You are on page 1of 8

Corrosion Science 57 (2012) 154161

Contents lists available at SciVerse ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

Characterization of passive lms on shape memory stainless steels


C.A. Della Rovere a,, J.H. Alano a, R. Silva a, P.A.P. Nascente a, J. Otubo b, S.E. Kuri a
a b

Department of Materials Engineering, Federal University of So Carlos, Rodovia Washington Luis, Km 235, 13565-905 So Carlos, SP, Brazil Division of Mechanical Engineering, Technological Institute of Aeronautics, Praa Marechal Eduardo Gomes, 50. Vila das Accias, 12228-900 So Jos dos Campos, SP, Brazil

a r t i c l e

i n f o

a b s t r a c t
Passive lms formed on three FeMnSiCrNi(Co) shape memory stainless steels (SMSSs) were studied based on polarization tests, EIS, XPS and MottSchottky analyses. The test results were compared with those of a type 304 austenitic stainless steel. The results indicated that silicon plays an important role in the passive lm properties of FeMnSiCrNi(Co) SMSSs. Anodic passive lms formed on Fe MnSiCrNi(Co) SMSSs are highly protective and consist of a chromium oxyhydroxide with incorporation of silicon in the chemical form of a silicate. MottSchottky analyses suggested that passive lms on FeMnSiCrNi(Co) SMSSs are less defective and thicker than those on austenitic stainless steel. 2012 Elsevier Ltd. All rights reserved.

Article history: Received 5 September 2011 Accepted 21 December 2011 Available online 3 January 2012 Keywords: A. Stainless steel C. Acid corrosion C. Passive lms B. Polarization B. EIS B. XPS

1. Introduction Since the discovery of the shape memory effect (SME) in FeMnSi alloys in the early 1980s [1], attention has focused strongly on this new class of shape memory alloys (SMAs) because they combine low production costs, excellent workability and good weldability. These alloys also show great promise as a possible alternative to expensive TiNi SMAs for constrained recovery applications in various industrial sectors, such as the construction, chemical and petrochemical industries [2]. The SME in FeMnSi SMAs results from the stress-induced c ? e martensite transformation and its reversion upon heating [13]. Despite their unique properties, FeMnSi alloys are currently used only in a few practical applications. Their poor corrosion resistance and low recoverable strains (less than approx. 2% without treatment) are the main limiting factors for the use of FeMnSi SMAs in engineering applications [4,5]. Much attention has focused on improving corrosion resistance through suitable additions of alloying elements while maintaining or even improving the SME [6,7]. In this context, chromium (Cr), nickel (Ni), cobalt (Co) and other elements have been added successfully to FeMnSi SMAs, thereby improving their SME and corrosion resistance [79]. The newly developed FeMnSiCrNi(Co) SMAs, called shape memory stainless steels (SMSSs), have been used as an alternative choice to welding in pipe couplings [10,11].

Nevertheless, the current literature contains very few detailed studies concerning the electrochemical corrosion behavior of FeMnSiCrNi(Co) SMSSs, and no attempt has been made to understand passive lms formed on these low chromiumFe alloys with substantial silicon (about 56 wt.%) content. In this work, the passive lm properties of three FeMnSiCrNi(Co) SMSSs were studied in a 0.5 M H2SO4 solution via polarization tests, electrochemical impedance spectroscopy (EIS), X-ray photoelectron spectroscopy (XPS) and MottSchottky analyses. The role of alloying elements in the composition and electrochemical characteristics of the passive lms is discussed here in relation to the corresponding passive lm on a standard corrosion-resistant alloy. 2. Material and methods FeMnSiCrNi(Co) SMSSs were prepared by the vacuum induction melting (VIM) process, using high-purity Fe, Mn, Si, Cr, Ni and Co. Table 1 describes the chemical composition of the alloys. Mn, Si, Cr, Ni, Co and Mo were determined by inductively coupled plasma optical emission spectrometry (ICPOES). C was determined by pyrolysis using a LECO CS-444 carbon determinator. The ingots were hot-forged at 1280 C into a bar shape, solution-treated at 1050 C for 1 h, and quenched in a water bath at room temperature (25 C). Samples were carefully machined from the bars for optical and scanning electron microscopy observation, electrochemical measurements and XPS analysis. For optical (OM) and scanning electron microscopy (SEM), the samples were carefully wet ground with up to #2000-grit silicon carbide (SiC) paper, polished with alumina 1 lm and then etched

Corresponding author. Tel.: +55 16 33518507; fax: +55 16 33518258.


E-mail address: carlosdrovere@hotmail.com (C.A. Della Rovere). 0010-938X/$ - see front matter 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.corsci.2011.12.022

C.A. Della Rovere et al. / Corrosion Science 57 (2012) 154161 Table 1 Chemical composition (mass%) of FeMnSiCrNi(Co) SMSSs and SS 304. Alloy SMSS A SMSS B SMSS C SS 304 Fe Bal. Bal. Bal. Bal. Mn 14.19 10.34 8.26 1.76 Si 5.30 5.31 5.25 0.39 Cr 8.81 9.92 12.80 19.28 Ni 4.65 4.87 5.81 7.64 Co 11.84 Mo 0.25 C 0.008 0.006 0.009 0.029

155

using Villelas reagent (95 ml of ethanol, 5 ml of hydrochloric acid and 1 g of picric acid) to reveal the microstructure. For electrochemical measurements, the samples were mounted in polyester resin after establishing the electric contact, taking special care to avoid the presence of crevices. The exposed area was 1.0 cm2. All the electrochemical measurements were taken at 25 C in a newly prepared naturally aerated solution of 0.5 M H2SO4. A standard three-electrode electrochemical cell was used with a platinum counter electrode and a saturated calomel reference electrode (SCE) connected to a Solartron system potentiostat (model 1287A)/frequency response analyzer (model 1260A). Three sets of measurements of each sample were taken and the average value was considered. Before the potentiodynamic polarization and EIS measurements, the samples were wet ground with #600-grit silicon carbide (SiC) paper, washed in distilled water, and immersed in the electrochemical cell. The samples were then subjected to open circuit conditions until a steady-state potential was reached. This procedure was accomplished in 30 min and the potential value obtained was considered the open-circuit potential (EOC). Potentiodynamic polarization was carried out at sweep rates of 1 mV/s, starting from a potential of 200 mV below EOC to a potential of 1200 mVSCE. The EIS measurements were taken with a 10 mV amplitude signal at an applied frequency ranging from 100 kHz to 1 mHz. The EIS data were collected in two experimental conditions: (1) after EOC stabilization, and (2) at 500 mVSCE in the passive region after 1 h of anodic passive lm growth, as described below. Prior to the XPS and MottSchottky analyses, an anodic passive lm was grown according to the following procedure: (1) the samples were polished mechanically to a mirror nish using alumina 1 lm, (2) cathodically polarized at 800 mVSCE for 5 min, and (3) passivated at a constant potential of 500 mVSCE in 0.5 M H2SO4 solution for 1 h to attain a steady state, corresponding to a nearly constant thickness of the passive lm. After polarization, the samples were immediately dipped in deionized water, dried with nitrogen gas and transferred to an ultra-high vacuum (low 107 Pa range). This procedure was performed in a laboratory atmosphere. The XPS analysis was performed in a Kratos Analytical spectrometer (model XSAM HS) using a non-monochromatic Mg Ka (hm = 1253.6 eV) X-ray source with an emission current of 5 mA and a voltage of 12 kV. No sputter-cleaning was performed prior to analysis. Gaussian and mixed Gaussian/Lorentzian functions, the Shirley background subtraction method, and a leastsquares routine were used for tting. The binding energies were referenced to the adventitious hydrocarbon C 1 s level set at 284.8 eV. The sensitivity factors for quantitative analysis were referenced to SF1s = 1.0. MottSchottky plots were obtained at a frequency of 1 kHz, using an amplitude signal of 10 mV and a step rate of 25 mV in the cathodic direction. To compare the electrochemical behavior of the FeMnSiCr Ni(Co) SMSSs with that of a standard corrosion-resistant alloy, samples of type-304 (SS 304) austenitic stainless steel were solution-treated at 1050 C for 1 h, quenched in a water bath and used in the electrochemical measurements.

Fig. 1. Representative micrographs of FeMnSiCrNi SMSSs: (a) OM and (b) SEM images.

3. Results and discussion Fig. 1a shows a representative optical micrograph of the Fe MnSiCrNiCo SMSSs. The microstructure of these alloys is composed mainly of austenite (c), some twins inside austenite grains and stacks of e martensite plates, as illustrated in the SEM micrograph in Fig. 1b. Maji et al. [12] also reported similar microstructures in FeMnSiCrNi shape memory alloys.

Fig. 2. Potentiodynamic polarization curves obtained from FeMnSiCrNi(Co) SMSSs and SS 304 in 0.5 M H2SO4 solution.

156

C.A. Della Rovere et al. / Corrosion Science 57 (2012) 154161

The EOC of the alloys was monitored for 30 min after immersion in 0.5 M H2SO4 solution. All the alloys initially showed a rapid change in potential towards a more noble potential, followed by an almost linear increase until the potential stabilized. The EOC values of the alloys were in the order of SS 304 (356 mVSCE) > SMSS C (417 mVSCE) > SMSS B (445 mVSCE) > SMSS A (464 mVSCE), indicating that the FeMnSiCrNi(Co) SMSSs are more active than the SS 304. Fig. 2 shows the potentiodynamic polarization curves of the FeMnSiCrNi(Co) SMSSs and SS 304 in 0.5 M H2SO4 solution. Note that the FeMnSiCrNi(Co) SMSSs and SS 304 showed a very different anodic behavior in the active dissolution region, which may be attributed to the chemical composition and microstructure of the alloys, although their respective curves showed similar passive regions. A more detailed study of the polarization curves of FeMnSiCrNi(Co) SMSSs can be found elsewhere [13]. In view of earlier studies [1215], the protectiveness of passive lms on FeMnSiCrNi(Co) SMSSs appears to be fairly high considering the low Cr content (913 wt.%) of these alloys. This behavior has been ascribed to the addition of Si [1214]. However, this assumption has not been conrmed and requires clarication. Because the protectiveness of passive lms on stainless steels is closely associated with the lms structure, thickness, composition, and semiconducting properties, the passive lms formed on FeMnSiCrNi(Co) SMSSs were analyzed by EIS measurements, XPS and MottSchottky plots.

Fig. 3 shows the Bode plots of the FeMnSiCrNi(Co) SMSSs and SS 304 in 0.5 M H2SO4 solution at their respective EOC and at 500 mVSCE. It is evident that the FeMnSiCrNi(Co) SMSSs at the EOC present three distinct regions, as follows: (1) In the region of high frequencies (between 1 kHz and 100 kHz), they exhibit a |Z| value of about 1.5 X cm2 and values of phase angle close to 0, indicating that their impedance is dominated by the electrolyte resistance. (2) In a small range in the region of intermediate frequencies, the Bode magnitude displays a linear slope that is close to 1 and the values of phase angles reach their maximum values, which are characteristic responses of a capacitive behavior. (3) At frequencies of 0.110 Hz, the |Z| values are in the order of 10100 X cm2 and the |Z| vs. frequency is constant and is associated with phase angle values lower than 0, indicating that their impedance is very low. Note that, for the FeMnSiCrNi(Co) SMSSs, the values of |Z| at a xed frequency of 0.1 Hz (which usually corresponds to the polarization resistance of the alloys) lie between 6 and 110 X cm2. In contrast, at the same frequency, the |Z| value of SS 304 is about 20 k X cm2, which indicates that the corrosion resistance of FeMnSiCrNi(Co) SMSSs is much lower than that of SS 304. On the other hand, FeMnSiCrNi(Co) SMSSs at 500 mVSCE show an almost ideal capacitive behavior inside a broad frequency range (between 0.1 Hz and 5 kHz) with values of phase angles close to 90 and high values of |Z| in the region of lower frequencies. With regard to the Bode magnitude plots, it can be seen that the values of |Z| at a xed frequency of 0.01 Hz lie in

Fig. 3. Bode plots obtained from FeMnSiCrNi(Co) SMSSs and SS 304 in 0.5 M H2SO4 solution: (a) at the EOC and (b) at 500 mVSCE.

Fig. 4. Nyquist plots obtained from FeMnSiCrNi(Co) SMSSs and SS 304 in 0.5 M H2SO4 solution: (a) at the EOC and (b) at 500 mVSCE.

C.A. Della Rovere et al. / Corrosion Science 57 (2012) 154161

157

the order of 80350 k X cm2 and 200 k X cm2 for FeMnSiCrNi (Co) SMSSs and SS 304, respectively. These results suggest that the passive lms formed on FeMnSiCrNi(Co) SMSSs offer high corrosion resistance. Despite the signicant differences between the Bode plots of FeMnSiCrNi(Co) SMSSs and SS 304 at their respective EOC, these plots did not present signicant differences at 500 mVSCE in either type of material. However, the Nyquist plots in Fig. 4 shows different capacitive semi-arcs, i.e., at the EOC, the FeMnSiCrNi (Co) SMSSs showed capacitive semi-arcs with much smaller diameters than that of SS 304. In contrast, at 500 mVSCE, the diameters of the capacitive semi-arcs of FeMnSiCrNi(Co) SMSSs were the same as that of SS 304, indicating that the protection afforded by the passive lms formed on them is similar to that of SS 304. It can also be noted that, in both experimental conditions, the diameters of the capacitive semi-arcs of FeMnSiCrNi(Co) SMSSs depend on the alloys Cr and Mn contents. To obtain quantitative conrmation of the experimental EIS data, the EIS plots were analyzed using the complex and single equivalent electric circuits (EEC) presented in Fig. 5. For FeMn SiCrNi(Co) SMSSs at the EOC, a good t was obtained by the single Randles EEC. The use of a constant phase element (CPE) instead of an ideal capacitor was necessary due to the distribution of relaxation times associated with heterogeneities (roughness and surface defects) at the surface of the electrodes. The impedance of a CPE is given by ZCPE = [Q (jx)n]1, where Q is a parameter related with the interfacial capacitance by Eq. (1) [16]; j is the current; x is the frequency and n is an adjustable parameter that always lies between 0.5 and 1. When n = 1, the CPE describes an ideal capacitor. For 0.5 < n < 1, the CPE describes a distribution of dielectric relaxation times in the frequency domain, and when n = 0.5, the CPE represents a Warburg impedance with a diffusional character [17]. At the EOC, the FeMnSiCrNi(Co) SMSSs are in the active state, so R1 and CPE1 are associated with the charge transfer resistance and double layer capacitance of the metal/electrolyte interface, respectively.
1 =n 1 1 n1=n C1 Q 1 R e R1

polarization resistance (the sum of R1 and R2) of the passive lms formed on FeMnSiCrNi(Co) SMSSs at 500 mVSCE is similar to that of SS 304. Considering that the corrosion protection afforded by a layer is directly related to its resistivity, it is concluded from the tted values of EIS plots at 500 mVSCE that the Cr-enriched barrier layer is the one that renders SS 304 and FeMn SiCrNi(Co) SMSSs corrosion resistant. Taking into account the lower Cr content (913 wt.%) of FeMnSiCrNi(Co) SMSSs than that of SS 304, the high resistance of their barrier layer suggests the incorporation of Si, since the Si oxide has a high resistivity [20]. The low values of resistance of the outer layer (R1) can be attributed to selective dissolution of Fe and to the low stability of Fe oxyhydroxides in acid solution [18,21], which leads to the formation of an outer layer with numerous defects and offers no protection against corrosion. The capacitance (C2) values obtained from EIS plots at 500 mVSCE can be used to estimate the passive lm thickness, based on Eq. (2), which is valid for the parallel plate capacitor model of a homogenous oxide layer [22]:

d2

e0 er 107 nm
C2 1 cm

On the other hand, for SS 304 at the EOC and the other EIS measurements at 500 mVSCE after growth of the passive lm, the materials are in the passive state, so the tting was carried out using a complex EEC (Fig. 5b) that takes into account the bilayer structure of the passive lm [18]. This EEC yielded the best tting parameters and capacitance values consistent with the passive lm thickness. The physical signicance of the elements of the complex EEC can be described as follows: the R1//CPE1 combination represents the properties of the outer layer/electrolyte interface and the R2// C2 combination represents the barrier layer properties. The term Re represents the electrolyte resistance in both EECs. Table 2 presents the numerical values determined for all the parameters of the EECs. Note that the charge transfer resistance (R1) of FeMnSiCrNi(Co) SMSSs at the EOC is very low, indicating that the corrosion resistance of these alloys is low in acid media. This behavior can be attributed to the low Cr and high Mn contents in the FeMnSiCrNi(Co) SMSSs, as discussed in more detail in previous works [13,19]. On the other hand, the

where d2 is the thickness of barrier layer, er is the dielectric constant of the passive lm, e0 is the vacuum permittivity (8.854 1014 F cm1) and C2 is the capacitance (in F cm2). However, it should be noted that is difcult to obtain an accurate thickness value of the passive lm when the dielectric constant is not well established, although a value of 15.6 has been used [23]. Nevertheless, since the capacitance is inversely proportional to the layer thickness, the capacitive response of the alloys can give an indication of how the thickness of the passive lm changes. In this sense, it can be observed that the capacitance (C2) values decreased in the following order: SS 304 > SMSS A > SMSS B > SMSS C, suggesting that the passive lm thickness increases in the inverse order. This behavior can also be attributed to the incorporation of Si in the passive lms formed on FeMnSiCrNi(Co) SMSSs. As the Pilling-Bedworth ratio of Si and its oxide is greater than that of Fe, Cr and their oxides, the incorporation of Si promotes a thickening of the oxide lm [24]. With respect to the outer layer, it can be noted that its thickness is at least one order of magnitude smaller than the thickness of the inner layer, since the values of Q1 are numerically almost 10 to 20 times higher than C2. However, a more accurate calculation of the thickness of the outer layer requires determining the interfacial capacitance from the parameter Q1, using Eq. (1). Fig. 6(a) shows the XPS survey spectra of the passive lms formed in 1 h on FeMnSiCrNi(Co) SMSSs at 500 mVSCE in 0.5 M H2SO4 solution. The Fe, Cr, Si, Mn, Ni, Co, C and O XPS core level peaks are easily identiable. To elucidate the nature of the FeCr oxide and the role of alloying elements in the passive lms, high-resolution spectra were acquired for O and C 1 s ionizations, as well as for Fe, Cr, Si, Mn, Ni and Co 2p ionizations. Fig. 6(b) shows high-resolution spectra of Fe 2p. The Fe 2p3/2 ionization reveals the presence of metallic Fe (706.8 eV), Fe(II) oxyhydroxides (709.1 eV) and Fe(III) oxides (711.2 eV) and hydroxides (712.5 eV).

Fig. 5. Equivalent electric circuits (EEC) used to analyze the EIS plots: (a) single Randles EEC and (b) complex EEC.

158

C.A. Della Rovere et al. / Corrosion Science 57 (2012) 154161

Table 2 Values of impedance parameters determined for FeMnSiCrNi(Co) SMSSs and SS 304 at the EOC and at 500 mVSCE in 0.5 M H2SO4 solution. Alloy SMSS A SMSS B SMSS C SS 304 SMSS A SMSS B SMSS C SS 304 Potential EOC Re (X cm2) 1.27 1.05 1.36 2.34 2.01 1.36 2.25 2.26 Q1 ((lF cm2)1/n) 2.10 10 6.27 102 2.08 102 68.15 55.08 37.10 29.88 28.64
3

n 0.852 0.892 0.893 0.895 0.855 0.885 0.895 0.906

R1 (X cm2) 5.76 11.15 107.10 9.03 26.33 16.75 13.33 9.22

C2 (lF cm2) 2.98 2.15 2.12 1.40 2.64

R2 (k X cm2) 61.57 79.34 188.25 379.12 211.41

(v2) 2.7 104 2.3 104 4.4 104 8.7 104 2.7 103 4.3 103 3.1 103 1.8 103

500 mVSCE

Fig. 6. (a) XPS survey spectra, (b) high-resolution XPS spectra of the Fe 2p, (c) Cr 2p and (d) Si 2p of the passive lms formed on FeMnSiCrNi(Co) SMSSs at 500 mVSCE in 0.5 M H2SO4 solution for 1 h.

Fig. 6(c) shows high-resolution spectra of Cr 2p. Note that the Cr 2p3/2 ionization consists mainly of two peaks: the one at the lowest binding energy (574.0 eV) corresponds to metallic Cr, whereas the one at approximately 577 eV is attributed to the presence of Cr(III) oxides (576 eV) and hydroxides (577.5 eV). These spectra are in good agreement with the literature for passive lms on FeCr alloys [25]. The chemical state of Si in the passive lms is not well established. While some authors [26,27] have reported the presence of Si(IV) (at 103.5 eV) in the passive lm only as a layer of SiO2, Ioka et al. [28] observed Si in more than one chemical state, forming a passive lm composed of SiO2 and SiO. These authors attributed the binding energy of 102.3 eV to Si(II) (SiO). However, Robin et al. [29] pointed out that Si is present as Si(IV), but in the

chemical form of an FeCr mixed silicate (SiO4 4 ) and not as SiO2. Fig. 6(d) shows high-resolution spectra of Si 2p. Note that the deconvolution of these spectra reveals two main peaks, one corresponding to metallic Si at approximately 99.1 eV and the other to oxidized Si at 101.9 eV, the same binding energy as that reported by Robin et al. [29]. The chemical states of Mn were obtained from three peaks in Mn 2p3/2 ionization, which were attributed to metallic Mn (638.6 eV), Mn(II) (642.0 eV) and Mn(IV) (644.4 eV) [30]. The Ni 2p3/2 ionization did not reveal the presence of Ni in oxidized form, but only in metallic form (852.8 eV). For the FeMnSiCrNi(Co) SMSS C, the chemical states of Co were obtained from two peaks in Co 2p3/2 ionization. The rst peak (778.2 eV) was attributed to metallic Co and the second (in the range of 780.5781.2 eV) to

C.A. Della Rovere et al. / Corrosion Science 57 (2012) 154161 Table 3 Cationic composition of the passive lms (at.%) formed in 1 h on FeMnSiCrNi (Co) SMSSs at 500 mVSCE in 0.5 M H2SO4 solution. Alloy SMSS A SMSS B SMSS C Fe(II) + Fe(III) 28.66 26.77 17.14 Mn(II) + Mn(IV) 3.35 1.32 1.29 Si(IV) 37.85 41.45 41.21 Cr(III) 30.14 30.46 37.37 Co(II) + Co(III) 2.99

159

density of charge carriers (NA for acceptors and ND for donors), E is the applied potential, EFB is the at band potential, kB is Boltzmanns constant (1.38 1023 J K1), and T is the absolute temperature. The space charge capacitance, CSC, is obtained from C = 1/ xZ00 , where x is the angular frequency and Z00 is the imaginary part of the impedance. The donor density (ND) is determined from the slope of the experimental 1/CSC2 vs. E plots, the at band potential (EFB) by the extrapolation to 1/CSC2 = 0, and the thickness of the space charge layer (dSC) is calculated by Eq. (4) for an n-type semiconductor [36]:

Co(II) and Co(III). The oxidation states of cobalt were not clearly identied due to the negligible difference in the binding energies between various co-oxides, such as CoO, Co2O3 and Co3O4 [31]. The elements found in the metallic state were attributed to the metal substrate and the chemical composition of the passive lm was determined considering the Fe, Mn, Si, Cr and Co cations and calculating their respective cationic percentages. Table 3 shows the cationic composition (at.%) of the passive lms formed in 1 h on FeMnSiCrNi(Co) SMSSs at 500 mVSCE in 0.5 M H2SO4 solution. The results indicate that the passive lms on FeMn SiCrNi(Co) SMSSs are composed mainly of Si(IV), Cr(III) and Fe(II) + Fe(III). Note that the Fe content in the lms was lower than that of the metal substrates, indicating a preferential dissolution of this element. However, the Si content (about 40 at.%) in the lms was nearly fourfold higher than in the metal matrixes (10 at.%), and the Cr content in the lms was nearly threefold that of the alloy matrixes, indicating a remarkable enrichment of these elements in the passive lms. Mn and Co contents in the lms were much lower than in the metal substrates, also indicating a preferential dissolution of these elements. According to the literature [32], the corrosion resistance of stainless steels and high-chromium alloys (Cr > 12 wt.%) in H2SO4 solution is attributed to the formation of a passive lm that is composed mainly of hydrated Cr(III) oxyhydroxides. Olsson and Landolt [18] reported that the Cr(III) content in passive lms on stainless steel normally amounts to 5070 at.% in acid solution. On the other hand, the passive lms formed on low-chromium Fe-alloys (Cr < 10 wt.%) consist mainly of Fe oxyhydroxide, and the protectiveness of these lms is not as high as that formed on FeMnSiCrNi(Co) SMSSs [1215]. Therefore, the XPS results are interpreted as evidence that the high protectiveness of passive lms formed anodically on FeMnSiCrNi(Co) SMSSs in 0.5 M H2SO4 solution results from a protective lm consisting of a Cr(III) oxyhydroxide with Si(IV) incorporated in the chemical form of a silicate, which may also be responsible for the changes in the electronic properties, as shown below. However, further analysis is necessary to determine the exact nature of this silicate. It is generally accepted that the corrosion resistance of a material is related to the electronic properties of its passive lm [33]. Thus, to gain a better understanding of the electrochemical corrosion behavior of FeMnSiCrNi(Co) SMSSs it is very important to evaluate the electronic properties of their passive lms. The analysis of the curves of capacitance as a function of the electrode potential based on the MottSchottky theory is the most common method for determining the semiconductive behavior and the density of charge carriers in passive lms [23,34,35]. According to this theory, the space charge capacitance of a p-type or n-type semiconductor is given by Eq. (3):

dSC

 1 2er e0 KBT 2 E EFB q qN D

Fig. 7 shows the MottSchottky plots for the passive lms formed in 1 h on FeMnSiCrNi(Co) SMSSs and on SS 304 at 500 mVSCE in a 0.5 M H2SO4 solution. Like the SS 304, the passive lms formed on FeMnSiCrNi(Co) SMSSs behave as n-type and p-type semiconductors above and below the at band potential, respectively. However, the electronic behavior of the passive lms differed noticeably in terms of the at band potential (EFB), the donor density (ND), and the thickness of the space charge layer (dSC). Table 4 presents the values of EFB, ND and dSC determined from the MottSchottky plots. Note that the ND of the FeMnSiCrNi (Co) SMSSs is lower than that of SS 304, indicating that the ionic and electronic conductivity of the passive lms on FeMnSiCr Ni(Co) SMSSs is lower than those of SS 304. Moreover, the passive lms on FeMnSiCrNi(Co) SMSSs were also found to be thicker than that on SS 304 because the thickness of the space charge layer is directly associated with the thickness of the passive lm [36,37]. These results are in good agreement with the EIS measurements and with the polarization curves.

Fig. 7. MottSchottky plots for the passive lms formed on FeMnSiCrNi(Co) SMSSs and the SS 304 at 500 mVSCE in 0.5 M H2SO4 solution for 1 h.

1 C2 SC

  2 KBT E EFB q er e0 qNq

Table 4 MottSchottky parameters for the passive lms formed on FeMnSiCrNi(Co) SMSSs and on SS 304. Alloy SMSS A SMSS B SMSS C SS 304 EFB (mVSCE) 105.8 128.1 96.6 47.0 ND (1020 cm3) 6.6 6.6 5.4 16.3 dSC (108 cm) 9.85 9.52 10.97 6.73

where CSC is the space charge capacitance of a passive lm, er is the dielectric constant of the passive lm (which is assumed to be 15.6 [23]), e0 is the vacuum permittivity (8.854 1014 F cm1), q is the elementary charge (+e for electrons and e for holes), Nq is the

160

C.A. Della Rovere et al. / Corrosion Science 57 (2012) 154161

Fig. 8. Schematic representation of the replacement of two Fe(III) cations by two Si(IV) cations, eliminating one oxygen vacancy.

According to the literature [18,35], passive lms formed on stainless steel in H2SO4 solutions can be described as a bilayer structure: a highly hydrated outer layer of Fe oxyhydroxides and an anhydrous inner layer of FeCr mixed oxides with a spinel structure. It is also known that this spinel is rich in Cr because Fe concentrates preferentially near the outer oxide surface. Moreover, it has been generally assumed that the Cr-rich inner layer has a ptype semiconductive behavior, while the Fe-rich outer layer has an n-type semiconductive behavior, which explains the dual semiconducting properties of the passive lms formed on FeMnSiCr Ni(Co) SMSSs and on SS 304. As for the donor density of the two types of alloys, Table 4 shows that the ND of FeMnSiCrNi(Co) SMSSs is almost three times lower than that of SS 304. Considering that the electron donors in semiconducting passive lms are non-stoichiometric defects in the space charge region (such as oxygen ion vacancies and/or cation interstitials), high ND values are strong indicators that the passive lm is highly disordered or non-stoichiometric, and a decrease in ND means a depletion of these types of defects [38,39]. The XPS results indicated that Si(IV) cations were incorporated in the passive lms formed on FeMnSiCrNi(Co) SMSSs. The fact that the donor density decreased (in comparison to that of SS 304) suggests that the incorporation of Si(IV) cations suppresses the formation of additional cation interstitials and/or increases the number of oxygen ions, which cancel out the oxygen ion vacancies and make the lms less defective and hence more protective. Fig. 8 shows a schematic representation of the replacement of two Fe(III) cations by two Si(IV) cations, eliminating one oxygen vacancy to maintain electroneutrality in the passive lm. 4. Conclusions Anodic passive lms formed on FeMnSiCrNi(Co) SMSSs in a 0.5 M H2SO4 solution were characterized in the present work. The conclusions drawn from this study are as follows: 1. The polarization curves suggested that the passive regions of FeMnSiCrNi(Co) SMSSs and SS 304 were similar. 2. The EIS measurements indicated that the protectiveness of the anodic passive lms formed on FeMnSiCrNi(Co) SMSSs is similar to that of SS 304.

3. The XPS analysis revealed that a considerably high Si content (about 40 at.%) is present in the anodic passive lms formed on FeMnSiCrNi(Co) SMSSs. 4. The high protectiveness of anodic passive lms formed on Fe MnSiCrNi(Co) SMSSs results from a protective lm consisting of a Cr(III) oxyhydroxide with incorporation of Si(IV) in the chemical form of a silicate. 5. The MottSchottky analysis suggested that Si(IV) acts as a dopant in the anodic passive lms formed on FeMnSiCrNi(Co) SMSSs, making them less defective and thicker than the lms formed on SS 304.

Acknowledgements The authors gratefully acknowledge PPGCEM/UFSCar (Postgraduate Program in Materials Science and Engineering of the Federal University of So Carlos) and the Brazilian research funding agency CNPq (National Council for Scientic and Technological Development) for their nancial support of this work. References
[1] A. Sato, E. Chishima, K. Soma, T. Mori, Acta Metall. 30 (1982) 11771183. [2] K. Otsuka, C.M. Wayman (Eds.), Shape Memory Materials, Cambridge University Press, Cambridge, 1988. [3] S. Kajiwara, Mater. Sci. Eng. A 273275 (1999) 6788. [4] H. Li, D. Dunne, ISIJ Int. 37 (1997) 605609. [5] D.F. Wang, Y.R. Chen, F.Y. Gong, D.Z. Liu, W.X. Liu, J. Phys. IV, Colloque C8 (1995) 527530. [6] H. Otsuka, H. Yamada, T. Maruyama, H. Tanahashi, S. Matsuda, M. Murakami, ISIJ Int. 30 (1990) 674679. [7] L.J. Rong, D.H. Ping, Y.Y. Li, C.X. Shi, Scripta Metallurgica et Materialia 32 (1995) 19051909. [8] X. Huang, S. Chen, T.Y. Hsu, X. Zuyao, J. Mater. Sci. 39 (2004) 68576859. [9] H.B. Peng, Y.H. Wen, B.B. Ye, N. Li, Mater. Sci. Eng. A 504 (2009) 3639. [10] D.Z. Liu, D.F. Wang, W.Y. Ji, W.X. Liu, Proceedings of the 2nd International Conference on Shape Memory and Superelasitc Technologies, California, 1997, p. 329. [11] J.C. Li, X.X. L, Q. Jiang, ISIJ Int. 40 (2000) 11241126. [12] B.C. Maji, C.M. Das, M. Krishnan, R.K. Ray, Corros. Sci. 48 (2006) 937949. [13] C.A. Della Rovere, J.H. Alano, J. Otubo, S.E. Kuri, J. Alloys Compd. 509 (2011) 53765380. [14] H. Moriya, S. Kimura, K.P. Ishizaki, K.P. Hashizume, S. Suzuki, H. Suzuki, T. Sampei, J. Phys. IV, Colloque C4 (1991) 433437. [15] O.P. Modi, M.N. Mungole, K.P. Singh, Corros. Sci. 30 (1990) 941947.

C.A. Della Rovere et al. / Corrosion Science 57 (2012) 154161 [16] G.J. Brug, A.L.G. van den Eeden, M. Sluyters-Rehbach, J.H. Sluyters, J. Electroanal, Chem. Interfac. 176 (1984) 275295. [17] A.K. Shukla, R. Balasubramaniam, Corros. Sci. 48 (2006) 16961720. [18] C.-O.A. Olsson, D. Landolt, Electrochim. Acta 48 (2003) 10931104. [19] Y.S. Zhang, X.M. Zhu, S.H. Zhong, Corros. Sci. 46 (2004) 853876. [20] J.F. Shackelford, W. Alexander (Eds.), CRC Materials Science and Engineering Handbook, 3rd ed., CRC Press, Boca Raton, 2001. [21] S. Haupt, H.H. Strehblow, Corros. Sci. 37 (1995) 4354. [22] Z. Feng, X. Cheng, C. Dong, L. Xu, X. Li, Corros. Sci. 52 (2010) 36463653. [23] A. Di Paola, Electrochim. Acta 34 (1989) 203210. [24] E. McCafferty, Introduction to Corrosion Science, Springer, New York, 2010. [25] P. Marcus, I. Olefjord, Corros. Sci. 28 (1988) 589602. [26] H. Kajimura, N. Usuki, H. Nagano, Proceedings of the Symposium on Passivity and its Breakdown, Paris, 1997, p. 332. [27] K. Hio, T. Adachi, T. Yamada, Y. Tsuchida, K. Nakajima, Y. Hosoi, Mater. Trans. 42 (8) (2001) 17231730. [28] I. Ioka, J. Mori, C. kato, M. Futakawa, K. Onuki, J. Mater. Sci. Lett. 18 (1999) 14971499.

161

[29] R. Robin, F. Miserque, V. Spagnol, J. Nucl. Mater. 375 (2008) 6571. [30] W. Fredriksson, K. Edstrm, C.-O.A. Olsson, Corros. Sci. 52 (2010) 25052510. eu  niene [31] S. Surviliene , V. Jasulaitiene , A. C , A. Lisowska-Oleksiak, Solid State Ionics 179 (2008) 222227. [32] K. Asami, K. Hashimoto, S. Shimodaira, Corros. Sci. 18 (1978) 151160. [33] J.W. Schultze, M.M. Lohrengel, Electrochim. Acta 45 (2000) 24992513. [34] N.E. Hakiki, S. Boudin, B. Rondot, M. Da Cunha Belo, Corros. Sci. 37 (1995) 18091822. [35] M.J. Carmezim, A.M. Simes, M.F. Montemor, M. Da Cunha Belo, Corros. Sci. 47 (2005) 581591. [36] V.A. Alves, C.M.A. Brett, Electrochim. Acta 47 (2002) 20812091. [37] J. Xu, L. Liu, X. Lu, S. Jiang, Electrochem. Commun. 13 (2011) 102105. [38] G. Goodlet, S. Faty, S. Cardoso, P.P. Freitas, A.M.P. Simes, M.G.S. Ferreira, M. Da Cunha Belo, Corros. Sci. 46 (2004) 14791499. [39] A. Fattah-alhosseini, F. Soltani, F. Shirsalimi, B. Ezadi, N. Attarzadeh, Corros. Sci. 53 (2011) 31863192.

You might also like