Electrochemical Corrosion Behavior of 300M Ultra High Strength Steel in Chloride Containing Environment

Available online at www.amse.org.cn

Acta Metall. Sin.(Engl. Lett.)Vol.23 No.4 pp301-311 August 2010

Electrochemical corrosion behavior of 300M ultra high

strength steel in chloride containing environment

Min SUN, Kui XIAO, Chaofang DONG and Xiaogang LI ∗Corrosion and Protection Center, School of Material and Engineering, University of Science and TechnologyBeijing, Beijing 100083, China

Manuscript received 26 February 2010; in revised form 23 April 2010

The electrochemical corrosion behavior of 300M ultra high strength steel in chloridecontaining environment was investigated by potentiodynamic polarization technique,electrochemical impedance spectroscopy (EIS) and scanning electron microscopy(SEM). The results show that uniform corrosion occurs on 300M steel during the elec-trochemical measurements because no anodic passivation phenomenon is observed onpolarization curves within the measurement range. The tests also show that 300Msteel is highly susceptible to chloride containing solution, which is characterized bycorrosion current density increasing with the addition of chlorides, and corrosionpotential shifting towards positive direction and corrosion resistance decreasing, pos-itively suggesting that chloride ions speed up the corrosion rate of 300M steel. Mean-while corrosion products on the 300M steel surface formed during the salt spray testare too loose and porous to effectively slow down the corrosion rate. Additionally, aschematic structure of uniform corrosion mechanism can explain that 300M steel hasbetter property of stress corrosion cracking (SCC) resistance than stainless steels.KEY WORDS 300M ultra high strength steel; Electrochemical corrosion;

Chloride

1 Introduction

300M steel, as a kind of ultra high strength steel, is always used as high rate ofloading[1,2], i.e., the airplane landing gear and pressure vessel because of possessing adesirable combination of high tensile strength and fatigue strength, high hardness, andtherefore it is always faced with the chloride containing environment during the servicelife, which can cause corrosion, especially SCC.

Corrosion is one of the most important failures which threaten the safe usage of 300Msteels, especially the SCC and failure cracking (FC). Graca et al.[1] have studied that therupture occurred suddenly at a pressure level lower than expected for the proof pressurewhen 300M steel pressure vessel failed during hydrotest, and hydrogen assisted SCC wasthe mechanism responsible for the failure. Up to date, most studies have been focusedon microstructure and the mechanical properties of 300M steel[3−5], while very few papers

∗Corresponding author. Professor, PhD; Tel.: +86 10 62333975-509; Fax: +86 10 62334005.

E-mail address: [email protected] (Xiaogang LI)

· 302 ·

studied about the corrosion behavior in bulk solutions by electrochemical technique. It iswell known that some aggressive ions, like chloride ions, can be adsorbed on the surfaceand increase corrosion rate[6,7].

The corrosion rate of steels is enhanced by chloride ions, and the effects of chloride ionson stainless steel and magnesium alloy have been studied and mature mechanisms havebeen summarized[8]. Chloride ions can be adsorbed on the passive surface and penetratethe oxide film through pores or defects easier than other ions, such as SO2−

4 . Through flowof current[9,10] chloride ions can transfer into the pits, forming concentrated solutions ofFe2+, Ni2+, and Cr3+ chlorides, which, by hydrolysis, account for an acid solution.

Dıaz et al.[11] have studied the effect of the chloride content on high strength steelwires in NaOH solutions and found that the typical feature corresponding to chlorideadsorption can be seen in the passivity domain, because the passivity current decreases asthe chloride increases. In the absence of chlorides, the steel surface remains passive at 70%of the ultimate tensile strength, and no pitting potential and the zero current potentialin the reverse curve defined by the reduction product γ-Fe2O3/Fe3O4. However, in thepresence of very small amounts of chlorides, pitting occurs under this aerated condition.The pitting potential decreases as the [Cl−]/[OH−] ratio increases. In fact, deep selectivecorrosion patterns have been measured by AFM profiling in 0.05 M NaCl solution[12].

Jegdic et al.[13] have studied the influence of chloride ions on austenitic 304 stainlesssteel in aqueous sulphuric acid solution and found that, addition of NaCl accelerates thehydrogen evolution reaction on the passive surface, while the same reaction on the baresurface is somewhat inhibited by NaCl. On the other hand, presence of NaCl acceleratesthe anodic reaction on the bare surface, and it activates the dissolution of the passive layer,so that the passivation currents increase with addition of NaCl.

However, few investigations on the influence of chlorides have been focused on ultrahigh strength steels. The present paper studied the corrosion behaviors of 300M ultrahigh strength martensite steels in chloride containing environment by polarization curvemeasurement and EIS. In addition, the influences of chloride concentration on the corrosionform, corrosion rate and morphology of 300M were also investigated.

2 Experimental

2.1 Material and test solutionsThe working electrodes for electrochemical measurement were cut from a bulk of 300M

ultra high strength steel. The chemical composition (wt pct) of the steel is C 0.40, Mn0.64, Si 1.66, S 0.0013, P 0.009, Ni 1.90, Cr 0.71, Mo 0.37, V 0.008 and Fe balanced.

Fig.1 is the metallographic microstructure of the 300M ultra high strength steel. Themicrostructure of the steel contains martensite, lower bainite and retained austenite, whichseems to be in agreement with the above reference literature findings[9]. The specimenswere embedded in epoxy resin leaving a working area of 1.0 cm2. The specimen preparationwas carefully controlled to ensure that there was no grooving and bubble at the epoxy/steelinterface. Prior to the tests, all the working surfaces were subsequently polished to 1200-grade emery paper, and rinsed with deionized water and methanol.

Solutions were prepared from deionized water with 0, 0.1, 0.3, 0.5 mol/L NaCl con-centrations, respectively, and used to simulate the atmospheric environment. All the tests

· 303 ·

were carried out at room temperature. Thevalue of pH for the testing solution can re-markably influence the corrosion rate andcorrosion form. In this paper, all of the NaClsolution is near-neutral, with a pH value of6.8.

2.2 Electrochemical measurementsThe conventional electrochemical mea-

surements, including corrosion potential, po-tentiodynamic polarization curves and EIS,were performed in a three-electrode cellthrough a PAR 2273 workstation. 300M steelwas used as the working electrode, a plati-

Fig.1 Metallographic microstructure of the300M ultra high strength steel.

num sheet as the counter electrode and a saturated calomel electrode (SCE) as the referenceelectrode.

Prior to the electrochemical measurements, the working electrode was cathodicallypolarized at −1.3 V vs. SCE for 3 min to remove the air-formed oxides on the surface ofelectrode. The corrosion potential was then recorded for 1 h. EIS technique was used tostudy the corrosion behavior of 300M steel in NaCl solutions, which was measured at theopen-circuit potential and frequency was ranged from 100 kHz to 0.01 Hz with an appliedac perturbation of 10 mV. The impedance data were analyzed by a commercial ZSimpWinsoftware package. After that, the polarization curve was measured at a potential scan rateof 0.5 mV/s. The potential was ranged from −0.25 mV (vs. the open potential EOC) to0.3 mV (vs. the reference potential ER).

2.3 Salt spray testCorrosion resistance of 300M ultra high strength steel was evaluated by salt spray test,

which was used to simulate the atmospheric environment. Salt spray test was followed byStandard ASTM B117-97. Working materials were tested under (35±2) C temperature,5 wt pct neutral sodium chloride solution and continuous spraying 4, 8, 12, 16, 20 days. Allthe electrode samples were tested by electrochemical techniques, including potentiodynamicpolarization curves and EIS, which were used to study the electrochemical behavior of 300Mwith corrosion products. Solutions were prepared from deionized water with 0.2 wt pctNaCl concentrations.

3 Results and Discussion

3.1 Effect of chloride on electrochemical corrosion for 300M steel3.1.1 Open circuit potential

Fig.2 shows the variations of open circuit potential with immersion time of 300M steelelectrodes in NaCl solutions with different concentrations at the room temperature. Ob-serve that open circuit potentials of the electrodes in deionized water and 0.1 mol/L NaClsolution shift positively first and then decrease to reach a relatively steady value of about−0.58 V vs. SCE and −0.61 V vs. SCE, meanwhile open circuit potentials of electrodesin 0.3 mol/L and 0.5 mol/L NaCl solutions slowly reach a steady-state value of −0.62 V

· 304 ·

vs. SCE and −0.64 V vs. SCE. With the increase of Cl− content, there is a slight changeof open circuit potential, less than 15 mV. The open circuit potentials shift toward neg-ative direction, implying that the working electrodes are electrochemically unstable andsusceptible to anodic dissolution with increase of Cl− content.

3.1.2 Polarization curveAfter measurements of open circuit potential on the 300M steel electrodes for 1 h,

the potentiodynamic polarization technique was used to investigate the electrochemicalbehavior of 300M steel in NaCl solutions with different concentrations. Fig.3 shows thepotentiodynamic polarization curves of 300M steel in NaCl solutions.

0 500 1000 1500 2000 2500

-0.8

-0.7

-0.6

-0.5

-0.4

Pot

entia

l / V

vs.

SC

E

Time / s

Deionized water 0.1 mol/L NaCl 0.3 mol/L NaCl 0.5 mol/L NaCl

-8 -7 -6 -5 -4 -3 -2 -1

-0.90

-0.75

-0.60

-0.45

-0.30

-0.15

0.00

0.15

P

oten

tial /

V v

s. S

CE

lg(i /A cm-2)

Deionized water 0.1 mol/L NaCl 0.3 mol/L NaCl 0.5 mol/L NaCl

Fig.2 Time dependence of open circuit poten-tials of the 300M steel in NaCl solu-tions.

Fig.3 Polarization curves of the 300M elec-trodes measured in NaCl solutions.

The polarization curve was scanned from cathodic to anodic potentials. Notice fromthe anodic polarization curves that, the current density increases rapidly with the poten-tial. At low current densities, the cathodic reaction is a predominantly charge transfercontrolled process. All electrodes are in active state and unable to be passivated in themeasured potential range. It is conformed that the anodic reaction is controlled by acti-vated dissolution process. Furthermore, with the increase of chloride concentration, thepolarization curves shift toward the right-down direction and the anodic current densityincrease.

On the cathodic branch of the polarization curve, the cathodic current density of thecorrosion system should be attributed to the reduction of oxygen dissolved in the solution,and the cathodic current density decreases with the chloride concentration increasing. Thepresence of cathodic limiting diffusion current means that all the cathodic reactions areoxygen diffusion-controlled processes. When the potential is below −0.85 V, the cathodiccurrent density shifts towards right obviously, because hydrogen reduction occurs besidethe oxygen reduction. The anodic dissolution of 300M ultra high strength steel can beexpressed as

Fe → Fe2+ + 2e (1)

This work shows that diffusion and reduction of oxygen dominate the cathodic process,as demonstrated by the presence of cathodic limiting diffusive current density. So thecathodic reaction can be expressed as

O2 + 2H2O + 4e → 4OH− (2)

· 305 ·

When the corrosion potential is below −0.85 V, the reduction of hydrogen occurs. Inneutral solutions hydrogen evolution occurs predominantly through the reduction of H2Orather than hydrogen ions:

2H2O + 2e → H2 + 2OH− (3)

The values of the electrochemical parameters corrosion potential Ecorr, corrosion currentdensity icorr, anodic tafel slope ba and cathodic tafel slope bc in NaCl solution are shownin Table 1. Notice that with the increase of chloride concentration, the corrosion potentialEcorr decreases from −0.448 V vs. SCE to −0.637 V vs. SCE and this result is remarkableat higher chloride concentration. Meanwhile the chloride ions can induce the increase of thecorrosion current density icorr and the decrease of anodic Tafel slope ba. In deionized watericorr is only 7.24 µA, which is remarkably lower than the values in NaCl solutions. 300Msteel has the lowest corrosion rate in deionized water, because there is no accelerating effectof chloride ions. It is thus reasonable to assume that chloride ions accelerate corrosion rateof 300M steel obviously.

Table 1 Values of parameters observed from polarization curves

Solutions Ecorr/V vs. SCE icorr/µA ba/V vs. SCE bc/V vs. SCE

Deionized water −0.448 7.24 0.2141 0.40780.1 mol/L NaCl −0.552 10.4 0.0607 0.74880.3 mol/L NaCl −0.587 11.5 0.0564 0.59400.5 mol/L NaCl −0.637 14.8 0.0480 0.8718

After measurement of polarizationcurves, the surface morphologies of elec-trodes were observed visually. It is foundthat there are trace amounts of dark corro-sion products on surface of the 300M elec-trodes. No evidence of localized attack isobserved after corrosion products removed.

3.1.3 EISThe EIS results are presented in the form

of Nyquist plots, in which the imaginary im-pedance Zim is plotted against the real im-pedance Zre. Fig.4a shows the Nyquist plotsobtained on electrodes in deionized waterand NaCl solutions with different concentra-tions at the open circuit potential. Noticethat there is a similar impedance feature forall the electrodes, i.e., one depressed semicir-cle over the whole frequency range, and theyare all characteristic of a capacitive behavior,indicating that there is one interfacial reac-tion process occurring over the measurementfrequency range.

0 500 1000 1500 2000 2500

0

250

500

750

1000

1250

1500

1750

2000

Zre / cm2

-Zim

/

cm2

Deionized water 0.1 M NaCl 0.3 M NaCl 0.5 M NaCl

(a)

Fig.4 Nyquist plots of impedance in NaClsolutions with different concentrationsand electrochemical equivalent circuits.

· 306 ·

In the monitoring technique based on the theory of electrochemical impedance, thesolution resistance Rs is estimated from impedance measured in the high frequency range,while the sum of polarization resistance and the solution resistance are estimated from theimpedance in the low frequency range. It is well known that the polarization resistance isinversely proportional to the corrosion rate in aqueous solutions. Thus this technique isvery useful for monitoring the atmospheric corrosion rates.

Notice from Fig.4a that the Nyquist impedance semicircles decrease with the increaseof chloride concentration. While for electrode in deionized water, the semicircle size ismuch larger than that in NaCl solutions, implying that the polarization resistance alsodecreases with addition of chloride ions, which in turn increases the corrosion rate. Thisagrees with the previous polarization work.

The electrochemical equivalent circuit for a simple corroding electrode is representedby a semi-circle, with the parallel connection of the charge-transfer resistance Rct and thedouble-layer capacitance Cdl. When potential E is the only status variable during electrodeprocess, the charge-transfer resistance Rct is equal to linear polarization resistance Rp.

The simple linear relation that defines the corrosion current density forms, icorr isestimated from Rp using the Stern-Geary relationship

icorr =B

2.303Rp(4)

where B=f(βa, βc), and Tafel anodic βa and cathodic βc slopes are taken as positive kineticparameters for determining icorr of a corroding or oxidizing metallic material. The valueof B is less than 1 V. Rp is the linear polarization resistance, which can be evaluated frompolarization curves. Eq.(4) predicts that the corrosion current density is very sensitiveto changes in the polarization resistance. This icorr expression is simple but essential incorrosion measurement, since icorr can be converted to corrosion rate in units of mm/y,which are more convenient for engineering purposes[17].

The electrochemical equivalent circuit for 300M electrodes in NaCl solutions with differ-ent concentrations are shown in Fig.4b with the parallel connection of Rct and Q. Electro-elements in Fig.4b have the following meanings[17]: Rs is the electrolyte resistance betweenthe working and reference electrodes; Rct is the polarization resistance and in inverselyproportional to icorr, and Rct is equal to Rp because there is only one time constant; Q isthe constant phase angle element (CPE) in place of the interfacial capacitance Cdl whenthe dispersion effect exists, which is caused by the roughness of working electrode surface.

The parameters observed from Fig.4 are shown in Table 2. Observe that in deionizedwater Rs is thousands of times of that in NaCl solutions, and Rct decreases with theincrease of chloride concentration, indicating that the corrosion rate increases according tothe Eq.(4). It agrees with previous potentiodynamic polarization studies.

Table 2 Values of parameters observed from Nyquist plots

QSolutions Rs/Ω·cm2

Y0/Ω−1·cm2·s−n nRct/Ω·cm2

Deionized water 192.3 0.000473 0.7012 22440.1 mol/L NaCl 27.48 0.001594 0.7684 12630.3 mol/L NaCl 5.476 0.002695 0.6611 12570.5 mol/L NaCl 2.592 0.001186 0.7742 1019

· 307 ·3.2 Electrochemical behavior of 300M steel

with corrosion products after salt spraytestAfter the salt spray test with different

days, the polarization curves of 300M steelwere measured in 0.2 wt pct NaCl solu-tions, as shown in Fig.5. Notice that withthe processing of salt spray test, polarizationcurves move to the right-down side. Addi-tionally, comparing Fig.3 with Fig.5, it canbe seen that the corrosion current in Fig.5 is2–3 orders of magnitudes than that in Fig.3,even the polarization curves are measured inlower chloride concentration solutions. Thecorrosion potentials shift towards negativeand the corrosion currents become larger,meaning that the corrosion rate of 300M steelis speeded up after salt spray test. But thedifferences between all of the polarizationcurves are small. It can be concluded thatcorrosion products have formed and adheredto the metal surface, acting as the functionof passive film on the stainless steel. But theprotection of the corrosion products againstchloride ions is not effective as passive film,so the corrosion rate is accelerated with theprocessing of the salt spray test.

Fig.6 shows the Nyquist plots of 300Msteel after the salt spray test with differenttime. For all the Nyquist plots, there is adepressed semicircle at the high frequencywhich is characterized by a capacitive be-havior, and another incomplete semicircle ispresent at the low frequency which is relatedwith the corrosion products on the metal sur-face.

-5 -4 -3 -2 -1 0 1 2

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

P

oten

tial /

V v

s. S

CE

4 days 8 days 12 days 16 days 20 days

lg(i /A cm-2)

Fig.5 Polarization curves of 300M ultra highstrength steel in 0.2 wt pct NaCl aftersalt spray test.

0 50 100 150 200 250 300 350

0

25

50

75

100

125

150

175

200

Zre / cm2

-Zim

/

cm2

4 days 8 days 12 days 16 days 20 days

(a)

Fig.6 Nyquist plots of impedance for 300Multra high strength steel in 2 wt pctNaCl after salt spray test and electro-chemical equivalent circuits.

The electrochemical equivalent circuit for the 300M electrodes after salt spray test in0.2 wt pct NaCl solutions is shown in Fig.6b, Qf is the CPE and Rf is resistance causedby the film of corrosion products. The other electro elements in Fig.6b have the samemeanings as described in Fig.4b.

With the processing of the salt spray test, the diameters of capacitive semicircles de-crease. Corrosion products formed on the metal surface during the salt spray test have twoeffects: one is that the corrosion products can slow down the exchange of reaction ions tosome extent, for example, chloride ions permeates into the substrate/corrosion productsinterface and the iron ions permeate out of the corrosion products into the solutions. The

· 308 ·

other one is that corrosion products are too loose to stop the permeation of chloride ionsinto the iron/corrosion products interface. Chloride ions continue to destroy the substrateintegrity and aggravate the corrosion rate.

3.3 Analysis of corrosion mechanism for 300M ultra high strength steelMorphologies of 300M steels after anodic polarization in NaCl solutions with different

concentrations are shown in Fig.7. It is obviously seen that corrosion morphologies becomemore serious with increasing concentrations of chloride ions. As shown in Fig.7a, littlecorrosion product has formed on the whole surface in deionized water. After polarizationin 0.1 mol/L NaCl solutions, a thin layer of corrosion products covers the whole surfaceand a number of fine cracks lie among them, shown in Fig.7b. Block corrosion productscan be seen in Figs.7c and 7d, and the sizes of them are approximately close to each other.From the even micrograph in Fig.7, it can be calculated that uniform corrosion occurs on300M steel in NaCl solutions. The SEM results show that the presence of chloride ionsincreases corrosion rate of 300M steel, and this phenomenon is in complete agreement withthe results obtained from electrochemical measurements.

In the environment containing appreciable concentrations of Cl−, corrosion productsare formed on 300M steels surface and block the diffusion of corrosion media containingCl−, which act as the effective barriers against diffusion of matters through the film andhence decreases the active dissolution current. When the oxide-film is present, corrosionoccurs through outward diffusion of cations and inward diffusion of ions. But it can beseen that, from the corrosion morphologies of 300M shown in Fig.7, corrosion products aretoo loose and porous to effectively block the chloride ions penetrating into the interface

Fig.7 Morphologies of 300M steels after anodic polarization in NaCl solutions with different con-centrations: (a) deionized water; (b) 0.1 mol/L; (c) 0.3 mol/L; (d) 0.5 mol/L.

· 309 ·

of corrosion products/substrate, so Cl− penetrates the oxide film through pores or defectseasier than do that in stainless steel; hence more corrosion pits initiate on the interface ofthe corrosion products/metal and evenly distribute on the surface, so the corrosion formis uniform corrosion macroscopically.

There is more evidence that uniform corrosion occurs on 300M ultra high strength steelduring dynamic polarization test and salt spray test. Fig.8 shows the lateral morphologyof 300M steel after salt spray test and corrosion products removed. Notice that the lateralside is irregularity, with many low holes. According to define of pitting, all the holes arenot deep enough to be defined pits. The schematic structure of uniform corrosion of the300M steel exposed to chloride solution is shown in Fig.9.

Fig.8 Lateral morphology of 300M steel aftersalt spray test.

Fig.9 Schematic structure of uniform corro-sion mechanism for 300M steel exposedto chloride solution.

From the lateral morphology and schematic structure of uniform corrosion for 300Msteel, it can explain that the diameters of Nyquist diagrams for 300M steels after salt spraytest dramatically decrease.

Comparing the difference of the corrosion morphologies between 300M steel and stain-less steels, stainless steels remain essentially passive in environments containing Cl−. Fromthe perspective of the oxide-film theory, breakdown of passivity by Cl− occurs locallyrather than generally. It is evident, because of the possibility of passive-active cell forma-tion, that deep pitting is much more common with passive metals than with nonpassivemetals[10,18,19].

Once a pit initiates, a passive-active cell is set up of 0.5–0.6 V potential difference.Minute anode of active metal is formed and surrounded by large cathodic areas of passivemetal. The resultant high current density accompanies a high corrosion rate of the anode(pit) and, at the same time, polarizes the alloy surface immediately surrounding the pitto values below the critical potential. High current densities at the anode cause highrates of metal penetration. Through flow of current, Cl− transfer into the pit formingconcentrated solutions of Fe2+, Ni2+, and Cr3+ chlorides, which, by hydrolysis, accountfor an acid solution, as shown in Fig.10. A pit stops growing only if the surface within thepit is again passivated[20,21].

It has been well known that pits and defects play a role in SCC initiation. In general,SCC is observed in alloy-environment combinations that result in the formation of a film

· 310 ·on the metal surface, making the alloy desir-able for resistance to uniform corrosion. SCCcan initiate at pits that form during exposureto the service environment by a breakdownin the protective film.

The pit geometry is important in deter-mining the stress and strain rate at the baseof the pit. Generally, the aspect ratio be-tween the penetration and the lateral corro-sion of a pit must be greater than about 10before a pit acts as a crack initiation site.A penetration to lateral corrosion ratio of 1corresponds to uniform corrosion, and a ra-tio of about 1000 is generally observed fora growing stress-corrosion crack. As in the

Fig.10 Schematic structure of passive-activecell responsible for pit growth in stain-less steel exposed to chloride solution.

case of a growing crack, the pit walls must exhibit some passive film forming capabilityin order for the corrosion ratio to exceed 1[22]. Comparing Fig.9 and Fig.10, the stainlesssteel is more susceptive to SCC.

Considering the 300M steel in NaCl solutions, it can be calculated that, even thoughpassive film can not form and 300M steel are less corrosion resistance than stainless steels, ithas a better property desirable for resistance to SCC. In this aspect, 300M steel is suitablefor applying in the aggressive environment with corrosive media and tensile stress whereSCC tends to occur.

4 Conclusions

Uniform corrosion occurs on 300M steel during electrochemical measurement in NaClsolutions with different concentrations. There is no anodic passive potential region on thepolarization curves. It is clear that the 300M steel is highly susceptible to NaCl solutions,which is characterized by corrosion current density increasing with addition of chlorideions, and the increase of chloride concentration shifts corrosion potential towards positivedirection and decreases the corrosion resistances, positively suggesting that the chlorideions speed up the corrosion rate of 300M steel.

Meanwhile the corrosion rate of 300M steel with corrosion products are acceleratedafter salt spray test, and corrosion products on the 300M steel surface formed during thesalt spray test are too loose and porous to effectively slow down the corrosion rate. Theschematic structure of uniform corrosion mechanism can well explain that the 300M steelhas better property of SCC resistance than stainless steels.

Acknowledgements—This work was supported by the Chinese National Science and Technology Infrastruc-ture Platforms Construction Project (Grant No.2005DKA10400) and Beijing Scienceand Technology Project (No.D09030303790901).

REFERENCES

[1] M.L.A. Graca, C.Y. Hoo and O.M. Silva, Eng Fail Anal 16(1) (2009) 182.

[2] R.O. Ritchie and R.M. Horn, Metal Mater Trans A 9(3) (1978) 331.

[3] Y. Tomita and T. Okawa, Mater Sci Eng A 172(1-2) (1993) 145.

· 311 ·

[4] C.F.G. aballero, M.J. Santofimia, C. Garcıa-Mateo, S. Kaliappan, H. Shaikh, and P.V. Sivaprasad,Mater Design 30(6) (2009) 2077.

[5] H. Garbacz, M. Lewandowska, W. Pachla and K.J. Kurzydlowski, J Microsc-Oxf 223(3) (2006) 272.[6] P.J. Antony, S. Chongdar, P. Kumar and R. Raman, Electrochim Acta 52(12) (2007) 3985.[7] A. Igual Munoz, J. Garcıa Anton, J.L Guinon and V. Perez Herranz, Corros Sci 49(8) (2007) 3200.[8] H. Altun and S. Sen, Mater Design 25(7) (2004) 637.[9] L.C. Chang and H.K.D.H. Bhadeshia, Mater Sci Eng A 184(1) (1994) L17.

[10] R.W. Revie and H.H. Uhlig, Corrosion and Corrosion Control, An Introduction to Corrosion Scienceand Engineering, 4th ed. (A John Wiley & Sons, Inc., Hoboken New Jersey, 2007) p.350.

[11] B. Dıaz, L. Freire, X.R. Novoa and M.C. Perez, Electrochim Acta 54 (2009) 5190.[12] J. Sanchez, J. Fullea, C. Andrade, J.J. Gaitero and A. Porro, Corros Sci 50 (2008) 1820.[13] B. Jegdic, D.M. Drazic and J.P. Popic, Corros Sci 50(5) (2008) 1235.[14] W.R. Osorio, L.C. Peixoto, L.R. Garcia and A. Garcia, Mater Corros 60(10) (2009) 804.[15] C.O.A. Olsson and D. Landolt, Electrochim Acta 48(9) (2003) 1093.[16] Y.F. Cheng, Bull Electrochem 21(11) (2005) 503.[17] X.L. Zhang, Z.H. Jiang, Z.P. Yao, Y. Songa and Z.D. Wub, Corros Sci 51(3) (2009) 581.[18] M. Pisarek, P. Kedzierzawski, M. Janik-Czachor and K.J. Kurzydlowski, J Solid State Electrochem

13(2) (2009) 283.[19] T.P. Hoar and W.R. Jacob, Corros Sci 66(7) (1967) 341.[20] P.C. Pistorius and G.T. Burstein, Corros Sci 33(12) (1992) 1885.[21] K.V.S. Ramana, T. Anita, S. Mandal, S. Kaliappan, H. Shaikh and P.V. Sivaprasad, Mater Design

30(9) (2009) 3770.[22] S.D. Cramer, ASM Handbook: Corrosion (ASM International, United States of America, 1987) p.302.