**3. Results**

#### *3.1. Surfaces of Coupons Before Cleaning*

Photos of coupon surfaces were taken directly after removal from the test solution at the end of the 8-week immersion period (see Figure 1). The T1 and T3 test condition coupons had orange and black corrosion products on the surfaces, the T4 coupons were relatively free of corrosion products and black films were present on the remaining treatments (i.e., T2, T5 and T6).

The surfaces of individual test replicate coupons set aside for biofilm/corrosion product examination were observed using the SEM (Figure 2), and analysed by EDS (Figure 3). The coupon from the uninoculated test (T1) was covered in a general corrosion product, which appears to be a form of iron oxide. The images of the coupon exposed to the orange tubercle biofilm inoculum (T2) showed a biofilm present, with phosphorus and a relatively high sulfur content detected by the EDS (Figure 3). The coupon from the *D. desulfuricans* inoculum under aerobic incubation (T3) visually had a thicker biofilm on the surface compared to the coupon from the *D. desulfuricans* inoculum under anaerobic incubation (T6) (Figure 2). The biofilm of the T3 coupon also had much more sulfur present (relative to iron) compared to the T6 coupon, according to EDS (Figure 3). For the defined mixed microbial species tests (T4 and T5), the consortia including *D. desulfuricans* (T5) visually had a much denser biofilm compared to the sparse biofilm of T4 (Figure 2). No sulfur was detected in the mixed consortia without *D. desulfuricans* (T4) (Figure 3), while a small amount of sulfur in the biofilm was detected in the defined microbial consortia when the *D. desulfuricans* were present (T5) (Figure 3). The iron:sulfur ratios shown in Figure 3 provide an indication of the relative amount of sulfur detected in surface product analysis.

**Figure 1.** Photos of uncleaned and cleaned coupons after 8-weeks immersion in half-strength marine broth for the six different microbial testing conditions, together with 3D surface profiles of cleaned coupons showing surface and pit morphology.

**Figure 2.** SEM images of biofilms/corrosion products on surfaces of coupons tested after 8-weeks incubation in the six different test conditions. All of these coupons underwent fixation (dehydration and 2% glutaraldehyde immersion) prior to imaging.

**Figure 3.** EDS spectra of the biofilms/corrosion products formed on DH36 coupons tested in the six different treatments after 8-week incubation.

The XRD spectra peaks of individual phases were identified by comparing the diffraction pattern to a known standard from the diffraction powder database of the International Centre for Diffraction Data (Diffrac Eva software, version 4.1). Unfortunately, the XRD spectra obtained (Figure S1) provided ambiguous results for corrosion products/biofilms, with the three peaks observed in each scan matching that of iron. It is possible that the intensity of iron dominated the scanning spectra or that the corrosion products formed were amorphous in nature.

## *3.2. Corrosion Evaluation*

Example photos of the surfaces of test coupons after removing any biofilms and corrosion products are shown in Figure 1. The key difference observed was that while the uninoculated samples (T1) showed signs of general corrosion and occasional pitting, most of the coupons tested in different biotic conditions appeared to suffer from sparse localised pitting corrosion. SEM scans of the cleaned surfaces (Figure 4) confirmed that the uninoculated control (T1) coupon had a combination of general and localised corrosion attack, as did the coupon tested in the *D. desulfuricans* aerobic test configuration (T3). The coupons from the other tests (T2, T4, T5 and T6) contained localised corrosion but were relatively free from general corrosion. Finally, the diameters of pits observed for coupons tested with *D. desulfuricans* (T3, T5 and T6) appeared to be generally larger than those for the other test configurations.

**Figure 4.** SEM images of surfaces of cleaned test coupons after 8-week immersion in the six different test conditions.

To further characterise the morphology of localised corrosion, 3D optical profiling was performed on one side of one of the cleaned coupons for each test condition (see examples in Figure 1). The pit density, which was calculated for all treatments based on the number of pits (>10 μm) per mm2, was ~2 (pit/mm2) for T1, (uninoculated treatment), and T2, T4 and T5 (mixed consortia inoculum), and ~1 (pit/mm2) for T3 and T6 (*D. desulfuricans* alone). The coupons with the largest pits were from tests containing *D. desulfuricans* (T3, T5 and T6, see details in Figure 5). Apart from the coupon from T4 (four isolates) which had one pit deeper than 30 μm, only the coupons from tests with *D. desulfuricans* (T3 and T6) had any pits greater than 30 μm depth (five and six pits, respectively). The deepest pit found was 50.2 μm for the coupon tested in *D. desulfuricans* under anaerobic conditions (T6), compared to 27 μm for the deepest pit for the uninoculated control (T1). The average pit volumes (of five deepest pits) were nearly an order of magnitude greater for the tests containing *D. desulfuricans* (T3, T5 and T6) compared to the tests lacking *D. desulfuricans*.

**Figure 5.** Summary of the corrosion attack on steel samples from the six different treatments. Averages of 5 highest values of relative volume and pit depth as well as average corrosion rate are shown with statistical analyses. Significant differences (*p* < 0.05) were shown by different series of superscript letters on each parameter.

The average corrosion rates of the DH36 coupons after 8 weeks of incubation in the different test scenarios were determined via mass loss (Figure 5). The highest corrosion rates among the six treatments obtained (*p* < 0.05) were for the T3 (*D. desulfuricans* tested aerobically) and T1 (uninoculated) test conditions. For all the other test conditions very low average corrosion rates were measured; always much less than the uninoculated control, indicating a form of corrosion inhibition. Given that the localised attack of the uninoculated control (T1) was relatively minor, the relatively high weight loss for this test condition will have been dominated by general/uniform corrosion.
