Semiconductivity Conversion of Passive Films on Alloy 800 in Chloride Solutions Containing Various Concentrations of Thiosulfate

Semiconductive properties of the passive ﬁlms formed on Alloy 800 in 0.1 M chloride solutions containing various concentrations of thiosulfate are investigated using Mott-Schottky analysis. The results indicate that when the concentration of the thiosulfate ions increases, the semiconductivity is converted from the n-type to the p-type. This conversion is postulated to be due to the incorporation of sulfur into the passive ﬁlm, which changes the vacancy type. The breakdown behavior of the passive ﬁlms is correlated with the semiconductivity of the ﬁlm in this system.

Fe-Cr-Ni alloys, such as Alloys 600, 690, 800 and stainless steels have often been used in demanding applications in water and steam, including nuclear power plants and steam generators (SGs) of pressurized water reactor systems. 1,2 The extensive use of Fe-Cr-Ni alloys has led to numerous studies on the properties of their passive films. The composition, thickness and the semiconductivity of a passive film are critical to the corrosion behavior of the alloy. 3,4 Chloride and thiosulfate are common ions existing in power-generating systems. 5 It has been reported previously that the ratio of chloride concentration to thiosulfate concentration had a remarkable effect on passive film degradation. [6][7][8] There is maximum pitting susceptibility at a certain chloride-to-thiosulfate ratio for austenitic stainless steels. 9 At a higher ratio, the passive layer breakdown potential is significantly lowered, 5 because chloride ions induce metastable pitting and a small amount of thiosulfate can stabilize metastable pits by generating adsorbed sulfur within the metastable pits. However, this effect becomes less significant at a lower concentration ratio. 6 Newman et al. 7 investigated pitting corrosion of 304 SS in 0.25 M NaCl solution containing various concentrations of thiosulfate. They also found that additions of 0.01 to 0.02 M thiosulfate lowered the pitting potential by more than 300 mV, while additions of more than 0.5 M thiosulfate inhibited corrosion. They claimed that, at a higher concentration ratio, the passive film was broken down by chloride ions and stabilized by reduction of thiosulfate ions. At a lower concentration ratio, the chloride accumulation would be prevented and the decomposition of thiosulfate within the pits would tend to neutralize any acid generation S 2 10 The variation of the break-down potential of the passive film in chloride-thiosulfate solutions is likely associated with the semiconductive properties of the passive film, 11 but the detailed solution chemistry affecting the semiconductivity is not clear. Sato 12 pointed out that passive film of n-type oxide appears to be more stable against anodic polarization than that of p-type oxide in chloride-free solution. In p-type oxide, the electron acceptor levels introduced by anion adsorption and the dislocation-induced electron levels in the passive film will lower the transpassive potential. In this investigation, passive films formed on Alloy 800 in solutions containing chloride and various thiosulfate concentrations were investigated at room temperature to simulate the start-up transient conditions for nuclear reactors following prolonged shut downs when the temperatures are relatively low. * Electrochemical Society Active Member. c Present address: Retired. z E-mail: jingli.luo@ualberta.ca

Experimental
The test specimens were prepared using Alloy 800 SG tubing (Sandvik, heat number 516809, outer diameter of 15.88 mm, wall thickness of 1.13 mm). The outer surface was used for the tests (the inner surface and cross sections were sealed with epoxy resin). Prior to each measurement, the test surface was polished with wet silicon carbide papers (320, 600, 800 and 1200 grits), rinsed with water, then dried in a desiccator for 24 hours. The chemical composition (wt%) of Alloy 800 was: C(0.017), Si(0.46), Mn(0. Potentiodynamic polarization was conducted at 21 • C by sweeping the potential in the positive direction at 0.1667 mV/s in deaerated solution. A three-electrode cell was used with the Alloy 800 as the working electrode, a saturated calomel electrode (SCE) as the reference electrode (RE) and a platinum electrode as the counter electrode (CE).
For the Mott-Schottky analysis (M-S), the scan direction was from high to low potential with a stepping interval of 50 mV. An AC signal with a frequency of 1000 Hz and a peak-to-peak magnitude of 10 mV was superimposed on the potential.
For the scanning electrochemical microscope (SECM) experiments, an electrochemical cell was mounted on the SECM stage (from CHI, USA). An ultramicroelectrode of Pt (radius of 10 μm) with the side sealed in glass was used as the SECM probe. 13 Scanning electron microscope (SEM) and energy-dispersive X-ray analysis (EDX) were used to obtain the surface morphologies and elemental distributions, respectively.
Time-of-flight secondary ion mass spectrometry (ToF SIMS) analysis was carried out using a ToF SIMS IV instrument (ION-ToF Gmbh). The information depth of ToF SIMS analysis was limited to the top 1 to 20 monolayers. Ions from mass 1 (hydrogen) to ∼9000 amu (for cluster ions) were detected with resolutions ranging from 1 ppb to ppm concentrations, depending on the element. In the current work, the analysis source used was Ga + , operating at 15 kV; the sputtering source was Cs + , operating at 1 kV. The samples were immersed in the solution until the SIMS test to minimize atmospheric oxidation. Before the test, specimens were cleaned with distilled water.   (2) As the thiosulfate concentration increases to 0.1 M, the pitting potential increases. (3) As the thiosulfate concentration continues to increase to 0.5 M, the breakdown potential also increases but the corrosion form changes from pitting to transpassive dissolution.

Results and Discussion
In region (1), in plain chloride solution, metastable pits occur and chloride ions lead to film break down but the film can be repassivated if there is no thiosulfate. In chloride solution containing a small amount of thiosulfate, the pitting potential lowers to about 100 mV, because thiosulfate can adsorb onto the metastable pits, and stabilize them. 11 As discussed before, Newman et al. 7 found the same phenomenon for 304 SS in 0.25 M NaCl solution containing various concentrations of thiosulfate. They claimed that the combined effect of these two ions was due to the breakdown of passive film by chloride ions and accumulation of thiosulfate at rapidly dissolving pit precursors to form S 0 which would stabilize metastable pits and accelerate anodic dissolution. This explanation can also be applied to our system. In region (2), as the thiosulfate concentration further increases, no transient peaks can be seen on the polarization curve. The pitting potential increases when the thiosulfate concentration increases to 0.1 M. In this case, it can be assumed that the adsorption of chloride ions is inhibited by thiosulfate and therefore, there are insufficient adsorbed chloride ions on the surface to break down the film. In region (3), when the thiosulfate concentration increases to 0.3 M, transpassive dissolution occurs at a high potential. Newman et al. claimed that, as the thiosulfate concentration in chloride solution further increased, the chloride accumulation would be prevented. They claimed that pits were generated in this situation, and the electromigration and decomposition of thiosulfate within the pits would tend to neutralize any acid generation, as mentioned above, this would inhibit pitting corrosion. 7,10 This may be unlikely under this condition because the adsorption of thiosulfate is dominant so that there is no opportunity for chloride ions to adsorb onto the surface to damage the passive film. Therefore, no pits are initiated at a lower chloride-to-thiosulfate concentration ratio, as shown in the polarization curve and surface morphology (Figures 1a  and 2). This is in agreement with Marcus's conclusions that thiosulfate does not damage the passive film without the aid of chloride ions; 14 but it can accelerate anodic dissolution of bare metal. This effect is related to the uncapped ends which are free-radicals containing lone, unpaired electrons that render them very reactive, especially toward bare metal surfaces. 15 Figures 2a-2e show the SEM morphologies after the polarization tests. As the thiosulfate concentration increases, the corrosion form changes from pitting to transpassive dissolution. In chloride-only solution, many small pits appear on the surface (Figure 2a). However, in solution containing a small amount of thiosulfate, the pit size is larger (Figure 2b), indicating that the addition of small amount thiosulfate leads to a rapid pitting propagation rate. As the thiosulfate concentration further increases, no pits were found on the surface (Figure 2d and 2e), and the corrosion form is transpassive dissolution. Figure 3 shows the EDX elemental distributions of one typical pit formed in 0.1 M chloride + 0.01 M thiosulfate. Sulfur is enriched within the pit, i.e., thiosulfate is reduced within the pit to form adsorbed sulfur. 3 Figure 4a shows the SECM image when the electrode was polarized at 0.3 V SCE (metastable pitting occurs) in 0.1 M chloride solutions. Some active spots are possibly related to grain boundary, triple point and/or inclusions. 8 After the SECM image was scanned half way, 0.075 M thiosulfate was added into the chloride-only solution. The surface reactivity increased after the addition of 0.075 M thiosulfate and some "active spots" with high current (blue color) appear in the image, indicating that thiosulfate stabilizes the metastable pits and makes them develop into stable pits. However, if 0.5 M thiosulfate is added into 0.1 M chloride solution, no pits are observed (Figure 4b), suggesting that there is no combined effect when the concentration of thiosulfate is high. The probe approach curves (PACs) shown in Figure 5 also indicate that spot B is active whereas spots A and C are passive. Figures 6a and 6b show the M-S results and the slopes of the linear region can be related to defect density by: 16 where q is the elementary charge of electrons, ε is the dielectric constant of the oxide, ε 0 is the vacuum permittivity, A is the geometrical surface area, N is the defect density. Positive and negative slopes are for n-type and p-type semiconductors, respectively. The value of ε for the system under investigation is unknown and can be considered as constant because the quantity of the incorporated sulfur atoms is very low. The results show that when the thiosulfate concentration increases from 0 to 0.1 M, the slope does not change significantly and the semiconductivity of the passive film is n-type, suggesting that oxygen vacancies and/or interstitial cations are the major defects. 16 When the thiosulfate concentration is higher than 0.1 M, the slopes change from positive values to negative ones, indicating that the semiconductivity of the passive film is p-type and cation vacancies are the major defects. Further increase in the thiosulfate concentration from 0.3 M to 0.5 M results in more cation vacancies. Figure 7 shows the sulfur intensities as a function of depth on samples passivated at corrosion potential in the chloride solution without thiosulfate or with various concentrations of thiosulfate passivated at the corrosion potential. S intensity was determined by SIMS analysis using Cs + sputtering the surface. The results show that as the     When the thiosulfate concentration is higher than 0.1 M, the adsorption of thiosulfate is enhanced. Thiosulfate can be reduced on the film surface 7 and incorporated into the passive film. When the oxygen vacancies are filled with sulfur ions, the electrons in the n-type semiconductor are consumed. When the thiosulfate concentration further increases to 0.3 M, the electrons are completely consumed; further increase in the incorporated sulfur atoms will generate holes. Therefore, the semiconductivity of passive film is changed from n-type to p-type; the passive film cannot be broken down and transpassive dissolution is observed at high potentials. Further increase in the thiosulfate concentration from 0.3 M to 0.5 M results in more incorporated sulfur as shown in Figure 7, therefore more holes are generated, resulting in the further decrease of the slope.

Conclusions
Semiconductivities change from n-type to p-type is due to sulfur incorporation into oxygen vacancies. N-type semiconductors suffer pitting corrosion and p-type semiconductors suffer transpassive dissolution when the film breaks down in this system.