel reinforcement were studied. The authors used aportable multi-channel instrument which also performed zero resistance ammetrv.47

The electrochemical noise data were collected with a microvoltmeter and a computerwhich collected time records consisting of 1024 potential readings taken at onesecond intervals. These data were converted to the frequency domain and thensubjected to a statistical analysis. Moosavi et al.47 presented their results for areinforced concrete block exposed to a marine medium containing active SRB in twonoise plots: a time record and a potential distribution chart showing the populationand the magnitude of the potential fluctuations. The data were collected at Ecorr. Thetime record after 218 days for the covered rebar reveals events that could be due tosudden rupture of the protective oxide film followed by immediate repassivation(Fig. 15). The events recorded for the exposed rebar showed fluctuations, however.their magnitude was about a factor of 10 lower (Fig. 16). King et al.19 interpretedtheir results obtained with noise measurements for steel pipes in environments

5 90Steel in concrete after 250 daysin sewage system

1. control 752. flushed

-60J

r..JN--4- 45 Cu

S 2

15

3 1-3 -2 -1 0 1 2 3 4

LOG f (Hz)Fio. 14. Bode-plots for mild steel embedded in concrete after 250 days exposure in acorrosion chamber fed with sewer gas; curve 1: control, curve 2: flushed periodically with

sewage.-U

" ._

Electrochemical techniques applied to MIC 265

+-0530~ TIME RECORD: NP26

-.0530'

+.06"/6.'"

.i . .°

-1737

-.2943138.96 39.03 39.10 39.17 39.24

TIME (hours)

POTENTIAL DISTRIBUTION (-re and +ve)1000 " 1000

(b)

MAGNITUDEPOPULATION

100 100

uJ

z IS10 III10

I I

I Iitl

1 I0.001 0.01 0.1 1 10 100

MILLIVOLTSFto. 15. Time record after 218 days for covered reinforcements in media containing activc

cultures of SRB.47

containing SRB as being indicative of film formation and breakdown as the probablemechanism of corrosion. Similar conclusions had been reached from the analysis oftheir zero resistance ammetry data.19 In the analysis of noise data such as thoseshown in Figs 15 and 16 it has to be considered that the magnitude of the noisefluctuations depends on the total impedance of the system. A corroding metal

undergoing uniform corrosion with fairly high corrosion rates might be less noisythan a passive metal which shows occasional bursts of noise due to localizedbreakdown of the film followed by rapid repassivation.

The experimental approach of Iverson reported in 196848 differs from thatdescribed above insofar as the test specimen is coupled to a Pt electrode and the noise

is measured across a resistor between the two electrodes. In this case the testelectrode was polarized to an unknown extent and the measured noise might be aresult of this polarization. This limitation of Iverson's approach was not reported byDexter et al.3 in their recent review which seemed to suggest that all noise

266 F. MANSFELD and B. LITMLE

measurements are carried out in this manner. Nevertheless, Iverson et al.49 haveobtained potential noise data in the laboratory for mild steel in a trypticase-seawaterculture of the marine strain of Desulfovibrio and in the field for a gas transmissionline. The authors concluded that breakdown of the iron sulfide film may beaccompanied by the generation of electrochemical noise. They also were suggestingfurther development of the noise technique for monitoring purposes.

Large signal polarization techniquesIn the experimental techniques described above, measurements are carried out

without the application of an external signal (Ecorr, dual-cell technique, electro-chemical noise) or with the application of only a very small perturbation (Rp, EIS).The large signal polarization techniques employ potential scans which can rangefrom several hundred mV to several V. The recording of a polarization curveprovides an overview of the types of reactions which occur for a given corrosionsystem such as charge transfer or diffusion controlled reactions, passivity, transpass-

+.0054 TIME RECORD: NP20

(a)

+.0027•

4 . " " "

.- 1= % '14:

-.0054, I I I36.26 36.33 36.39 36.46 3653

TIME (hours)

POTENTIAL DISTRIBUTION (-ve and +ve)100: o100

MAGNITUDE (b)

POPULATION

10100

10 0.

I I

I

(00 0.01 0.1 1 10 100MILLIVOLTS

FIG. 16. Time record after 218 days for exposed reinforcements in media containing activecultures of SRB. 47

Electrochemical techniques applied to MIC 267

ivity and localized corrosion phenomena. The effects of alloying or the addition ofinhibitors can be recognized in the changes of the polarization curves. While therecording of potentiodynamic polarization curves seems to be very simple, especiallyif commercially available software is used, it has to be remembered that theexperimental conditions have to be tailored to the electrochemical characteristics ofthe system under study to obtain optimum results. The choice of the correct sweeprate and the elimination of the ohmic drop during the experiment are importantconsiderations as discussed elsewhere. 5° The use of excessive sweep rates as in the'corrosion behavior diagrams' obtained by the rapid scan potentiodynamic polariz-ation technique 3 reflects the effects of the experimental conditions on the shape ofthe polarization curves, but does not result in any meaningful mechanistic data.Polarization over wide potential ranges can drastically alter the properties of thesurface under investigation. Therefore separate samples should be used for therecording of anodic and cathodic polarization curves and each sample should only beused once. Quantitative information which can be obtained from polarization curvesincludes the values of the Tafel slopes ba and bc, the corrosion c.d. iorr, the values ofthe diffusion limited c.d. for hydrogen reduction i1j 1 , and oxygen reduction i1 , .o andparameters related to passivity such as the critical potential Ecrit and critical c.d. icrifor passivation and the passive c.d. /pass" Additional parameters such as the pittingpotential Epit and the protection potential Eprot are related to the susceptibility of ametal to localized corrosion. Since the recording of polarization curves is a well-established technique covered in many text books and in ASTM Standard Recom-mended Practices such as G3, G5, G59 and G61. no further discussion concerningthe details of experimental procedures is needed here.

Numerous investigators have used the recording of polarization curves todetermine the effects of microorganisms on the electrochemical properties of metalsurfaces and the resulting corrosion behavior. In most of these studies comparisonshave been made between polarization curves recorded in sterile media with thoseobtained in the presence of bacteria and fungi. Early studies of MIC of mild steelwere reviewed by Costello, who in 1969 provided a survey of the literature. 5 1

Recently Tiller 52 reviewed the European research effort on microbial corrosionbetween 1950 and 1984. His paper contains 21 references to the use of polarizationcurves for the interpretation of the impact of SRB on mild steel. A further review ofpolarization techniques applied to MIC has been given by Duquette and Ricker 53

who among others cited the work of Salvarezza and Videla 54 and Obuekwe et al.55 56

While there is no general agreement concerning the details of the corrosionmechanisms, most authors suggest that the observed increase of corrosion rates isdue to the influence of the microorganisms on the rates of the anodic and/or cathodicreactions involved in the corrosion process.

The complexity of the naval brass/polluted seawater interface was demonstratedby Deshmukh et al.57 who discussed the influence of sulfide pollutants based on theresults obtained with potentiodynamic polarization curves. The shape of the anodicpolarization curve was drastically changed in the presence of sulfides or SRB (Fig.17). Ec,,rr became substantially more negative. An active-passive transition andhysteresis in the reverse scan were observed. An interesting application of polariz-ation techniques has been given by de Mele 58 in an evaluation of the biodeteriorationof implant material including aluminum in human plasma and silver in saline solutioncontaining Na2S.

268 F. MANSFELD and B. LITTLE

-200.

-40

15 SR8 INOCULATED

....... ... .. ....- IE + 0 P

. . . . . . . .

-70...0............. SULFIDE

..... STERILE-700 - _ I I IIII l 'l I I il

0.001 0,01 0.1 1.0 1100

CURRENT DENSITY (mA/cm 2)

Fi. 17. Potentiodynamic polarization curves for naval brass in seawater media. 7

Polarization techniques have also been used in attempts to determine themechanism by which microorganisms induce localized corrosion in the forms ofpitting or crevice corrosion. In most cases the pitting potential Epit has beendetermined in the presence and absence of bacteria. At this point it has to beconsidered that the pitting potential Epit provides only the tendency for pitting, butdoes not give information concerning the rate at which pits propagate. For Al alloysin sterile 3% NaC! it was found that A] 2024 has a more noble Eri, than Al 1100, Al7075 or Al 6061, however the localized corrosion rate was the highest for Al 2024.The extent of pitting cannot be predicted from Epit data alone.

Salvarezza et al.596(' have used several isolates obtained from jet fuel storage tanksludges to obtain information concerning the influence of different microorganismson the corrosion process of Al alloys in fuel-water systems. The authors concludedthat Epit could be used to determine the importance of each biological species in thecorrosive attack of the Al alloys. Rosales et al.6 1 - 4 have determined the influence ofmicrobial contaminants in aircraft fuel, alloying elements and surface heterogen-eities on the nucleation and propagation of pitting of Al 7075 and have concludedthat the electrochemical tests reproduced the corrosion problems which wereobserved in service conditions.

The experimental conditions used to record a pitting scan can affect the value ofEpit If the scan rate is too high, Epit will usually be higher than the value determinedunder steady-state conditions. Likewise the experimental value of Eprot becomesmore negative if the c.d. at which the scan is reversed is increased. Because of theseproblems. Epit has been determined sometimes by a constant-potential technique inwhich a series of increasing anodic potentials is applied and the current is measuredas a function of time. If the current decreases continuously at an applied potential E,it is concluded that E is more negative than Epit.By increasing E stepwise, Epit can bedetermined. Very few applications of this very useful approach can be found in theliterature for the study of MIC phenomena. Videla65 has used this technique in astudy of the action of Cladosporium resinae growth on the electrochemical behavior

Electrochemical techniques applied to MIC 269

of aluminum. Current transients at different potentials were compared for Al 2024 in0.01 M NaCI and in 0.01 M NaCI + 0.05 M citric acid (pH = 2.5). These data suggestthat Ei t in 0.01 M NaCI was between -0.55 and -0.50 V(SCE) and was morepositive than -0.55 V in the solution of pH = 2.5. These results confirm that Epitdepends on pH. It should be noted that a printing error 65 which was copied in theDexter et a/. review3 might make it difficult to analyse Videla's data.6 5 Curve a in Fig315 was measured at -0.55 V as in Fig. 4 rather than at -0.05 V as stated in the figurecaption for Fig. 3.

Ringas and Robinson"' have performed electrochemical tests on several stainlesssteels and on mild steel in three cultures of SRB. In all cases the pitting resistance waslower in cultures of SRB. Figure 18 shows potentiodynamic polarization curves witha reversed scan for SS 316 in sterile medium and in strain 8303.66 In the sterilemedium, the alloy was passive and displayed a large passive range. The reverse scanshowed that active pits were not formed. The shape of the anodic polarization curvewas very different in the bacterial culture. Ecorr was more active, an active-passivetransition was observed and the passsive c.d. was higher suggesting that the passivefilm formed in the bacterial culture was less protective. The reverse scan showed thatactive pits had been formed which did not repassivate. Similar results were shown forSS 409 and 430.( ' The authors concluded that sulfide-induced pits were initiated atmore negative potentials than those initiated by chlorides. Additional studies bythese authors6 7 involved total immersion in cultures of SRB and surface characteriz-ation with SEM. The results of this study agreed with those from the electrochemicaltests."'

SUMMARY AND CONCLUSIONS

The authors of this review would expect the reader who has followed theirdiscussion to this point to pose the obvious question: 'Are electrochemical tech-niques helpful for the study of MIC?' A general answer by the authors would be:

1. ST E I E ; ; "'RA IN 8303

1.5

1.1

0.7

~0-0.1

_ I I I i I I

101 102 103 104 105 106

CURRENT DENSITY (nA/cm 2)Fic,. 18. Pitting scans of SS 316 in sterile medium and strain 8303.

270 F. MANSFELD and B. Llrft.E

'Yes'. The details of the positive answer depend, however, on the particular interestsof the person asking the question. Obviously in an area of corrosion as complicatedas MIC, one single technique cannot be expected to provide all the answers which areneeded for the establishment of mechanisms. As in other fields of corrosion, inputsare needed from other disciplines. A case in point would be the phenomena of stresscorrosion cracking or corrosion fatigue, where electrochemical tests have providedanswers to explain the results of experiments based on the concepts of fracturemechanics. The most productive use of electrochemical techniques in basic researchin the field of MIC would be in providing answers to questions posed by the experts inthe fields of microbiology and corrosion.

In judging the usefulness of electrochemical techniques for the evaluation ofMIC. it should not be forgotten that many billions of dollars which have beeninvested in the equipment of oil refineries, chemical plants and other facilities areprotected from corrosion including MIC by electrochemical monitoring techniquessuch as the linear polarization technique. This very simple concept has foundworldwide acceptance as a monitoring tool in which the exact details of the corrosionmechanism might sometimes not be too well understood, but where the experienceof the responsible corrosion engineer suggests the steps which are needed to avoid acostly corrosion problem. With further investments in basic research concerning theapplications of electrochemical techniques in the study of MIC combined withmicrobiological and surface analytical techniques, similar success should evolvequickly in basic research and practical applications.

,Atr'knorredrtmnti-NORl)A Contribution Number 333:021:901: approved for public release: distri-bution is tnlimited.

REFERENCES

I. S. M. (;Inrt IAKov. B. J. Lrrrii- and P. WAiNER. (orrosion 42. 689 (1986).2. B. .1. l[Iiv-, R. RAN. P. WAGNR, Z. I.EWANDOWSKI, W. C. LI- . W. G. (HAR.\KIIS and F.

MANS- t n, Electrochemical behavior of stainless steels in natural scawater. 'orroonm9), Papcr No.W0150.

3. S. C. Dr-.x ti., 0. W. SIa-ir 1. I). J. DuQuFI F-v and H. VIDELA. Use and lir.itations of eLctrochcmical

techniques for investigating microbiological corrosion. Corroesion/89. Paper No. 616.4. R. F. HADt ,.V. The influence of Sporovibrio desudfuricanx on the Current and Potential Behavior of

Corroding Iron. National Bureau of Standards Corrosion Conference 1943).5. J. N. WANKt. N and C. I. P. SPRuIr. U Aanre 169. 928 (1952).6. A. MoI.I.IA, A. "lTRvts, E. TrAvERsO, G. Vl-N[ICRf.. V. Scorto, (, At .i\msio, G. MIrWINsaRo, 1L'

MONuiINi. G. .I: ('AROIts and R. Di.I EP'IANF. Interaction between biotouling and oxygen reductionrate (in stainless steel in seawater. Prot. 6itl Int. (ong. Marine (orroson arid IotlintL. Athe ns.(ireece. p. 269 (1984),

7. V. Storro. R. DECINtro and G. MARCENARO. ('orros. Sci. 25. 185 (1985).8. R. JotiNsr.s and E. BAR~AL. Corrosion 41. 296 (1985).

. R. JOINSEN arid E. BARtDAi. The effect of a microbiological slime layer on stainless steel in naturalseawater. (orrosion/86. Paper No. 227.

10. H. A. VIDIA. M. F. .. DE MItt and G. BRANKEVtCiI, Microfouling of several metal surfaces in

polluted seawater and its relation with corrosion. Corrosion/87. Paper No. 365.II. A. MOLuCA. G. VENIURA, E. TRAV Rso and V. SroTro, Int. Biode'erior. 24. 221 (1988).12. S. C. Dt XER and G. Y. GAO, Corrosion 44. 717 (1988).13. A. MottICA, A. Trrvs. E. TRAVE.RSO. G. VENtRA. G. D)FCAROtLSs and R. 1IDt I rritNi, Corrosion

45.48 (19s9).14. F. MANSFL.D and B. J. Lxt.r. Discussion of Ref. 12 and reply by S. C. I)ElR and (N. Y. (imo.

(orro.ion 45. 786 (1989).

I' -i =d m 'mli~~mm mlmm m

Electrochemical techniques applied to MIC 271

15. Z. 1.EWAN[)OWSKI. W. C. LEE,~ W. G. Cir.RAKIAjS and B. LIrrLE. Dissolved oxygen and pHmicroelectrode measurenments at water immersed me~tal surfaces. Corn.,;ionI89. Paper No. 93.

16. F. NtANSUFtL, C. H. TSAI, H. Sum.j B. I-trtuE R. RAY and P. WAGNER. Results of exposure oftansssteels and titanium to natural seawater. Corrosiont/9f,. Paper No. 90109.

17. F. MANSFIldo1, C. H. TSAI. H. Siimi and B. J. I-irtiEi, An EIS evaluation of stainless steels exposed toseawater. 175th meeting of the Electrochemical Society. Los Angeles. CA. Paper No. 57 (1989).

IN . K. KASAItAtiA and F. KAJIVAMA. Biologically induced corrosion. Proc. lt. (onj.. NACE-8, p. 171( 198h).

91. R. A. Kisc;. B. S. SKERRYi, 1). C. A. MOORE, J. F. D. St~yi and J. L. DAWSON, Proc. Int. Conf.VA)h-8 p,. 83.

"'0. A. E. SMoROIN.lj. N. M. AGAER. M. M. (hsE~ov and A. V. AitAKiltEtinovA. Prot. Metals 22. 391(1986).

211. R. C. SAst VARFiZA. M. F. L. OF, Miiit and H. A. VIDELA. Corrosion 39. 26 (1993).22. G. H. Bloott.mi(~il~iO r~. Mills and Boon. London (1971).23. X. /.IIANu. R. A. BU(IiANAN. E. E. StiANSHUR) and N. J.- E. LDowt~iNG. Eleet roehemical responses o1

stiructujral materials to rnicrobiologicallv induced corrosion. Corrosionl8o), Paper No. 512.24. F. MANSI-t t) Ihe polarizatin resistance technique for measuring corrosion current,.. Advances tit

'rioi cieijice aid Fcltol og.' ol. 6. p. 163. Plenum Press, New York (1476).25 F \ NI - I . Corrosion 29. 397 (1973).2(1. S. M. (ii-tm ii.Akov. I . R. t..t)E and F. MNSFisrri. (Corrosion 37. 696 (1981).27. H- Sitti and . 0 i. l'OLFIT-A Comp~uter P

t rogramt jor tite, Artahsis of Pokarizunon Cuovesitith ( omntmitsaon Of Ille ()hmo lDrum. Corrosion and Environmental Effects Laboratory. ( SC

'S M. titm'RN and A. L- 61 ixum J1. ClCCII-001i. SOC. 104. 33 (1957).21 ). t- Nt iNs. P). I). Ntonot s. J1. M. HrsoN. (. G. GEEy and D. C. Wi:IrF. Corrosiont 42. 204

I \ NSI I t t n. A. Posts\ 1.1. Stint, .1. D~iiN R. lI ANDER and C. L. CHEN. Corrosion monitoringmid co nt rol tn ot concrete 'c\ecr pipe'.. ( mriom oni9,0 . Paper No. 113.

iB, J, 111-1 1t. P'. W SMiNt R. S. %I. C LR( tiAKitv . MI. Wm t cit and R. Nmn cimt tI Conrrosiont 42.533 (1986).I I Asstri tI. ( onrnini%ioit 29. 403 (1973).

33S. l).xm s. Y'. \I \ssi Ami and J1. (tSit Rli ( orrm.\ Sot. 28. 1041 ( 1988).34. BJ .. 1 111 t iid P.), \\ \,N.i. . An elcetroche mical evaluation if inicritbioloc'ica lv tinduced corrosion

h\ tiso i ron ix diii n b acteria . ( orrot ion/SI,. Paper No. 1 22.11S . W \(,-, ii .nd B. J. 1.1 ]-11 1 . A.pplications. of a technique for thie investigation of miicrolbiologicallxinduced corrosion. ( (orosionVNo. Paper No. 12 1.

3(,, 13 .1 li 11 . P). W \.im t and S. MI. (inR( i..sov. A quanitatio'.c investigation of mechanisms formien ial corroion . Pto , /tlit ( opt. Rfiohnpzi a//v Intduced Co~rroson')1 .1 ( 1>8. p. 2019 (1986).

;7 l'tm I irnt . .Sminj?. IS. Bam banntes. f raine N10a, 1)9) .

\18 \-NI xi~t i .( ommni 37. 301 (I lI)X It- 1 NI Nsit t I),\ W. Ktxoti, and S. l1xxi. ( orroimi 38. 571 (982).

40. \%, J1 LOI N/ and F<.s xii t 1i1. (orrm.~ 21, 047 ( 11)81).41. K. Ji i it im. W. J, Lim N/. NI. W. KisONiti and F. MNsitN to 1. J. cmm trothem. Som. 135. 332 19088).42. 1-. NI '.ssit i I). Corroinh 44. 558 (1f988).43. N. J, E )ii' tiNt,. M. FutaNud iN. 1). C. Wi iti. C. H. 1,1t and C. LxI-NDN. The effect of

ntierobitli ta fly influ ,need co irrosion n ii tn tless %teel weldinents' in a rtifieial seawater. (Cnrro-ston/59. Paper No. 187.

44. N. J. E. Ditwi INiG. J. 0t [.tNi(. . . Li mait:. A. Ti, ii) and 1). C. Wii . Coirrosioin 44. 969

(1988).41. F". NIANSrn.. NM. Ki/Nmto and S. issi. ( orropon 38. 478 (1482).46. F. .MAxNSFthm and H. Stmi. J. eleo'crrocwm. 'Sot. 131. 1171 (1988).47. A. N. IMOOxAVI. 3. L. [DA'WSON and( R. A. KiNi,. 'The effect oft sulphate-reducing bacteria on the

corrosiotn oif reinforced ciointrete . lroc. Iii. Con0/. 'Biologwallv Indiai'd Coirro%an' NACF( . p 241

(1986).18, W. P1. Ivi.Rsos, J. el/(ctroc/it'tt Sot 115. 617 (1968).49 W. P. IV[ isliN. G. 3. ()t so-, and L. F. Ill RNut~, The role oft phosphorus and hydrogen sullhde in the

anaerobic cotrriisiiin oif iron and the possible detection of this corrosion by an eleetrochemical noisetechnique. Prom, lnt. ( on. Jbolo~i',eallv Induced C orrosioii, NAl -. p 154 (1986).

212 F. MANSFEIlD and B. LITt.FE

50. F. MANSFEID, Corrosion 44, 856 (1988).51. J. A. COSTELLO, Int. Biodeterior. Bull. 5, 101 (1969).52. A. K. TIIl ER, A review of the European research effort on microbial corrosion between 19510 and

1984. Proc. Int. Conf. 'Biologicall' Induced Corrosion', NA CE-8. p. 8 (1986).53. D. J. DQuEurrE and R. E. RICKER, 3lectrochemical aspects of microhiologically induced corrosion.

Proc. /it. Cont 'Biologically Induced Corrosion'. NACE-8. p. 121 (1986).54. R. C. SALVAREZZA and H. A. VIDEIA, Corrosion 36, 551) (1981).55. C. 0. OBUEKWE. D. W. S WESYLAKE. J. A. PIAMBECK and F. D. (OOK, Corrosion 37. 4u! (i"Sl).56. C. 0. 0Bt'EKWE. D. W. S. WESFLAKE, J. A. PLAMBECK and F. D. CooK, Corrosion 37, 632 (19I1).

57, M. B, DrswtlHKH, 1. AKITAR and C. P. Dir, Influence of sulphide pollutants of bacterial origin oncorrosion behaviour of naval brass. 2nd Int. Symnp. on Industrial and Oriented Basic Eluitrochem.,Sacst, India. p. 6.19.1.

58. N1. F. L. DE MELE, Biodeterioration of implanted materials. Proc. ArgentineIUSA Workshop onBiodeterioration (CONICET-NSF), LaPlata. Argentina, Aquatec, Quimica. pp. 119-133 (1986).

59. R. C. S..Xl.VAREZZA, M. F. L. DE MEtE and H. A. VIDELA. Int. Biodeterior. Bull. 15. 125 (1979).61). N1. F. L. DE MELE, R. C. SALVAREZZA, and H. A. VIDELA, It. Biodeterior. Bull. 15, 39 (1979).01. E. R. DE SCHI.,PPAREI.tI and B. M. ROSALES, Proc. 5th /it. Cong. o'i Marine Corrosion and Fouling.

Barcelona, Spain. p. 1 (19801).62. 1. N1. RosAtEs and E. R. DL S(HIAPPARELLI. Materials Performance 19, 4 (1980)).03. B, M. ROSALES. Corrosion measurements for determining the quality of maintenance in jet fuel

storage. Proc. Argentinc/USA Workshop on Biodeterioration (CONICET-NSF). LaPlata. Argen-tina. Aquatec Quimica, p. 135 (1986).

(4. E. S, At I YON and B. M. ROSALES, Corrosion 44, 638 (1988).65. H. A. VIDs, The action of Clado.sporiuo resinae growth on the electrochemical behavior of

aluminum. Proc. bit. Conf. 'Biologically Induced Corrosion', NACE-8. p. 102 (1986).w. C. RINs and F. P. A. RoBINsON. Corrosion 44. 386 (1988).

67. C. RINi.sS and F. P. A. ROBINSON. Corrosion 44, 671 (1988).

-11.3 0

0 UL

S. . I

£