Physicochemical performance of FeCO₃ films influenced by anions

Abstract

The corrosion film plays an important role in the further electrochemical processes of steel in CO₂ corrosion. Thus, the physicochemical performance of an FeCO₃ film was investigated using the Mott–Schottky electrochemical technique, potentiodynamic polarization, electrochemical impedance spectroscopy (EIS), X-ray diffraction (XRD) and scanning electron microscopy (SEM). The performance of the FeCO₃ film was dependent on the aqueous solution. In 0.5 mol L⁻¹ NaHCO₃ solution, an n-type semiconducting behavior was found for the FeCO₃ film. The dense microstructure and lower interstitial cation or anion vacancy doping was in favor of strong corrosion resistance of the film. On the contrary, a p-type semiconducting behavior of the FeCO₃ film was exhibited in both 0.5 mol L⁻¹ NaCl solution and 0.5 mol L⁻¹ Na₂SO₄ solution. Higher cation vacancy doping was found, and the integrity of the microstructure was damaged, which decreased the transfer resistance of electron and mass. As a result, the protective ability of the FeCO₃ film was decreased. The physicochemical mechanism for the semiconducting properties of the FeCO₃ film was explained using the Point Defect Model (PDM).

---

1. Introduction

CO₂ corrosion of steel has been one of the most important problems in the petroleum industry and CO₂ capture process,^1,2 which has caused thinning, perforation or fracture of tubular goods.

Though many different electrochemical mechanisms have been proposed to explain the CO₂ corrosion of steel, the overall reaction is presented as eqn (1):

Fe + CO₂ + H₂O → FeCO₃ + H₂ (1)

where FeCO₃ precipitates on the surface of the steel to form a corrosion film, which determines the further corrosion behavior of the steel.³ When referring to the cause of the changes in the corrosion rate, the pitting development, the erosion behavior, etc, many studies have been conducted on such films.^4–6

The protective ability strongly depends on the properties of the FeCO₃ film on the surface of the steel. The first requirement for corrosion inhibition is the integrity of the corrosion film. Thus, the mechanical performances of the corrosion film, including fracture toughness,⁷ wear resistance,⁸ Young’s modulus and adhesion strength,⁹ have been investigated, and their relationships to the protective behavior were also discussed.

The electrochemical behavior of the corrosion film is another important factor in anti-corrosion. The electrochemical characteristics were monitored during corrosion film formation. They indicated that the CO₂ corrosion film promoted electrode potential and mass transferring resistance.^10,11 The passive film on the iron base alloy presented semiconductor properties; for example, an n-type semiconductor formed on the surface of X70 steel in carbonate/bicarbonate solution.¹² A deep understanding of the physicochemical process of the film in electrochemical corrosion can be realized from the semiconducting properties.

Literature has shown that there is a correlation between the corrosion behavior of a metal and the semiconducting properties of a passive film,^13–15 but in the CO₂ corrosion field, reports are scarcely found. Mott–Schottky measurement is a powerful method for studying semiconductor characteristics. The well known equations are listed as eqn (2) for an n-type semiconductor and eqn (3) for a p-type semiconductor, according to the Mott–Schottky theory:

where ε_r is the relative dielectric constant of the specimen, which is roughly estimated to be 8.0,¹⁶ ε₀ is the permittivity of free space (8.854 × 10⁻¹⁴ F cm⁻¹), e is the electron charge (1.602 × 10⁻¹⁹ C), A is the sample area (cm²), N_D and N_A are the donor density (cm⁻³) and acceptor density (cm⁻³), E_FB is the flat band potential (V), k is the Boltzmann constant (1.38 × 10⁻²³ J K⁻¹), and T is the absolute temperature (K).

As our previous work reported, the corrosion of an oil tube was severely affected by groundwater chemistry.¹⁷ Generally, anions play an important role in the corrosion. In Nesic’s review, the species of anions in the formation solution are Cl⁻, HCO₃⁻ and SO₄²⁻. For these reasons, the purpose of our work is to clarify the influence of Cl⁻, HCO₃⁻ and SO₄²⁻ on the semiconducting properties of a FeCO₃ film in order to understand the different corrosion behaviors of tubular steel in CO₂ storage oil and gas wells.

2. Experimental

Material

The experimental material is N80 steel with the chemical composition (wt%) 0.24 C, 0.22 Si, 1.19 Mn, 0.013 P, 0.004 S, 0.036 Cr, 0.021 Mo, 0.028 Ni, and 98.248 Fe. The sample, a plate with a 10 × 10 mm² surface and 3 mm thickness, was machined from the oil tube directly. The pretreatment of the sample included polishing with 800-grit silicon carbide paper, degreasing with acetone, washing with distilled water and drying with a blow drier, sequentially.

Film preparation

Distilled water was deoxygenated by using continuous N₂ bubbling for an adequate period of time (more than 6 h). The pretreated steel samples were hung separately in a high temperature and high pressure autoclave. The distilled water was pumped into the autoclave till all samples were submerged in the solution. N₂ was bubbled from the bottom of the autoclave for 1 h to further ensure that it was oxygen-free. CO₂ was introduced under pressure into the autoclave to maintain a total pressure of 8 MPa, and the temperature was increased to 90 °C. The corrosion film was prepared after 72 h.

Exposure and electrochemical testing

The samples were taken out of the autoclave and respectively exposed to three types of solution: (1) 0.5 mol L⁻¹ NaCl, (2) 0.5 mol L⁻¹ Na₂SO₄ and (3) 0.5 mol L⁻¹ NaHCO₃. All the solutions were prepared using analytical grade reagents and distilled water and were deoxygenated sufficiently with N₂ bubbling before the immersion experiment. The experiment was carried out in an airtight five necked flask at 90 °C. After 72 h exposure, the Mott–Schottky electrochemical technique and potentiodynamic polarization were carried out under the three conditions. A classic three electrode cell, including a platinum plate counter-electrode, a saturated calomel reference electrode inserted in a electrolytic bridge and a specimen of film-covered N80 steel (only 10 × 10 mm² surface exposure), was built for the electrochemical tests. An Autolab Model PGSTAT302N electrochemical potentiostat was used to carry out electrochemical measurements. The Mott–Schottky curve was measured at a frequency of 1 kHz and with a potential scan rate of 10 mV s⁻¹. The potentiodynamic polarization was carried out using a scan rate of 0.5 mV s⁻¹. EIS measurements were carried out in the 100 kHz to 0.00001 kHz frequency range at open potential, with a peak-to-peak amplitude of 10 mV.

SEM and XRD measurements

The phase composition analysis of the prepared film was carried out using a DX-2000 Rigakudmax X-ray diffractometer with a copper Kα X-ray source. The scanning range of 2θ started from 10° to 80°. A JSM-6490LV scanning electron microscope was used to observe the microstructure of the film.

3. Results and discussion

Prepared FeCO₃ corrosion film

A pure FeCO₃ corrosion film was prepared successfully in distilled water with dissolved CO₂ as expected, which was proven by XRD as shown in Fig. 1. All the peaks indicated that FeCO₃ (siderite) was the sole product. The formation of FeCO₃ is regarded as the secondary product of the anodic and cathodic electrochemical reactions when the concentrations of Fe²⁺ and CO₃²⁻ exceed the solubility limit in the local space near the surface of steel. In the various formation solutions, the ions hardly take part in the sedimentary film directly, except for Ca⁺.^1,18 Both CaCO₃ and FeCO₃ belong to the calcite structure, so the trace Ca displaces an equivalent number of Fe in the crystal lattice of FeCO₃ to produce (Fe, Ca)CO₃. This perturbation cannot change the properties of the corrosion film. Therefore, the pure FeCO₃ film certainly represents the corrosion process in the formation solution environment.

Fig. 1 XRD diffraction of the corrosion film prepared from water with dissolved CO₂ at 90 °C and 8 MPa.

As can be seen in Fig. 2, the corrosion film was composed of rhombic crystal grains. A compact structure was observed, because the crystal grains are arranged in an ordered fashion. The porosity of the film is very low (below 0.1%).¹⁸

Fig. 2 SEM image of the FeCO₃ film prepared from water with dissolved CO₂ at 90 °C and 8 MPa.

Physicochemical mechanism for semiconductor type

Mott–Schottky curves of FeCO₃ films in the three solutions are shown in Fig. 3, according to the dependence of E and C⁻². The electronic conducting behavior can be analyzed from the Mott–Schottky plots.

Fig. 3 Mott–Schottky plots of the FeCO₃ films exposed to different solutions.

A marked difference was observed from the three plots in Fig. 3. In NaHCO₃ solution, a positive slope was found, which indicated an n-type semiconductor film. However, a p-type semiconductor film was judged from the negative slope in both NaCl and Na₂SO₄ solutions. Thus, the semiconducting properties of the FeCO₃ film are determined by the anion species in the formation solution.

According to the electron band theory of a solid, if the number of electrons in the conduction band is more than that of holes in the valence band, the solid should be considered as an n-type semiconductor. In the opposite case, it belongs to the p-type semiconductors. The n-type semiconductor FeCO₃ film is attributed to the doping of an interstitial cation or anion vacancy. The doping of cation vacancies in the FeCO₃ film makes it present p-type semiconductor characteristics.

Referring to the PDM resulting from the research group of Macdonald,¹⁹ the physicochemical process of the FeCO₃ films in different solutions is described schematically in Fig. 4. PDM-II is suitable for explaining the corrosion film by precipitation of ferrous ions with carbonate ions. The semiconducting properties of a FeCO₃ film are determined by the complex process of generation, diffusion and annihilation of charge carriers.

Fig. 4 Physicochemical mechanism for the semiconductor type of FeCO₃ films.

It is well known that seven basic reactions occur in the steel/film/solution system, as seen in Fig. 4 labeled as (I) to (VII). The various anions have different influences on these reactions by their diverse adsorptive and reactive abilities. Reactions (I), (IV), (VI) and (VII) are present at the interface of the outer layer of the FeCO₃ film and solution (film/solution), while the other reactions occur at the interface of the steel and the inner layer of the FeCO₃ film (steel/film). Here, the symbols are cation vacancy, V_Fe²⁻, anion vacancy, V_O²⁺, interstitial cation, Fe_i²⁺, neutral vacancy, V_Fe, ferrous ion in a cation site, Fe_Fe, anion in anion site, [O]_o, hydration anion, X^x−·nH₂O, film forming anion, [O], aqueous ferrous ion, Fe(H₂O)_n²⁺, and film compound, Fe[O].

The V_Fe²⁻ is generated according to reaction (I) by adsorption of X^x−·nH₂O on the active Fe_Fe in the FeCO₃ film. Cation vacancies are generated by the catalysis of adsorbed species.²⁰ The subsequent step is the diffusion of V_Fe²⁻ from film/solution towards to steel/film. The V_Fe²⁻ is eliminated in reaction (II). The V_O²⁺ and Fe_i²⁺ are produced by reactions (III) and (V). After diffusion to the surface of the film, they disappear by reactions (IV) and (VI). Reaction (VII) means the dissolution of the film.

Guo et al. believed that FeCO₃ had strong adsorption to anions via Coulombic as well as Lewis acid–base interactions.²¹ Among HCO₃⁻, SO₄²⁻ and Cl⁻, the former is softer than the latter two in Lewis acidity on the basis of the “hard and soft acid and base” concept,²² so the adsorption ability of HCO₃⁻ for the same base, Fe²⁺, in the film is weaker than SO₄²⁻ and Cl⁻. Literature has reported that Cl⁻ and SO₄²⁻ played an important role in the corrosion products of iron to form complex compounds.^23,24 It could be supposed that transient products FeCO₃Cl⁻_ads and FeCO₃SO₄²⁻_ads are present in the film due to the adsorption of Cl⁻ and SO₄²⁻ on Fe_Fe in the film, where the subscript “ads” means adsorption state. Using XPS, the adsorption on the film was found to occur by M–O–X bonds.²⁵ The amount of V_Fe²⁻ caused by reaction (I) in NaHCO₃ solution is much less than in Na₂SO₄ and NaCl solutions.

The kinetics of reaction (IV) are also affected by the X^x−·nH₂O; thus, the amount of consumed Fe_i²⁺ in the Na₂SO₄ and NaCl solutions is more than in the NaHCO₃ solution. The annihilation of V_O²⁺ runs in two ways as reaction (VIa) and reaction (VIb). Reaction (VIa) indicates that the anion, HCO₃⁻, is in favor of FeCO₃ film formation, while reaction (VIb) cannot precipitate solid production by anions, such as SO₄²⁻ and Cl⁻. The adsorbed anion X^x− will be trapped by V_O²⁺. The anion occupies the V_O²⁺, leading to further reactions. When it meets with V_Fe²⁻:

Here, Null means disappearance of X_O^2−x and V_Fe²⁻. However, when it reacts with neighboring Fe_Fe:

Therefore, HCO₃⁻ takes part in the film formation, resulting in a more dense structure. On the other hand, adding NaHCO₃ elevates the pH of the solution because of its strong base and weak acid performance, which mitigates the solubility of iron carbonate. However, SO₄²⁻ and Cl⁻ lead to a porous structure based on the soluble ion Fe(H₂O)_n²⁺ produced in several reactions. The anions adsorbed on the film undergo hydrolysis and reduce the local pH, which also causes the dissolution of the film. The microstructure can be confirmed from Fig. 5. The SEM images of the FeCO₃ film surface after 72 h immersion in NaHCO₃, Na₂SO₄ and NaCl solutions are shown in Fig. 5. The shape and arrangement of the crystals in the NaHCO₃ solution were similar to the initial FeCO₃ film prepared from distilled water with dissolved CO₂. However, the surface morphology became porous after immersion in the other two solutions, and the crystalline grain changed to spherical.

Fig. 5 Surface morphologies of the FeCO₃ film after immersion in different solutions.

Cl⁻ is a well known corrosion accelerator of metal in aggressive environments. The role of SO₄²⁻ is not well clarified. It has been found that SO₄²⁻ is more aggressive than Cl⁻, and SO₄²⁻ was the species responsible for a pitting attack on copper.²⁶ During the formation of artificial steel rust particles prepared from acidic aqueous Fe(III) solution, the influence of SO₄²⁻ is more important than Cl⁻.²⁴ Deng et al. suggested that SO₄²⁻ would supplant the adsorbed Cl⁻ and accelerate the effect of pitting initiation.²⁷ Thus, SO₄²⁻ is indeed a strongly aggressive anion to metal.

In NaHCO₃ solution, less V_Fe²⁻ is produced, so it is easy to meet the requirement of flux:

The FeCO₃ film is doped by Fe_i²⁺ and V_O²⁺, and thus it presents n-type semiconductor characteristics. In Na₂SO₄ and NaCl solutions, it is easy to generate V_Fe²⁻. Contrarily, the flux is abided by:

The FeCO₃ film is a p-type semiconductor doped by V_Fe²⁻.

The charge carrier concentration N_q, flat band potential E_FB and semiconductor type corresponding to each of the three solutions are listed in Table 1, where N_q includes donor density (N_D) for the n-type semiconductor and acceptor density (N_A) for the p-type semiconductor. All the parameters were calculated using linear fitting of the Mott–Schottky curves in Fig. 3, based on eqn (2) and (3).

Table 1 Electrochemical parameters calculated from the Mott–Schottky plots

Solution	N q (cm^−3)	E FB (VSCE)	Semiconductor type
NaHCO3	0.27 × 10^23	−0.31	n
Na2SO4	1.72 × 10^23	−0.26	p
NaCl	17.2 × 10^23	0.14	p

---

When the n-type or p-type semiconductor contacts the electrolyte, a conductivity band bending occurs in the semiconductor due to non-uniform distribution of charge at the interface zone. It can be described by E_FB. The E_FB is clearly environment-dependent. The E_FB of the n-type semiconductor in NaHCO₃ solution is more negative than that of the p-type semiconductor in Na₂SO₄ solution and NaCl solution.

The charge carrier concentration reflects the number of defects in the film. It further affects the conductivity of the film due to the transport of charge carriers. The lowest N_D, 0.27 × 10²³ cm⁻³, was obtained in the NaHCO₃ solution. N_A values of 1.72 × 10²³ cm⁻³ and 17.2 × 10²³ cm⁻³ were found in the Na₂SO₄ solution and NaCl solution, respectively. The lower N_A in Na₂SO₄ solution should be due to the greater radius of SO₄²⁻ than Cl⁻.

This indicated a less-doped semiconductor structure of the film in NaHCO₃ solution and a higher doping degree of the film in the other two solutions. The electron passes through the film by the way of flow of doping ions. The higher the doping degree of the film, the better the conductivity obtained for the film.

Fig. 6 shows the XRD results of the film after immersion in the three solutions. As can be seen, the composition is still FeCO₃ (siderite) in each case. According to the mechanism mentioned above, SO₄²⁻ and Cl⁻ play the role of producing charge carriers and dissolving the solid film. They do not participate directly in the substance exchange reaction. For this reason, no matter what the formation solution type in the wells containing CO₂, the corrosion scale is mainly composed of FeCO₃. The diffraction peak height of the film in the NaHCO₃ solution is stronger than the other two. This means that the crystal perfection in the NaHCO₃ solution is the best, which is clearly observed in Fig. 5.

Fig. 6 XRD patterns of the corrosion film after immersion in different solutions.

Relation of semiconducting properties and corrosion resistance

Oliveira et al. reported that the film gave rise to higher protection when the film changed from a p-type to n-type conductor.¹³ Different pitting susceptibilities were obtained for the p-type and n-type semiconducting oxide films on the surface of stainless steel.²⁸ In fact, there should not be a direct relationship between semiconductor type and corrosion resistance. The corrosion behaviors rely on two important factors, namely, potential and current. They correspond to flat band potential and charge carrier concentration.

Potentiodynamic polarization was used to determine the corrosion resistance behaviors of the FeCO₃ films after immersion in different solutions. The polarization curves are shown in Fig. 7, and the analysis results are shown in Table 2. After strong polarization, a high stable passivity was observed in the NaHCO₃ solution due to the wide passivation potential range. However, activation polarization was presented when a high potential was applied in the Na₂SO₄ and NaCl solutions, that is to say, passivity was broken down by the aggressive anions Cl⁻ and SO₄²⁻. Using the method of Tafel extrapolation, the corrosion currents were calculated from the plots in Fig. 7. They increase with in sequence of NaHCO₃, Na₂SO₄ and NaCl solutions. Therefore, HCO₃⁻ inhibits the CO₂ corrosion of N80 steel, but Cl⁻ and SO₄²⁻ lead to low corrosion resistance. Thus, the protection of the FeCO₃ film in different solutions decreases in the sequence of NaHCO₃, Na₂SO₄ and NaCl. During the electrochemical corrosion process, the anodic reaction, steel dissolution, occurs at the steel/film interface, but the cathodic reaction is always found at the film/solution interface, because the film has a more positive potential than the steel itself.²⁹ The link between the cathodic reaction and anodic reaction is an electron, i.e., the charge carrier in the film. A higher charge carrier concentration implies a more rapid electron transfer step. The lowest charge carrier concentration results in passivation in the NaHCO₃ solution. The charged particles do not diffuse easily. The higher charge carrier concentration decreases the IR (the potential drop of scale caused by flow current (I) and resistance (R) of the film) of the film, which accelerates the kinetics of corrosion in Na₂SO₄ and NaCl solutions.

Fig. 7 Polarization curves of the FeCO₃ film influenced by different solutions.

Table 2 Analysis of polarization curves of the FeCO₃ films in three solutions

Solution	i corr ^a (×10^−5 A cm^−2)	Characteristic
a i corr means corrosion current density.
NaHCO3	7.60	Passive state
Na2SO4	13.5	Active state
NaCl	31.6	Active state

---

Fig. 8 shows the Nyquist plots of the EIS and their equivalent circuits in the three solutions. In the Na₂SO₄ and NaCl solutions, one capacitive loop (semi-circle) was observed, but a Warburg impedance (oblique line) overlapped with one capacitive loop in the NaHCO₃ solution. Warburg impedance is attributed to the transfer resistance of the passive film to the conductive ions suggested in the PDM. One capacitive loop without Warburg impedance means that the electrochemical reaction at the interface between steel and film is the rate controlling step. The measured data were fitted well by the equivalent circuits shown. Table 3 lists the calculated values of the various electrochemical parameters. R_t, interface reaction resistance, is inversely proportional to the electrochemical corrosion rate. Therefore, the results also prove the higher protection of the corrosion film in NaHCO₃ solution compared with the other two solutions.

Fig. 8 EIS of FeCO₃ films in different solutions.

Table 3 Fitted electrochemical parameters in EIS by the equivalent circuits

Solution	R s ^a (Ω cm^2)	C dl ^b (Ω^−1 cm^−^2 s^−n)	n ^c	R t ^d (Ω cm^2)	Z w ^e (Ω^−1 cm^−2 s^−0.5)
a R s, solution resistance; b C dl, double layer capacitance; c n, frequency independent parameter value; d R t, interface reaction resistance; e Z w, Warburg impedance.
NaHCO3	2.62	3.66 × 10^−3	0.73	4047	3.91 × 10^−3
Na2SO4	3.79	1.42 × 10^−3	0.79	1139	—
NaCl	2.09	1.55 × 10^−3	0.80	999.5	—

---

Many works have attempted to discover the relationship between E_FB and pitting. The apparent phenomenon that lowers E_FB and mitigates the susceptibility of pitting nucleation was reported.³⁰ The decrease in E_FB increases the band gap energy and pitting potential. Pagitsas et al.³¹ mentioned that accumulation of V_Fe²⁻ due to the incomplete annihilation in reaction (II) prevents condensation, and then a large void occurs due to the large number of V_Fe²⁻. The voids weaken the bonding strength of the film to the steel; as a result, pitting initiates. The vacancy condensation is proportional to the flux of cation vacancies.²⁰ The adsorbates of some aggressive anions, resulting in cation vacancy generation at the deep donor level, are mainly responsible for pitting.^32,33 Cl⁻ and SO₄²⁻ adsorption answer for the possible pitting.

The quantitative laws between electronic structure and general and pitting corrosion should be further put forward in the future.

4. Conclusions

In order to discover the influence of anions on the physicochemical performance of a FeCO₃ corrosion film, the film, composed of rhombohedral FeCO₃ crystals, was first prepared under high temperature and high pressure.

After exposure to the NaHCO₃ solution, the FeCO₃ film presented an n-type semiconductive character according to the Mott–Schottky measurement. The flat band potential and donor density were −0.31 V_SCE and 0.27 × 10²³ cm⁻³. The compact microstructure was maintained. Warburg impedance was observed when EIS was measured. The lowest corrosion current was obtained by fitting the potentiodynamic polarization curve.

In the Na₂SO₄ solution, a p-type semiconductor with a flat band potential of −0.26 V_SCE and acceptor density of 1.72 × 10²³ cm⁻³ was obtained for the FeCO₃ film. A loose microstructure was observed. The high conductive ion concentration elevated the corrosion current.

Similarly to the FeCO₃ film in Na₂SO₄ solution, p-type semiconducting properties were found in the NaCl solution, but the flat band potential and acceptor density were enhanced to 0.14 V_SCE and 17.2 × 10²³ cm⁻³. Also, the integrity of the microstructure was damaged. The highest corrosion current was presented.

Acknowledgements

The authors thank financial support by the open fund (PLN1306) of the State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation (Southwest Petroleum University), National Natural Science Foundation of China (51374180) and Key Lab of Material of Oil and Gas Field (X151514KCL24).

References

1. S. Nešić, Corros. Sci., 2007, 49, 4308.
2. I. S. Molchan, G. E. Thompson, R. Lindsay, P. Skeldon, V. Likodimos, G. Em. Romanos, P. Falaras, G. Adamova, B. Iliev and T. J. S. Schubert, RSC Adv., 2014, 4, 5300.
3. J. K. Heuer and J. F. Stubbins, Corros. Sci., 1999, 41, 1231.
4. J. B. Sun, G. A. Zhang, W. Liu and M. X. Lu, Corros. Sci., 2012, 57, 131.
5. Z. Yin, W. Zhao, Z. Bai and Y. Feng, Surf. Interface Anal., 2008, 40, 1231.
6. D. Liu, B. Qiu, Y. Tomoe, K. Bando and X. P. Guo, Mater. Corros., 2011, 62, 1153.
7. Y. Zhang, X. Pang, S. Qu, X. Li and K. Gao, Int. J. Greenhouse Gas Control, 2011, 5, 1643.
8. G. F. Lin, M. S. Zheng, Z. Q. Bai and Y. R. Feng, J. Iron Steel Res. Int., 2006, 13, 47.
9. K. Gao, F. Yu, X. Pang, G. Zhang, L. Qiao, W. Chu and X. Lu, Corros. Sci., 2008, 50, 2796.
10. J. Han and J. W. Carey, J. Appl. Electrochem., 2011, 41, 1367.
11. Z. Q. Bai, C. F. Chen, M. X. Lu and J. B. Li, Appl. Surf. Sci., 2006, 252, 7578.
12. G. A. Zhang and Y. F. Cheng, Electrochim. Acta, 2010, 55, 316.
13. M. C. L. Oliveira, V. S. M. Pereira, O. V. Correa, N. B. Lima and R. A. Antunes, Corros. Sci., 2013, 69, 311.
14. Z. Feng, X. Cheng, C. Dong, L. Xu and X. Li, J. Nucl. Mater., 2010, 407, 171.
15. Y. Zhang, M. Urquidi-Macdonald, G. R. Engelhardt and D. D. Macdonald, Electrochim. Acta, 2012, 69, 1.
16. J. K. Xiao, Chin. J. Geochem., 1990, 9, 169.
17. C. Ren, X. Wang, L. Liu, H. Yang and N. Xian, Mater. Corros., 2012, 63, 168.
18. M. Gao, X. Pang and K. Gao, Corros. Sci., 2011, 53, 557.
19. D. D. Macdonald, Electrochim. Acta, 2011, 56, 1761.
20. Z. Jiang, X. Dai, T. Norby and H. Middleton, Corros. Sci., 2011, 53, 815.
21. H. Guo, Y. Li and K. Zhao, J. Hazard. Mater., 2010, 176, 174.
22. B. Ter-Ovanessian, C. Alemany-Dumont and B. Normand, Electrochim. Acta, 2014, 133, 373.
23. M. B. Nemer, Y. Xiong, A. E. Ismail and J. H. Jang, Chem. Geol., 2011, 280, 26.
24. H. Tanaka, N. Hatanaka, M. Muguruma, T. Ishikawa and T. Nakayama, Corros. Sci., 2013, 66, 136.
25. S. Ningshen, U. K. Mudali, V. K. Mittal and H. S. Khatak, Corros. Sci., 2007, 49, 481.
26. J. P. Duthil, G. Mankowski and A. Guisti, Corros. Sci., 1996, 38, 1839.
27. B. Deng, Y. Jiang, J. Liao, Y. Hao, C. Zhong and J. Lin, Appl. Surf. Sci., 2007, 253, 7369.
28. G. Bianchi, A. Cerquetti, F. Mazza and S. Torchio, Corros. Sci., 1972, 12, 49.
29. J. Han, S. Nešić, Y. Yang and B. N. Brown, Electrochim. Acta, 2011, 56, 5396.
30. M. A. Pech-Canul, M. I. Pech-Canul, P. Bartolo-Pérez and M. Echeverrí, Electrochim. Acta, 2014, 140, 258.
31. M. Pagitsas, A. Diamantopoulou and D. Sazou, Electrochem. Commun., 2001, 3, 330.
32. S. Yang and D. D. Macdonald, Electrochim. Acta, 2007, 52, 1871.
33. Y. F. Cheng and J. L. Luo, Electrochim. Acta, 1999, 44, 2947.

---