Influence of pre-deformation and oxidation in high temperature water on corrosion resistance of type 304 stainless steel

Accepted Manuscript Influence of pre-deformation and oxidation in high temperature water on corrosion resistance of type 304 stainless steel Lv Jinlong, Luo Hongyun, Liang tongxiang PII:

S0022-3115(15)30128-8

DOI:

10.1016/j.jnucmat.2015.07.041

Reference:

NUMA 49241

To appear in:

Journal of Nuclear Materials

Received Date: 2 May 2015 Revised Date:

21 July 2015

Accepted Date: 26 July 2015

Please cite this article as: L. Jinlong, L. Hongyun, L. tongxiang, Influence of pre-deformation and oxidation in high temperature water on corrosion resistance of type 304 stainless steel, Journal of Nuclear Materials (2015), doi: 10.1016/j.jnucmat.2015.07.041. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Influence of pre-deformation and oxidation in high temperature water on corrosion resistance of type 304 stainless steel Lv Jinlong2, Luo Hongyun1, Liang tongxiang2

RI PT

1. Key Laboratory of Aerospace Materials and Performance (Ministry of Education), School of Materials Science and Engineering, Beijing University of Aeronautics and Astronautics, Xueyuan Road 37, Beijing, China, 100191 2. Beijing Key Laboratory of Fine Ceramics, Institute of Nuclear and New Energy Technology, Tsinghua University, Zhongguancun Street, Haidian District, Beijing 100084, PR China

SC

Abstract: The passivation properties of deformed 304 stainless steels after immersion in borate buffer solution containing 0.2821 mol/L Cl- at 288 °C were investigated. The spinel and magnetite

M AN U

oxides were formed on all the samples. However, the hematite oxides reduced significantly with the increasing of strain. The sample with maximum strain possessed the poorest corrosion resistance. The hematite oxide could offer high corrosion resistance, while magnetite evidently deteriorated corrosion resistance. Moreover, the influence of the donors in outer layer of oxide

TE D

film on corrosion resistance was more important than that of the acceptors in inner layer.

AC C

EP

Keywords: Stainless steel; EIS; Raman spectroscopy; High temperature corrosion; Oxide films;

*Corresponding author.Tel.: +86 10 89792737;fax: +86 10 69771464 e-mail addresses: [email protected] 1

ACCEPTED MANUSCRIPT

1. Introduction The oxide layers on some austenitic stainless steels and alloys in high temperature water were studied with respect to the film formation mechanism. In addition, the oxide

RI PT

film formed on the surface of stainless steels was supposed to play a significant role in the process of stress corrosion cracking (SCC) in the primary circle of steam generator. Tapping et al. [1] investigated the morphology and compositions of oxide films

SC

formed on type 304 stainless steel exposed to 300 °C water at pH=10 and found that

M AN U

the outer layer in oxide was Fe-rich and the inner layer was Cr-rich. Stellwag et al. [2] also reported that the inner layer in oxide films consisted of a chromium-rich spinel which was covered by an outer layer of magnetite or iron-nickel spinel for austenitic stainless in high temperature water. Moreover, breakdown of the passive film was

TE D

related to crystallization of the initially amorphous and metastable passive film. Because transformation from initially amorphous passive film to crystallisation during high temperature could induce defects, which resulted in breakdown of the passive

EP

film. A study by Hakiki et al. [3] revealed a duplex structures which consisted of an

AC C

inner region composed of a few atomic layers of chromium oxide in contact with the metallic substrate and an outer region of iron oxides and hydroxides at the film/electrolyte interface for type 304 stainless steel and Fe-Cr alloys at room temperature. Moreover, the capacitance behaviour of the passive films was studied by the Mott-Schottky approach. The results showed that the films behaved as n-type and p-type semiconductors in the potential range above and below the flat band potential, respectively. This behaviour was due to the semiconducting properties of the iron

2

ACCEPTED MANUSCRIPT oxide and chromium oxide in passive films at room temperature [3]. Growth processes of the inner layer and the outer layer of the oxide films occurred at the metal/oxide and oxide/electrolyte interfaces, respectively. The growth rates of the

RI PT

oxide films were controlled by a transport of the layer-forming species through the layer, i.e. by the inward diffusion of oxygen as well as electrolyte species and the outward diffusion of metal cations [4]. In addition, the hydroxide ion could affect the

SC

type of oxides, especially in alkaline solutions. Structures of the passive film could be

M AN U

affected by solution temperature. For example, the oxide film formed on Inconel alloy 600 in aqueous 0.1 M Na2SO4 solution transformed from one-layer structure below 60 ◦C to two layers, i.e. an inner layer and an outer layer, above 100 ◦C. The metal ions were precipitated on inner layer film, which resulted in outer layer film [5]. According

TE D

to the analysis and structural investigation, the outermost part of the oxide film was the Ni0.75 Fe2.2504 inverse spinel, while in the intermediate part both Ni0.75 Fe2.25O4 and Fe304 inverse spinels were detected. Finally, the inner part of the oxide film consisted

EP

of mixed chromium oxides (Cr2O3+ FeCr2O4) for 316L stainless steel in a primary

AC C

type pressurized water reactor (PWR) environment [6]. Wada et al. [7] found that the dominant phase in the oxide film formed on 304 stainless steel exposed to 288 °C water containing 1 ppm O2 was magnetite type spinel. The study of Panter et al. [8] revealed that the oxide film formed on Alloy 600 in simulated PWR primary water was composed of outer loosely packed spinel particles and inner compact Cr-rich layer. Since the values of diffusivity of Ni and Fe ions were much higher than that of Cr ion [9].

3

ACCEPTED MANUSCRIPT Some investigations have been devoted to the effect of different solution environments on the thickness, composition and structure of the oxide film formed on stainless steels in high temperature water [5, 10]. Less hematite and more spinel

RI PT

oxides were formed on fresh sample as dissolved oxygen (DO) decreased, moreover, the base passive layer became more porous due to dissolution of Cr for 304 stainless steel in high temperature water [11]. Kumai and Devine [12] suggested that the outer

SC

oxide layer formed on 304 stainless steel was M3O4 when DO was below a critical

M AN U

value, then, as DO increased, M2O3 particles gradually formed with roughening of the preformed M3O4 particles. Moreover, the oxygen diffusion was probably enhanced by the local strain concentration by quantum chemical molecular dynamics study [13]. More detailed work was still needed to fully understand the effects of DO.

TE D

In fact, the effect of strain on passive film growth is also very important. The work presented here investigates the effect of pre-deformation on the chemical compositions and morphologies of the oxide films grown on type 304 stainless steel

EP

at 288 °C water by means of X-ray diffraction (XRD), Raman spectroscopy and

AC C

scanning electron microscopy (SEM). Special attention is paid to the influence of oxide structure on corrosion resistance by electrochemical tests.

2. Experimental

The chemical compositions (wt.%) of type 304 stainless steel are 0.05 C, 1.8 Mn,

0.84 Si, 17.35 Cr, 9.12 Ni, 0.024 P, 0.024 S and balance Fe. The as-received sample was annealed at 1050 °C for 1 h and water-quenched. The dog-bone-shaped samples with a gauge section 80 mm in length and 20 mm in width were strained to

4

ACCEPTED MANUSCRIPT engineering strain levels of 0%, 10%, 20%, 30% and 40%, respectively, at a strain rate of 1×10-4 s–1 using a testing system (SANS±100 kN) in laboratory air. The axial deformation level of the gauge length was measured by a contact clip-on extensometer.

RI PT

Samples were mechanically abraded with emery paper up to #3000 successively and degreased with ethanol before the exposure tests. Coupon type samples (10×10 mm2) were exposed to high temperature solution in autoclaves. The electrolyte solution used

SC

was 0.05 M H3BO3 (pH=9.2 adjusted with NaOH at room temperature) containing

M AN U

0.2821 mol/L Cl- by adding NaCl. The solution was prepared with analytical grade reagent and distilled water. The solution was deaerated by continuously bubbling with nitrogen gas (99.999%) for 4 h. Prior to the test, the autoclave was deoxygenized by purging with N2 gas for at least 2 h. DO in the inlet water is controlled at <5 ppb.

TE D

Then test was performed at 288 °C and 11.5±0.5 MPa for 28 days. Cu Kα (0.154056 nm) radiation at 40 kV and 40 mA at 4° / min was used for X-ray diffraction (XRD Rigaku Ultima IV) analysis. The surface morphology of the

EP

sample was examined by SEM. Raman measurements were performed using Jobin

AC C

Yvon (Laboratory RAM HR800) confocal micro-Raman spectrometer backscattered geometry through a 10×(NA=0.25) microscope objective. Ar+ laser emitting at a wavelength of 514.5 nm was used as a source of excitation (5.5 mW laser power available at sample). The number of grating in the Raman spectrometer was 1800 grooves/mm. The scattered light was analysed with a Dilor XY triple spectrometer and a liquid-nitrogen-cooled CCD multi-channel detector with the accuracy better than 1 cm-1. The Raman band of a silicon wafer at 520 cm-1 was used to calibrate the

5

ACCEPTED MANUSCRIPT spectrometer. The Raman spectra were recorded in the range of 200 to 1000 cm-1 and collected using three points of each sample in reproducible tests. For electrochemical measurements, a standard three-electrode electrochemical cell

RI PT

was used with very high density graphite counter electrode and a saturated calomel reference electrode (SCE) connected to a CHI 660B electrochemical workstation (Chenhua instrument Co. Shanghai, China) controlled by a computer and software.

SC

All the potentials in the paper were referred to SCE. Electrochemical impedance

M AN U

spectroscopy (EIS) measurements were carried out at the following immersion times: 60, 120 and 240 min for high temperature oxidized samples. The measurements were performed at the open-circuit potential (OCP) using a frequency range of 100 kHz to 10 mHz and a 5 mV amplitude of the AC signal. All the experiments were carried out

temperature.

TE D

in pH 9.2 borate buffer solution(0.075 M Na2B4O7 ·10H2O + 0.05 M H3BO3) at room

3. Results and discussion

EP

3.1 Surface oxide layers analysis

AC C

Fig. 1a shows XRD patterns of the samples with the engineering strain from 0% to 40%, and the microstructure of solution-annealed sample is mainly austenite. The intensities of austenite peaks gradually decrease with increase of the strain, while the intensities of α´-martensite peaks increase due to the large tensile strain. The slight diffraction peaks of ε-martensite can also be found. Obviously the volume fraction of α´-martensite increases with engineering strain. This result coincides with the previous report [14]. Fig. 1b shows XRD patterns of the oxide films formed on the

6

ACCEPTED MANUSCRIPT surface of all the samples after exposure to 288 °C borate buffer solution containing 0.2821 mol/L Cl-. The characteristic peaks suggest that they are spinel, magnetite and hematite. The high temperature oxides with similar structures were observed. For

(Ni0.2Fe0.8)(Fe0.95Cr0.05)2O4

on

the

top

of

a

outer layer

RI PT

example, two spinel oxide layers, including a ferrite-based chromite-based

inner

layer

(Ni0.2Fe0.8)(Cr0.7Fe0.3)2O4, were observed for 304 stainless steel in hydrogenated water

SC

at 260 °C [15]. Moreover, the hematite could be formulated as α-(Fe, Cr)2O3 with

M AN U

irregularly shape for 304 stainless steel in oxygenated high temperature [16]. These results support our XRD results above. Although the autoclave was deoxygenized by purging with N2 gas before immersion test. The hydroxide ion could affect the type of oxides, especially in alkaline solutions. Further experiments are needed for revealing

TE D

new form mechanism of hematite oxide and will be elaborated in future reports. The relative intensity of a characteristic peak in XRD is directly related to the proportion of its corresponding oxide structure. As strain increases up to 30%, the relative

EP

intensities of peaks corresponding to hematite are quite evident. However, the peak

AC C

intensities significantly decrease with the further increase of strain, which suggests the reduction of proportion of hematite. Obviously, different strain levels could affect the oxide films formed on the

surface of the 304 stainless steel. the Raman spectroscopy technique was used to further analyse compositions of oxide film on the samples. Generally, passive film is very complex and its analysis is rather difficult using Raman Spectroscopy, since different bonds may have the same frequency, even though their atoms and structures

7

ACCEPTED MANUSCRIPT are completely different. However, Raman spectra of passive films in steels have been investigated else where [17-19]. According to the Raman spectra, the compositions of passive film formed on the surface of the steel could be determined and its corrosion

RI PT

behaviour could be further understood. Fig. 2a-e show the Raman spectra of the oxide films of all the samples. The broad peak near 615 cm-1 indicates the bending vibration of hematite [18], while the peak at 664 cm-1 originates from magnetite [18]. It was

SC

inferred that Raman peaks around 300–330 cm-1 was Fe–O stretching vibrations of

M AN U

β-FeOOH [20]. However, β-FeOOH and γ-FeOOH have structural similarities with FeOCl, and their spectra look alike except for the peak due to O–H stretching frequency [21]. Moreover, the peak around 329–332 cm-1 was assigned to Fe–O stretching vibrations of FeOCl [19]. Therefore, it is very difficult to identify the oxide

TE D

type in Raman peaks around 300–332 cm-1 in Fig. 2a-e. However, their peak intensity is very low in the present study. A weak peak around 520–570 cm-1 in Fig. 2b is attributed to Cr2O3 which is diminished and/or overlapped with FeCr2O4, which will

EP

affect the corrosion resistance on the surface. However, the peak at 860 cm-1 due to

AC C

the Cr(III) and Cr(VI) mixed phase [19] is not observed. This is mainly attributed to chromium oxides formed in inner layer and is discussed in the next section. Despite nickel oxide is not detected in passive films by Raman spectroscopy technique, its presence can not be ruled out in the oxide film. The analytic results suggest that the structures of the oxide film analysed by XRD are in good agreement with those analysed by Raman spectroscopy. In the present work, it is found that the characteristics of oxide film are notably

8

ACCEPTED MANUSCRIPT influenced by the pre-strain. It may be relative to the different content of strain-induced α´-martensite, content of retained austenite and the dislocation densities in martensite and austenite, respectively.

RI PT

Fig. 3 shows SEM morphologies of the oxide films formed on all the samples. From Fig. 3a to c, it was found that samples were predominantly covered by small equiaxial Fe2O3 oxide particles with curving edges and big spinel oxide, while many

SC

large faceted particles with straight edges and planar faces can be found in Fig 3d and

M AN U

e. The growth of the outer oxide particles by precipitation mechanism was also supported by the previous works [22, 23].

In Fig. 3e, energy-dispersive X-ray detection (EDX) analysis showed that the chemical compositions of the big crystallites were 62.88 at% of O, 32.74 at% Fe, 1.70

TE D

at% Cr and 2.24 at% Ni, which indicated the outer layer was rich in iron and oxygen. The outer layer might consist mainly of iron oxides. The fine small and compact oxides close to the alloy base gave chemical compositions of 19.14 at% O, 56.66 at%

EP

Fe, 16.16 at% Cr and 6.72 at% Ni, which indicated that the compact inner layer was

AC C

rich in chromium and nickel. These results suggest that the formation of duplex layer oxides is possibly attributed to the different formation mechanisms of the inner and outer layers. The inner layer may be formed by solid-state growth processes controlled by oxygen diffusion, while outer layer may be formed by solid-state growth and dissolution–precipitation mechanisms in Fig. 3a-e.

3.2 EIS and Mott-Schottky measurements As above mentioned, the Fe2O3 is evidently deteriorated due to the increase of

9

ACCEPTED MANUSCRIPT strain, which could decrease the corrosion resistance of the samples. The electrochemical properties of the oxides on the surface of strained stainless steels were measured using EIS which has been widely applied for investigating the

RI PT

corrosion and passivation phenomena due to its non-destructive capability of corrosion monitoring [24]. The EIS measurements were performed at OCP for different immersion times in borate buffer solution. Fig. 4a, b and c show that the

SC

Nyquist plots at the various time display the similar feature. They are all composed of

M AN U

depressed semicircles. A significant decrease of semicircle is observed in sample with the largest strain. Fig. 4a´, b´ and c´ show the Bode plots of all the samples at the different immersion times. It is evident that a time constant appears on the basis of peak value of phase angle. Several models of equivalent circuits have been attempted

TE D

to fit these experimental data, and the most consistent results between experiment and fitting are obtained with the equivalent circuit shown in Fig.5a. Rs, Rt and Q represent the resistance of solution, the charge transfer resistance and the capacitance behaviour

EP

of the passive film, respectively. The constant phase angle element Q (CPE) is used to

AC C

replace the pure capacitance element C to reflect the non-ideal capacitance behaviour, including geometric origin, surface heterogeneity, surface roughness and so on. The impedance of a CPE is defined by the following equation

Z CPE =

1 Q(jω)n

(1)

where ω is the angular frequency (rad/s), j2 = -1 is the imaginary number and n is an adjustable parameter that always lies between 0.5 and 1. When n = 1, the CPE 10

ACCEPTED MANUSCRIPT describes an ideal capacitor. For 0.5 < n < 1, the CPE describes a distribution of dielectric relaxation times in frequency domain, and when n=0.5 the CPE represents a Warburg impedance with diffusion character. This equivalent circuit is the most

RI PT

suitable to describe the corrosion mechanism produced in the interface electrolyte/passive film/metal [25].

Fig. 5b and c show the fitted values of the elements. The smallest impedance is

SC

observed for sample with 40% strain, which indicates deteriorated property of its

M AN U

passive film. This may be due to the less hematite oxide on the sample as shown in Fig. 1b and Fig. 2e, considering conductivity and dopants in hematite are lower than those in magnetite. Moreover, with the increasing of immersion time, all impedance spectra change little, except for sample with 20% strain. The Q is biggest for sample

TE D

with 40% strain, which indicates an increase of defects on the surface film and the poor protective properties of the passive film. It is worthwhile to note that all CPE values change little after 4 h immersion at OCP in the aerated borate buffer solution.

EP

The oxygen diffusion was probably enhanced by the local strain concentration [13],

AC C

which facilitate to form more Fe2O3. However, in present work, the compact hematite was evidently decreased and the magnetite became the main part in oxide film due to large strain. Moreover, Kumai and Devine also suggested that the oxide film at higher solution temperature could be partly oxidized to loose or non-protective oxides at higher potentials when the dissolved oxygen level in the water was increased [26]. This could be attributed to more oxygen vacancies in oxide film for sample with 40% strain, which will be proved by following Mott-Schottky experiment. The

11

ACCEPTED MANUSCRIPT conductivity of hematite is very low compared to magnetite because of the very low defect concentration [27]. Considering the defects or dopants in magnetite are much more than those in hematite [28], it can be concluded that oxides with relatively more

RI PT

hematite will have a relatively high corrosion resistance compared with the oxides without or with a relatively thin hematite layer [29]. The present more hematite oxides on samples with small strain highly hinders the transport rate of electrons and ions

SC

through the oxide, resulting in a higher corrosion resistance. As discussed above, the

M AN U

effect of doping concentration in oxide film on corrosion resistance is very important and will be discussed below. Then, the obtained results will further support the above conclusion.

Based on Mott-Schottky theory [30], the space charge capacitances of the n-type

-2 C -2 = C H-2 + C SC =

2 kT (E - E fb ) ε S ε0 qN D e -2 kT (E - E fb ) εS ε0 qN A e

(2)

(3)

AC C

EP

-2 C -2 = CH-2 + CSC =

TE D

and p-type semiconductor are given by Eq. (2) and (3), respectively

where ε0 is the vacuum permittivity (8.854×10-12 F m-1), εs is the dielectric constant of the specimen, e is the electron charge (1.6×10-19 C), k is the Boltzmann constant (1.38×10-23 J K-1), ND and NA are the donor or acceptor concentrations, respectively, T is the absolute temperature and Efb is the flat-band potential. For p-type semiconductor, C-2 versus E should be linear with a negative slope which is inversely proportional to the acceptor concentration. On the other hand, n-type semiconductor yields a positive slope which is inversely proportional to the donor concentration. The

12

ACCEPTED MANUSCRIPT dielectric constant εs is assumed as 12 for the passive films on stainless steels [30]. The Mott–Schottky plots of the oxide films formed on different samples are showed in Fig. 6a-c. The results show that there are changes on the slope of the

RI PT

Mott–Schottky plots at a potential around -0.3 V. The slopes of the Mott–Schottky plots determine the donor concentration ND and the acceptor concentration NA. The values of donor and accept concentrations are summarized in Fig. 6a´-c´. They are of

SC

the order of magnitude of 1021 ~ 1022 cm−3, which are much higher than that in ref [4].

M AN U

The concentration of the donor is lower than that of the acceptor in Fig. 6a´-c´. The slight difference between the acceptor and donor concentrations is observed for

samples with 10% and 30% engineering strain. However, the total acceptor and donor concentrations are minimum for the sample with 20% strain, which indicates a decrease in defects on the passive film.

TE D

The n-type oxide films Fe2O3 and Fe(OH)3 [3] and p-type oxide films Cr2O3 [31] and Fe3O4 [31] were probably formed on the stainless steels. The enrichment in

EP

chromium was generally more pronounced in the inner layer of the passive films [3]. According to point defect model (PDM) [32], the passive film includes a number of

AC C

point defects, such as interstitial cations (donors), oxygen vacancies (donors) and/or cation vacancies(acceptors). The PDM provided a deterministic description of passivity breakdown on metals and proposed that cation vacancies generated at the film/solution interface moved to the metal/film interface. Therefore, the increased acceptor concentration in the passive film may be regarded as the rising of chromium cation vacancies in stainless steels in Fig. 6a´-c´. The acceptor concentration is always more than donor concentration in Fig. 6a´-c´. The values of NA obtained in the present work first decreases and then increases with the increase of strain. This indicates that 13

ACCEPTED MANUSCRIPT chromium cation vacancies which occupy the main part in acceptor in oxide film first decreases and then increases with the increasing of strain. The minimum acceptor concentration for the sample with 20% engineering strain could improve the corrosion resistance. Whereas the fluctuated trend is observed for donor concentration.

RI PT

Comparing with sample without strain, the ND value of sample with 40% strain significantly increase, although the NA value of the former is higher than that of the latter. Considering results above and lower corrosion resistance of sample with 40%

SC

strain, It can be inferred that donor concentration in outer oxide significantly decreases corrosion resistance. It indicates that deteriorated stability of passive film

M AN U

on the sample with 40% strain depends mainly on the outer layer of the oxide and its donor concentration. This is attributed to more oxygen vacancies in Fig. 6a´-c´ and less content of hematite in Fig.2e in the sample with 40% strain. The magnetite contains more defects and reduce the corrosion resistance. At the same time, we must

TE D

point out that the donor concentration in the outer layer consisted of iron interstitial cations and oxygen vacancies. It is difficult to distinguish them. However, the oxygen vacancy is the main part and more oxygen vacancies lead to incompact iron oxide on

EP

the outer layer. These results agree with those in Fig. 1b and Fig. 2. The incompact

AC C

outer iron oxide deteriorates the corrosion resistance of the sample with 40% engineering strain in borate buffer solution.

4. Conclusions

The present work investigated the effect of pre-deformation on the chemical

compositions and the morphologies of the oxide films formed on type 304 stainless steel in 288 °C borate buffer solution containing 0.2821 mol/L Cl-. The corrosion resistance was investigated by electrochemical impedance techniques in borate buffer

14

ACCEPTED MANUSCRIPT solution. The main conclusions are as follows. (1) The inner layer of the oxide films consisted of the mixed Cr oxides, while the outer layer of the oxide film contained spinel, magnetite and hematite. When

RI PT

strain value exceeded a critical value, the preformed hematite tended to diminish. (2) The minimum acceptor and donor concentrations could improve the corrosion resistance for the sample with 20% engineering strain .

SC

(3) Comparing with hematite, higher dopants in magnetite facilitated the transport

M AN U

rate of electrons and ions through the oxide, resulting in a lower corrosion resistance. Therefore, the corrosion resistance of sample with 40% strain significantly decreased.

(4) Comparing to acceptors in inner layer of oxide, more donors in outer layer oxide

strain.

Acknowledgments

TE D

could significantly decrease the corrosion resistance of the sample with 40%

EP

This work was financially supported by National Natural Science Foundation of China (Grant No. 91326203)

AC C

Reference

[1] R.L. Tapping, R.D. Davidson, E. McAlpine, D.H. Lister, Corros. Sci. 26 (1986) 563–576. [2] B. Stellwag ,Corros. Sci. 40 (1998) 337–370. [3] N. B. Hakiki, S. Boudin, B. Rondot, M. Da Cunha Belo, Corros. Sci. 37(1995)1809–1822. [4] D. J. Kim, H. C. Kwon, H. P. Kim, Corros. Sci. 50 (2008) 1221–1227 [5] J. J. Park, S. Pyun, S. B. Lee, Electrochim. Acta 49 (2004) 281–292.

15

ACCEPTED MANUSCRIPT [6] M. Da Cunha Belo, M. Walls, N. E. Hakiki, J. Corset, E. Picquenard, G. Sagon, D. Noël, Corros. Sci.40(1998) 447-463. [7] Y. Wada, A. Watanabe, M. Tachibana, K. Ishida, N. Uetake, S. Uchida, K.Akamine, M. Sambongi, S. Suzuki, K. Ishigure, J. Nucl. Sci. Technol. 38 (2001) 183–192.

RI PT

[8] J. Panter, B. Viguier, J.M. Cloué, M. Foucault, P. Combrade, E. Andrieu, J. Nucl.Mater. 348 (2006) 213–221. [9] J. Robertson, Corros. Sci. 29 (1989) 1275–1291.

SC

[10] M. C. Sun, X. Q. Wu, Z. E. Zhang, E. H. Han, Corros. Sci. 51 (2009) 1069–1072.

M AN U

[11] W. J. Kuang, X. Q.Wu, E. H. Han, Corros. Sci. 63 (2012) 259–266. [12] C. S. Kumai, T. M. Devine, Corrosion 63 (2007) 1101–1113.

[13] N. K. Das, K. Suzuki, Y. Takeda, K. Ogawa, T. Shoji, Corros Sci.50 (2008)1701-1706. [14] S. Curtze, V. T. Kuokkala, A. Oikari, J. Talonen, H. Hanninen, Acta Mater. 59 (2011)

TE D

1068–1076.

[15] S. E. Ziemniak, M. Hanson, Corros Sci. 44 (2002) 2209–2230

EP

[16] W. J. Kuang, X. Q. Wu, E.H. Han, L. Q. Ruan, Corros Sci. 53 (2011) 1107–1114 [17] W. Chen, R.G. Du, C. Q. Ye, Y. F. Zhu, C.J. Lin, Electrochim. Acta 55 (2010) 5677–5682

AC C

[18]T. K.Yeh, Y. C. Chien, B. Y. Wang, C. H. Tsai, Corros Sci. 50 (2008) 2327–2337 [19] S. Ramya, T. Anita, H. Shaikh, R. K. Dayal, Corros Sci. 52 (2010) 2114–2121 [20] J. Gui, T. M. Devine, Corros Sci. 32 (1991) 1105–1124. [21] J. H. Choy, J. B. Yoon, K. S. Han, J. de Physique IV France 07 (1997) 335–336. [22] B. Stellwag, Corros. Sci. 40 (1998) 337–370 [23] D. H. Lister, R. D. Davidson, E. Mcalpine, Corros. Sci. 27 (1987) 113–140 [24] K. M. Ismail, Electrochim. Acta 52 (2007) 7811–7819.

16

ACCEPTED MANUSCRIPT [25] C. M. Abreu, M. J. Cristóbal, R. Losada, X. R. Novoa, G. Pena, M. C. Perez, J. Electroanal. Chem. 572 (2004) 335–356. [26] C. S. Kumai, T.M. Devine, Corrosion 61 (2005) 201–218.

RI PT

[27] J. H. Kennedy, K. W. Frese Jr., J. Electrochem. Soc. 125 (1978) 723–726. [28] R. M. Cornell, U. Schwertmann, The iron oxides: Structure, properties, reactions, occurrences and uses.seconded.,Wiley-VCH,2003. New York.

SC

[29] J. Wielant, V. Goossens, R. Hausbrand, H. Terryn, Electrochim. Acta 52 (2007)7617–7625.

M AN U

[30] W. P. Gomes, D. Vanmackelbergh, Electrochim. Acta 41 (1996)967–973.

[31] C. Sunseri, S. Piazza, F.D. Quarto, J. Electrochem. Soc. 137 (1990) 2411–2417. [32] D. D. Macdonald, A. Sun, Electrochim. Acta 51(2006) 1767–1779.

Figure caption

TE D

Fig. 1. (a) The X-ray diffraction patterns of 304 stainless steels with different engineering strain levels; (b) X-ray diffraction patterns of the oxide films formed on 304 stainless steel with different strain after exposure to 288 °C borate buffer solution containing 0.2821 mol/L Cl-.

EP

Fig. 2. The Raman spectra of the oxide films formed on 304 stainless steel with (a) 0%, (b)10%, (c)

AC C

20%, (d) 30% and (e) 40% strain after exposure to 288 °C borate buffer solution containing 0.2821 mol/L ppm Cl-.

Fig. 3. SEM morphologies of the oxide films formed on 304 stainless steel with (a) 0%, (b)10%, (c)20%, (d) 30% and (e) 40% strain after exposure to 288 °C borate buffer solution containing 0.2821 mol/L Cl-. Fig 4. Nyquist plots under immersion in borate buffer solution at OCP after (a) 1 h, (b) 2 h, (c) 4 h, respectively. Bode plots obtained at OCP after (a´) 1 h, (b´) 2 h, (c´) 4 h, respectively.

17

ACCEPTED MANUSCRIPT Fig.5. (a) The electrochemical equivalent circuit for EIS fitting. The fitted values of (b) Rt and (c) Q by the corresponding electron-circuit. Fig. 6 The Mott–Schottky plots for oxidized samples with different strains after (a) 1 h, (b) 2 h, (c) 4

AC C

EP

TE D

M AN U

SC

(a´) 1 h, (b´) 2 h, (c´) 4 h at OCP in the aerated borate buffer solution

RI PT

h at OCP in the aerated borate buffer solution; The donor ND and the acceptor NA concentration after

18

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT