The Effect of H2S Pressure on the Formation of Multiple Corrosion Products on 316L Stainless Steel Surface

H2S gas when exposed to metal can be responsible for both general and localized corrosion, which depend on several parameters such as H2S concentration and the corrosion product layer formed. Therefore, the formation of passive film on 316L steel when exposed to H2S environment was investigated using several analysis methods such as FESEM and STEM/EDS analyses, which identified a sulfur species underneath the porous structure of the passive film. X-ray photoelectron spectroscopy analysis demonstrated that the first layer of CrO3 and Cr2O3 was dissolved, accelerated by the presence of H2S-Cl−. An FeS2 layer was formed by incorporation of Fe and sulfide; then, passivation by Mo took place by forming a MoO2 layer. NiO, Ni(OH)2, and NiS barriers are formed as final protection for 316L steel. Therefore, Ni and Mo play an important role as a dual barrier to maintain the stability of 316L steel in high pH2S environments. For safety concern, this paper is aimed to point out a few challenges dealing with high partial pressure of H2S and limitation of 316L steel under highly sour condition for the oil and gas production system.


Introduction
Southeast Asia is a rich source of natural gas and petroleum, especially Brunei, Vietnam, Malaysia, and Indonesia. From GlobalData statistic, Malaysia and Indonesia will contribute around 80% and 70% of the Southeast Asia's total crude oil and natural gas production from eight planned projects in 2025, respectively. In brief, once the well is successfully drilled and installation is completed, the product must be transported to a facility where it can be treated, stored, processed, and refined through the pipeline system. erefore, the environmental conditions of the production and pipeline systems are required for the prediction of materials life cycle and its maintenance requirements. us, any corrosion that occurs inside oil and gas pipeline systems is a serious factor that could lead to system failure. Corrosion not only causes economic losses but also greatly affects the safety and protection of the oil and gas resource. Hence, in upstream oil and gas applications, the most desirable alloy should be made of a strong material with good localized corrosion resistance, low cost, and suitable mechanical characteristics.
Austenitic stainless steel type 316L (UNS S31603) has excellent corrosion resistance to the electrochemical properties of the passive film that forms on its surface. e passive film mainly consists of iron (Fe), chromium (Cr), and nickel (Ni) oxides. Cr promotes the formation of protective surface oxide, and Ni enhances the stability of the oxide film. us, stainless steel alloys with a higher composition of Cr and Ni prevent the iron from rusting and provide heat resistant properties [1]. Besides, 316L steel is a low carbon alloy that contains molybdenum (Mo), which also makes it more corrosion resistant, especially in highly sour conditions [1]. e physicochemistry of passive films directly influences the film properties. erefore, many studies have investigated the effect of hydrogen sulfide corrosion at different hydrogen sulfide partial pressures (pH 2 S) [2], pH [3], and temperature [4] on the material surface. However, this area remains underexplored due to the cross effects between the parameters of the study [3]. Ding et al. [1] investigated the electrochemical behavior of 316L steel in Cl − solutions under different pH 2 S. Additionally, Gao et al. [4] found an increased presence of iron oxide (Fe 3 O 4 ) and iron sulfide in the corrosion product layer on substrate after performing H 2 S corrosion tests involving increased exposure times at 120°C. When both H 2 S and Cl − were present, the passive film would become more sensitive to temperature, therefore significantly affecting the corrosion rate, the corrosion mechanism, and the properties of the material surface. Although various corrosion reactions could occur, the corrosion rate generally increases in the presence of H 2 S until sulfide saturation was achieved [2,5]. erefore, the authors aim to determine the effect of different pH 2 S exposure on the formation of multiple corrosion product layer of the 316L steel surface. e surface analysis of 316L steel was performed using a multianalytical technique consisting of X-ray photoelectron spectroscopy (XPS), field emission scanning electron microscopy (FESEM), and energy dispersive X-ray spectrometry (EDS). e microstructure characterization and crystalline phase characterization of 316L steels were performed using transmission electron microscopy (TEM). e metal ions that had dissolved in the test solution after 7 days were characterized via inductively coupled plasma-mass spectrometry (ICP-MS). Finally, the reaction mechanism occurring on the surface of 316L steel was explained in terms of the dissolution of CrO 3 and Cr 2 O 3 on the steel protective film at 3 bar pH 2 S after 7 days of testing.

Materials and Method.
For the experiment, a commercial austenitic stainless steel type 316L with a chemical composition listed in Table 1 was used. en, optical emission vacuum spectrometer analysis was performed on the stainless steel using the point-to-plane excitation technique, following ASTM E 1086 Standards [6]. e 316 L steel sample was machined into a dimension of 20 × 20 × 3 mm 3 , abraded with a series of silicon carbide papers (up to 1200 grit), polished with a diamond spray (up to 1.0 μm), rinsed with distilled water, degreased with acetone, dried at room temperature, and then stored in a desiccator. Each procedure for preparing the 316L steels was performed following ASTM G1-03 Standards [7]. All experiments were carried out in NACE TM0177 Solution A [8]. e solutions were added to an autoclave and deoxygenated by purging with nitrogen gas for two hours before the beginning of each experiment. All experiments were conducted at 60°C and pH 3.0 ± 0.5 to simulate the crude oil conditions in Southeast Asia especially around Malaysian region (30°C-90°C and pH 2-6) [9].

Experimental Procedure and Weight-Loss Measurement.
Experiments were carried out in a 5 L autoclave using experimental conditions which are summarized in Table 2. A schematic diagram of the corrosion test apparatus was set up for the 0 bar and 3 bar of pH 2 S and is shown in Figures 1(a) and 1(b), respectively, in Sour lab, at DNVGL Singapore laboratory. e samples were weighed using an analytical balance for the initial weight (W i ), before being immersed vertically in test solution. en, the 316L steel sample with a holder was attached to the top lid and sealed in the autoclave before immersing it vertically in the test solution. Besides, 3.5 L of the NACE A solution and 1.5 L vapor were used to allow the water vapor to expand in high pressure and temperature conditions. After that, a deoxygenation process was conducted by purging the autoclave with nitrogen gas for two hours before each test. e samples were heated at 60°C and pressurized with premixed gases (3 bar of H 2 S and 27 bar of N 2 gas) equaling a total pressure of 30 bar until a stable pressure was reached. e pH and temperature of the solution were measured and recorded at the start and end of each experiment as shown in Table 2. After 7 days of immersion, the samples were taken out from the solution, cleaned, and preserved with nitrogen according to ASTM G1-03 Standards [7]. e final samples were weighed again to obtain the final weight (W f ) based on which weight loss was determined. After that, the corrosion rate was calculated using the following equation: where K � constant, W � weight loss in mg, ρ � density of the metal in g/cm 3 , A � area in cm 2 , and T � time of exposure in hours. en, the procedure was repeated with H 2 S-free conditions for the control sample in this experiment.

Characterization of Corrosion Products.
Morphological images characterization of all 316L steels were performed using a Zeiss SUPRA 55VP field emission scanning electron microscope (FESEM) and a high-efficiency in-lens detector for clear topographic imaging in high-vacuum mode, in conjunction with energy dispersive X-ray spectroscopy (EDS) to characterize the elemental composition of the samples. e surface analysis was performed using X-ray photoelectron spectroscopy (XPS) (X-ray microprobe PHI Quantera II, Ulvac-PHI, Inc.) with a monochromatic Al-Kα (hv � 1486.6 eV) X-ray source. Before deconvolution, the charging effects were corrected using a Kratos charge neutralizer system in order to minimize the carbon charging effect. In this case, all spectrum is corrected and referred to adventitious carbon binding energy at 284.8 eV. A TEM thin foil was prepared using the focused ion beam (FIB) technique (Helios Nanolab 600i, FEI). Also, detailed subsurface microstructure characterization was performed on a transmission electron microscope (TEM) (Tecnai G2 F20, FEI) with an EDS detector operated at 200 kV. Finally, the elemental analyses of the corrosion product dissolution in the test solutions were carried out on an inductively coupled plasma-mass spectrometer (ICP-MS) PerkinElmer Sciex Elan 9000.

e Effect of H 2 S Pressure on the Corrosion Rate.
e average corrosion rate of 316L steels after immersion in test solution at 0 bar and 3 bar pH 2 S were calculated using weight loss measurement. e 316L steel at 3 bar pH 2 S had higher corrosion rate compared to that in 0 bar pH 2 S of 316L steel. e corrosion rate increased by 90.1% from 0.07 mm/yr at 0 bar pH 2 S to 0.74 mm/yr at 3 bar pH 2 S due to aggressiveness of sulfide ions in test solution [1]. e data show that overall mass loss is symmetrical with H 2 S pressure, and sulfide ions play a significant role in determining the kind of corrosion scales and reducing the surface protection of 316L steel [1].

FESEM-EDS.
e surface morphology of the austenitic 316L steels immersed in test solution without H 2 S at 60°C for 7 days is shown in Figure 2(a). Under an environment free of H 2 S, the 316L steel surface showed scratch marks because of polishing, but no grains or pits are shown. Besides, the elemental composition of 316L steel free of H 2 S was done using EDS, as shown in Figure 2(a). e surface of the 316L steel in the H 2 S-free conditions showed that the passive film was formed in the presence of Cr. However, when the 3 bar pH 2 S was applied, the surface of the 316L steel shows a cracked surface, which indicates local breakdown of the passive layer, as shown in Figure 2 e sulfide compounds were detected on the 316L steel surface as a result of applying pH 2 S. Due to local breakdown of the protective layer, microcracking was observed, as shown in Figure 2 erefore, corrosion products are formed after 3 bar pH 2 S exposure, demonstrating that the protectiveness of the passive film had very much degraded. e EDS analysis in Figure 2(b) shows the presence of a sulfur element, suggesting the presence of metal sulfides layer on the 316L steel surface [10]. However, Cr signals were depleted in the protective layer after exposure with H 2 S due to the dissolution of Cr in the test solution, as reported by Beverskog and Puigdomenech (1997) [11].

X-Ray Photoelectron Spectroscopy (XPS).
e chemical composition of the 316L steel passive films that formed under different pH 2 S conditions after 7 days of immersion in test solution was investigated using XPS measurements, as shown in Figure 3. e major peaks present in the spectra for both 316L steels corresponded to C, O, Fe, Cr, Ni, Mo, and S. Generally, austenitic stainless steel is self-passivated, as chromium oxide (CrO 3 ) instantly forms when it is exposed to air, but the nature of the passive layer changes when exposed to an aqueous solution [12]. Apart from that, the element contribution analysis of Cr and Mo was performed on the sample without exposure to H 2 S as shown in Figure 4.
Meanwhile, the elemental contribution analysis of the specimen after immersion in the test solution in H 2 S-containing conditions yielded Ni, Fe, Mo, and S, as shown in Figure 5.
e Ni 2p narrow spectra consisted of three compounds, namely, Ni(OH) 2 (857.4 eV), NiO (855.7 eV), and NiS (853.9 eV), as shown in Figure 5(a). e dominant peaks of Ni 2p 3/2 in the spectrum corresponded to NiO, indicating that NiO could stably exist in the passive film. However, Figure 5(b) shows the narrow scan spectra of Fe 2p 3/2 obtained on the 316L steel sample, which was immersed in the H 2 S solution at 3 bar pH 2 S for 7 days. Two deconvolution signals were observed, attributed to Fe 3 O 4 (710.1 eV) and FeS (712.8 eV). From the reactivity series, Fe was more active than Ni and could participate in the passive film formation in the form of Fe 3 O 4 , Fe 2 O 3 , Fe(OH) 2 , FeOOH, and Fe(OH) 3 [13]. However, iron sulfide (FeS) also appeared in the film due to the high affinity of sulfides with iron in the H 2 S solution [14]. Moreover, the possibility of Fe species dissolved preferentially in the H 2 S-containing solution and affected the migration of S 2− into the passive film, preventing the formation of Fe 2 O 3 on the passive film [13]. Apart from that, the Mo 3d 3/2 spectra consisted of MoO 3 (3d 3/2 ; 235.5 eV and 3d 5/2 ; 232.4 eV), MoO 2 (3d 3/2 ; 230.2 eV), and bulk Mo from metal (226.9 eV), as shown in Figure 5(c).   [15]. Although the H 2 S molecule has a chemical structure similar to H 2 O, the polarizability of the sulfides such as S 2− or HS − is higher than that of halides (Cl − ) and OH − when adsorbed onto the passive film and could appear in that film [16]. Since chromium sulfides were not detected, S 2could only unite with Fe 2+ and Ni 2+ . erefore, from the deconvolution of S 2p shown in Figure 5(d), four peaks were fitted by the narrow spectra at the binding energy of FeS 2 (2p 3/2 ; 162.2 eV), NiS (2p 3/2 ; 162.9 eV), FeS 2 (2p 1/2 ; 163.7 eV), and NiS (2p 1/2 ; 164.5 eV) corresponding to disulfide (S 2 2− ), S 2− , and S, respectively [16]. Based on the reactivity series of metals, nickel is less reactive than Cr and Fe, but nickel oxide could be porous and could easily entrap S 2within the porous oxide. However, traces of NiO and/or Ni(OH) 2 were also found in the passive film, similar to that reported by Luo [17].
e O 1s narrow scan spectra for austenitic 316L steel for both 0 bar and 3 bar H 2 S-containing conditions are shown in Figure 6. From the deconvolution of O 1s, several oxide  4 eV), formed via the lattice nonconserving reaction through diffusion vacancy or interstitial vacancy filling based on the cation-anion reaction [18]. Besides, the Cr content in the passive film was very low, thus having no corresponding XPS spectra and revealing a decreased stability of this species at high H 2 S-Cl − concentration in the test solution.
Based on the XPS deconvolution in the atomic concentration analysis of Table 3

Inductively Coupled Plasma-Mass Spectrometry (ICP-MS).
Metal ions of 0 bar and 3 bar pH 2 S after immersion in the test solutions for 7 days were evaluated via ICP-MS, as shown in Figure 7. Y-axis in Figure 7 shows metal ions concentration after 7 days immersion in the test solution, while x-axis shows the main elemental composition in the 316L steel with two plots for each element which represents a partial pressure of H 2 S. In brief, both solutions (0 and 3 bar pH 2 S) showed no Mo, but Fe and Cr were selectively dissolved. e highest concentration of Fe dissolved in the test solution suggests that Fe is the main component of the 316L stainless steel. e selective dissolution of Cr became more pronounced as 3 bar pH 2 S was exposed. One point to note is that metals that rely on Cr 2 O 3 for corrosion protection may be at risk of having the Cr 2 O 3 dissolve into the solution with increased pH 2 S. is issue is a considerable concern since an increase in the concentration of HS − , S 2-, or Cl − anions could accelerate the Cr 2 O 3 film dissolution. In comparison to the 0 bar pH 2 S sample, the sample exposed to 3 bar pH 2 S had significantly increased the concentrations of Fe and Cr. Castle and Qiu (1990) [20] reported that Cr dissolves into solution as trivalent ions, while Fe dissolves as divalent ions. Additionally, the adsorbance of S on the metal surface produces metal sulfides that make the oxide film more   Table 3 shows that the Cr content was very low for the sample exposed to 3 bar pH 2 S, indicating that Cr dissolution had indeed happened. Pourbaix [11] [22,23]. Overall, the dissolution process, which is the release of alloying elements in the electrolyte from the bulk or the film, was investigated using ICP-MS after 7 days of exposing the sample in NACE A solution. e main elements to dissolve were iron and chromium, which were detected in the electrolyte in the same proportion as the chemical composition of the 316L steel (Table 1).

Transmission Electron Microscopy (TEM).
Transmission electron microscopy (TEM) was conducted to further investigate the structure of passive films. TEM micrograph with the point EDS of both 316L steels was recorded as shown in Figure 9. 316L steels were prepared using FIB with a platinum coating layer, an oxide layer, and a substrate of 316L steel. e layer formed in both samples showed different patterns and multilayered in the 3 bar pH 2 S 316L steel (Figure 9(c)) compared to that in the 0 bar pH 2 S 316L steel (Figure 9(a)). ese films acted as layers that barred the reaction products from the corrosive environment and took the form of mono-or multilayer oxygen or other chemical species adsorbed onto the metal surface. e four-point EDS (marked as P1, P2, P3, and P4) was analyzed in both 0 bar and 3 bar pH 2 S conditions. P1 represents the outer layer/ interface, P2 and P3 represent the inner layer, and P4 is the sample substrate. is analysis focused on elements that had       However, Mn and Cr elements were not detected at P1 in 3 bar pH 2 S, but the S element was present in the outer layer of the 3 bar pH 2 S sample as shown in Figure 9(d). e Cr signal disappeared at the outer layer, due to formation of Cr(OH) 3 , hence suggesting that more Cr(OH) 3 was dissolved in the test solution [24]. us, one of the primary elements that may contribute to P1 in the 3 bar pH 2 S is MoO 3 and a result that is in good agreement with the XPS results. Moreover, points P2 and P3 represent the inner part of the multilayer that gives strong signals of Fe, Cr, and Ni for 0 bar pH 2 S and Fe, Ni, and S for the EDS analysis of 3 bar pH 2 S. In addition, Ni enriched about 51.42% at P2 and 53.43% at P3 for the 3 bar pH 2 S sample (Figure 9(d)). As the EDS and XPS results of 0 bar pH 2 S are linked, the primary elements that contributed to P2 and P3 were MoO 3 , MoO 2 , Cr 2 O 3 , and CrO 3 . However, it has been proposed that Mo could thicken the passive film, increase the surface affinity towards oxygen, and decrease the propensity for Cl − to adsorb or form an extra secondary protective film [18]. Nevertheless, the primary elements that contributed to P2 and P3 at 3 bar pH 2 S were NiO, Ni(OH) 2 , NiS, FeS, Fe 3 O 4 , and FeS 2 , consistent with the EDS and XPS results. Since Ni did not participate in the passive film formation, it would be segregated by the oxides, and it consequently enriched underneath the passive film in the form of NiO and Ni(OH) 2 , as also verified by the XPS results in Figure 5(a). Finally, EDS P4 showed a strong signal attributed to the Fe element in both 0 bar and 3 bar pH 2 S conditions. e atomic composition showed Fe enriched by about 61.01% and 74.07% at 0 bar pH 2 S and 3 bar pH 2 S, respectively, both of which correspond to the bare metal. Meanwhile, the enlargement of TEM micrograph of 3 bar pH 2 S 316L steel in Figure 9(c) shows that the porousness of the oxide layer was precipitated at the upper layer of the film. Generally, passive films are formed with a highly disordered "barrier" layer adjacent to the substrate and film, comprising a precipitated phase that may incorporate anion and/or cation from the solution [25]. However, the contribution of ions in the NACE A solution, i.e., NaCl, CH 3 COOH, and H 2 O in the presence of H 2 S, affected the oxide layer of the 316L steel samples, forming multilayer corrosion products after 3 bar partial pressure H 2 S exposure as shown in Figure 10.

3.6.
e Formation and Dissolution of 316L Corrosion Products. A passive film is envisaged to be a bilayer or multilayer structure comprising a point defective, nanocrystalline barrier layer and a porous outer layer that is formed by the hydrolysis of cations, depending upon the local conditions [25]. Corrosion products formation and dissolution are a continuous process and, generally, the film will have defects (vacancies and interstitials). At 3 bar pH 2 S condition, a decrease in corrosion resistance due to the introduction of sulfide ions (H 2 S, HS − , and S 2-) to 316L steel was associated with the increase in proton reduction, leading to the anodic dissolution. Sulfides in the passive film are reported to have more defects than oxides [16], and thus, the film is less protective and stable in a high sour environment. In brief, Cr and Mo oxides were first generated under H 2 S-free conditions following the reactions in equations (2)-(3) for Mo oxides and the reactions in equations (4)-(6) for the Cr oxides: After long-time passivation, a stable passive film forms on the 316L steel surface with a dual-layer structure, as shown in Figure 9(a). e amorphous outer layer consists of CrO 3 , Cr 2 O 3 , MoO 3 , and MoO 2 in H 2 S-free conditions, as proven by the XPS surface analysis (Figure 4). In addition, MoO 2 and MoO 3 metals are also formed after exposure to 3 bar pH 2 S via a reaction with hydroxyl, similar to the reactions in equations (2)-(3), respectively. e passivation of Mo at higher pH 2 S conditions could mainly be associated with the formation of MoS 2 (equation (7)) [26]. However, the replacement of MoS 2 by MoO 2 could be predicted by the reaction with water as shown in equation (8) [26,27]: Since the corrosion product of Cr compound did not form after 3 bar pH 2 S exposure due to dissolution of Cr in the test solution to produce Cr 3+ (Figure 8), as per the reactions are shown in equations (9)-(11): e depletion of Cr in the passive film resulted in the creation of a porous-precipitated layer because of the vacancies of cations and anions, as shown in Figure 10. us, depletion of Cr induces a higher S content in the passive film and results in passive film degradation. Besides, Ni(OH) 2 , NiO, and NiS are formed after exposure to 3 bar pH 2 S, respectively, based on the reactions in equation (12)- (14): Ni(OH) 2 ⟶ NiO + H 2 O (13) Aside from that, Fe 3 O 4 , FeS, and FeS 2 formations were slightly different compared to other elements in 3 bar pH 2 S conditions. e OH − from the solution reacted first with ionic Fe in the outer passive film layer to form iron hydroxide followed by oxidation to form Fe 3 O 4 . e main reactions are described by the following equations [28]: In an H 2 S environment, HS − or S 2could also incorporate into the film to form FeS and FeS 2 based on the reactions in equations (18) and (19) [29]: Fe 2+ + 2HS − ⟶ FeS 2 + H 2 Fe 2+ + S 2− (19) erefore, the protectiveness of the 316L passive film could decrease, and the corrosion resistance of the 316L also deteriorates at higher H 2 S conditions. Hence, different materials could be subjected to further study and are analyzed at a range of different scales for safety concern, relating to the oil and gas production system.

Conclusions
From this study, the effect of H 2 S pressure on the formation of corrosion products on the 316L steels was investigated. e corrosion rate was increased at 3 bar pH 2 S, and the morphology of corrosion product experienced cracked surface and local breakdown of barrier layer compared to H 2 S-free conditions. Similarly, the TEM results have shown that the 316L steel at 3 bar pH 2 S condition exhibits more porous in the corrosion product layer. XPS results indicated the signals (peaks) of passive films which are Mo, MoO 3 , NiO, Ni(OH) 2 , NiS, Fe 3 O 4 , and FeS 2 . However, in H 2 S-free conditions, the only passive film was observed at MoO 2 , MoO 3 , Cr 2 O 3 , and CrO 3 signals. Finally, Cr signals in XPS analysis were decreased after exposure to 3 bar pH 2 S, and it has been confirmed by ICP-MS that the chromium was dissolved into solution. erefore, 316L could not be sustained in high partial pressure of H 2 S due to local breakdown of the crystalline layer, and the protectiveness of the passive film had very much degraded.

Data Availability
e Pourbaix diagrams of Cr with pH 2 S-free and pH 2 Scontaining conditions in aqueous at T � 25°C [23,24] and data supporting (inductively coupled plasma-mass spectrometry (ICP-MS) analysis) are from previously reported studies and datasets, which have been cited. e processed data are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.