Effect of Hydrogen Sulﬁde Ions on the Passive Behavior of Type 316L Stainless Steel

The effect of hydrogen sulﬁde ions (HS − ) on the passivity of type 316L stainless steel was investigated in pH 8.4 boric acid-borate buffer solution. Galvanostatic polarization of a silver microelectrode covered with Ag 2 S generated both OH − and HS − above the stainless steel surface. During potentiostatic polarization of the stainless steel, the passivity-maintaining current density increased with increase in the concentration of HS − in the vicinity of the surface. The impedance of the stainless steel at a constant frequency decreased during polarization in the presence of HS − , while it was sustained after dilution of HS − . Electrochemical impedance spectroscopy (EIS), Mott-Schottky (M-S) analysis and scanning electrochemical microscopy (SECM) showed that a defective and n-type semiconductive passive ﬁlm was formed in the solution containing HS − . Auger electron spectroscopy (AES) revealed that metal cations and oxygen vacancies in the passive ﬁlm on the stainless steel increased when it was formed in a HS − -containing solution. The series of changes in passive ﬁlm properties is thought to be due to adsorption of HS – on the ﬁlm surface during the polarization.©TheAuthor(s) Corrosion resistance of stainless steel is thought to be dependent on degradation of the passive ﬁlm, which is important to understand a precursor process involved in localized corrosion such as pitting corrosion and to estimate the long-term performance of the material.

Corrosion resistance of stainless steel is thought to be dependent on degradation of the passive film, which is important to understand a precursor process involved in localized corrosion such as pitting corrosion and to estimate the long-term performance of the material. Inclusions of sulfides such as manganese sulfide (MnS) are known to provide pitting corrosion sites of stainless steel. [1][2][3] As for the roles of MnS in pitting corrosion, it has been generally agreed that electrochemical and/or chemical reactions of MnS release S species (SO 4 2-, HSO 3 -, S 2 O 3 2-, S and S 2-). The released S species change the composition of the local solution contiguous to the inclusion and lead to a decrease of pH near the micro-area. The decrease in pH and the presence of aggressive S species result in transition of the passive surface to a transpassive state, causing exposure of the substrate to the solution, which is the initiation of pitting corrosion. [1][2][3][4][5] Eklund suggested that the dissolution of MnS gives rise to acidification of the solution by producing sulfate ions: 1 MnS + 4H 2 O = Mn 2+ + SO 4 2− + 8H + + 8e − [1] MnS + 2H + = Mn 2+ + H 2 S [ 2 ] H 2 S = S + 2H + + 2e − [3] Solution acidification is also caused by production of thiosulfate ions: 2 2MnS + 3H 2 O = S 2 O 3 2− + 2Mn 2+ + 6H + + 8e − [4] 2H + + MnS = Mn 2+ + S + H 2 [5] In both cases, elemental sulfur is finally formed. On the other hand, Wraglén proposed that elemental sulfur formed by MnS dissolution leads to further acidification as follows: 5 MnS = S + Mn 2+ + 2e − [6] S + 3H 2  Many possible explanations for the detrimental effects of various S species causing initiation and/or propagation of pitting corrosion on stainless steel have been presented. Most previous studies have focused on the overall processes, including destabilization of the passive film, removal of the film, and initiation and/or propagation of pits. Since degradation of the passive film is the initial process of pitting corrosion, it is important to contemplate the change in passivity or passive film until depassivation. When stainless steel is exposed to an aqueous solution, a small amount of MnS on the stainless steel surface dissolves because its solubility in water is 4.7 ppm at 291 K. 6 MnS + 2H + = Mn 2+ + H 2 S (in acidic solution) [2] MnS + H 2 O = Mn 2+ + OH − + HS − (in neutral or alkaline solution) [9] Furthermore, the dissociations of H 2 S and HSin aqueous solutions are as follows: HS − = S 2− + H + . [11] The values of pK a for Eqs. 10 and 11 are 7.05 and 19.0, respectively, at 298 K. 7 The dissociation of HSis negligibly small and H 2 S generates mainly protons during its dissociation. Thus, the primary dissolution reaction of MnS momentarily increases pH of the local solution near the MnS. However, little attention has been given to the effect of HSon the passivity of stainless steel. The use of a liquid-phase ion gun (LPIG) is a microelectrode technique, a type of scanning electrochemical microscopy (SECM), which is effective for controlling the release of infinitesimal anions to a local space in the solution. 8 Recently, we have developed an LPIG to release a ppm-order amount of HSby cathodic polarization of a silver microelectrode covered with a silver sulfide (Ag 2 S) layer. 9 The total amount of HSduring the operation of the LPIG is in a safe order. It is possible to control the concentration of HSin the vicinity of the LPIG by its polarization. Application of the LPIG to other metal surfaces is expected to elucidate the mechanism and/or kinetics of depassivation of the stainless steel surface in a solution containing HS -. This study is the first study in which the LPIG was applied to type 316L stainless steel as a generator of HS -. The effect of HSon degradation of the passive film is discussed.

Experimental
Materials.-Type 316L stainless steel (782560, Nilaco) embedded in epoxy resin with a surface area of 0.07 cm 2 was used as a specimen. The procedure for fabrication of the LPIG microelectrode is basically the same as that reported previously. 9 A silver microelectrode was prepared from a silver wire with a purity of 99.9% and a diameter of 500 μm by embedding in a 1 mm diameter glass capillary with epoxy resin. The microelectrode was polarized at -0.7 V SSE in deaerated 0.1 mol dm -3 Na 2 S solution until the electric charge of 10 mC was consumed. Tungsten wire with a purity of 99.9% and silver wire with a purity of 99.9%, both with diameters of 100 μm, were embedded in epoxy resin with a diameter of 25 mm as a substrate for estimating pH and hydrogen sulfide ion concentration [HS -], respectively. A platinum microelectrode with a diameter of 30 μm was used as a tip electrode of an SECM. All electrodes were mechanically ground with SiC papers down to 4000 grit and then rinsed with distilled water.
Electrochemical setups and conditions.- Figure 1 shows schematic illustrations of the experimental setups used in this study. In all measurements, an acrylic cell with a solution volume of 100 cm 3 was used. A platinum plate was used as a counter electrode, and a silver/silver chloride electrode in contact with KCl-saturated solution was used as a reference electrode. The microelectrode of the LPIG or SECM was positioned above the substrate electrode with an interelectrode distance of 125 μm or 20 μm using a stepping motor X-Y-Z stage (SGSP20-35, Sigma Koki) and an optical microscope.
Operations of the LPIG were carried out in deaerated pH 8.4 buffer solution (0.15 mol dm -3 H 3 BO 3 and 0.15 mol dm -3 NaB 4 O 7 ) with an interelectrode distance of 125 μm. The LPIG microelectrode was galvanostatically polarized at -3 μA using a battery-driven current source (SS7012, HIOKI). The potential of the LPIG microelectrode was measured by an electrometer (R8240, Advantest). Figure 1a shows an LPIG setup for investigating the passivity of the stainless steel specimen. After immersion for 600 s, the LPIG microelectrode was polarized for 0, 1900, 1950 or 2400 s to generate S species. After the LPIG operation, potentiostatic polarization of the specimen electrode at 0.4 V SSE and following electrochemical impedance spectroscopy (EIS) or Mott-Schottky (M-S) analysis were conducted using a potentiostat (SP-150, Biologic). In EIS, electrode potential was perturbed by ±10 mV in a frequency range from 10 4 to 10 -1 Hz. The M-S analysis was promptly conducted at a frequency of 15 Hz and at a potential of 0.4 V SSE and stepwise-shifted potentials to -0.4 V SSE . A software package (EC lab V, Biologic) was used to fit curves of the impedance data. On the other hand, for estimating solution pH and/or [HS -] during the LPIG operation, both the tungsten and silver microelectrodes were located as substrates with an interelectrode distance of 125 μm and connected to different electrometers with the same reference electrode (Figure 1b). For the calibration of tungsten microelectrode potential to pH, the following deaerated solutions were used: 0.15 mol dm -3 sulfuric acid (pH 0.9), 0.04955 mol dm -3 phthalic acid-phthalate buffer (pH 4.0), 0.02489 mol dm -3 phosphoric acid-phosphate buffer (pH 6.9), 0.15 mol dm -3 boric acid-borate buffer (pH 8.4) and 0.025 mol dm -3 sodium hydroxide (pH 10.0). After monitoring the rest potential for 3600 s in solutions of various pH values, the calibrated potential of the tungsten microelectrode was obtained as a function of solution pH.
After the LPIG operation for 600 s in the vicinity of the specimen surface, the specimen was polarized at 0.4 V SSE for 100 s in deaerated pH 8.4 buffer solution and the specimen surface was monitored by SECM in a solution containing 1 × 10 -3 mol dm -3 hydroxymethylferrocene (FcMeOH). Figure 1c shows the SECM setup used in this study. Most of the parts were similar to those used previously by Fushimi. 10 The platinum tip electrode and specimen substrate electrode were connected to a bipotentiostat (HAL-1512 mM2, Hokuto Denko) and polarized independently at E t = 0.6 V SSE and E s = -0.2 V SSE , respectively, for a tip generation/substrate collection (TG/SC) mode. Simultaneously, the tip microelectrode was scanned in an area of 3000 μm square with stepwise of dx = 30 μm and dy = 30 μm, respectively, and intermissions of 0.5 s and 5 s, respectively, with an interelectrode distance of 20 μm.
Surface analysis.-The surface of the stainless steel specimen was analyzed by an Auger electron spectroscope (AES; JAMP-9500F, JEOL) with an electron beam (10 keV, 15 nA). Ar + sputtering at an etching rate of 3.2 nm min −1 equivalent to silica was used for obtaining a depth profile of the local specimen surface with an electron beam diameter of 30 μm. In all experimental tests, consistency was confirmed more than 3 times by repetition with different specimens under the same conditions. Figure 2 shows the electrode potential of the tungsten microelectrode as a function of solution pH. It is obvious that the potential and pH have a linear relation:

Estimation of pH and [HS -] during LPIG operation.-
The slope is almost in agreement with that reported when a tungsten microelectrode with a diameter of 25 μm was used to estimate various pH values. 11 In this study, the pH value in the interelectrode space during the LPIG operation was estimated by Eq. 12.   Figure 3 shows changes in electrode potentials of the LPIG microelectrode E LPIG , tungsten microelectrode E W , and silver microelectrode E Ag during the operation of the LPIG microelectrode. Before the operation, the value of E LPIG remains constant. This implies that the LPIG microelectrode is relatively stable and does not release HSduring that period. However, when the LPIG microelectrode is galvanostatically polarized at -3 μA, E LPIG changes to a negative potential. E W and E Ag also shift to negative potentials. It has been shown that Ag 2 S is reduced as follows: 12 The equilibrium potential of Eq. 13 is a function of pH and [HS -] as follows: The reduction of Ag 2 S increases pH as well as [HS -]. The value of pH, converted from Eq. 12 by substituting the value of E W , of the solution is sustained at ca. 8.5 before polarization of the LPIG. However, pH rapidly reaches a constant value of ca. 9.5 after the onset of polarization, and this value is sustained during the polarization. This means that local alkalization in the vicinity of the tungsten microelectrode is in a steady state. It is thought that the mass of OHgenerated from the LPIG microelectrode is balanced between the interelectrode space and bulk solution. During the local alkalization, hydrogen gas did not evolve on the LPIG, whereas E LPIG was sustained at −0.7 V SSE . On the other hand, it is possible to estimate the value of [HS -] by substituting the values of pH and E Ag into Eq. 14. When the LPIG was polarized cathodically, [HS -] reached ca. 1.5 × 10 -3 mol dm -3 within 100 s and then gradually increased and reached ca. 4.0 × 10 -3 mol dm -3 of 2400 s. When the polarization of the LPIG microelectrode was stopped and the LPIG microelectrode was pulled up to the bulk solution, the values of pH and [HS -] immediately decreased, suggesting that the products, OHand HS -, in the interelectrode space are diluted. The space is so small that products from the LPIG accumulated and the buffering effect of the solution did not act effectively to keep the pH in the space. Nevertheless, the operation of the LPIG microelectrode is useful for concentrating HSin the local space of the substrate. The concentration of HSin the narrow space between the electrodes can be ca. 1 mm from the LPIG microelectrode center seems to be lower than that under the LPIG microelectrode. Assuming that the total amount of HSgenerated from the LPIG microelectrode is ca. 9.3 × 10 -9 mol with a 100% current efficiency at -3 μA for 600 s, [HS -] in a bulk solution with a volume of 0.1 dm 3 is 9.3 × 10 -8 mol dm -3 . In the following experiments, polarization of the LPIG microelectrode was carried out at an interelectrode distance of 125 μm for 100, 150 or 600 s, corresponding to local [HS -] of 1.5, 2.2 and 2.8 × 10 -3 mol dm -3 , respectively.
Passive behavior of type 316L stainless steel in the presence of HS -.- Figure 4 shows a dynamic polarization curve of the type 316L stainless steel electrode in deaerated pH 8.4 buffer solution. An activepassive transition is not observed and the anodic current reaches a passivity-maintaining current at potentials lower than ca. 0.5 V SSE , implying that the specimen surface is spontaneously passivated before the polarization. At a potential higher than 0.5 V SSE , the anodic current increases and a peak is observed at 0.7 V SSE , which is attributed to the oxidation of metal cations and/or alloying elements in the passive film or stainless steel substrate. [13][14][15][16] At a potential higher than 0.8 V SSE , the anodic current decreases and reaches a secondary passivation.
The LPIG microelectrode in the vicinity of the stainless steel substrate was galvanostatically polarized at -3 μA in pH 8.4 buffer solution. Figure 5 shows the changes in electrode potentials E LPIG and  E 316L of the LPIG and the stainless steel, respectively. In all cases, before the LPIG operation, E LPIG does not shift, whereas E 316L shifts to a positive potential. This means that the LPIG is relatively stable without releasing HS -. The stainless steel surface is in a passive state and the passivity seems to be gradually improved. When the LPIG microelectrode is cathodicaly polarized, i.e., in an LPIG operation, however, E LPIG immediately shifts to -0.7 V SSE , indicating that Ag 2 S is reduced and generates OHand HS -. The rest potential E 316L of stainless steel is gradually shifted to a negative potential. It seems that the products from the LPIG accelerate the anodic reaction of the stainless steel surface. In the presence of OHand HSgenerated from the LPIG, potentiostaitic polarization of the specimen electrode at 0.4 V SSE allowed a relatively large anodic current to flow. Figure 6 is a double logarithmic plot of current density of the stainless steel specimen and time during the potentiostatic polarization when the polarization was started after the LPIG operation for 0, 100, 150 or 600 s. These operation periods correspond to [HS -] of 0.0, 1.5, 2.2 and 2.8 × 10 -3 mol dm -3 , respectively, in the narrow space between the local specimen and the microelectrode. It is clear that the current density decreases exponentially with time. The slope in the absence of HSis ca. -1, indicating that a high field mechanism is adopted for the formation of a passive film on the surface. 17 However, the slope becomes less steep in the presence of HSand the effect increases with increase in [HS -]. The current density flowing at 100 s also increases with increase in [HS -], indicating that a more conductive passive film is formed on the specimen surface in the solution with HSthan that formed without HS -.
Following polarization of the stainless steel specimen electrode at 0.4 V SSE for 100 s with or without HS -, impedance measurement was carried out at the same potential at 15 Hz. At this frequency, the capacitive property of the electrode surface was dominated the impedance response in an EIS measurement as discussed later. Figure 7 shows the change in impedance |Z| during potentiostatic polarization at 0.4 V SSE . It is obvious that the value of |Z| gradually increases with polarization time, indicating that the stability of the passive surface improves during the polarization. The value of |Z| is also dependent on [HS − ] and becomes smaller in a solution containing larger [HS − ]. Even after stopping the LPIG operation, the increasing tendency of |Z| does not change regardless of [HS − ], and |Z| without HS − is higher than that with HS − . This implies that the passive film formed in solution with HS − is less stable than that formed without HS − . Although the polarization affects stabilization of the film, dilution of HSin the solution does not seem to stabilize the passive film after the film has been meta-stabilized by the presence of HS -. The passive film is in a relatively unstable state. After the impedance measurement, EIS was immediately carried out in a frequency range from 10 4 to 10 −1 Hz at the same potential at 0.4 V SSE , The specimen in this study has a relatively small area, and the measurements needed to be repeated several times at lower frequencies. A small discrepancy was observed at frequencies lower than 1 Hz, although Kramers-Kronig transformation diagram was satisfied at most frequencies. Figure 8 shows Bode plots of the stainless steel specimen. Although there is some scattering in the data, the plot is fitted with a so-called Randles-type R ct C-R el equivalent electronic circuit, where R ct and R el are solution resistance and charge transfer resistance, respectively, and C is capacitance. Since the capacitance of a passive film/electrolyte interface consists of capacitance of the space charge layer C SC and capacitance of the Helmholtz layer C H in series, Assuming that the value of C H is 0.1 mF cm -2 for austenitic stainless steels in alkaline solutions, 18,19 the C SC value was close to the measured value of C. Hence, C is considered to be C SC in this paper. The values of R el , R ct and C as a function of [HS − ] are shown in Table I. The values of R el are almost constant because [HS − ] is less than 10 -6 mol dm -3 at maximum and is too low to change the solution conductivity. With increase in [HS − ] during the passivation, however, C increases and R ct decreases, clearly corresponding to the increase in passivation-maintaining current and the decrease in |Z| shown in Figs. 6 and 7, respectively. It is thought that an electronically damaged passive film was formed by the presence of HS − .    Figure 9 shows an M-S plot of the stainless steel specimen after passivation in the solution with or without HS − . The capacitance was measured at 15 Hz as was the capacitance shown in Figure 7. This frequency is seen in Figure 8 to be in the region dominated by a capacitive response. Though a negative slope is observed at potentials higher than 0.3 V SSE , due to the continuous growth of the film, a positive slope in the M-S plot means that the specimen has an n-type semi-conductive property. Regardless of the [HS -] in the solution, a linear relationship is observed at potentials from -0.15 to -0.05 V SSE . The Mott-Schottky equation of an n-type semiconductor is defined as follows: [16] where ε is the dielectric constant, ε 0 is the vacuum permittivity constant, e is the elementary charge, N D is the donor density, E fb is the flatband potential, k is the Boltzmann constant and T is the absolute temperature. The values of E fb and N D are shown in Table II. E fb is independent of [HS − ], meaning that the structure and/or chemical composition of the passive film on stainless steel is not affected by the presence of HS − in the solution during passivation. On the other hand, the value of N D increases with increase in [HS − ]. Since semi-conductivity is associated with the band structure of a space charge layer formed in a passive film and the surface state at the electrolyte/film interface, N D is correlated with the concentrations of oxygen vancancies and interstitial metal ions in the film. 20 The increase in N D implies that the presence of HS − during the passivation induces more donor levels in the passive film. However, the increased  concentration of dopants is not large enough to affect the structure and chemical composition of the film. Figure 10 shows an SECM tip current image of the stainless steel specimen surface, which was polarized at 0.4 V SSE for 100 s in a solution containing 2.8 × 10 -3 mol dm -3 of HS − using the LPIG microelectrode. The Ag 2 S layer on the LPIG microelectrode with a diameter of 500 μm was located at almost the center of a 3 mm diameter stainless steel specimen at a distance of 125 μm. TG/SC mode SECM was carried out with polarization of the tip and substrate electrodes at 0.6 V SSE and −0.2 V SSE , respectively, in deaerated pH 8.4 buffer solution containing 1 × 10 -3 mol dm -3 FcMeOH as a redox mediator. The anodic current flowing through the tip electrode corresponds to oxidation of FcMeOH, which is associated with surface reactivity of the substrate. Since a passive film on a stainless steel substrate has an n-type semiconductive property, the reactivity is strongly related to donor density or thickness of the passive film. Several studies have shown that a higher tunneling current will flow in the case of a thinner and/or a more defective oxide film than in the case of a thicker and/or a less defective oxide film. 10,[21][22][23] In the image, the passive film on the stainless steel can be distinguished from the insulating epoxy resin by a bright circle surrounded by four dark corners of the edge. The striped pattern and dispersed black spots in the image are considered to be the result of a tip movement artifact and residual abrasive particles, respectively. Moreover, the brightest circle is at the center of the specimen surface. The diameter of the circle is ca. 500 μm, which coincides with the diameter of the Ag 2 S layer on the LPIG microelectrode. This indicates that a relatively reactive part on/in the passive film is induced by the LPIG operation. The products generated from the LPIG might cause the formation of a more defective film and/or a thinner film partially in the vicinity of the LPIG microelectode tip compared with the area away from the LPIG microelectrode tip. Futhermore, a slightly dark arc is seen on the outside of the brightest circle of the stainless steel specimen. The diameter of the arc is almost the same as the diameter of the glass sheath of the LPIG microelectrode. This region seems to be slightly less reactive than the stainless steel surface distant from the LPIG microelectrode location. Figure 11 shows an AES depth profile of the stainless steel specimen after polarization at 0.4 V SSE for 100 s in a solution with or without the LPIG operation, in which the estimated [HS -] was 2.8 × 10 -3 mol dm -3 . It was clear that the specimen surfaces were covered with oxide of the passive film. Assuming that the passive film-substrate interface is located at the transition with a half of the atomic concentration of oxygen, the passive film formed in a solution with HSshows the same thickness as that of the passive film formed in a solution without HS -. It was also observed that a very small amount of elemental S is contained in the outmost passive film of the sample formed in the solution with HS -. The atomic concentrations of Fe ( Figure 12a) and Cr (Figure 12b) are larger in the passive film formed in a solution with HS − than in the passive film formed in a solution without HS -, while the atomic concentrations of Ni ( Figure 12c) and O (Figure 12d) in the film with HS − are smaller than those in the passive film without HS − , though the difference is relatively small. Metallic Ni seems to be depleted in the film but to have accumulated in the film/substrate interface. These results coincide with results of previous studies. [24][25][26][27]

Discussion
The effect of HS − on the passivity of type 316L stainless steel was investigated by using an LPIG microelectrode. Electrochemical and surface analyses revealed that the presence of HS − makes a passive film defective and conductive by increasing oxygen vacancies and metal cations.
It has been widely proposed that the presence of aggressive anions changes the passivity and leads to localized corrosion on stainless steels. [28][29][30][31][32] Three main depassivation models have been proposed in the presence of aggressive anions. 33 The adsorption model 34,35 is associated with the absorption of aggressive anions on the passive film. The adsorbed anions transfer metal cations to the electrolyte by forming a complex of metal cations on the film. As a result, the passive film is thinned and removed. According to the penetration model, 36,37  the depassivation of metal is due to the transfer of aggressive anions through the passive film to the metal surface. The adsorbed and/or contaminated anions introduce higher ionic conductive paths through the film and lead to a rapid release and removal of metal cations. The passive film breakdown model 38,39 is related to mechanical breakdown of the film. The adsorption of aggressive anions on the passive film reduces surface tension, resulting in a mechanical break. All of the proposals are based on the prior adsorption of aggressive anions on a passive film followed by the depassivation process of the film. The adsorption of an anion is associated with polarizability when it is adsorbed on metal cations. 40 The higher polarizability the anion has, the stronger is the adsorption of the anion on cations of the passive film. Polarizability of sulfides such as S 2− or HS − is approximately twotimes higher than that of halides and OH − . 40 Therefore, the adsorption of HS − might be prior to that of OH − on a stainless steel surface in a solution containing HS -.
When the stainless steel surface was polarized in a solution containing HS -, adsorbed HS − was preferentially incorporated into the formed passive film. It was confirmed from AES and M-S analyses that elemental S existed on the outermost passive film. If the outermost S is present as HSor S 2-, oxygen anion transfer from the electrolyte to the outermost lattice of the film should be inhibited. In order to maintain a charge balance in the film against the existence of HSor S 2-, metal cations and/or oxygen vacancies need to be produced at substrate/film and/or passive film/electrolyte interfaces, resulting in the concentrations of metal cations and oxygen vacancies becoming larger in the film. This is also supported by results in Figures 10 and  12 showing that the HSadsorption results in an increase of donor density in the film.
Though it is hardly forming S 0 from HSin pH 9.5, the polarization at potentials higher than −0.594 V SSE enables to induce the oxidation of HSto SO 4 2-. 41 Increase in local concentration of HSin the vicinity of the stainless steel specimen surface might lead to a change an electrochemical potential of the surface. The relatively small impedance of the stainless steel surface formed in the solution with HSwas due to the existence of HSin solution during the passive film formation. However, an increase in impedance was also observed during the anodic polarization continuously even after stopping HSenrichment by the LPIG operation. Thus, it is suggested that no oxidation of HSoccurs on the stainless steel surface, though further study of the ionic state of S species during a polarization of specimen surface is needed for further understanding about oxidation states of sulfide ions and local acidification of the local solution.
On the other hand, the concentration of Ni decreased in the film when it was formed in the solution containing HS − . Olefjord et al. reported that Ni does not participate in passive film formation. 24 Castle et al. suggested that metallic Ni is enriched at the substrate/passive film interface during passivation because it is relatively noble compared with other metallic elements in stainless steel. 26 It is thought that the presence of other enriched metallic elements in a film forming in the solution containing HSmakes it difficult for metallic Ni to be concentrated at the interface. In any case, the passivity of the stainless steel surface is strongly affected by the adsorption of HS − .
A deteriorating effect of HS − on the passivity of stainless steel has been reported. 30 As far as we know, however, this study is the first study to examine the role of HS − in the initiation of localized corrosion of type 316L stainless steel using an LPIG. HS − adsorption is one of the important conditions causing instability of a passive film, which would be a trigger of localized corrosion of the steel. When the specimen surface in which a defective part had been formed by using the LPIG was polarized in a solution containing a high concentration of Cl -, we confirmed the formation of pits in the defective part on the surface. It is thought that pitting was preferentially and locally initiated and/or propagated from the defective part of the passive film when the surface was exposed to the Cl --containing solution. Passive film degradation of type 316L stainless steel with an LPIG was investigated at higher potentials in various solutions. The critical conditions for degradation and depassivation will be reported in the near future.

Conclusions
The effect of hydrogen sulfide ions (HS − ) on passivity of type 316L stainless steel was investigated in pH 8.4 buffer solution using the LPIG technique. Galvanostatic polarization of the Ag/Ag 2 S LPIG microelectrode generated locally both HS − and OH − on the stainless steel surface. The passivity of the stainless steel became relatively unstable due to the formation of a more defective n-type semiconductive passive film with HS − than that formed without HS -. AES revealed an increase of metal cations and oxygen vacancies in the passive film formed in a solution containing HS -. The adsorption of HS − during passivation of the stainless steel surface could lead to the formation of a defective passive film. Change in the stability of a passive film due to the presence of HS − would be a trigger for the initial depassivation of stainless steel.