Pitting initiation on 304 stainless steel in a chloride-contaminated pore solution under alternating temperature conditions

2.1 Materials

304 stainless steel samples with a size of Ф 10 mm × 10 mm were taken, and a copper wire was welded to one end of the samples. The samples were sealed by applying epoxy resin, and the other ends were left exposed for electrochemical measurements. Before testing, the exposed surface was ordinally ground to grade 1200 with emery paper, washed with alcohol and deionized water. For avoiding crevice corrosion, the edges between the exposed stainless steel surface and epoxy resin coating were covered by applying the silica gel. Saturated calcium hydroxide (Ca(OH)₂) solutions (Feng et al. 2014; Lu et al. 2019) or cement extracts (Feng et al. 2011a, b) were widely applied to simulate the concrete pore solutions. In the present study, 0.1 mol/L sodium hydroxide (NaOH) solution (pH 13) was used to simulate the pore solutions (Abreu et al. 2004a, b, 2006; Deus et al. 2012), to eliminate the change in Ca(OH)₂ solubility and the variation in solution pH at different temperatures. Pore solutions containing 0.5 mol/L NaCl were prepared to accelerate the pitting initiation on the stainless steel.

2.2 Experimental procedures

In the Cl⁻ contaminated pore solutions, different temperature conditions, including constant room temperature (22 °C), 50 °C, 60 °C, 70 °C and 22–50 °C, 22–60 °C and 22–70 °C alternating temperature cycles, were chosen to investigate the influence of alternating temperature on pitting initiation on the 304 stainless steel grades. The alternating temperature cycles were conducted by utilizing in-house equipment (Figure 1). In the alternating temperature cycles, the pore solutions were continuously cooled for 1 h and then were maintained at a higher temperature for 1 h (Figure 2), which was defined as one alternating temperature cycle (2 h). The average heating and cooling rates of the pore solutions in alternating temperature cycles were 1.5 °C/min and 1.26 °C/min, respectively.

Figure 1:

A schematic illustration of the alternating temperature system.

Figure 2:

Cycles of the alternating temperature of the pore solution.

For electrochemical tests, a 304 stainless steel sample was used as the working electrode, a saturated silver/silver chloride (Ag/AgCl) electrode and a carbon electrode were adopted as the reference electrode and counter electrode, respectively. Electrochemical impedance spectroscopy (EIS) and Mott–Schottky plots were performed using a Gamry Potentiostat (Gamry instrument) after the samples were exposed to the pore solution for 48 h (24 cycles in the alternating temperature solution). Electrochemical tests were performed at the low temperature (22 °C) for the samples which experienced the alternating temperature cycles, and the same measurements were carried out at the constant temperatures for the ones designed in various constant temperature conditions. EIS measurements were carried out by applying an alternating current disturbance signal of 10 mV versus an open circuit potential, in the range 10⁵ Hz to 10⁻² Hz. Identical to earlier studies (Feng et al. 2014, 2016), Mott–Schottky plots were tested at 1000 Hz from −1.2 to 1.2 V_Ag/AgCl, with a potential step of 50 mV. Every electrochemical measurement was tested on three samples under each temperature condition. Moreover, the surface characterization of the stainless steel samples was examined using a scanning electron microscope (SEM, ZEISS), and the surfaces of the samples were analyzed using a PHI-5702 X-ray photoelectron spectrometer (PerkinElmer, USA), following immersion for 48 h in the Cl⁻ contaminated pore solutions at various temperatures.

3.1 Electrochemical impedance spectroscopy

Figure 3 displays the EIS of 304 stainless steel grades in the Cl⁻ contaminated pore solutions at various temperatures. The impedances of the samples significantly decreased with temperature in the constant temperature pore solutions. The impedance of the sample at 70 °C was almost two orders of magnitude lower than that of its counterpart at 22 °C (Figure 3(b)). For the samples that experienced the alternating temperature cycles, the impedance values dramatically declined with the amplitude of the alternating temperature. For example, the impedance of the sample in the 22–70 °C alternating temperature solution was about 1/10 that of the one at the 22–50 °C temperature cycle. Simultaneously, in comparison to the stainless steel in the constant temperature solutions, the impedance of the sample experienced the alternating temperature was closer to that of the one in solution at the higher temperature of the temperature cycle. The impedance of the sample in the 22–70 °C pore solution was close to that of its counterpart at 70 °C, while it was lower than that of the sample at 22 °C (Figure 3(b)).

Figure 3:

EIS results of AISI 304 stainless steel in 0.1 mol/L NaOH + 0.5 mol/L NaCl solution at various temperature conditions: (a) Nyquist plots, (b) and (d) Bode plots.

An equivalent circuit (Figure 4) was introduced to analyze the EIS results, where R_s represents the solution resistance; R_mp is the resistance of defects (micropores) in the passive film; R_p is the polarization resistance and accounts for the oxidation process in the defects of the passive films. CPE_f and CPE_dl are the capacitances of the passive films and double layer capacitances of the oxidation process in the film defects, respectively (Feng et al. 2011a, b, 2019a, b; Gadadhar and Balasubramaniam 2008; Mohammadi et al. 2011). The tested EIS data excellently fitted the equivalent electrical circuit (Figure 3).

Figure 4:

Equivalent electrical circuit for modeling the impedance data.

Table 1 shows the fitting values of the parameters, and Figure 5 displays the fitting passive film resistances and polarization resistances. The R_mp results demonstrated that passive films were more severely degraded by the higher temperature or higher amplitude of the alternating temperature. The R_p values decreased with the temperature at constant temperature conditions, while those of their counterparts under alternating temperature conditions decreased with the amplitude of the temperature change. This situation indicated that the activity of stainless steel increased with temperature and the amplitude of the temperature change. Simultaneously, the R_p of the samples that experienced the alternating temperature was even lower than that of the one at the higher temperature of the alternating temperature cycle. For example, the R_p of the sample at 20–50 °C alternating temperature was noticeably lower than that of its counterpart at 50 °C. The fitting results of R_mp and R_p appear to indicate that the alternating temperature showed a more significant degradation on the passive film when compared with the higher constant temperature in the alternating temperature cycle (Table 1).

Figure 5:

Fitting results of EIS data for AISI 304 stainless steel immersed into 0.1 mol/L NaOH + 0.5 mol/L NaCl solution for 48 h under various temperature conditions.

According to previous studies, the thickness of the passive film (d) was a parameter for assessing its protective effect, which can be calculated from its capacitance (Feng et al. 2019a, b; Gadadhar and Balasubramaniam 2008), as follows:

where ε₀is the dielectric constant in a vacuum, ε is the relative dielectric constants of the passive films, and S is the area of the stainless steel substrate surface. Therefore, assuming the value of ε is constant, the higher CPE_f value implies a thinner passive film. As Figure 5 and Table 1 show, both the values of CPE_f and CPE_dl increased with the solution temperature and the amplitude of the alternating temperature, which confirms that the passive film on the stainless steel became thinner as the solution temperature or the range of the temperature changed. Furthermore, the CPE_f and CPE_dl of the samples that experienced the alternating temperature conditions were evidently higher than their counterparts in the constant temperature solutions, which also suggested that the former temperature condition presents a more noticeable degradation effect on the passive film relative to the corresponding higher constant temperature of the alternating temperature cycles.

3.2 Semiconducting properties

The semiconducting behavior of 304 stainless steel grades were tested under different temperature conditions, in 0.1 mol/L NaOH + 0.5 mol/L NaCl solution at 48 h. Two peaks around −0.1 and 0.4 V_Ag/AgCl can be observed (Figure 6). Negative slopes and p-type semiconducting films were observed when the potentials were lower than −0.95 V_Ag/AgCl (R₁), which was attributed to the capacitance response of the Cr₂O₃ inner layer (Feng et al. 2016; Hamadou et al. 2010). In the −0.65 ∼ −0.10 V_Ag/AgCl range (R₂), positive slopes were observed. This scenario was ascribed to the capacitance response of the outer iron oxides, which presents an n-type semiconducting film (Feng et al. 2014; Hamadou et al. 2010). As potentials further increased and exceeded −0.10 V_Ag/AgCl, negative slopes were again observed, which was thought to be the beginning of passive film breakdown (Feng et al. 2014, 2016; Hamadou et al. 2010).

Figure 6:

Mott–Schottky plots of AISI 304 stainless steel in 0.1 mol/L NaOH + 0.5 mol/L NaCl solution at 48 h under various temperature conditions.

The secondary peaks around 0.4 V_Ag/AgCl were observed in the Mott–Schottky plots (R₃) (Figure 6). Similar situations were also reported for the passive film on 2205 and 304 stainless steel grades in concrete pore solutions (Feng et al. 2014; Luo et al. 2012). Simões et al. (1990) reported that iron (II, III) oxide (Fe₃O₄) and γ-ferric oxide (γ-Fe₂O₃) are stable components at low and high potential, respectively, in the passive film. Li et al. (2009) further confirmed that Fe²⁺ is oxidized to Fe³⁺ in the passive film on carbon steel, as the potential exceeds 0.3 V_SCE (∼0.35 V_Ag/AgCl). In a previous study, the secondary peaks between 0.3 and 0.6 V_SCE were ascribed to the oxidization of Fe²⁺ ox in Fe₃O₄ in the passive films (Feng et al. 2014). Therefore, the secondary peaks in the Mott–Schottky plots (Figure 6) should also be ascribed to the oxidization of Fe²⁺ in Fe₃O₄ in the present study.

According to previous studies (Hamadou et al. 2010; Li et al. 2007; Luo et al. 2012; Simõs et al. 1990), the capacitances of n-type and p-type semiconducting films are expressed by Eqs. (2) and (3), respectively, after ignoring the capacitance of the Helmholtz layer (C_H).

where C_SC represents the space-charge capacitance, N_A and N_D separately represent the acceptor and donor densities. ε₀ and ε (ε = 12 (Hamadou et al. 2010; Li et al. 2007; Luo et al. 2012; Simõs et al. 1990)) are the same as those in Eq. (1). E_FB is the flat-band potential. k, T, and q are the Boltzmann constant, the absolute temperature, and the electron charge, respectively. The value of kT/q was ignored because it amounted to only ∼25 mV.

According to Eqs. (2) and (3), the donor density (N_D) and acceptor density (N_A) were calculated, and the results are presented in Figure 7. Clearly, N_D and N_A increased with the increasing temperature for samples in these constant temperature solutions. N_D and N_A also increased with the range of the alternating temperature for the samples that experienced the alternating temperature cycles. The values of N_A were higher than those of the N_D under the same temperature conditions, which was consistent with earlier studies (Feng et al. 2014, 2019a, b; Simõs et al. 1990). A previous study (Feng et al. 2012) suggested that more frequent and higher fluctuations were observed on the carbon steel with higher N_D passive films, when the sample was potentiostatically polarized at a potential in the passivation region. Simultaneously, more metastable pits can be observed in the anodic polarization curves for stainless steels when their passive films contained higher N_A and N_D (Feng et al. 2016). Therefore, these results confirmed that the higher values of N_D and N_A meant the lower stability and less protectiveness of the passive films. Furthermore, as Figure 7 shows, the values of N_D for the samples that experienced the alternating temperature conditions were close to those of their counterparts under the constant higher temperatures. For instance, the values of N_D for the samples that experienced the 22–50 °C, 22–60 °C, and 20–70 °C cycles were approaching those of the samples under 50 °C, 60 °C, and 70 °C constant temperature conditions, respectively. However, the values of N_A for the samples that experienced the former temperature condition were much lower than those of their counterparts in the higher constant temperature pore solutions. Many previous studies have confirmed that the passive film on stainless steel presents a double-layer structure. The outer layer is an iron oxides layer and presents n-grade semiconducting characteristics, while the inner is a chromium oxides layer and exhibits p-type semiconducting behavior (Feng et al. 2020; Kirchheim et al. 1990; Olsson and Landolt 2003). Therefore, the carrier density results (Figure 7) suggest that both the high constant temperature and alternating temperature cycle slightly doped the outer iron oxides layer, and the former temperature condition much more heavily doped the inner chromium oxides layer than the latter temperature condition. Moreover, the N_A and N_D of the samples that experienced the 22–60 °C and 20–70 °C alternating temperatures were generally close to those of the samples in the constant 50 °C and 60 °C temperature solutions, respectively.

Figure 7:

Carrier densities of the passive film on AISI 304 stainless steel in 0.1 mol/L NaOH + 0.5 mol/L NaCl solution at 48 h under various temperature conditions.

The space-charge layer thickness (W) was calculated according to Eq. (4) (Feng et al. 2014, 2019a, b):

where N is the carrier or donor density. The results of W are displayed in Figure 8. Clearly, the value of W decreased with the temperature or the amplitude of the alternating temperature for samples in constant temperature or alternating temperature pore solutions, respectively. Moreover, the W for the samples that experienced the alternating temperatures were always thinner than those of their counterparts at room temperature (22 °C), and were thicker than those of the samples at the higher temperature of the alternating temperature cycle (Figure 8). This situation, together with EIS results (Figure 5), suggested that the thickness of the passive film declined with solution temperature and the amplitude of the alternating temperature. Simultaneously, the reductions in the thickness of the passive films under alternating temperature conditions were less than its corresponding higher constant temperature in the alternating temperature cycle. For instance, the passive film on samples that experienced the 22–50 °C alternating temperature was always thicker than that on its counterpart in the 50 °C pore solution. However, the passive film thickness on the sample in the 22–60 °C alternating temperature solution was close to that of the sample in the 50 °C solution, and the W under 20–70 °C alternating temperature condition was approaching that at the 60 °C solution (Figure 8).

Figure 8:

Thickness of the space-charge layers of AISI 304 stainless steel in 0.1 mol/L NaOH + 0.5 mol/L NaCl solution at 48 h under various temperature conditions.

3.3 Surface characterization of the stainless steel samples

The surface of the stainless steel was examined after 48 h in different temperature pore solutions, and the results are presented in Figures 9 and 10. Many pitting initiation sites were observed on the surface of the samples in the constant temperature solutions (Figure 9). In addition, both the number of pitting initiation sites and their size increased with the temperature of the solution, which is in line with the results from previous studies (Dong et al. 2011; Ernst and Newman 2002). Ernst and Newman (2002) applied the foil technique to study the influence of the solution temperature on the pitting behavior of 304 stainless steel in a NaCl solution, and they noticed that the repetitive undercutting of the steel surface resulted in the increment of the width of the pit opening, then the width linearly increased with time. However, the growth of pitting depth was under diffusion control with a salt film being precipitated at the pit bottom, then the pitting depth linearly increased with the square root of time (Ernst and Newman 2002). Therefore, both the width of the pit opening and its depth increased with time, and the growth rate of the width was much higher than that of its depth with increasing solution temperature. Dong et al. (2011) investigated the pitting characteristics of 2205 stainless steel in a Cl⁻ contaminated solution at different temperatures, and they found that the number of pits strikingly increased when the solution temperature increased. In the present study, more pitting initiation sites were observed on the surface when the solution temperature changed from 22 °C to 70 °C (Figure 9). The size of the pitting initiation sites also increased with the increasing solution temperature. Furthermore, in comparison, the number of pitting initiation sites on the samples that experienced the alternating temperature was much higher than those on their counterparts under the constant temperature conditions (Figures 9 and 10). For instance, the number of pitting initiation sites on the stainless steel sample in the solution that experienced the 20–50 °C cycle (Figure 10(a)) was more than twice that on the sample in the constant 50 °C pore solution (Figure 9(b)), and the density of pitting initiation sites on the former sample was even higher than that on the counterpart in the constant 70 °C pore solution (Figure 9(d)). This situation suggested that the alternating temperature more significantly promoted pitting initiation on the stainless steel in the Cl⁻ contaminated pore solution when compared to its corresponding higher constant temperature in the alternating temperature cycles. In addition, the size and the number of pitting initiation sites on the stainless steel increased with the range of the alternating temperature. Simultaneously, cracks were observed in passive films on the sample that experienced 24 cycles of the 22–70 °C alternating temperature (Figure 10(d)).

Figure 9:

Corrosion state of AISI 304 stainless steel in 0.1 mol/L NaOH + 0.5 mol/L NaCl solution under constant temperature conditions after 48 h: (a) 22 °C, (b) 50 °C, (c) 60 °C, and (d) 70 °C.

Figure 10:

Surface characterization of AISI 304 stainless steel in 0.1 mol/L NaOH + 0.5 mol/L NaCl solution under alternation temperature conditions after 48 h (24 cycles): (a) 22–50 °C alternating temperature cycle, (b) 22–60 °C alternating temperature cycle, (c) 22–70 °C alternating temperature cycle, (d) cracks in the passive films on the stainless steel experienced 22–70 °C alternating temperature cycles.

3.4 XPS of the stainless steel surface

The stainless steel surfaces were analyzed by x-ray photoelectron spectroscopy. Typical full survey spectra and spectra of Fe 2p_3/2, Cr 2p_3/2, Ni 2p_3/2, as well as O 1s are presented in Figure 11. Besides Cr and Fe cations, a small quantity of Ni was also detected on the stainless steel surface.

Figure 11:

Typical XPS full survey spectra and the high-resolution spectra of Fe 2p_3/2, Cr 2p_3/2, Ni 2p_3/2, and O 1s of the 304 stainless steel, after 48 h of immersion in 50 °C chloride contaminated pore solution: (a) full survey spectra, (b) Fe 2p_3/2, (c) Cr 2p_3/2, (d) Ni 2p_3/2, and (e) O 1s.

In the high-resolution spectra, four composition types, including metallic state (Fe met, 706.9 eV), iron(II) oxide (Fe²⁺ ox, 709.6 eV), iron(III) oxide (Fe³⁺ ox, 710.8 eV), and iron(III) oxi-hydroxide (Fe³⁺ hy, 712.3 eV) were observed in the Fe 2p_3/2 signal (Elsener et al. 2011; Feng et al. 2014). Metallic chromium (Cr met, 574.0 eV), chromium(III) oxide (Cr³⁺ ox, 576.5 eV), and chromium(III) hydroxide (Cr³⁺ hy, 578.0 eV) were observed in the Cr 2p_3/2 spectrum (Addari et al. 2008; Feng et al. 2014). Simultaneously, three states, including the lattice oxygen (O²⁻, 530.2 eV), surface OH⁻ groups (531.8 eV, originated from the Fe³⁺ hy and Cr³⁺ hy), and absorbed water (H₂O, 533 eV) were assigned to the O 1s spectrum (Feng et al. 2014; Sosa et al. 2003). In the spectrum of Ni 2p_3/2, only metallic nickel can be observed (Figure 11(d)). The detected composition of Ni was in line with the results from previous studies (Lothongkum et al. 2003; Ma et al. 2018), in which Ma et al. (2018) and Lothongkum et al. (2003) noticed that nickel was not oxidized and the intensity of the Ni 2p_3/2 signal depended on the content of nickel in the stainless steel substrate.

Figure 12 displays the atomic percent of Fe, Cr, Ni, and the individual composition of Fe 2p_3/2, Cr 2p_3/2, and O 1s in the passive film on the sample in constant temperature pore solutions. The atomic percent of Fe and Cr seems to decrease and increase, while the content of Ni increased with temperature in the constant temperature solutions (Figure 12(a)). Simultaneously, the content of Fe²⁺ ox and Fe³⁺ ox did not display obviously change, while the content of Fe met increased and that of Fe³⁺ hy decreased, as the solution temperature increased (Figure 12(b)). In general, the ratio of (Fe³⁺ ox + Fe³⁺ hy)/Fe²⁺ ox increased with solution temperature. In the high-resolution spectra of Cr 2p_3/2 (Figure 12(c)), the content of Cr³⁺ ox declined, and the contents of Cr met and Cr³⁺ hy, as well as the Cr³⁺ hy/Cr³⁺ ox ratio, increased with the solution temperature. Furthermore, the content of OH⁻ increased while the content of O²⁻ decreased with the increasing temperature of the pore solution. The increment of the contents of Ni (Figure 12(a)), Fe met (Figure 12(b)), and Cr met (Figure 12(c)) could be ascribed to the decline in the passive film thickness. Lothongkum et al. (2003) and Zeng et al. (2018) confirmed that Ni concentrated at the film/substrate interface. Castle and Qiu (1990) applied the inductively coupled plasma source mass spectrometry (ICP-MS) and XPS to study the passive films on different types of stainless steels. Their results also suggested that Cr was enriched while Ni was depleted in the passive film, and Ni was enriched in the metallic phase underneath the passive film. The fact that the depletion of nickel in the passive film and only metallic Ni was detected, was attributed to the selective dissolution of Ni in the solution, and the Ni ions in the passive film were below the detection limit of the XPS spectrometer, i.e., <1% (Castle and Qiu 1990). On the other hand, Feng et al. (2014) and Jung et al. (2012) reported that the signals of Fe met and Cr met in the XPS spectra came from the stainless steel substrate. The passive film on the sample became thinner as the solution temperature increased, then the content of Ni, Fe met, and Cr met increased with the solution temperature (Figures 5 and 8).

Figure 12:

XPS results of passive films on the stainless steel in the constant temperature solutions at 48 h: (a) atomic percent profile of Fe, Cr, Ni; (b), (c), and (d) the different components of Fe 2p_3/2, Cr 2p_3/2, and O 1s, respectively.

The Fe²⁺ ox was derived from magnetite (Fe₃O₄) in the passive film on stainless steel, as shown in earlier studies (Feng et al. 2014; Freire et al. 2010; Simõs et al. 1990). The increment of (Fe³⁺ ox + Fe³⁺ hy)/Fe²⁺ ox ratio suggests that the magnetite was gradually oxidized in the external films with the increasing solution temperature, which is in line with the secondary peaks (R₃) observed in the Mott–Schottky plots (Figure 6). Moreover, Freire et al. (2012) noticed that the Fe³⁺/Fe²⁺ ratio increased with the decreasing solution pH value. Feng et al. (2014, 2016 investigated the concentration of doping densities (N_A and N_D), the composition of the passive films on deformed stainless steels, as well as the current densities potentiostatic polarization at 0.5 V_SCE (in the passivation region). The results indicated that the passive current density and doping densities increased, while the (Fe³⁺ox + Fe³⁺hy)/Fe²⁺ox ratio in the passive decreased with the magnitude of deformation of the stainless steels. Furthermore, the number of pitting also increased with doping densities (N_A and N_D) (Feng et al. 2014). The authors (Feng et al. 2014, 2016; Freire et al. 2012) suggested that the Fe³⁺/Fe²⁺ ratio and doping density are reliable indicators for the stability of the passive films. Therefore, as Figure 12(b) shows, the increasing (Fe³⁺ ox + Fe³⁺ hy)/Fe²⁺ ox ratio indicates that the stability of passive films on 304 stainless steel grades decreased with solution temperature, which is consistent with the semiconducting property results (Figures 6 and 7). On the other hand, the content of Cr³⁺ ox decreased, while the content of Cr³⁺ hy and the ratio of Cr³⁺ hy/Cr³⁺ ox increased, when the solution temperature increased (Figure 12(c)). This scenario reflected the fact that the Cr³⁺ ox gradually transformed into Cr³⁺ hy in the passive film. The increment of the OH⁻ content (Figure 12(d)) further confirmed the transformation of Cr³⁺ ox as the temperature increased. The higher Cr³⁺ hy/Cr³⁺ ox ratio means more unstable passive films on the stainless steel, according to earlier studies (Dai et al. 2020; Feng et al. 2019a, b). Therefore, the semiconducting properties (Figure 7), together with (Fe³⁺ ox + Fe³⁺ hy)/Fe²⁺ ox ratio (Figure 12(b)), Cr³⁺ hy/Cr³⁺ ox ratio (Figure 12(c), and O²⁻ content (Figure 12(d)), suggest the decreasing stability of the passive films on 304 stainless steel grades, as the solution temperature increased.

Figure 13 displays the XPS results of 304 stainless steel grades that experienced the alternating temperature cycles. As the results show, the content of Cr and Ni increased, and the content of Fe decreased with the increasing amplitude of the alternating temperature (Figure 13(a)), which is similar to the atomic percent profile observed on the samples in the constant temperature solutions (Figure 12(a)). The content of Fe²⁺ ox (Figure 13(b)) and Cr³⁺ ox (Figure 13(c)) decreased, while the content of OH⁻ (Figure 13(d)) increased with the amplitude of the alternating temperature. Simultaneously, the (Fe³⁺ ox + Fe³⁺ hy)/Fe²⁺ ox ratio (Figure 13(b)) and Cr³⁺ hy/Cr³⁺ ox ratio (Figure 13(c)) increased with the range of the alternating temperature. This situation suggested that the Fe²⁺ (mainly came from the magnetite) gradually oxidized to Fe³⁺ in the external films, and the Cr³⁺ ox changed to Cr³⁺ hy in the inner layer, with the increasing range of the alternating temperature. The XPS results (Figure 13), together with the electrochemical results (Figures 6 and 7), reflected the decreasing stability of the passive films, as the amplitude of the alternating temperature increased (Feng et al. 2014; Freire et al. 2010; Simõs et al. 1990).

Figure 13:

XPS results of passive films on the stainless steel experienced alternating temperature cycles at 48 h: (a) atomic percent profile of Fe, Cr, Ni; (b), (c), and (d) the different components of Fe 2p_3/2, Cr 2p_3/2, and O 1s in the passive film.

For comparison the degradation effects of the constant temperature and alternating temperature, the (Fe³⁺ ox + Fe³⁺ hy)/Fe²⁺ ox ratios and Cr³⁺ hy/Cr³⁺ ox ratios of the passive film on samples under various temperature conditions were plotted together (Figure 14). Generally, the higher (Fe³⁺ ox + Fe³⁺ hy)/Fe²⁺ ox ratio and Cr³⁺ hy/Cr³⁺ ox ratio meant the lower stability of the passive films, basing on earlier studies (Feng et al. 2014; Freire et al. 2010, 2012; Simõs et al. 1990). As the results show, the (Fe³⁺ ox + Fe³⁺ hy)/Fe²⁺ ox ratio and Cr³⁺ hy/Cr³⁺ ox ratio of the passive film on stainless steel in 22–60 °C cycle solution were higher than those of their counterparts in the 50 °C solution, while the two ratios of the samples in the 22–70 °C cycle solution were even higher than those of the one in the 70 °C solution.

Figure 14:

(Fe³⁺ ox + Fe³⁺ hy)/Fe²⁺ ox ratio and Cr³⁺ hy/Cr³⁺ ox ratio of the passive films on the stainless steel samples in the Cl⁻ contaminated pore solutions under different temperature conditions.

3.5 Effect of alternating temperature on pitting initiation on 304 stainless steel

The 22–50 °C, 22–60 °C, and 22–70 °C alternating temperature cycles more significantly degraded the passive films than their corresponding higher constant temperature (50 °C, 60 °C, 70 °C) in the alternating temperature cycles, respectively, as the results showed in the EIS (Figures 3 and 5), XPS (Figure 14), and surface characterization (Figures 9 and 10). However, the semiconducting property results suggested that the former temperature conditions presented a less degradation effect than their constant temperature counterparts (Figures 6–8). Thus, it was difficult to determine whether the alternating temperature cycle or the higher temperature in the temperature cycle had the stronger degradation effect on the passive films, in the present study. However, it was easy to conclude that the degradation of the passive film under the 22–60 °C and 22–70 °C alternating temperature conditions were more serious than that of their counterparts in the constant 50 °C and 60 °C conditions, respectively, as all the results showed in Figures 3, 5, 6–10, and 14.

Average temperature was widely used and the change of temperature was ignored in lifetime estimation for concrete structures (Lu et al. 2011). For discussing the reliability of the application of average temperature, the average temperature of the alternating temperature cycles, 36 °C, 41 °C, and 46 °C for the 22–50 °C, 22–60 °C, and 22–70 °C cycles, respectively, was introduced to help analyze the degradation effect of the different temperature conditions. It has been confirmed that the passive films were more seriously degraded by the increasing solution temperature (Feng et al. 2020; Li et al. 2007), which means that 50 °C and 60 °C can more significantly degrade the passive films than the average 41 °C and 46 °C, respectively. In the present study, all the experimental results suggested that the alternating temperatures, 22–60 °C and 22–70 °C more seriously degraded the passive film than the constant 50 °C and 60 °C. Therefore, it can be concluded that the alternating temperature cycles (22–60 °C and 22–70 °C) show a more serious degradation on the passive film, when compared to the average temperature (41 °C, 46 °C) of the alternating temperature cycles. These results suggest that besides the high average temperature, the alternating temperature should also be considered when designing or assessing the lifetime of concrete structures in a tropical marine environment.

Alternating temperature cycles and high constant temperature both heavily doped the passive film (Figure 7) and reduced its thickness (Figures 8, 12(a), and 13(a)), as well as promoted the oxidation of magnetite (Figures 12(b) and 13(b)) and the transformation of Cr³⁺ ox into Cr³⁺ hy (Figures 12(c) and 13(c)). In addition, the alternating temperature cycles caused cracks (Figure 10(d)) in the passive film on the stainless steel, resulting from a significant thermal expansion coefficient difference between the iron oxides layer, chromium oxide layer, and stainless steel substrate (Polanco et al. 2004; Robertson and Manning 1990).

Figure 15 illustrates the pitting initiation on the stainless steel under constant temperature and alternating temperature conditions, respectively, in the Cl⁻ contaminated pore solution. A double-layer structural passive film formed on the surface when the stainless steel was immersed in the pore solution at room temperature (22 °C). The chromium oxide and iron oxide layers were separately doped with a low concentration of accepter and donor in the inner layer and the outer layer (Hakiki et al. 1995; Kirchheim et al. 1990; Olsson and Landolt 2003) (Figure 15(a)). According to the point defect model (Macdonald 1992, 1999; Macdonald et al. 1992), the oxygen vacancies were related to the donors, and the cation vacancies were identified as the acceptors in the passive film. The dissolution rate of the film was enhanced by the increasing solution temperature at the iron oxide/solution interface (Macdonald 1999, Macdonald et al. 1992). Therefore, when the pore solution temperature increased, the thickness of the outer iron oxide layer decreased, and the donor density in the iron oxide layer and the acceptor density in the chromium oxides layer increased, respectively (Figure 15(b)). Simultaneously, the Fe²⁺ in magnetite was oxidized to Fe³⁺ in the outer layer, while the Cr³⁺ ox was transformed into Cr³⁺ hy in the inner layer. Compared to the sample in the room-temperature (22 °C) solution (Figure 15(a)), the passive film became more unstable in a higher temperature solution (Figure 15(b)), then pitting initiation more frequently occurred at the film/metal interface (Figure 15(c)) (Macdonald 1992, 1999; Macdonald et al. 1992), for the higher carry densities (Figure 7), thinner thickness (Figures 8, 12(a) and 13(a)), and less stable composition (Figures 12(b), (c) and 13(b), (c)).

Figure 15:

Illustration of passive films and pitting initiation on the stainless steel in the Cl⁻ contaminated pore solution under different temperature conditions: (a) passive film formed on the stainless steel in a low constant temperature pore solution, (b) passive film and (c) pitting initiation on the stainless steel in a high constant temperature pore solution, (d) passive film, and (e) pitting initiation on the samples in an alternating temperature pore solution.

For the samples that experienced the alternating temperature cycles, besides these negative effects of high temperature, the double-layer structural passive film on the stainless steel suffered from the adverse effect of the alternating temperature, which can result in cracks in the film, as Figure 15(d) shows, according to an earlier study [13]. In the Cl⁻ contaminated pore solution, as shown in Figure 15(e), pitting initiation often preferentially occurred at sulphide inclusions and passive films’ cracks. Therefore, as shown in Figures 9 and 10, the stainless steel samples that experienced the alternating temperatures were more sensitive to pitting than their counterparts under the constant temperature conditions (with the average value of the alternating temperature cycle), and the amount of pitting initiation under the former temperature condition generally was higher than that on their counterparts at the latter temperature conditions.