Glutaraldehyde Enhancement by D-limonene for Mitigating Biocorrosion of Carbon Steel by An Oileld Biolm Consortium

Microbiologically inuenced corrosion (MIC) is one of the corrosion threats in the oil and gas industry. It is caused by environmental biolms. Glutaraldehyde is a popular green biocide for mitigating biolms and MIC. This work investigated the enhancement of glutaraldehyde ecacy by food-grade green chemical D-limonene in the biolm prevention and MIC mitigation using an oileld biolm consortium. After 7–d incubation at 37 o C with enriched articial seawater (EASW) in 125 mL anaerobic vials, the 100 ppm glutaraldehyde + 200 ppm D-limonene combination treatment reduced the sessile cell counts on C1018 carbon steel coupons by 2.1 − log, 1.7 − log, and 2.3 − log for SRB (sulfate reducing bacteria), APB (acid producing bacteria), and GHB (general heterotrophic bacteria), respectively in comparison with the untreated control. The treatment achieved 68% weight loss reduction and 78% pit depth reduction. The 100 ppm glutaraldehyde + 200 ppm D-limonene combination treatment was found more effective in biolm prevention and MIC mitigation than glutaraldehyde and D-limonene used individually. Electrochemical tests data The experiment results indicated 200 ppm D-limonene enhanced 100 ppm glutaraldehyde in the biolm prevention against Consortium II oileld MIC of carbon steel. The treatment with 200 ppm D-limonene + 100 ppm glutaraldehyde exhibited the 2.1 − log, 1.7 − log, and 2.3 − log reduction for SRB, APB, and GHB, respectively, in comparison with the untreated control. The treatment achieved higher ecacy in biolm prevention and MIC mitigation than using glutaraldehyde and D-limonene individually. The electrochemical tests corroborated weight loss and pit depth data trends.


Introduction
Biocorrosion, more commonly known as microbiologically in uenced corrosion (MIC) in the literature, is one of the root causes of equipment failures, especially for pipelines in the oil and gas industry. MIC accounts for 20% of total corrosion losses (Dall'Agnol and Moura 2014), which amounts to approximately 30-50 billion US dollars per year worldwide (Javaherdashti 1999). Microbial bio lms cause MIC. Biocides are routinely used to treat bio lms to mitigate the risk of MIC failures. The desired biocide should have broad-spectrum e cacy and safety to the environment and eld operators.
SRB are microbes that thrive in an anaerobic environment with sulfate. They can survive exposure to oxygen but without growth (Bade et al. 2000). SRB uses the dissimilatory sulfate reduction process to couple the oxidation of an energy molecule such as organic carbon or dissolved hydrogen gas. When there is a local shortage in organic carbon, or the sessile cells are in the resting (non-growth) state, SRB can switch to elemental iron as the electron donor in energy production. This mechanism is the BCSR theory (biocathode theory) (Gu et al. 2018). Lactate oxidation reaction provides an energy source, as shown in Reaction (1). SRB produces HS − from sulfate respiration in reaction (2). HS − then combines with a proton to become H 2 S or release a proton to become sul de, as shown in Reactions (4) and (5), respectively. HS − can also react with Fe 2+ in Reaction (3) to form FeS as presented in Reaction (6), which appears as a black color product when it is precipitated out in the culture medium beyond saturation (Jia et al. 2019 (6) The sulfate reduction reaction happens in the cytoplasm of SRB with enzyme catalysis, but the electrons from Fe 0 (insoluble) oxidation are released extracellularly. This means that extracellular electrons need to be transported to the cytoplasm for sulfate reduction. This electron transport process is known as the extracellular electron transfer (EET) process (Gu et al. 2018). Mesophilic SRB produces the maximum corrosion rate at a mild temperature, such as 37 o C (Rasol et al. 2014 (Wen et al. 2009) demonstrated that ethylene amine disuccinate (EDDS), which is a green chemical, enhanced glutaraldehyde mitigation of Desulfovibiro vulgaris (a common SRB strain) bio lm on carbon steel. A high concentration of EDDS (2,000 ppm by mass) is needed because some of it is consumed by ions that it chelates in the liquid.
D-amino acids have been found to enhance THPS (tetrakis hydroxymethyl-phosphonium sulfate) against D. vulgaris, and also a recalcitrant corrosive oil eld bio lm caodenamed Consortium II (Jia et al. 2017). More recently, Jia et al. reported that a 14 − mer peptide with its 12 − mer core sequence derived from a sea anemone's Equinatoxin II at 18 ppb (w/w) enhanced Consortium II considerably. Because glutaraldehyde is a potent crosslinker of peptides (Migneault et al. 2004), it is not compatible with Damino acids and Peptide A. Thus, it is desirable to nd other green chemicals that can enhance glutaraldehyde, which is the most popular green biocides together with THPS in the oil and gas industry.
D-limonene (C 10 H 16 ) is a biodegradable antimicrobial agent found in some "green" household cleaners (Van vuuren and Viljoen 2007;Settanni et al. 2012;Chikhoune et al. 2013). It is a common terpene that is extracted from oils of citrus fruits such as orange, lemon, and lime. It is considered safe for human use with slight irritation when in direct contact with the skin (Kim et al. 2013). Even though it is widely applied in the food industry as a avoring agent and it is sold as a nutritional supplement (Dabbah et al. 1970), its application in bio lm treatment and MIC mitigation needs to be investigated since there is a lack of literature data.
This project compared glutaraldehyde e cacy and D-limonene e cacy and the e cacy of combining the two biocides in mitigation Consortium II and its MIC against C1018 carbon steel. Consortium II was grown on carbon steel coupons in EASW. Weight loss, sessile cell count, and electrochemical tests were used to assess biocide e cacy.

Metal coupons and microbes
The test coupon was cut from C1018 carbon steel. Its elemental composition is listed in Table 1.
Coupons were protected with inert Te on paint except for the top 1 cm 2 square test area. Prior to incubation, coupons were sequentially abraded up to 600 grit, degreased with pure isopropanol, and air dried under ultraviolet light for at least 0.3 h.  (Jia et al. 2017), a corrosive oil eld bio lm consortium, was used in all the tests in this work. It was grown in enriched arti cial seawater (EASW) with composition listed in Table 2. The culture medium was sterilized and then deoxygenated using N 2 . L-cysteine (at 100 ppm) was added as a scavenger of O 2 . The culture medium's initial pH was adjusted to 7.0 before inoculation in a N 2 -lled glovebox. Each 125 mL vial was lled with 100 mL of EASW and 3 coupons. The inoculum was 1 mL 3-d old seed culture.

Corrosion Analysis
After sessile cell counting, the coupons were clean using a fresh Clarke's solution for weight loss measurements (Jia et al. 2019) and pit depth scanning using a pro lometer (Jia et al. 2019). The coupon surfaces were further examined under scanning electron microscopy (Jia et al. 2019).

Electrochemical measurements
All the electrochemical measurements were executed in 450 mL glass cells (300 mL EASW plus 3 mL seed culture). OCP (open circuit potential), LPR (linear polarization resistance), and EIS (electrochemical impedance spectroscopy) were measured daily during the 7 − d incubation period (Jia et al. 2019). Potentiodynamic polarization (PDP) was scanned at the end of the 7 − d incubation (Jia et al. 2019). The counter electrode was a thin platinum sheet (1 cm 2 ) and the reference electrode was a saturated calomel electrode (SCE). A C1018 coupon (1 cm 2 ) in an epoxy cake served as the working electrode. A VersaSTAT 3 potentiostat (Princeton Applied Research, Oak Ridge, TN, USA) was employed.

Results
Sessile cell enumeration Figure 1 shows the sessile cell counts after the 7 − day incubation. The SRB, APB, and GHB cell counts in the bio lm without any biocide treatment were 6.9 − log, 4.5 − log, and 3.8 − log, respectively. The 100 ppm glutaraldehyde + 200 ppm D-limonene treatment resulted in the lowest sessile cell counts with 2.1, 1.7, and 2.3 − log reductions for SRB, APB, and GHB, respectively, in comparison with the no treatment control. In comparison, the e cacies using 100 ppm glutaraldehyde alone, and 100 ppm D-limonene alone, were lower. The 100 ppm glutaraldehyde treatment had 1.1, 0.7, and 1.0 − log reductions, whereas 100 ppm D-limonene provided 0.8, 1.0, and 0.8 − log reductions for SRB, APB, and GHB, respectively, in comparison with the no treatment control. The e cacy results for 100 ppm glutaraldehyde + 50 ppm Dlimonene exhibited no signi cant improvements compared with 100 ppm glutaraldehyde alone treatment. It means that 50 ppm of D-limonene was insu cient to enhance 100 ppm glutaraldehyde.

Corrosion analyses
Weight loss measurements after the 7 − d incubation are presented in Fig. 2. The coupons in the abiotic control vial exhibited a small and practically negligible weight loss of 0.2 ± 0.05 mg/cm 2 . The maximum weight loss of 3.13 ± 0.08 mg/cm 2 was observed for coupons without treatment (biotic control). The 100 ppm D-limonene treatment gave a slightly smaller weight loss of 2.9 ± 0.05 mg/cm 2 followed by 100 ppm glutaraldehyde treatment, which had 2.7 ± 0.06 mg/cm 2 . These data suggested that 100 ppm D-limonene or 100 ppm glutaraldehyde was not very effective in the MIC mitigation. The various combinations of glutaraldehyde + D-limonene led to considerable improvement in weight loss prevention. The combination of 50 ppm, 100 ppm and 200 ppm D-limonene with 100 ppm glutaraldehyde yielded weight losses of 1.60 ± 0.08 mg/cm 2 , 1.4 ± 0.02 mg/cm 2 and 1.0 ± 0.08 mg/cm 2 , respectively, considerably lower than those for the 100 ppm glutaraldehyde alone treatment and the 100 ppm D-limonene treatment, showing synergy for the two anitmicrobials used together. The 100 ppm glutaraldehyde + 200 ppm D-limonene treatment yielded the largest weight loss reduction of 68%. SEM images after the 7 − d incubation are presented in Fig. 3. It shows a large pit (approximately 30 µm in surface diameter) and numerous small pits on the untreated coupon. In comparison with other treatments, 100 ppm glutaraldehyde + 200 ppm D-limonene yielded the least pitting, in agreement with the least weight loss shown in Fig. 2.
Maximum pit depths are reported in Fig. 4. The biotic coupons exhibited a (maximum) pit depth of 24.5 µm, largest among all coupons. With 100 ppm of individual biocides, pit depths were found to be 21.7 µm and 18.0 µm for D-limonene and glutaraldehyde, respectively. For the glutaraldehyde + D-limonene combinations, the pit depths were 13.1 µm, 10.1 µm, 5.4 µm for 100 ppm glutaraldehyde in combination with 50 ppm, 100 ppm, and 200 ppm of D-limonene. The 100 ppm glutaraldehyde + 200 ppm D-limonene achieved a pit depth reduction of 78% compared with the untreated control. The pit depth data here were coherent with the weight loss trend. Figure 5 presents OCP during the 7 − d incubation. Theoretically, a more negative OCP suggess that the working electrode is less likely to be reduced or more likely to be oxidized, which also means having a higher thermodynamic tendency to be corroded. The OCP trend in Fig. 5 obviously does not follow the weight loss trend in Fig. 2, because the data do not show the 100 ppm glutaraldehyde + 200 ppm Dlimonene OCP curve on top and the no-treatment OCP curve at the bottom. This discrepancy is common for MIC of carbon steel. Because for a complicated MIC system, an electron acceptor (oxidant) in the culture medium can exhibit a large thermodynamic tendency to grab electrons from the C1018 working electrode, but the process can be kinetically hindered due to a high activation energy for the reduction of the oxidant. Kinetic electrochemical tests have repeatedly proven that kinetic electrochemical data are in agreement with weight loss and pit depth data trends, despite misleading OCP trends (Jia et al. 2019). This R p fall shared the same pattern with 100 ppm glutaraldehyde + 100 ppm D-limonene treatment. This could mean that initially, D-limonene treatment was effective, but the effect waned over time. The R p data here provided useful transient corrosion behavior corresponding to biocide treatment. This transient information is not available from the cumulative weight loss or pit depth data. It means that this simple electrochemical method is a useful tool in biocide e cacy assessment.

Electrochemical tests
EIS was scanned daily. Nyquist and Bode plots on the day 7th are shown in Fig. 7 and Fig. 8, respectively.
The 100 ppm glutaraldehyde + 200 ppm D-limonene treatment provided the largest semi-circle diameter, which means the highest corrosion resistance. The EIS tted parameters are given in Table 3 using the equivalent circuit model illustrated in Fig. 9, in which R s , R f, and R ct represent solution resistance, lm resistance (bio lm and corrosion products), and charge transfer resistance, respectively. The R f + R ct combination can be used to describe corrosion resistance (Jia et al. 2019). In Table 3, R f + R ct values indicate that the 100 ppm glutaraldehyde + 200 ppm D-limonene treatment out-performed other treatments. It provides an approximately 200% increase in corrosion resistance compared to the untreated coupon, much greater than using 100 ppm glutaraldehyde alone or 100 D-limonene alone. The R f + R ct (scanned using alternate current in EIS) trend corroborated R p (scanned with direct current in LPR) and weight loss trends. PDP curves at the end of the 7 − d incubation are shown in Fig. 10. The tted Tafel are listed in Table 4.
The i corr value was 38.4 µA/cm 2 for the untreated coupon. For individual biocide treatments, the i corr value for 100 ppm D-limonene alone was 19.7 µA/cm 2 , and it was 14.1 µA/cm 2 for 100 ppm glutaraldehyde alone. For the combination treatments, the i corr values were 13.7 µA/cm 2 for 100 ppm glutaraldehyde + 50 ppm D-limonene, 12.3 µA/cm 2 for 100 ppm glutaraldehyde + 100 ppm D-limonene, and 7.9 µA/cm 2 for 100 ppm glutaraldehyde + 200 ppm D-limonene. The 100 ppm glutaraldehyde + 200 ppm D-limonene treatment achieved the best outcome, con rming the weight loss data.

Discussion
Consortium II caused MIC of C1018 carbon steel in deoxygenated EASW. After 7-d incubation, the sessile cell counts of the untreated coupons for SRB, APB and GHB were 6.9 − log, 4.5 − log, and 3.8 − log, respectively. They were reduced to 4.8 − log, 2.8 − log, and 1.5 − log, respectively with the 100 ppm glutaraldehyde + 200 ppm D-limonene combination treatment in the bio lm prevention test. The combination treatment was much more superior to the individual 100 ppm glutaraldehyde alone, 100 ppm D-limonene alone treatments and 100 ppm glutaraldehyde + 100 ppm D-limonene treatment. The 100 ppm glutaraldehyde + 200 ppm D-limonene treatment reduced sessile cell counts of SRB, APB and GHB by 2.1, 1.7, and 2.3 − log, respectively, compared to the untreated biotic control. The antimicrobial activity of D-limonene has been discussed in the literature (Van vuuren and Viljoen 2007;Settanni et al. 2012); therefore, the reduction of sessile cell by cocktail treatment demonstrated the synergy between Dlimonene and glutaraldehyde to achieve higher antimicrobial e cacy. The weight loss and pit depth were also considerably reduced, which were corroborated by LPR, EIS and PDP data. The 100 ppm glutaraldehyde + 200 ppm D-limonene treatment led to a 68% reduction in weight loss, and 78% reduction in pit depth after the 7-d incubation. The i corr data after the 7 − d incubation indicated a 79% reduction achieved by the 100 ppm glutaraldehyde + 200 ppm D-limonene treatment, not far from the 68% reduction in weight loss. Thus, D-limonene enhanced glutaraldehyde in MIC mitigation alternatively to D-amino acid based enhancer (Wen et al. 2009;Xu et al. 2012).

Conclusion
The experiment results indicated 200 ppm D-limonene enhanced 100 ppm glutaraldehyde in the bio lm prevention against Consortium II oil eld MIC of carbon steel. The treatment with 200 ppm D-limonene + 100 ppm glutaraldehyde exhibited the 2.1 − log, 1.7 − log, and 2.3 − log reduction for SRB, APB, and GHB, respectively, in comparison with the untreated control. The treatment achieved higher e cacy in bio lm prevention and MIC mitigation than using glutaraldehyde and D-limonene individually. The electrochemical tests corroborated weight loss and pit depth data trends.

Declarations
Tables     Variations of OCP during the 7−d incubation.

Figure 6
Rp values vs. time during the 7−d incubation.

Figure 8
Bode and phase-frequency plots after 7−d incubation. Equivalent circuits used for tting EIS spectra.

Figure 10
PDP curves after 7−d of incubation.