Role of NaCl, CO 2 , and H 2 S on Electrochemical Behavior of 304 Austenitic Stainless Steel in Simulated Oil Industry Environment

: The electrochemical behavior of 304 austenitic stainless steel (304ASS) was studied by different methods such as potentiodynamic polarization, EIS, SEM, and Raman spectroscopy. Potentiodynamic polarization data suggest that 304 ASS could be more susceptible to corrosion due to the presence of H 2 S. The coexistence of H 2 S and Cl − -type ionic species in 304 ASS lead to a decrease in the corrosion resistance as compared to the H 2 S-free condition. It is seen that CO 2 helps form a passive layer on the metallic surface, which eventually decreases its corrosion rate. Raman spectroscopy analysis shows that the passive layer developed under different condition consists of FeCO 3 , FeS 2 , Fe 2 O 3 , Fe(OH) 2 , etc. SEM images further conﬁrm that elemental S − and Cl − can inﬁltrate the passive ﬁlm and cause the passive ﬁlm to deteriorate.


Introduction
During the gas and oil production, CO 2 corrosion can cause pipeline failures [1][2][3][4][5][6][7], and this problem can become a threat when emission gasses are compressed into deep geological layers (Carbon Capture and Storage (CCS)) during combustion processes [8,9]. Commonly, steels used in industrial pipeline form layers of FeCO 3 (siderite) [10][11][12][13][14][15][16][17]. Initially, a corrosive environment has been formed due to the dissolution of CO 2 . Since the solubility of FeCO 3 in water (pK sp = 10.54 at 25 • C) is low, and as a result of the anodic iron dissolution, a siderite corrosion layer is formed on the surface of the alloy [12,13]. CO 2 corrosion sensitively depends on alloy composition, environmental factors such as temperature, partial CO 2 pressure, flow conditions, and protective corrosion scales in the processing of geothermal energy. [18][19][20][21][22][23][24][25][26]. The concentration of H 2 S has a significant impact on the defensive ability of the sulfide film formed. Due to the increase of H 2 S concentration, a very unstable film formed at a pH level of 3-5 that does not have any contribution in combating corrosion. With the increasing absorption of H 2 S, the film formed is very loose at a pH range of 3-5 and does not contribute to the inhibiting effect of corrosion [27]. The pH of the solution played an important role on the nature and composition of the corrosion product. Iron dissolves at a lower pH level (1)(2), and no precipitation of FeS occurs on the metal surface, as the solubility of FeS is very high in that pH range. Here, the acceleration of Fe dissolution is due to H 2 S. The formation of a protective film of iron sulfide (FeS) on the surface of the electrode is due to the inhibitive effects of H 2 S at a pH level between 3 and 5 [28]. As the solubility of H 2 S in water is three times more than that of CO 2 gas, the acid formed by H 2 S dissociation is around three times weaker than carbonic acid (H 2 CO 3 ) [29]. Therefore, the effect of H 2 S gas on reducing the pH of the solution is nearly the same as that of CO 2 gas. Unlike dissolved CO 2 , to become an acid, dissolved H 2 S does not need to undergo the slow hydration process. H 2 S reduces the solution pH in an H 2 S-driven environment when it serves as a weak acid such as carbonic acid. Similar to carbonic acid (H 2 CO 3 ), it may also increase the corrosion rate by providing an additional cathodic reaction generated by the following equation [30].
However, this direct reduction of H 2 S is only possible if the amount of H 2 S is too high, which means that either it needs to be the H 2 S controlled system (sour regime) or the CO 2 / H 2 S mix system [10]. In comparison, a high H 2 S concentration is often correlated with elemental sulfur, and relatively little is understood about the dynamic interactions that take place in the presence of the elemental form of sulfur. In the presence of low H 2 S concentration in an environment of CO 2 , the formation of the carbonate scale (FeCO 3 ) is affected by the iron sulfide (FeS) film [30]. This aspect of the environment or system is of concern, because it would appear that the iron sulfide film is removed more easily from the pipe wall than the iron carbonate scales. The protective film can be removed in the turbulent situations, which lead to an increase in the rate of corrosion. In the CO 2 /H 2 S system, the kinetics of film formation is complex and still not well known. Under these circumstances, the face of the surface scale will not only depend on the corresponding solubility of iron carbonates and iron sulfides but also on the mechanisms of the twoscale formation [31]. The aim of this work is to provide a systematic understanding of the corrosion behavior of 304 stainless steel in CO 2 and CO 2 -H 2 S environments in the presence of NaCl. By a combination of potentiostatic measurements and potentiodynamic polarization, in the presence and absence of H 2 S, the influence of pH value on the corrosion behavior of 304 SS was assessed. EIS was deployed to investigate the synergistic effect of CO 2 and CO 2 -H 2 S. SEM was employed to see the changes of the surface nature after corrosion in various environments.

Materials and Methods
The test was carried out on the polished 304 ASS specimens. Sample surfaces were polished using 120, 180, 1/0, and 2/0 grade emery paper and finally by a cloth polisher. A simulation environment was deployed in the lab scale similar to oil corrosion industries. Throughout electrochemical tests, H 2 S (200 ppm), CO 2 (200 ppm), or a mixed gas (H 2 S (200 ppm)/CO 2 (200 ppm): 1/1) was purged into the simulated solutions of deionized water, sodium chloride and acetic acid under supervision; a constant 40 mL/min flow rate was maintained to attain point of saturation.

Potentiodynamic Polarization
Potentiodynamic polarization and an Electrochemical Impedance Spectroscopy (EIS) test of all the samples were conducted in four different test solutions with varying compositions. The solutions were: 3.5% NaCl with three different pH levels (pH at 3, 7, 10), 3.5% NaCl + saturated CO 2 with three different pH levels (pH at 3, 7, 10), and 3.5% NaCl + saturated CO 2 + H 2 S with three different pH levels (pH 3, 7, 10). After performing a potentiodynamic polarization experiment, the samples were examined for visible effect of CO 2 and sodium thiosulfate with the help of Scanning Electron Microscope (SEM).
Potentiodynamic polarization experiments were carried out on the range of −0.6 to 0.6 V vs. SCE with a scan rate of 1 mV/s on all the samples using a Gamry Potentiostatic PC/750 (Metrohm AG, Herisau, Switzerland). All the four samples were polished with emery paper and later cloth polished. After polishing, with the help of acetone and ethanol solution, all the 304 ASS samples were cleaned. A standard three electrode system was deployed to measure the i corr and E corr values of the bar sample. Here, test specimens were used as a working electrodes, graphite was used as an auxiliary electrode and a saturated calomel electrode (SCE) is reflected as the reference electrode. All the tests were performed at room temperature in previously mentioned solutions. The surface area was varied for all the samples that are exposed to the solutions.

Electrochemical Impedance Spectroscopy (EIS)
For Electrochemical Impedance Spectroscopy (EIS), a three-electrode corrosion cell was used. The experiments were performed using 100 mV AC voltage RMS (root-meansquare) and 0 V DC in the frequency range of 0.01-10 kHz at 10 cycles/decade. Nyquist and Bode plots were obtained by measuring the phase angle and absolute impedance at each frequency point. Impedance data were analyzed with the help of a suitable equivalent electrical circuit utilizing the real and imaginary components using a simplex fit model supplied with the EIS software for impedance data interpretation. Gamry Echem Analyst software was used here.

SEM Analysis
In order to study the surface morphology of 304 SS after the potentiodynamic polarization experiment in different solution, scanning electron microscope (SEM) was used. The surface characterization studies of all the samples were carried out using Model JEOL JSM-6360 SEM (JEOL Ltd., Akishima, Tokyo, Japan). In this present study, 200× magnification was used.

Raman Spectroscopy
Raman spectroscopy helps with determining the composition of steel surfaces and hence corrosion behavior. The Raman spectrum of the samples was measured using a Raman spectrometer LabRAM HR800 (HORIBA Jobin Yvon, Paris, France) with an incident laser light at a wavelength of 514.5 nm. Raman bands were calibrated using the 520 cm −1 band from Si. The Raman spectra were taken at three appropriate points on each sample in the range of 100 to 2000 cm −1 .

Potentiodynamic Polarization Analysis
The potentiodynamic polarization experiments of a 304 stainless steel sample were done in different simulated solutions, and the obtained results are shown in Figure 1. The i corr and E corr values obtained from these curves by Tafel's extrapolation methods are tabulated in Table 1. It is noticeable that the i corr value of the 304 SS sample in 3.5% NaCl with saturated CO 2 solution is lower compared to any of the other solutions, 3 × 10 −8 A/cm 2 , 1 × 10 −7 A/cm 2 , and 2 × 10 −7 A/cm 2 , considering pH values of 10, 7, and 3, respectively. It is also found that the values of i corr can be ranked from low to high: in CO 2, NaCl, and CO 2 + H 2 S solutions. In 304 SS, the processes of corrosion are predominantly controlled by the formation of a thick oxide film that inhibits the diffusion of destructive ions into the materials. It is clear that in case of CO 2 solution at higher pH, it forms a stable iron carbonate film, which also has a positive effect on the E corr value of this sample. In CO 2 + H 2 S solution with pH ≈ 3, H 2 S and Cl− were thought to cause 304 SS pitting corrosion at the inclusion of MnS [32]. Ikeda et al. [33] in their research reported that with 9-25 wt% Cr, the alloy becomes nearly resistant to CO 2 corrosion, as it is in our case with 304 SS (16-18.5 wt% Cr) possessing the lowest i corr in CO 2 solution. Furthermore, in H 2 S + CO 2 solution, 304 SS displays higher i corr compared to the other two solutions. It is believed that the synergistic effect of CO 2 and H 2 S is causing relatively more serious corrosion rather than helping in prevention from protective corrosion products.

Solution pH icorr (A cm −2 ) Ecorr (V)
3.5%. NaCl + CO2 Figures 2 and 3 show a Bode magnitude plot and Bode phase plot, respectively. The Bode magnitude plot of all the samples shows two distinct regions. In the low and higher frequency region, the Bode magnitude plot exhibits constant log ‫|‪Z‬|‬ values vs. log (f) with a phase angle near 0 0 . This is due to the response of the solution resistance. The log ‫|‪Z‬|‬ of a CO2-containing sample stabilized at a higher frequency range, whereas the CO2 + H2S containing sample stabilized at the lowest log ‫|‪Z‬|‬ value. In the wide-ranging (1 kHz-100 mHz) frequency, the spectra display a linear slope of about -1. This shows the capacitive characteristic of the surface film [34,35]. In the low-frequency range, the phase angle of three samples decreases, indicating the contribution of surface film resistance to the impedance.   Figures 2 and 3 show a Bode magnitude plot and Bode phase plot, respectively. The Bode magnitude plot of all the samples shows two distinct regions. In the low and higher frequency region, the Bode magnitude plot exhibits constant log |Z| values vs. log (f) with a phase angle near 0 0 . This is due to the response of the solution resistance. The log |Z| of a CO 2 -containing sample stabilized at a higher frequency range, whereas the CO 2 + H 2 S containing sample stabilized at the lowest log |Z| value. In the wide-ranging (1 kHz-100 mHz) frequency, the spectra display a linear slope of about -1. This shows the capacitive characteristic of the surface film [34,35]. In the low-frequency range, the phase angle of three samples decreases, indicating the contribution of surface film resistance to the impedance.   Figure 4 shows the Nyquist plots of 304 stainless steel under different solutions conditions. In the Nyquist plot, a higher radius of imaginary versus real resistance is related to the corrosion resistance of the alloy. The 304 SS sample in CO 2 solution shows the highest area under the curve compared to other sample, which indicates a higher polarization resistance value. The higher polarization resistance value is related to the higher corrosion resistance of the sample, which is also supported by the corrosion data, i.e., i corr value. The 304 SS sample in CO 2 + H 2 S exhibits the lowest area under the curve. So, naturally, its corrosion resistance is weaker when compared to other samples. For all the samples, the area under the curve decreases with the decreasing of pH. This illustrates that in the proposed environment, the formed film has been disbanded and the protection is lost, which in turn corrodes the surface beneath at a higher rate.

EIS Study
Additionally, the EIS results were analyzed and fitted with appropriate equivalent circuits (EC) using a Gamry Potentiostatic PC/750 (Metrohm AG, Herisau, Switzerland) in order to gather more information about the corrosion mechanism. Multiple equivalent circuits were used to analyze the EIS data. The equivalent circuit with a minimum error value was chosen in this study and can be seen in Figure 5, which has been reported by other research as well [36].
The equivalent circuit parameters in Figure 5 are the solution resistance (R s ), the capacitance of passive film (Q f ), the resistance of passive film (R f ), the capacitance of an electrical double layer at interfaces (Q dl ), and the charge transfer resistance (R t ). The constant phase element, CPE (Q), is usually used for uneven current distribution at the surface or increased surface roughness. The impedance of CPE is given by: where Y 0 is a modulus, j 2 = −1, ω is the angular frequency, and n is an empirical exponent between 0 and 1. The property of CPE depends on the value of n. CPE acts as a pure capacitor for n = 1, as a resistor when n = 0, and as a Warburg impedance when n = 0.5. Table 2 represents the values of equivalent circuit parameters for each solution. In 3.5% NaCl solution, the smaller impedance arc represents that the corrosion resistance decreases due to Cl − concentration. It has been seen that the value of R f decreases in a CO 2 + H 2 S-   Figure 4 shows the Nyquist plots of 304 stainless steel under different solutions conditions. In the Nyquist plot, a higher radius of imaginary versus real resistance is related to the corrosion resistance of the alloy. The 304 SS sample in CO2 solution shows the highest area under the curve compared to other sample, which indicates a higher polarization resistance value. The higher polarization resistance value is related to the higher corrosion resistance of the sample, which is also supported by the corrosion data, i.e., icorr value. The 304 SS sample in CO2 + H2S exhibits the lowest area under the curve. So, naturally, its corrosion resistance is weaker when compared to other samples. For all the samples, the area under the curve decreases with the decreasing of pH. This illustrates that in the proposed environment, the formed film has been disbanded and the protection is lost, which in turn corrodes the surface beneath at a higher rate.   Additionally, the EIS results were analyzed and fitted with appropriate equivalent circuits (EC) using a Gamry Potentiostatic PC/750 (Metrohm AG, Herisau, Switzerland) in order to gather more information about the corrosion mechanism. Multiple equivalent circuits were used to analyze the EIS data. The equivalent circuit with a minimum error value was chosen in this study and can be seen in Figure 5, which has been reported by other research as well [36]. Additionally, the EIS results were analyzed and fitted with appropriate equivalent circuits (EC) using a Gamry Potentiostatic PC/750 (Metrohm AG, Herisau, Switzerland) in order to gather more information about the corrosion mechanism. Multiple equivalent circuits were used to analyze the EIS data. The equivalent circuit with a minimum error value was chosen in this study and can be seen in Figure 5, which has been reported by other research as well [36]. The equivalent circuit parameters in Figure 5 are the solution resistance (Rs), the capacitance of passive film (Qf), the resistance of passive film (Rf), the capacitance of an electrical double layer at interfaces (Qdl), and the charge transfer resistance (Rt). The constant phase element, CPE (Q), is usually used for uneven current distribution at the surface or increased surface roughness. The impedance of CPE is given by: where Y0 is a modulus, j 2 = −1, ω is the angular frequency, and n is an empirical exponent between 0 and 1. The property of CPE depends on the value of n. CPE acts as a pure capacitor for n = 1, as a resistor when n = 0, and as a Warburg impedance when n = 0.5.   Figure 6 represents the SEM images of the samples after a corrosion test in different electrolytes with the same pH value of 7. From the micrograph, it is quite clear that the pit formation is higher in the CO 2 + H 2 S environment. This is because the iron sulfide (FeS) film, which is formed due to the presence of H 2 S, interferes with the formation of the carbonate scale (FeCO 3 ). This iron sulfide film can be more easily removed from the pipeline than the iron carbonate scale, which causes an increase in the corrosion rate. Whereas, a clear protective layer can be seen from the SEM images, which is due to the formation of FeCO 3 in an only CO 2 environment. In a chloride environment, a uniform corrosion and little bit of pitting formation can be seen. The corrosion data also support the same.

Raman Spectroscopy
In Figure 7a, the sample corroded in NaCl shows a Raman peak in two bands (at 250 and 1250 cm −1 ), due to the formation of FeOOH. This peak of band is also shown at 400 cm −1 due to the formation of magnetite (Fe 3 O 4 ). A low-density peak is also shown at 540 cm −1 , which indicates the formation of Fe(OH) 2 [37,38]. Figure 7b presents the Raman spectroscopy curve of the sample corroded in NaCl solution saturated with CO 2. FeCO 3 and Fe 3 C are the typical compositions of corrosion product during the CO 2 corrosion process shown in peaks in two bands (375, 1400 cm −1 and 400, 600 cm −1 ), while Fe 3 C is not the corrosion product but the remainder of ASS304 after the selective dissolution of austenite [39,40] . FeCO 3 is very protective in nature. In the absence of H 2 S, FeCO 3 could nucleate and grow rapidly around Fe 3 C and form a FeCO 3 scale, which gives better bonding stress to the steel and fastened Fe 3 C on the steel surface.

Raman Spectroscopy
In Figure 7a, the sample corroded in NaCl shows a Raman peak in two bands (at 250 and 1250 cm −1 ), due to the formation of FeOOH. This peak of band is also shown at 400 cm −1 due to the formation of magnetite (Fe3O4). A low-density peak is also shown at 540 cm −1 , which indicates the formation of Fe(OH)2 [37,38]. Figure 7b presents the Raman spectroscopy curve of the sample corroded in NaCl solution saturated with CO2. FeCO3 and Fe3C are the typical compositions of corrosion product during the CO2 corrosion process shown in peaks in two bands (375, 1400 cm −1 and 400, 600 cm −1 ), while Fe3C is not the corrosion product but the remainder of ASS304 after the selective dissolution of austenite [39,40]. FeCO3 is very protective in nature. In the absence of H2S, FeCO3 could nucleate and grow rapidly around Fe3C and form a FeCO3 scale, which gives better bonding stress to the steel and fastened Fe3C on the steel surface. Figure 6. Surface morphology of 304 stainless steel after corrosion in (a) untreated 304 SS, (b) 3.5 % NaCl + CO 2 environment, (c) 3.5 % NaCl + CO 2 + H 2 S environment, (d) 3.5% NaCl solution. Figure 7b presents the Raman spectroscopy curve of the sample corroded in NaCl solution saturated with CO 2 and H 2 S. During the dissolution of H 2 S in the solution, a porous and loose iron sufide layer formed due to the absorbance of HS − ion, which delayed the formation of FeCO 3 . As a result, the remained Fe 3 C was effortlessly wrinkled off in the dynamic aqueous phase. Furthermore, the FeS layer that is porous and cracked iron may contain little Fe 3 C, and it also effortlessly spalled off due to the poor bond stress of the scale to steel matrix.
It is well known that the iron sulfides formed in H 2 S-saturated solutions using electrochemical techniques [41]. The Raman spectrum of the sample after corrosion is in good agreement with the spectra attributed to the iron sulfide.  . Raman spectroscopy after a potentiodynamic polarization test in pH7 for three different solutions: (a) 3.5% NaCl solution, (b) 3.5 % NaCl + CO2 environment, (c) 3.5% NaCl + CO2 + H2S environment. Figure 7. Raman spectroscopy after a potentiodynamic polarization test in pH7 for three different solutions: (a) 3.5% NaCl solution, (b) 3.5 % NaCl + CO 2 environment, (c) 3.5% NaCl + CO 2 + H 2 S environment.

Conclusions
The corrosion behavior of 304 ASS steel has been studied under simulated sea environmental conditions in the presence of CO 2 and H 2 S. Both H 2 S and Cl − ions present in sea water can decrease the corrosion resistance of the 304 SS sample. However, in a CO 2-containing environment, due to the formation of stable iron carbonate, there was a positive effect on the corrosion resistance of the 304 ASS sample. The pH also plays an important role. The corrosion rate was in general high at lower pH level. So, lower pH and a combination of CO 2 + H 2 S with 3.5% NaCl has the most fatal effect on the corrosion resistance property of the 304 ASS sample. Raman spectroscopy analysis shows that the passive film formed under a saturated CO 2 environment is composed of FeCO 3 and is considered as the primary reason for this decreasing corrosion rate.