influence of alternating, direct and superimposed alternating and direct current on the corrosion of...

Desalination 216 (2007) 103–115

Influence of alternating, direct and superimposed alternatingand direct current on the corrosion of mild steel

in marine environments

Srinivasan Muralidharana*,b, Dae-Kyeong Kima, Tae-Hyun Haa, Jeong-Hyo Baea,Yoon-Cheol Haa, Hyun-Goo Leea, J.D. Scantleburyc

aUnderground Systems Group, Korea Electrotechnology Research Institute,28-1, Seongju-dong, Changwon, 641-120, Republic of Korea

Tel. +82 (55) 280-1362; Fax: +82 (55) 280-1476; email: [email protected] Protection Division, Central Electrochemical Research Institute, Karaikudi 630 006, Tamilnadu, India

Corrosion and Protection Centre, University of Manchester, M60 1QD, UK

Received 2 September 2005; Accepted 29 November 2006

Abstract

The corrosion of mild steel in marine environments was studied with respect to induced alternating current (AC),direct current (DC) and the superimposed AC+DC sources. Corrosion rate determination at the different applied AC,DC and AC+DC current densities was carried out by conventional weight loss method for the exposure period of24 h. The quantitative estimation of leaching of iron into solution was done by using inductively coupled plasmaspectrometry (ICP). Mild steel specimens were subjected to surface examinations after treatment with various AC,DC and AC+DC current densities. Surface analysis was done by optical microscopy. The salient features of thisinvestigation are AC source accelerate the corrosion of mild steel in the presence of marine environments, but theextent of corrosion was increased when it was superimposed with DC. Especially, upto the corrosion current density,the AC field is not a great threat for mild steel and uniform corrosion was observed, But after the corrosion currentdensity, AC accelerate the pitting corrosion of mild steel in marine media. Surface micrographs showed severe pittingat the higher AC sources. The concentration of iron was increased above the corrosion current density. Theelectrochemical measurements coupled with surface examination and solution analysis proved to be a very effectivetool by means of characterizing the AC corrosion effect of mild steel in marine environments.

Keywords: AC corrosion; Mild steel; Marine environments

1. Introduction

It is well known that unidirectional currentsare involved during the process of metallic

*Corresponding author.

corrosion. However, under certain conditions,alternating currents can also cause corrosionalthough at a lower rate than direct currents. Areview of literature reveals that a number ofinvestigations have been carried out on the AC

c

doi:10.1016/j.desal.2006.11.0210011-9164/07/$– See front matter © 2007 Published by Elsevier B.V.

S. Muralidharan et al. / Desalination 216 (2007) 103–115104

corrosion of iron, lead, copper, aluminium andother metals [1–3]. Beside, the corrosion ofmetals the dissolution of metals by AC has alsoattracted the attention [4]. The accelerated corro-sion of metallic structures due to the presence ofstray AC currents has been reported by Kulman[5] and Chin [6].

In rapidly growing demand for energy requiresthe construction of an increasing number of highvoltage, high power transmission lines and layingof high pressure pipelines of large diameterssituated nearby marine environments. AC corro-sion mechanism is not yet understood and reliabletechniques for the determination of the corrosionrisk are not yet available. It becomes moredifficult to install metallic pipelines with ade-quate distance from overhead high-voltage ACpower transmission lines and/or AC powered railtransit systems and also away from marine areas.Therefore the risk of AC corrosion is increasingon the pipelines. Actually, several cases of ACcorrosion on buried steel pipelines due to inducedAC arising from high voltage AC power trans-mission lines have been reported [7]. The ACcorrosion could occur even when the instant-offpotential satisfied the conventional “potentialbased” cathodic protection (CP) criteria such asthe !0.85 V vs. Cu/CuSO4 polarized potentialcriterion. A number of laboratory and fieldstudies have been performed to investigate theinfluence of AC induced voltage on the corrosionbehaviour of corroding systems. The influence ofAC in acidic and sulfate solutions has beenreported [8–14]. However, more research isessentially needed for the influence of AC onchloride environments, since the marine environ-ments have a high concentration of chloride ions,it is expected that AC may accelerate thecorrosion.

This investigation is a systematic and detailedstudy on the influence of AC induced corrosionof mild steel in natural seawater. Since the sea-water used here contains 18,500 ppm of chloride,studies were also carried out in NaCl solution

containing the same amount of chloride. In orderto realize the individual influence of AC alone,experiments were also conducted on DC andsuperimposed AC+DC simultaneously in both themedia.

2. Experimental

2.1. Materials

Mild steel obtained from Dongil IndustriesCo., Ltd., South Korea was used and the com-position in wt% is follows: C (0.43), Si (0.22),Mn (0.72), P (0.013), S (0.015), Ni (0.05), Cr(0.10), Cu (0.12) and Fe balance. The mild steelspecimen of size 2 cm (l) × 2 cm (b) × 0.5 cm(thickness) was used. The total geometrical areaof the specimen was 11 cm2 but the actualexposed area was 5.5 cm2 and the other areaswere used for taking electrical connection andproperly insulated from the corrosive environ-ments. They were then mechanically polished andthen successively with finer SiC papers (240,400, 600, 800, 1200 grit). Before and after eachimmersion experiment, individual specimenswere cleaned with solutions containing concen-trated HCl, SnCl2 and Sb2O3 followed by a rinsewith de-ionized water and acetone. Finally, all thespecimens were dried under stream of hot air.Natural seawater (Jinhae sea, South Korea) andAR grade NaCl (DC Chemicals Co Ltd.,) wasused as aggressive solutions. Solutions werepurged with nitrogen gas to remove the dissolvedoxygen.

2.2. Methods

2.2.1. DC polarisation techniqueA conventional three electrode cell assembly

is used here. The working electrode was mildsteel specimen. Platinum foil and a saturatedcalomel electrode served as the auxiliary andreference electrodes respectively. The area of thecounter electrode is much larger than the area of

S. Muralidharan et al. / Desalination 216 (2007) 103–115 105

the working electrode. This will exert a uniformpotential field on the working electrode. Potentio-dynamic polarization studies were carried outboth in 18,500 ppm NaCl solution and naturalseawater environments. A constant quantity of thetest solution was taken in the polarization cell.The working, counter and reference electrodeswere assembled and connections were made. Thetest solution were continuously stirred using amagnetic stirrer to avoid the concentration polari-zation. A time interval of about 10 to 15 min wasgiven for each system to attain a steady state.Both cathodic and anodic polarization curveswere recorded potentiodynamically using aSolartron-1480 multi-channel electrochemicalinterphase controlled by computer with Corrwaresoftware. Procedures and parameters specified inASTM standard G5-87 were followed and all theexperiments were carried out at a room tempera-ture of 25±1EC. From the Tafel curves, actual icorrvalues for mild steel in NaCl solution and naturalseawater were derived.

2.2.2. Corrosion rate determination by weightloss measurements

Immersion experiments were carried out toquantify the corrosion rate of mild steel in NaClsolution and natural seawater medium under theinfluence of AC, DC and superimposed AC+DCsources. The weight of the mild steel samples wastaken before and after immersion using OhausExplorer 4-digit electronic balance for gravi-metric weight loss measurements. The AC signalwas applied between the working electrode (mildsteel) and the counter electrode (platinum gauge).A Solartron 1480 Multistat electrochemical mea-surement unit coupled with a Solartron 1255-Bfrequency response analyzer and multimediacomputer was used to supply the required ACwaveform. An oscilloscope was used to adjust aswell as to monitor the AC signal applied. Immer-sion studies were carried out at the corrosioncurrent density in both the media for the duration

of 24 h. Two current density values were chosenon lower and higher sides from the actual corro-sion current density value and the experimentswere conducted under the influence of AC, DCand the superimposed AC + DC sources. Blankexperiment without applying any current densitywas carried out in both the media under naturalimmersion corrosion condition. Experiments werecarried out both in natural seawater and in18,500 ppm NaCl at an ambient temperature of25±1EC. The values obtained are reasonable andreproducible from the triplicate set of experi-ments for each AC source and the averagecorrosion rate was made in mpy. Corrosion rate inmpy was made by using difference in weightlosses between initial and final weight of thespecimens for the exposure period of 24 h.

2.2.3. Solution analysisThe solution analysis was carried out with an

idea to quantify the leaching characteristics ofmild steel. It is expected that during the immer-sion studies the metal dissolution expectedreleasing considerable amount of metal ions fromthe material. In this study, mild steel specimenswere immersed in test solutions at various sourcecurrent densities. The concentration of metal ionspresent in the test solution was determined aftertreatment with AC, DC and the superimposedAC+DC sources at a fixed duration of 24 h. Ablank experiment was also performed in both themedia. The solution analysis was carried outusing an inductively coupled plasma-atomicemission spectrometry (ICP, Applied ResearchLaboratory, USA). Triplicate experiments werecarried out for each system and the averagevalues are noted.

2.2.4. Surface examinationAn optical microscopy was used to examine

the nature of corrosion on the mild steel surfacebefore and after immersion in various test solu-tions. The mild steel specimens were immersed in


various solutions for a fixed duration of 24 hunder the influence of AC, DC and superimposedAC+DC sources. At the end of 24 h the speci-mens were taken out from the solution, thor-oughly washed with distilled water, cleaned withacetone and dried. Micrographs were recorded atthe different applied current densities in NaClsolution and natural seawater environments. Therepresentative micrographs for each currentsource at the icorr were given. The microscopicexamination of the specimen was carried outusing an Olympus-GX 71 Inverted System metal-lurgical microscope coupled with a computer.Micrographs were recorded uniformly at themagnification of 40×.

3. Results and discussion

3.1. DC polarization measurements

Fig. 1 shows the typical polarization curvesfor mild steel in NaCl solution and naturalseawater medium. It shows that mild steel has noactive/passive transition when immersed in achloride solution and also a linear behaviour wasobtained in the Tafel region. The analysis of thesecurves in Corrware software gives the corrosion

Fig. 1. Potentiodynamic polarization curves for mild steelin NaCl solution and natural seawater.

current density (icorr). The icorr values for mild steelin NaCl solution and natural seawater environ-ment were found to be 1.22 mA/cm2 and0.94 mA/cm2 respectively. With respect to icorr,two current density values were chosen on thelower and higher sides. The blank experimentwithout applying any current density (i.e. naturalimmersion corrosion condition) is denoted as i =0. Table 1 gives the applied source currentdensity values in mA/cm2 for NaCl solution and

Fig. 2. Sinusoidal waveform of AC source applied at60 Hz in NaCl solution.

Fig. 3. Sinusoidal waveform of AC source applied at60 Hz in natural seawater.


Table 1Applied source current density values for mild steel in 18,500 ppm NaCl solution and natural seawater

S. no. Systemstudied

Details of applied current density Source current density values (mA/cm2)

NaCl Seawater

1 Blank No applied current density 0 02 i n icorr Applied current density far less than from the

corrosion current density0.0018 0.0018

3 i < icorr Applied current density less than from thecorrosion current density

0.018 0.018

4 i = icorr Applied current density equal to the corrosioncurrent density

1.220 0.940

5 i > icorr Applied current density greater than fromcorrosion current density

1.270 1.270

6 i o icorr Applied current density far greater than fromthe corrosion current density

1.810 1.810

natural seawater environments. In all the cases,the frequency of the AC modulation was held at60 Hz. Figs. 2 and 3 show the sinusoidal wave-form of AC source applied at 60 Hz across thecell in 18,500 ppm NaCl solution and in naturalseawater respectively.

3.2. Influence of AC source on corrosion of mildsteel in marine environments

3.2.1. Determination of corrosion rateThe influence of AC source on the corrosion

of mild steel for the exposure period of 24 h inboth the media is given in Table 2. It was foundthat the corrosion rate is increased as the ACsource current density level is increased. It isinteresting to note that, upto the four currentdensity levels [i = 0, i n icorr, i < icorr and i = icorr),a steady increase in the corrosion rate wasnoticed. After that, i.e. above i = icorr, a suddensteep increase in the corrosion rate is observed inNaCl solution. For example, the corrosion ratevalues up to the AC source level i = icorr is foundto be 30 mpy only, but when the AC source levelis increased beyond i = icorr (i.e., i > icorr and i oicorr), the corrosion rate was found to be 92 mpy.

Here, it is noticed that there was a 3-fold increasein the corrosion rate values. It indicates thatabove the corrosion current density the ACinduced corrosion is very a great threat for themild steel in chloride environments. In seawater,there is a regular increase in the corrosion ratevalues as the current density is increased from 0to i o icorr. The corrosion rate values are found tobe in the ranges from 10 mpy to 80 mpy, which isless when compared to the corrosion rate in NaClsolution. This may be due to the presence of otheradditional constituents in natural seawater whichrendering the attacking anions inactive. Onlychloride is taken care of when preparing18,500 ppm NaCl solutions. But in natural sea-water other additional constituents like Ca, Mg,K, S, Br etc., and few heavy elements inhibit thecorrosion of mild steel and thereby showed lesscorrosion rate. The higher magnitude of corrosionrate for mild steel in both environments above thecorrosion current densities may be due to thedestabilization of the passive layer on the mildsteel surface. The growth of the passive layeroccurs as a result of electrochemical reaction andthis passive layer is tends to depassivate at thehigher current densities and accelerate thecorrosion of mild steel [15].


Table 2Corrosion rate (mpy) of mild steel in both the media under various electrical field from the immersion studies

Source Current density

AC DC Super imposed AC + DC

NaCl Seawater NaCl Seawater NaCl Seawater

i = 0 16±0.001 10±0.002 22±0.003 12±0.002 33±0.002 16±0.002i n icorr 17±0.003 11±0.002 33±0.007 28±0.003 50±0.003 32±0.007i < icorr 19±0.005 20±0.004 192±0.009 122±0.005 100±0.005 80±0.006i = icorr 30±0.005 28±0.006 362±0.001 320±0.007 566±0.006 482±0.007i > icorr 50±0.006 42±0.006 384±0.001 358±0.008 600±0.006 578±0.008i o icorr 92±0.001 80±0.006 486±0.002 443±0.008 833±0.008 803±0.007

3.2.2. Surface examinationFigs. 4a and b show micrographs for a mild

steel surface in natural seawater and NaCl solu-tion respectively at icorr for the exposure period of24 h under the influence of an AC source.Micrographs for mild steel in NaCl solution up toicorr show the uniform corrosion on the mild steelsurface. On the other hand, small numerous pitswere observed at the higher AC source currentdensities. The changes in the surface morphologyof mild steel by the AC source is indicating thatAC sources have a strong influence on the corro-sion kinetic parameter, i.e., corrosion currentdensity. Recently Goidanich et al. reported thatthe AC affects the kinetic parameters such asTafel slopes and exchange current densities forcarbon steel, copper and zinc in 1 M sulphatesolutions [16]. The initiation of pitting at the icorrand above icorr is due to the metal solution inter-face that could be altered by the AC signal andthereby affect the corrosion kinetics. Mostprobably both reversibility and faradaic rectifi-cation aspects have to be considered to under-stand the AC-induced corrosion [16]. Here theirreversibility of the chemical reactions occurringat the interface causes a change in the doublelayer compositions and a modification of themetal surface. But a definite trend was observedbetween AC corrosion immersion studies andsurface morphology examinations.

3.2.3. Solution analysis by the ICP techniqueThe solution analysis by ICP technique was

carried out to measure the exact leaching of Fefrom mild steel in both the environments. Theinfluence of AC source on the leaching of iron atvarious applied current densities in both themedia is given in Fig. 5a. It was observed that theamount of leaching of Fe in NaCl solution isgreater than natural seawater at all the currentdensities. The concentration of Fe at the icorr wasfound to be 21 ppm for natural seawater and23 ppm for the NaCl solution. In both the media,the amount of leaching was slightly increasedwhen the current density is increased. The trendof reduction in the amount of leaching of Fe byvarious current densities follows the order:

(i = 0) > (i n icorr) > (i < icorr) > (i = icorr) > (i > icorr)

> (i o icorr)

The same order was already noticed in theimmersion studies too.

3.3. Influence of DC source on corrosion of mildsteel in marine environments

3.3.1. Determination of corrosion rateThe influence of DC on the corrosion rate of

mild steel in NaCl solution and natural seawater


(a) (b)

Figs. 4a and b. Micrographs ofmild steel for AC source at i =icorr. 40×. (a) Natural seawater;(b) NaCl solution.

(c) (d)

Figs. 4c and d. Micrographs ofmild steel for DC source at i =icorr. 40×. (a) Natural seawater;(b) NaCl solution.

(e) (f)

Figs. 4e and f. Micrographs ofmild steel for AC+DC sources ati = icorr. 40× (a) Natural seawater;(b) NaCl solution.

medium is given in Table 2. The measuredcorrosion rates are directly proportional to the DCapplied current densities in both the media. Itwas found that large variation in the corrosionrates between lower and higher current densities.Between the two solutions, NaCl showed morecorrosion rate than natural seawater. For examplethe maximum corrosion rate for mild steel innatural seawater is 443 mpy but for NaCl solutionis 486 mpy. Interestingly, the extent of corrosionby DC showed more than that of AC in both thesolutions. Under the same experimental condi-tions and solutions, DC system showed a 12-fold

increase in the corrosion rate rather than ACsystem at icorr. DC sources accelerate the cor-rosion in chloride solutions [17,18] in two waysas follows:

Fe + H2O ø Fe.H2Oads (1)

Fe.H2Oads + Cl! ø FeCl!ads + H2O (2)

FeCl!ads + OH! ø FeOH+ + Cl! + 2e (3)

FeOH+ + H+ ÷ Fe2+ + 2e (4)or


Fe + Cl! ø FeCl!ads (5)

FeCl!ads + H2O ø FeOH!ads + H+ + Cl! (6)

FeOH!ads ø FeOHads + e (7)

FeOHads ø FeOH+ads + e (8)

FeOH+ads ø Fe2+ + OH! (9)

In either way, mild steel subjected to severecorrosion in the chloride environments in thepresence of DC sources. The corrosion rate wasfound lower in seawater than NaCl solution. Saltssuch as carbonates, hydroxides, sulphates etc., ofCa2+ and Mg2+ are rather insoluble and wouldnear to the metal surface and the presence ofthese ions therefore decrease the corrosion rate ofmild steel in seawater.

3.3.2. Surface examinationFigs. 4c and d show micrographs for a mild

steel surface in natural seawater and NaClsolution respectively at icorr for the exposureperiod of 24 h under the influence of a DCsource. When the DC current density applied atthe electrode surface corresponds to icorr, a pittingcorrosion was started in both media. After that,the specimens underwent severe pitting andseveral pits grow with time. Later pits tend tomerge and overlap each other showed larger pitson the mild steel surface. As evidenced fromimmersion studies, higher corrosion rates arenoticed at the higher current densities, it is clearlyreflected in the surface morphology also. So inDC corrosion studies also a definite trend wasobserved between immersion and surface mor-phology examination.

3.3.3. Solution analysis by the ICP techniqueThe amount of leaching of iron into the solu-

tion during the DC corrosion studies was carriedout by ICP technique both in NaCl solution and

natural seawater environments and is given inFig. 5b. It was found that the leaching of Fe inNaCl solution is found to be more than naturalseawater at all the current densities. For example,the concentration of Fe at icorr was found to be480 ppm for natural seawater and 579 ppm forNaCl solution. The amount of leaching of Fe wasalmost doubled at the higher current densities.This is to be observed in both the media. Theextent of leaching of Fe due to DC source wasfound more when compared to the AC source.

3.4. Influence of superimposed AC+DC source oncorrosion of mild steel in marine environments

3.4.1. Determination of corrosion rateThe influence of superimposed AC + DC

sources on the corrosion of mild steel for theexposure period of 24 h in both the media isgiven in Table 2. It is observed from Table 2 thatthe corrosion rate of mild steel in both the mediawere maximum at i o icorr and lower at i n icorr.The influence of superimposed AC+DC sourceon the corrosion of mild steel was found to bevery severe when compared to the individualinfluence of either AC or DC alone. Largervariation was observed in the corrosion ratevalues between superimposed AC upon DC andthe individual AC or DC. For example at icorr, inNaCl solution, the corrosion rate values are foundto be 566 mpy, 362 mpy and 30 mpy respectivelyfor (AC+DC), DC and AC. Here it is concludedthat the extent of corrosion of mild steel due tothe superimposed AC+DC in NaCl solution is 1.5times than DC and 18 times than AC. Tan andChin [19] reported that superimposed AC + DCincreased the corrosion of mild steel in aqueousNa2SO4 solution. In this study, the superimposedAC favoured the active dissolution potentialregion than the passive region. It appears that thepassivity was destroyed in two ways. Firstly, thedestruction of passivity of mild steel by super-imposed AC + DC sources. Secondly the passive


(a) (b)

(c)

Fig. 5. Concentration of Fe vs. source current density.(a) AC; (b) DC; (c) AC+DC. " NaCl. O Seawater.

region was disappears completely due to thehigher concentration of chloride ions in the solu-tions. The breakdown of passivity by chlorideions has been attributed to the ability of chlorideions to penetrate and to dissolve the passive oxidefilms on mild steel surface. This results in com-plex current fluctuations at the mild steel surfaceand thereby accelerating the corrosion vigorously.The trend in reduction of corrosion rate betweenthe various types of sources in natural seawaterenvironment follows the order: AC > DC > (AC+ DC).

3.4.2. Surface examinationFigs. 4e and f show micrographs for mild steel

surface in natural seawater and NaCl solution

respectively at icorr for the exposure period of 24 hunder the influence of superimposed AC + DCsources. The optical micrographs strongly suggestthe significant evidence of severe pitting corro-sion of mild steel surface in both the media. Here,micrographs strongly suggest that the super-imposition of AC on DC influence the morphol-ogy of the specimens undergoing dissolution ofmetals. Superimposed AC + DC sources wereapplied especially at icorr and above, an irregularshaped pits of various diameters and very deepwere developed (Figs. 4e and f). The presence ofchloride ions in solutions removed the passivityon localized areas on mild steel surfaces resultingpitting starts at higher current densities. In chlo-ride solutions, the attack on mild steel behaved


like “etching” rather than “pitting”. As explainedearlier in weight loss measurements, it is a factthat chloride ions had the ability to penetrate anddisintegrate the oxide film on the metal surfaceand again superimposed AC+DC sources isapplied, the protective film could be easilydestroyed by perturbing the potential and virtu-ally eliminated any oxide film and developspitting corrosion at the mild steel surface. A closeagreement was noticed between weight lossmeasurements and surface morphological studiesin NaCl solutions and natural seawaterenvironments.

3.4.3. Solution analysis by the ICP techniqueThe influence of superimposed AC + DC

sources on the leaching of iron at various currentdensities in NaCl solution and natural seawaterenvironments is given in Fig. 5c. As alreadyobserved in other cases, here also, the leaching ofFe into NaCl solution is more than natural sea-water. As already evidenced from immersionstudies and optical microscopic studies, in thistechnique also more leaching of Fe was noticed athigher source current density values and therebyindicating more corrosion rate values andaccelerating pitting on the steel surface. The con-centration of Fe at the corrosion current densitylevel was found to be 639 ppm for natural sea-water and 919 ppm for NaCl solution. Thereduction of the amount of leaching by thedifferent types of sources follows the order: AC >DC > (AC + DC). It is a fact that the amount ofleaching of Fe is directly proportional to the cor-rosion rate, the superimposed AC+DC showedmore corrosion rate than either AC or DC sourcesalone.

4. Proposed mechanism of action of ACcorrosion at the various current densities

Below the corrosion current density (i < icorr )— AC sources uniformly corrode the mild steel

in both the media. The corrosion rate was alsofound to be less at i < icorr. Micrographs supportthe uniform corrosion at this current density inboth the media. The schematic of the uniformcorrosion of mild steel at the i < icorr is shown inFig. 6(a). Uniform corrosion will tend to occurwhen some mild steel surface regions becomeanodic for a short period but at their location andthat of the cathodic regions constantly change.General rusting of mild steel will take place whenthere is a uniform supply of oxygen availableacross the surface of the steel and there is auniform distribution of defects in the oxide film.Therefore, overall attack takes place at a numberof anodic sites whose positions may changeleading to general rusting across the surface.

At the corrosion current density (icorr) — Wheniappl was held at icorr, the dissolution rate and sur-face morphology of the mild steel in NaCl andseawater were significantly changed. Surfacemicrographs strongly suggest the possibility ofpitting corrosion at icorr. This is illustrated sche-matically in Fig.6(b). Here mild steel has under-gone pitting in NaCl solution. Rapid dissolutionoccurs within the pit while oxygen reductiontakes place on adjacent surfaces. This process isself-stimulating and self-propagating. The rapiddissolution of metal within the pit tends to pro-duce an excess of positive charge in this arearesulting in the migration of chloride ions tomaintain electrical neutrality. Thus, in the pitthere is a high concentration of FeCl2 and as aresult of hydrolysis, a high concentration ofhydrogen ions. Both hydrogen and chloride ionsstimulate the dissolution of most metals andalloys, and the entire process accelerates withtime. Since the solubility of oxygen is virtuallyzero in concentrated solutions, no oxygen reduc-tion occurs within a pit. The cathodic oxygenreduction on the surfaces adjacent to pits tends tosuppress corrosion. In that case pits cathodicallyprotect the rest of the metal surface.

Although Fig. 6(c) indicates how a pit growsthrough self-stimulation, it does not immediately


(a) (b)

(c)

Fig. 6. Mechanism of action of AC corrosion at three stages. (a) i < icorr; (b) i = icorr; (c) i > icorr.

suggest how this process is initiated. Micrographs(Figs.4e and f) showed larger pits when immersedin NaCl solution in the presence of AC sources.At i o icorr the rate of metal dissolution is momen-tarily high at one particular point, chloride ionswill migrate to this point. Since chloridestimulates metal dissolution, this change tends to

produce conditions that are favourable for furtherrapid dissolution at this point. Here pitting corro-sion is predominantly taking place along withuniform corrosion.

Above the corrosion current density (i > icorr)— Higher corrosion rates were observed wheniappl was held at beyond icorr due to the fact that


pits once initiated on the mild steel surface at i >icorr the propagation is taking place as a resultreduction of pH in the bulk solution.

In the case of pitting of iron in a slightly alka-line chloride solution a copious anodic productionof positively charged Fe2+ attracts negative anionssuch as Cl! followed by hydrolysis as below:

Fe2+ + 2H2O + 2Cl! xxxxxv Fe(OH)2 + 2 HCl(10)

This will cause local pH reduction and, as aresult, a self-propagating or autocatalytic natureof growth of pits. The acid chloride solutionfurther accelerates anodic dissolution, which inturn further concentrates chloride in the pit. Aninsoluble Fe(OH)3 corrosion products collects atthe pit mouth when Fe2+ diffuses out of the pitinterior to the exterior where it is oxidized to Fe3+

and precipitates in the neutral bulk solution. Thecap impedes easy escape of Fe2+ but is suffi-ciently porous to permit migration of Cl! into thepit, thereby sustaining a high acid chloride con-centration in the pit. Anodic polarization of thepit interior occurs by coupling to the exteriorpassive cathode surfaces. Cathodic reduction of adissolved oxidizer such as oxygen consumes theelectrons liberated by the anodic pit reaction.

The importance of the cathodic reaction tosustain pitting should be understood. Pit growthcannot continue without a cathodic reductionreaction to consume the electrons liberated by(i.e., to polarize anodically) the pit anodic reac-tion. Thus, pits are widely spaced in aerated saltsolutions because oxygen has limited solubility,and a large surrounding area is needed to provideenough reduction capability to support the centralpit anode. Any pit initiating within the cathodicarea of a larger pit is suppressed by cathodicprotection. The greater solubility of Fe3+ yieldsgreater reduction rates on a smaller surface area,and pits can survive much closer to one another.In advanced stages, pits may become deepenough that reduction of H+ on the pit walls near

the outer surface is possible, while the pit bottomstill sustains anodic dissolution. Thus, hydrogenbubbles emanating from pits have been observed.

5. Conclusions

The following conclusions were drawn fromthe present investigation:

1. The extent of corrosion of mild steel inNaCl and in natural seawater environments wasfound to be less in AC when compared to the DCsources.

2. When alternating current (AC) signals aresuperimposed on a direct current (DC) at theirrespective corrosion current density (i = icorr) orabove the corrosion current density (i > icorr), ACsignal causes localized pits and simultaneouslyincrease the corrosion rate. Similarly severe pit-ting on the surface was observed, especiallysuperimposition of AC with DC at higher currentdensity region (i o icorr).

3. AC accelerates the corrosion of mild steelafter the corrosion current density values whereasDC and (AC + DC) sources initiate the corrosioneven below the corrosion current density values.

4. The electrochemical measurements coupledwith surface examination and solution analysisproved to be a very effective tool to characterizethe AC corrosion effect of mild steel in NaCl andnatural seawater environments and a definitecorrelation was observed among these techniques.

Acknowledgements

One of the authors (S.M.) thanks CSIR andCECRI, India, for permission to pursue a PDF atKERI, South Korea. S.M. also thanks KOFST,Korea, for financial assistance through the BrainPool Program.

References

[1] S.Z. Fernandes, S.G. Mehendale and S. Venkata-chalam, J. Appl. Electrochem., 10 (1980) 649.


[2] H.C. Koerts and J.M. Wetzer, Literature survey oncorrosion of underground metallic structures inducedby alternating currents, Arnhem, European CopperInstitute, 2001.

[3] R.A. Gummow, R.G. Wakelin and S.M. Segall, ACcorrosion — a new threat to pipeline integrity? Proc.1st International Pipeline Conference (IPC), Calgary,Canada, 1996.

[4] H.S. Song, Y.G. Kim, S.M. Lee and Y.T. Kho,Competition of AC and DC current in AC corrosionunder cathodic protection, Corrosion 2002, PaperNo. 02117.

[5] F.E. Kulman, Corrosion, 17 (1961) 34.[6] D.T. Chin and S. Venkatesh, J. Electrochem. Soc.,

126 (1979) 1908.[7] Y. Hosokawa, F. Kajiyama and T. Fukuoka,

Corrosion, 60 (2004) 408.[8] S.S. Abd El Rehim, Surface Technol., 25 (1985) 223.[9] J.L. Wendt and D.T. Chin, Corr. Sci., 25 (1985) 901.

[10] J.L. Wendt and D.T. Chin, Corr. Sci., 25 (1985) 889.

[11] D.S. Dunn, M.B. Bogart, C.S. Brossia and G.A.Cragnolino, Corrosion, 56 (2000) 470.

[12] Y. Hosokawa, F. Kajiyama and T. Fukuoka, Corro-sion, 60 (2004) 408.

[13] F. Kajiyama and Y. Nakamura, Corrosion, 55 (1999)200.

[14] F. Bolzoni, S. Goidanich, L. Lazzari and M.Ormellese, Laboratory test results of AC interferenceon polarized steel, Corrosion 2003, Paper No. 03704,Houston, TX, 2003.

[15] M.L. Mateo, T. Fernandez Otero and D.J. Schiffrin,J. Appl. Electrochem., 20 (1990) 26.

[16] S. Goidanich, L. Lazzari, M. Ormellese and M.Pedeferri, Influence of AC on corrosion kinetics forcarbon steel, zinc and copper, Corrosion 2005, PaperNo. 05189, Houston, TX, 2005.

[17] W.J. Lorenz, H. Yamakov and H. Fisher, Bun.Burnsenger. Phys. Chem., 67 (1963) 932.

[18] A.J. Arvia and J.J. Podesta, Corr. Sci., 8 (1968) 203.[19] T.C. Tan and D.T. Chin, J. Appl. Electrochem., 18

(1998) 831.

influence of alternating, direct and superimposed alternating and direct current on the corrosion of...

Documents