The X80 steel composition is listed in Table 1. Dogbone coupons were too heavy to measure milligram weight loss accurately in this work. Thus, three square coupons, each with a 1 cm² unpainted top surface (all other surfaces were covered with a polytetrafluoroethylene paint), were incubated without shaking in each anaerobic bottle with 150 ml SRB broth to obtain one MIC weight loss data point. Square coupons (1 cm² exposed surface) were also used as working electrodes in electrochemical glass cells. Dogbone specimens were used to test the mechanical properties. The dimensions of the dogbone coupons (Figure 1) were based on the ASTM E8/E8M standard (ASTM-E8/E8M-13a, 2013). The dogbone coupons were polished to 1,200 grit by the supplier. Each dogbone coupon was painted with polytetrafluoroethylene, except for a middle section with a width of 6 mm and length of 22 mm which was exposed to the SRB broth on all four sides. The top surfaces of all the square coupons (including the abiotic control and electrode coupons) were sequentially polished with 180, 400 and 600 grit abrasive papers. After that, all the coupons were cleaned with pure isopropanol and dried under UV light for 20 min.

D. vulgaris (ATCC 7757), a common SRB strain in MIC research, was selected for this research. The culture medium was ATCC 1249 medium, which is a modified Baar’s medium for sulfate reducers. The culture medium pH was adjusted to pH 7 using a NaOH solution. The culture medium was sterilized in an autoclave at 121°C. After autoclaving, the culture medium was deoxygenated using filter-sterilized N₂ sparging for more than 45 min. One hundred ppm (final concentration) L-cysteine was then added to the culture medium as an oxygen scavenger to reduce dissolved oxygen further and to mitigate possible slow oxygen leakage (Dou et al., 2019). Each anaerobic bottle was inoculated with 2 ml 3-day old SRB seed culture (grown in ATCC 1249 culture medium) before incubation at 37°C without shaking.

In this research, the X80 dogbone coupons were too large for SEM (scanning electron microscopy) imaging work and for obtaining accurate weight loss data. Thus, small X80 square coupons (1 cm² exposed top surface) were used to obtain the biofilm SEM images and weight loss data after 14 days of incubation 37°C. Each 450 ml anaerobic bottle contained 150 ml deoxygenated culture medium (fixed) with either 150 ml, 200 ml or 300 ml headspace (adjusted using inert glass beads or Epoxy blocks).

A SEM machine (FEI Quanta 250, Hillsboro, OR, United States) was used to observe the biofilm morphology on square coupons. Before the SEM observations, the cells and corrosion products/biomass were gentled rinsed with a PBS (phosphate buffered saline) solution for 15 s each for 3 times before sessile cell counting, and then soaked in 2.5% (w/w) glutaraldehyde biocide solution for 8 h at 10°C to immobilize the biofilm on each coupon. Then, the coupons were sequentially dehydrated with 50% (v/v), 70%, 80% 90%, and 95% ethanol sequentially for 10 min at each concentration and finally with 100% ethanol for 0.5 h. Subsequently, the coupon surfaces were sputter coated with Au to provide electric conductivity before the SEM observations.

After the 14-day incubation, the square coupons for weight loss data were cleaned with a fresh Clarke’s solution to remove biofilms and corrosion products before weighing. Each weight loss data point was the average of 3 replicate coupons from the same anaerobic bottle.

A potentiostat (Model VersaSTAT 3, Princeton Applied Research, Oak Ridge, TN, United States) was used to measure the electrochemical responses of the X80 working electrode (1 cm² surface) in SRB broth. Each glass cell contained 150 ml deoxygenated culture medium (fixed) with either 150 ml, 200 ml or 300 ml headspace (adjusted using Epoxy resin as space filler). Each bottle was inoculated with 2 ml 3-day old SRB seed culture for static incubation at 37°C. A saturated calomel electrode (SCE) was used as the reference electrode, and a thin platinum plate (10 mm × 10 mm × 1 mm) was used as the counter electrode. The abiotic control glass cell had 150 ml culture medium and 300 ml headspace without SRB inoculation. There was no need to vary the headspace for the abiotic control because there was no biogenic H₂S escape to the headspace.

Open circuit potential (OCP), LPR, EIS and potentiodynamic polarization analyses were performed. LPR was scanned at a rate of 0.1667 mV s⁻¹ in the range of −10 mV to +10 mV vs. OCP. EIS was performed at OCP by applying a sinusoidal signal of 10 mV (amplitude) in the frequency ranging from 10⁴ to 10^–2 Hz. Potentiodynamic polarization curves were measured at the end of the 14-day incubation from OCP to OCP −200 mV using one working electrode, and from OCP to OCP +200 mV using another working electrode in a replicate glass cell at a rate of 0.1667 mV s⁻¹. The corrosion potential (E _corr), corrosion current density (i _corr), and anodic and (absolute) cathodic Tafel slopes (β _a and β _c) were determined from a Tafel analysis of the polarization curves.

Dogbone coupons were immersed in anaerobic bottles with 150 ml culture medium and varied headspace volumes (150 ml, 200 ml and 300 ml), and each anaerobic bottle contained one dogbone coupon. The headspace variation was achieved using Epoxy resin as space filler because the bottle volume and liquid culture medium volume were the same, but the headspace volume varied.

The concentration of H₂S and total pressure in different anaerobic bottles were measured using a portable H₂S sensor (GAXT-H-DL, BW Technologies, Calgary, Alberta, Canada), a digital manometer (Xplorer GLX-PS-2002, PASCO scientific, Roseville, CA, United States), respectively. The H₂S sensor had an upper limit of 100 ppm (v/v). If a headspace sample had a higher concentration, dilution was required (Wang et al., 2020). A 125 ml anaerobic vial sealed with 1 atm air was injected with 10 ml headspace gas for 12.5X dilution. After mixing, a syringe was used to flush and flood the H₂S sensor’s port with 40 ml of the headspace gas before taking a meter reading.

Sessile cells on a dogbone coupon were counted using a hemocytometer under a 400X microscope. Each dogbone coupon had a total area of 4 cm² covered by the SRB biofilm. The biofilm biomass was first scrapped off the coupon and then suspended in a PBS buffer before counting. Because D. vulgaris cells were seen motile, they were easily distinguished from artifacts (Wang et al., 2020).

The biofilms and corrosion products on the dogbone coupon surfaces were removed using a fresh Clarke’s solution according to ASTM G1–03. After the removal, the maximum pit depth for each dogbone coupon was obtained under an InfiniteFocus Microscopy (IFM) machine (Model ALC13, Alicona Imaging GmbH, Graz, Austria).

After the pit depth analysis, tensile tests were performed on an electromechanical universal testing machine (E44.304, MTS system, MN, United States) on the same dogbone coupons.

The SEM biofilm images in Figure 2 show the surface morphologies of the D. vulgaris biofilms for different headspace volumes after the 14-day incubation. The short rod shape is typical for D. vulgaris, consistent with SEM images in other studies (Unsal et al., 2022; Wang et al., 2022). Figure 2 indicates that the number of sessile cells increased with increasing headspace volume. This qualitative information is consistent with quantitative sessile cell count data on dogbone coupons discussed below. A larger headspace led to less H₂S cytotoxicity in the broth and thus better SRB growth (Jia et al., 2019a).

The weight losses for 150 ml, 200 ml and 300 headspace volumes were 1.7 ± 0.17 mg cm⁻², 1.9 ± 0.33 mg cm⁻² and 2.3 ± 0.37 mg cm⁻², respectively (Figure 3). Although the neighboring weight loss data were close with fairly wide error bars as a result of the short-term test, the 150 ml and 300 ml weight losses had a p-value < 0.05, indicating that the weight increased with statistical significance when the headspace volume increased from 150 to 300 ml. These weight losses after SRB incubation were much larger than the 0.2 ± 0.05 mg cm⁻² abiotic carbon steel weight loss obtained after 14 days of incubation in the deoxygenated ATCC 1249 culture medium without SRB inoculation. The increasing SRB MIC weight loss trend corresponds to the increasing sessile cell trend observed in Figure 2, which is consistent with EET-MIC, in which more sessile cells harvest more electrons from elemental iron, leading to more severe corrosion (Jia et al., 2018).

The OCP trends for different headspace volumes during the 14-day incubation of X80 electrode in the SRB culture medium are shown in Figure 4A. A lower OCP indicates a higher tendency for the working electrode to lose electrons. Figure 4A does not consistently indicate that a higher headspace volume had a lower OCP. This is not surprising for complicated SRB systems (Tran Thi Thuy et al., 2020). After all, OCP only indicates corrosion tendency, but the actual corrosion outcome relies on corrosion kinetics. The same observation was made previously in a study on biogenic H₂S impact on carbon steel MIC by D. vulgaris in ATCC 1249 culture medium which included abiotic OCP (Jia et al., 2018).

Polarization resistance (R _p) from LPR scans in Figure 4B describes the transient corrosion kinetics during the 14-day incubation. R _p is inversely proportional to corrosion rate (Cai et al., 2022). Figure 4B shows a large drop of R _p in the first 3 days, suggesting that as biofilm established on the metal surface, corrosion rate increased. The abiotic R _p curve for X80 in the deoxygenated ATCC 1249 culture medium remained around 17–18 kΩ cm², much higher than the biotic R _p curves. Figure 4B also shows that R _p for the 300 ml headspace was the lowest, and R _p for 150 ml was the highest, indicating highest and lowest corrosion rate, respectively, which is consistent with weight loss data trend in Figure 3.

For EIS, the Nyquist and Bode plots of the abiotic and the biotic X80 coupons for different immersion times and different headspace volumes are shown in Figure 5. The abiotic EIS data in the deoxygenated ATCC 1249 culture medium were show the same trend with abiotic R _p trend. The Nyquist plots indicate a capacitive behavior. A larger diameter of the semi-circle in the Nyquist plot means a higher corrosion resistance in Figure 5A. The EIS data in Figure 5 were fitted with the equivalent electrical circuits in Figure 6. A simple one-time constant circuit was needed for the abiotic control EIS spectra, while the biotic EIS spectra required a two-time constant circuit. The fitted parameters are summarized in Table 2. The biotic impedance spectra for the three different headspace volumes (150 ml, 200 ml and 300 ml) fitted well with a two-time constant circuit model. The capacitors in the circuit model were not ideal capacitors. Thus, constant phase elements (CPE_s) were used instead with n values in Table 2 indicating how close the CPEs (constant phase elements) were to capacitors (n = 1). The biotic equivalent circuit in Figure 6B contains: 1) solution resistance (R _s), 2) a parallel combination of charge transfer resistance (R _ct) and CPE₁ (Q _dl) associated with the metal surface electric double layer, 3) a parallel combination of biofilm resistance (R _f) and CPE₂ (Q _f) associated with the biofilm/corrosion product layer on the X80 steel surface. The abiotic equivalent circuit in Figure 6A is simpler without R _f and Q _f.

Compared with charge resistance (R _ct) values, the film resistance R _f values were quite small. However, the R _f values became larger with the increasing headspace volume due to D. vulgaris becoming more corrosive, which is consistent with the increased sessile cell count (Figure 2). R _ct was rate limiting in this study because it was much larger than R _s and R _f. (R _ct + R _f) is often used as the equivalent to R _p in qualitative corrosion analysis. In this work, (R _ct + R _f) was the smallest for 300 ml headspace in Table 2, indicating the highest corrosion rate.

The Tafel plots of X80 are shown in Figure 7. The corrosion current densities from the Tafel analysis of the potentiodynamic polarization curves are listed in Table 3. After the 14 days of SRB incubation, the coupon for the 300 ml headspace had the highest corrosion current density (i _corr) of 74.8 μA cm⁻² (Table 3), compared to 19.1 μA cm⁻² (for 200 ml) and 4.8 μA cm⁻² (for 150 ml). The abiotic i _corr in the deoxygenated ATCC 1249 culture medium was 0.79 μA cm⁻², which was negligibly small. The corrosion current density trend here corroborates the R _p ⁻¹ trend in Figure 4B and (R _ct + R _f)⁻¹ trend in Table 2. Thus, all the electrochemical corrosion data trends, with the exception of OCP, are consistent with the weight loss trend, all pointing to more sessile cells for faster MIC, which is characteristic of EET-MIC.

Table 4 shows that the H₂S concentrations in the headspace gas phases for the anaerobic bottles (each containing one dogbone coupon) with headspace volumes of 150 ml, 200 ml and 300 ml were 8.50 × 10³ ppm (v/v), 7.75 × 10³ ppm, and 7.28 × 10³ ppm, respectively after the 14-day SRB incubation. The corresponding H₂S concentration in the liquid phase was estimated based on H₂S equilibrium at 37°C according to a published report (Ning et al., 2014). The dissolved [H₂S] values for the headspace volumes of 150 ml, 200 ml and 300 ml were 1.06 mM, 0.95 mM and 0.84 mM, respectively (Figure 8; Table 4). As expected, a larger headspace allowed more H₂S to escape from the liquid phase in order to reach a different H₂S equilibrium between the gas and liquid phases. Figure 8 also shows that the final broth pH values were 7.08, 7.26, and 7.54 corresponding to headspace volumes of 150 ml, 200 ml and 300 ml, respectively. The broth pH increased slightly with the increasing headspace volume, because a larger headspace allowed more H₂S to escape and this took away more protons from the broth as shown in Reaction (5). According to the following reaction (Jia et al., 2019a), HS − + H + ⇌ H 2 S (5)

The dissolved [H₂S] in the broth became lower with the increasing headspace volume while the pH value became higher (Table 4; Figure 8). All the pH values in this work were above and not far from 7. This is different from abiotic H₂S corrosion studies, in which researchers introduce exogenous H₂S to an aqueous solution and thus resulting in acidic pH, which is needed to cause appreciable abiotic H₂S corrosion (Sun and Nesic, 2007).

After the 14-day incubation, the sessile cell count was found to be higher in the anaerobic bottle with a larger headspace volume (Figure 9). The cell counts on coupons in the bottles with the headspace volumes of 150 ml, 200 ml and 300 ml were 6.5×10⁷ cells cm⁻², 2.3×10⁸ cells cm⁻² and 1.4×10⁹ cells cm⁻², respectively. The increasing sessile cell count trend agrees with the decreasing dissolved [H₂S] in Table 4. Decreased [H₂S] means less toxicity and thus better sessile cell growth (Jia et al., 2019b). Although the 300 ml headspace bottle had lower H₂S concentrations in both the gas and the liquid phases, its total amount (1.47 × 10^–4 mol) was higher than in the bottles with 150 ml and 200 ml headspace volumes. This was reasonable because less H₂S toxicity allowed better SRB growth and thus produced more H₂S in the total amount in the liquid and headspace of a sealed anaerobic bottle.

Coupon surface morphologies on dogbone coupons after the 14-day incubation with biofilms and corrosion products removed were examined under IFM. Figure 10A shows that the abiotic coupon surface exhibited polished coupon surface roughness (y-scale enlarged to show details). For the biotic dogbone coupons, the maximum pit depth increased with a larger headspace volume in Figures 10B–D. They were 2.6 μm, 4.2 μm and 6.2 μm for headspace volumes of 150 ml, 200 ml and 300 ml, respectively. The pit depth trend here is consistent with the weight loss data trend. In future studies, pit density should be investigated as well (Javed et al., 2020a; Javed et al., 2020b). With a larger headspace, there was a lower amount of dissolved [H₂S] and more sessile cells, which led to higher weight loss and deeper pits. The maximum pit depth increased by 58% when the headspace increased from 150 ml to 300 ml, while the broth volume was fixed at 150 ml.

Figure 11 shows the stress-strain curves of X80 dogbone coupons. The dogbone coupons were retrieved after they had been immersed in SRB bottles with fixed 150 ml culture medium volume and varied headspace volumes (150 ml, 200 ml and 300 ml) for 14 days at 37°C. The ultimate (tensile) strength is the maximum stress that a material can withstand before final failure (Thamma and Jantasorn, 2022). It is the highest point of the Y-axis in Figure 11. The ultimate strain (elongation at break) demonstrates the ability of a material to resist shape change before finally breaking (Tian et al., 2021). It is the largest value of the X-axis (strain) in Figure 11 (Sluzalec, 1992). Lowering of these parameters can reflect the mechanical property degradation of the material under different conditions such as different MIC severity.

The ultimate strength of abiotic control X80 carbon steel was 853 ± 3 MPa. The ultimate tensile strength values of the abiotic dogbone coupon, and biotic dogbone coupons from bottles with different headspace volumes were all quite close as shown in Figure 11. Compared with the abiotic dogbone, in the presence of SRB with headspace volumes of 150 ml, 200 ml 300 ml, the ultimate strength losses were 3%, 2% and 0%, respectively (Table 5). These values were rather small. On the other hand, ultimate strain was reduced in the presence of SRB. Compared with the abiotic dogbone, in the presence of SRB with headspace volumes of 150 ml, 200 ml 300 ml, the ultimate strain losses were 6%, 13% and 23%, respectively (Table 5). With an increased headspace, MIC severity increased, making X80 steel more brittle. The corrosion damage by SRB pitting was the main factor in its mechanical property degradation study. H₂S was unlike the driving force behind the relatively large ultimate strain loss, because in this work, more severe MIC corresponded with lower [H₂S] in the broth.