Cathodic protection and sulphate-reducing bacteria: a complex interaction in offshore steel structures

2.1 Introduction

Microbiologically influenced corrosion (MIC) is a specific type of electrochemical corrosion where the corrosion process is influenced by the activity of microorganisms (Javaherdashti 2011; Loto 2017; Skovhus et al. 2017). It is also known under other names such as microbial corrosion, biological corrosion and biodegradation (Skovhus et al. 2017). It is a relevant concern in a wide range of industries and infrastructures such as the oil, petroleum and gas industry (Skovhus et al. 2017), recirculated cooling systems (Ilhan-Sungur and Çotuk 2010), wastewater treatment and sewage plants (Stanaszek-Tomal and Fiertak 2016), marine infrastructures (Procópio 2019), chemical processing and power generation plants (Chajduk and Bojanowska-Czajka 2016; El-Shamy 2020) and underground pipelines (Li et al. 2000). MIC is believed to be responsible for more than 20 % of corrosion-related costs (Jia et al. 2019). Different metals such as carbon steel (Xu et al. 2012), stainless steel (Zhang et al. 2015), aluminium alloys (Dai et al. 2016), magnesium (Liang et al. 2022), copper (Reyes et al. 2008) and other materials such as concrete (Harbulakova et al. 2013) are susceptible to this type of corrosion. Titanium and its alloys are the most resistant against MIC (Javaherdashti 1999). MIC can be caused by the activities of biological agents, i.e. biological organisms and microorganisms: i) prokaryotes, i.e. Archaea (Davidova et al. 2012) and bacteria and ii) eukaryotes, such as fungi (Khan et al. 2021; Little and Lee 2009; Planý et al. 2021). In general, all microorganisms need both water and nutrients, including carbon (C), nitrogen (N), sulphur (S) and phosphorus (P), for their growth. Moreover, an electron donor and electron acceptor are needed for their energy metabolism (Khan et al. 2021; Little and Lee 2009). It is important to indicate that MIC typically interacts synergistically with other corrosion types such as pitting corrosion (Starosvetsky et al. 2001), crevice corrosion (Zhang et al. 2021), stress-corrosion cracking (SCC) (Wu et al. 2015b) and corrosion fatigue (Thomas et al. 1987). Figure 1 illustrates typical examples of the consequences of MIC. Corrosion pits (Figure 1a) and corrosion products, i.e. tubercles (Figure 1b) on a galvanised steel tube and rusticles (Figure 1c) on the Royal Mail Ship (RMS) Titanic, can be observed (Li et al. 2000; Little and Lee 2009).

Figure 1:

Examples of MIC consequences: (a) corrosion pits, (b) formation of tubercles and (c) rusticles. (a) Microbiologically influenced corrosion of underground pipelines under the disbonded coatings, Li et al., Metals and Materials International, 6, Springer Nature, 2000, reproduced with permission from Springer Nature Costumer Service Center (SNCSC); (b) and (c) Reprinted from Little and Lee (2009), with permission from John Wiley & Sons. Kirk-Othmer Encyclopedia of Chemical Technology. Copyright John Wiley & Sons, Inc. All rights reserved.

When bacteria are responsible for MIC, they can be classified into different groups according to different characteristics: i) shape and appearance of the bacterial cells, ii) temperature, iii) oxygen consumption, iv) energy source, v) carbon source, vi) electron donor and vii) electron acceptor. These categories are summarised in Table 1 (Javaherdashti 2017; Khan et al. 2021).

Besides the classification according to their features, the bacteria can also be categorised according to how they operate, which is already clear by their name. The most common bacteria involved in MIC are sulphate-reducing bacteria (SRB) (Hamilton 1985), sulphur-oxidising bacteria (SOB) (Little et al. 2000; Okabe et al. 2007), iron-oxidising bacteria (IOB) (Liu et al. 2016; Singh et al. 2018; Wang et al. 2014), iron-reducing bacteria (IRB) (Duan et al. 2008; Herrera and Videla 2009), acid-producing bacteria (APB) (Sowards and Mansfield 2014; Xu et al. 2016), nitrate-reducing bacteria (NRB) (Jia et al. 2017; D. Xu et al. 2013a) and methanogenic bacteria (Uchiyama et al. 2010). Table 2 summarises some specific failure cases where MIC occurred, which shows the corroded equipment and the involved microorganisms. More information on the analysis methods used to evaluate the specific corrosion cases can be found in the relevant references. Moreover, there are more dedicated review papers on MIC case studies (Olszewski 2007; Spark et al. 2020; Starosvetsky et al. 2001). Because the emphasis is on SRB, these will be described in more detail later.

The formation of a biofilm is one of the main features of microorganisms. A biofilm is a group of immobilised or sessile microorganism cells sticking together on a surface, in particular a metal surface, which consequently modifies the metal surface–solution interface (Chitra et al. 2014; Khan et al. 2021). Its formation occurs in different stages, which is schematically represented in Figure 2. The formation initiates with a conditioning film (stage a), being important in its nature and distribution on the surface, through the electrostatic arrangement of various proteins and other organic substances. The bacteria are still in their planktonic state, which means they are still ‘floating around’ within the solution. Subsequently, planktonic bacterial cells start to attach to the surface via the conditioning film. They start to form colonies and become sessile bacterial cells by the formation and excretion of extracellular polymeric substances (EPS) (stage b). These EPS are actually a self-produced matrix consisting of organic substances such as proteins, polysaccharides, lipids and nucleic acids, which bind the sessile bacteria together in a three-dimensional matrix and, furthermore, provide structural and functional integrity and mechanical stability. The EPS typically consist of two forms: on the one hand, soluble EPS, e.g. slimes and colloids, which are dissolved within the solution or weakly bound to the bacterial cells, and, on the other hand, bound EPS, e.g. condensed gels, being strongly attached to the bacterial cells. The formation of EPS results in the irreversible attachment or anchoring of sessile bacterial cells. A real biofilm layer is forming and growing in thickness by replication of different sessile species (stage c), which is possible through the consumption of available nutrients. As a consequence, it will start to function as a kind of net for trapping both organic and inorganic particles (stage d). The conditions beneath the biofilm, such as pH, dissolved oxygen concentration, etc., are significantly different compared to the conditions within the bulk solution. Finally, the maturation of a steady-state biofilm layer starts and some bacterial cells are able to detach and disperse (stage e) from the biofilm’s outer surface. Some areas, which are not covered by the biofilm, can eventually be recolonised by planktonic bacterial cells or even adjacent sessile bacterial cells (stage f) (Chitra et al. 2014; Hamilton 1985; Javaherdashti 2017; Khan et al. 2021; Lv and Du 2018).

Figure 2:

Different stages of biofilm formation. Microbiologically influenced corrosion: An engineering insight (2nd edition), Javaherdashti, Engineering Materials and Processes, Springer Nature, 2017, adapted with permission from Springer Nature Costumer Service Center (SNCSC).

2.2 Mechanisms

There are several possible mechanisms for MIC. Two mechanisms, i.e. concentration cell formation and metabolite MIC, are explained in more detail in the next sections. The corrosion mechanisms under the influence of SRB will be described in more detail later.

2.2.1 Concentration cell formation

As explained before, microorganisms colonise a metal surface through the formation of a biofilm. This results in the creation of concentration cells. A biofilm is believed to be rather non-uniform due to, among others, the presence of different bacterial species and the open structure. Therefore, it is assumed to consist of discrete microcolonies separated by interstitial voids, making the creation of gradients and concentration cells possible due to the straightforward mass transport of gases and matter (Lewandowski et al. 1995). Consequently, for example, oxygen concentration cells can be formed, when respiring aerobic bacteria form colonies on a metal surface, as they can deplete the oxygen concentration at the metal surface by consuming the oxygen for their energy metabolism. This favours the growth of anaerobic bacteria such as SRB. Consequently, the depleted oxygen region under the colony becomes an anode, while the surrounding area becomes a cathode. Biofouling, which involves the accumulation of microorganisms (and sometimes macroorganisms) and the formation of a biofilm, is sufficient by itself to establish oxygen concentration cells. An oxygen concentration gradient is indeed created through thickness from the outside towards the inside of the biofilm. Furthermore, tubercles, which are corrosion products containing magnetite (Fe₃O₄) and goethite (α-FeO(OH)), can also form on pipes (see Figure 1b). They can shield the metal surface from oxygen and, therefore, they can also establish oxygen concentration cells. The created oxygen concentration cells can give rise to localised corrosion types such as pitting corrosion and crevice corrosion due to local acidification within the deaerated zone (Iverson 1987; Javaherdashti 2017; Little and Lee 2009; Martins and Nunes 2006; Procópio 2019). Figure 3 illustrates how pitting corrosion of iron can result from the formation of an oxygen concentration cell (Jia et al. 2019). Besides oxygen concentration cells, metal concentration cells are also able to form by the EPS that are able to bind metals by their functional groups (Chitra et al. 2014).

Figure 3:

Pitting corrosion caused by the creation of an oxygen concentration cell, based on Jia et al. (2019).

Figure 4:

H₂SO₄ corrosion of a sewage pipe by the combined action of SRB and SOB, based on Little et al. (2000) and Wu et al. (2020).

3.1 Introduction

Sulphate-reducing bacteria (SRB) are the most commonly associated bacteria with anaerobic MIC and are considered to be strict anaerobes. They can cause serious corrosion issues at offshore structures such as marine steel pilings and offshore wind farms (Enning and Garrelfs 2014). Seawater is in general oxygenated, but other aerobic, e.g. iron-oxidising bacteria (IOB), and facultative anaerobic bacteria can reduce the oxygen content within the seawater environment to certain concentrations in which SRB are able to grow (Battersby et al. 1985). Furthermore, they are able to survive within the anoxic regions of a biofilm or slime layer containing other bacteria that consume oxygen (Battersby et al. 1985) or under macrofouling deposits and marine sediments such as seabed sediment and sea mud (Batt et al. 2002; Guezennec 1994; Zhu et al. 2007). Exposure to oxygen will alter the activity and behaviour of SRB, and the formed biofilm’s compactness and homogeneity (Cypionka 2000; Sigalevich and Cohen 2000; Wan et al. 2010). Other environments where SRB can be found are oil and gas fields (Cord-Ruwisch et al. 1987; Liu et al. 2018), sewers and sewage treatment plants (Oviedo et al. 2012; Stanaszek-Tomal and Fiertak 2016) and buried pipelines (Enning and Garrelfs 2014). They typically survive in a temperature range of 5–50 °C and a pH range of 5–10 (Javaherdashti 1999).

SRB obtain the energy for their metabolism by an electron donor–electron acceptor pair, i.e. via a redox reaction, as is generally the case for bacteria (Li et al. 2018). Their electron donors, energy sources and carbon providers are usually organic carbons such as pyruvate, lactate, acetate, propionate, alcohols and higher fatty acids (Cord-Ruwisch et al. 1987; Hamilton 1985; Little et al. 2000). Saturated and unsaturated hydrocarbons available in oil pipelines, among others, can also be used (Xu and Gu 2014). The SRB can typically be divided into two nutritional groups. On the one hand, those that do an incomplete oxidation of the organic sources with acetate as the end product and, on the other hand, those that do a complete oxidation of the organic sources into CO₂ (Cord-Ruwisch et al. 1987; Hamilton 1985). Moreover, hydrogenase-positive SRB such as Desulfovibrio vulgaris (D. vulgaris) and Desulfovibrio desulfuricans (D. desulfuricans), which contain the enzyme hydrogenase, can use molecular hydrogen (H₂) as an electron donor and they will use the oxidation of hydrogen for their growth. Therefore, it is clear that, in general, SRB can obtain electrons by the degradation of organic matter or H₂. Some SRB can also grow autotrophically through the use of CO₂ and H₂ (Xu and Gu 2014).

Besides its consumption and use as a possible substrate, hydrogen can also be produced by SRB as demonstrated by Martins and Pereira (2013) who tested the H₂ production by the hydrogenase-active D. vulgaris through dark fermentation of three different substrates.

The metabolism of SRB is based on anaerobic respiration, which makes use of an external oxidant as a terminal electron acceptor. This oxidant is not oxygen compared to aerobic respiration (Gu and Galicia 2012). The terminal electron acceptor is usually sulphate (SO₄ ²⁻) but can also be sulphite (SO₃ ²⁻), thiosulphate (S₂O₃ ²⁻) and even elemental sulphur (S) (Li et al. 2018). The latter acceptor is for example used by the Desulfuromonas acetoxidans (Hamilton 1985). Because of the different possible acceptors besides SO₄ ²⁻ and the fact that sulphide (S²⁻) is produced, a better description of these bacteria is sulphidogenic (Hamilton 1985).

When SO₄ ²⁻ is used as an electron acceptor, its reduction can happen through two different pathways, i.e. assimilated sulphate reduction (ASR) and dissimilatory sulphate reduction (DSR). The latter is considered to be the main pathway and is illustrated in Figure 5. The SO₄ ²⁻ is first activated by the reaction with adenosine triphosphate (ATP) resulting in the formation of adenosine phosphosulphate (APS) under the catalysis of ATP sulfurylase. This APS is further reduced to SO₃ ²⁻ by a two-electron transfer catalysed by APS reductase and subsequently SO₃ ²⁻ is reduced to S²⁻ by a six-electron transfer catalysed by sulphite reductase. Instead of the six-electron transfer, three two-electron transfers through trithionate (S₃O₆ ²⁻) and thiosulphate (S₂O₃ ²⁻) as intermediates are also possible (Hamilton 1985; He et al. 2023; Little et al. 2000; P. Zhang et al. 2015; Z. Zhang et al. 2022b).

Figure 5:

Dissimilatory sulphate reduction (DSR) by SRB, based on P. Zhang et al. (2015) and Z. Zhang et al. (2022b).

The generated S²⁻ can be further hydrolysed into H₂S (Little et al. 2000). H₂S is known to have an unpleasant odour, while at high concentrations it cannot be smelled and is extremely toxic when inhaled (Cord-Ruwisch et al. 1987; Li et al. 2000). It is reported that it causes the souring of oil reservoirs and contamination of fuel gas and fuel oil (Cord-Ruwisch et al. 1987; Hamilton 1985). It is also responsible for the massive killing of fish within marine environments and even the killing of sewer workers (Javaherdashti 2011). Furthermore, it can act as an inhibitor for SRB growth because of its intrinsic toxicity (Reis et al. 1992). When dissolved in an aqueous solution, H₂S becomes a weak acid and is known to be very corrosive to ferrous metals, inducing sour corrosion (Banakhevych et al. 2009; Li et al. 2000; King 2017). The general corrosion reaction is represented by Equation (6) (King 2017).

The formed iron sulphide corrosion products are in the more general form Fe_xS_y. Their structure, protectiveness, relative amount and corrosion rate are dependent on environmental parameters such as H₂S concentration, pH, temperature, composition of the solution and the presence of oxygen (Banakhevych et al. 2009; Huang et al. 2017; Jack et al. 1998; King 2017). They are for example able to plug injection wells and are typically recognised as voluminous black precipitates (Cord-Ruwisch et al. 1987). Figure 6 shows the surface of a pipe under a disbonded coating that is covered with these products (Li et al. 2000). Iron sulphides are typically electroconductive and they are more noble than iron, which makes them more cathodic (Lazzari and Pedeferri 2018). If the integrity and protectiveness of an iron sulphide layer are disrupted, a galvanic coupling between this layer, acting as the cathode, and the bare iron or steel spots, acting as the anode, can result in localised corrosion because of the large cathode and small anode area (Lazzari and Pedeferri 2018; Lv and Du 2018; Rebak and Perez 2017).

Figure 6:

Black iron sulphide corrosion products on the surface of a pipe underneath a disbonded coating. Microbiologically influenced corrosion of underground pipelines under the disbonded coatings, Li et al., Metals and Materials International, 6, Springer Nature, 2000, reproduced with permission from Springer Nature Costumer Service Center (SNCSC).

Liu et al. (2018) investigated the corrosion of X80 pipeline steel caused by the combined effect of SRB and CO₂ corrosion, also known as sweet corrosion (King 2017), within simulated CO₂-saturated oilfield-produced water. FeCO₃ was replaced by FeS as the dominant corrosion product when SRB-containing biofilms were present.

It should be mentioned that the problem of H₂S corrosion and iron sulphide corrosion products is, besides SRB-containing environments, also very relevant to industries where abiotic H₂S is present such as the oil and gas industry (Amani and Almodaris 2016; Asmara 2018; Marriott et al. 2016), geothermal industry (Banaś et al. 2007; Marshall and Hugill 1957) and the pulp industry (Chudnovsky 2008).

3.2 Corrosion mechanisms

In general, in anoxic and near-neutral systems, the oxidation of iron according to Equation (7) is a slow process because the proton (H⁺) reduction reaction, being dominant in the absence of oxygen, is a slow reaction and kinetically impeded at a pH value higher than 6, due to a low availability of protons. However, the presence of microorganisms such as SRB will accelerate the corrosion kinetics (Enning and Garrelfs 2014; Kato 2016).

Kuang et al. (2007) studied the influence of SRB on the corrosion of D36 carbon steel in seawater by potentiodynamic polarisation (PDP) and electrochemical impedance spectroscopy (EIS) as electrochemical measurement systems. They concluded by PDP that the anodic process and thus the corrosion rate were largely increased within SRB-inoculated seawater compared to seawater without SRB. This was caused by the accumulation of sulphide metabolite, rather than the active number of SRB. Furthermore, a decrease of the polarisation resistance (R_p) was observed by EIS.

There are several mechanisms proposed for MIC under the influence of SRB. Four mechanisms, namely the cathodic depolarisation theory (CDT), King’s mechanism (iron sulphides), biocatalytic cathodic sulphate reduction (BCSR) theory and extracellular electron transfer (EET) are explained in more detail here.

3.2.1 Cathodic depolarisation theory

A first mechanism is the cathodic depolarisation theory (CDT) already proposed by von Wolzogen Kühr and van der Vlugt in 1934 (Von Wolzogen Kuehr and van der Vlugt 1934). This mechanism is considered as the classical theory. It can be explained according to three main aspects, which are the metal, the solution and the microorganism, i.e. SRB, itself. The aspects (A, B and C) and reactions (Equation (8) – Equation (14)) are given below. The corrosion process of iron (Fe) occurs as follows (Javaherdashti 2011).

1. The metal (iron)

   (8) 4 F e → 4 F e 2 + + 8 e −   a n o d i c   r e a c t i o n ; i r o n   d i s s o l u t i o n

2. The solution

   (9) 8 H 2 O → 8 H + + 8 O H −   d i s s o c i a t i o n   o f   w a t e r

   (10) 8 H + + 8 e − → 8 H a d s      c a t h o d i c   r e a c t i o n

3. Microorganism (SRB)

   (11) S O 4 2 − + 8 H a d s → S 2 − + 4 H 2 O      c a t h o d i c   d e p o l a r i s a t i o n

   (12)   F e 2 + + S 2 − → F e S   s u l f i d e   p r e c i p i t a t i o n

   (13) 3 F e 2 + + 6 O H − → 3 F e O H 2   h y d r o x i d e   p r e c i p i t a t i o n

Overall reaction

(14) 4 F e + 4 H 2 O + S O 4 2 − → 3 F e O H 2 + F e S + 2 O H −

It is clear that the corrosion of the iron proceeds due to the removal and consumption of cathodically adsorbed hydrogen (H_ads) on the metal surface by the SRB and gets further metabolically oxidised. This oxidation is done by an enzyme called hydrogenase (Javaherdashti 2011). In general, hydrogenase is known to catalyse the reversible oxidation of molecular hydrogen (H₂) as given by Equation (15) (Stott 1993). After the water dissociation (Equation (9)), the produced protons (H⁺) are adsorbed on the metal surface through the cathodic reaction (Equation (10)), which results in the polarisation of the surface. Subsequently, they are removed by hydrogenase for the needed sulphate (SO₄ ²⁻) reduction, which results in depolarisation (Equation (11)) (Javaherdashti 2011). The depolarisation indirectly causes an acceleration of the corrosion process by the stimulation of the cathodic processes (King and Miller 1971; Lv and Du 2018). Therefore, iron is forced to dissolve at the anode (Equation (8)) (Little et al. 2000). The depolarisation can be further considered as a kind of equivalent for the decrease in the activation energy needed for hydrogen removal and, therefore, it bypasses the high activation energy needed for the recombination of hydrogen atoms (Li et al. 2018; Videla 2000). Equation (12) and Equation (13) show that iron sulphide (FeS) and iron hydroxide (Fe(OH₂)) corrosion products are, respectively, produced. Equation (14) represents the overall reaction, and the mechanism of CDT is schematically illustrated in Figure 7 (Mori et al. 2010).

(15) H 2 ⇌ 2 H ⇌ 2 H + + 2 e −

Figure 7:

Schematic illustration of the cathodic depolarisation theory (CDT), based on Mori et al. (2010).

Even though CDT is a well-known theory, it has some important shortcomings. Equation (8) and Equation (12) show that the ratio of the iron that is corroded (Fe²⁺) to the generated iron sulphide corrosion product (FeS) is 4:1. However, in practice, this ratio is typically in the range of 0.9–1 (Tiller 1983). This can be considered as a first shortcoming. A second shortcoming is the fact that the enzyme hydrogenase cannot play a role within the cathodic depolarisation by the removal of atomic hydrogen because it actually cannot work on atomic hydrogen but works on molecular hydrogen instead. It is further unclear how hydrogenase, present within the periplasm of the SRB cell, is able to ‘pull’ the hydrogen away from the extracellular cathode surface (Stott 1993). A third shortcoming is the unclear influence of iron sulphide corrosion products, which are more cathodic compared to iron and could probably play a role in iron dissolution (Videla 2000). Furthermore, it is clear that CDT is only applicable for hydrogenase-positive SRB containing the enzyme hydrogenase and, therefore, CDT is in general assumed to be a lithotrophic (see Table 1) concept (Enning and Garrelfs 2014).

3.2.2 King’s mechanism (iron sulphides)

A second mechanism, which tried to eliminate the shortcomings of CDT, was proposed by King and Miller in 1971 (King and Miller 1971). They focused more on the role of iron sulphide corrosion products as depolarising agents and absorbers of hydrogen within the corrosion mechanism; however, they still attributed a role to the enzyme hydrogenase (Stott 1993; Videla 2000). The iron sulphide is initially, when the concentration of Fe²⁺ is low, present under the form of mackinawite (Javaherdashti 2017; Rickard and Luther 2007) (tetragonal FeS_1-x: FeS_0.93-0.96), which limits the diffusion of Fe²⁺ and inhibits the corrosion by acting as a thin adherent and protective layer (Iverson 1987). Therefore, the corrosion rate in this case is still low and is typically influenced by the growth rate of the SRB (Hamilton 1985). Later, this layer ruptures or cracks by an increase of Fe²⁺ concentration into other forms of iron sulphide, which are typically greigite (Rickard and Luther 2007) (cubic Fe₃S₄) or smythite (Rickard and Luther 2007) (rhombohedral Fe₉S₁₁) and pyrrhotite (Javaherdashti 2017; Rickard and Luther 2007) (hexagonal Fe_1-xS: Fe_0.875-1S) (Videla 2000). A loss of the protective nature and a high corrosion rate typically follow this process. As such, the corrosion rate becomes independent on the SRB growth rate but becomes dependent on the iron sulphide corrosion product and its protectiveness. Bulky black iron sulphide precipitates are typically formed, which are non-adherent and cause in this specific case the cathodic depolarisation (Iverson 1987; Videla 2000). If sufficient iron sulphide is present, a galvanic corrosion cell is established with the iron sulphide as the cathode and iron as the anode, which is also illustrated in Figure 8a (King and Miller 1971). Moreover, the CDT is again schematically illustrated in Figure 8b (King and Miller 1971), which does not indicate the role of iron sulphide as the cathode. As already mentioned, iron sulphide is considered to be cathodic with respect to iron; however, it is not a permanent cathode and high corrosion rates remain by the continuous reformation of the iron sulphide by the precipitation of produced S²⁻ and HS⁻ and the removal of hydrogen by the enzyme hydrogenase (King and Miller 1971). This is because it is generally believed that the activity of the iron sulphide is reduced due to the bonding of atomic hydrogen within its crystal structure and, therefore, the hydrogen removal by hydrogenase can restore the sulphide’s activity (King and Miller 1971).

Figure 8:

Representation of King’s mechanism (a) and cathodic depolarisation theory (CDT) (b), based on King and Miller (1971).

Ocando et al. (2007) reported the relation between Fe²⁺, iron sulphides, H₂S and pH during the corrosion process of iron sheets by D. desulfuricans. Both pH and H₂S concentrations were measured by the use of microelectrodes. On the one hand, it was observed that the presence of Fe²⁺ within the SRB-inoculated medium caused more severe corrosion, which resulted in a dense mass of amorphous iron sulphide precipitates. Moreover, some intergranular corrosion attack was observed. On the other hand, within the SRB-inoculated medium without Fe²⁺, a clean biofilm structure and iron sulphides formed, which protected the iron sheet and avoided consumption of H⁺ by the corrosion process.

3.2.3 Biocatalytic cathodic sulphate reduction theory and extracellular electron transfer

A third mechanism, which is the biocatalytic cathodic sulphate reduction (BCSR) theory, was proposed by Gu et al. in 2009 (Gu et al. 2009). This theory uses as first the principle of bioenergetics to explain the MIC mechanism by SRB. As was mentioned before, lactate for example can be used as an electron donor for the SRB metabolism, as indicated by Equation (16). However, when there is a limitation in the diffusion of the lactate, there will be a lack of both electron donor, energy source, and carbon provider. As a consequence, the SRB must make use of another electron donor for the sulphate reduction and the creation of energy. Therefore, a limitation in the diffusion can be caused by the biofilm, because it will act as a barrier against mass transfer and will even consume some of the lactate resulting in a lower supply for the SRB sessile cells at the bottom of the biofilm (Li et al. 2018). This results in a local lactate or carbon source starvation of the sessile cells being situated the closest to the metal surface (Xu and Gu 2014). In this case, elemental iron (Fe) can be used as an electron donor, according to the oxidation Equation (17). Xu and Gu (2014) observed that the corrosion of carbon steel by D. vulgaris was more severe under carbon source starvation. The sulphate reduction reaction is given by Equation (18) (Zhang et al. 2015). Both the lactate oxidation (Equation (16)), iron oxidation (Equation (17)) and sulphate reduction (Equation (18)) have their redox potentials, denoted as E⁰ under the conditions of 25 °C, a pH of 7, 1 M solute concentration and 1 bar partial pressure for the gases, i.e. CO₂. Therefore, the cell potentials (E_cell) for the lactate oxidation/sulphate reduction pair and iron oxidation/sulphate reduction pair are 213 mV_SHE and 230 mV_SHE, respectively, resulting in a Gibbs free energy change (ΔG) of −164 kJ/mol and −178 kJ/mol, respectively, according to Equation (19) with n the number of involved electrons in the redox reaction, being eight, and F Faraday’s constant (96,485 C/mol). Both coupled lactate oxidation/sulphate reduction and iron oxidation/sulphate reduction are thermodynamically favourable; however, the latter releases more energy under the same conditions (Xu and Gu 2014). This shows that corrosion can occur spontaneously via the iron oxidation reaction. However, it is important to mention that in this case, although there is a driving force for corrosion, this does not imply that corrosion will occur because the kinetics can be unfavourable (Lv and Du 2018). The sulphate reduction needs in general a high activation energy and, therefore, it has a slow reaction rate. However, the biocatalysis provided by the SRB acting as a biocathode accelerates its kinetics. This idea of a biocathode is based on the microbial fuel cell (MFC) technology (Zhou et al. 2013). It should be noted that there is not actually a real cathode on which the sulphate reduction occurs because it happens within the SRB cell itself, i.e. within the cytoplasm, and, therefore, the SRB should be seen as a kind of imaginary biocathode (Lv and Du 2018).

(16) C H 3 C H O H C O O − + H 2 O → C H 3 C O O − + C O 2 + 4 H + + 4 e −   E 0 = − 430   m V S H E

(17) F e → F e 2 + + 2 e −     E 0 = − 447   m V S H E

(18)   S O 4 2 − + 9 H + + 8 e − →   H S − + 4 H 2 O   E 0 = − 217   m V S H E

(19) ∆ G = − n F E c e l l

Figure 9 illustrates the difference between the reaction mechanism when lactate (Figure 9a) or iron (Figure 9b) is used as electron donors. It is clear that the lactate oxidation (Equation (16)) happens within the SRB cytoplasm because the lactate is soluble and diffusible (Jia et al. 2019; Li et al. 2018). However, the iron is not because it is a solid-state electron donor (Philips et al. 2016). Therefore, iron cannot be incorporated within SRB cell and, consequently, its oxidation (Equation (17)) happens extracellularly (Kato 2016; Xu and Gu 2014). The sulphate reduction reaction (Equation (18)) on the other hand occurs intracellularly within the cytoplasm (Li et al. 2018). This implies that the released extracellular electrons of the iron oxidation, being used for the intracellular sulphate reduction, have to be transported through the SRB cell wall into its cytoplasm (Xu and Gu 2014).

Figure 9:

Sulphate reduction under biocatalysis with lactate (a) and iron (b) as electron donors, based on Li et al. (2018).

It is known that electrons cannot ‘swim’ freely within the aqueous solution from the iron to the SRB cytoplasm and, therefore, electron transfer across the cell wall is needed. This so-called extracellular electron transfer (EET), i.e. the electron transport between the metal surface and SRB cell, can happen in two ways: direct electron transport (DET) and mediated electron transport (MET) (Zhang et al. 2015). The SRB that are capable of doing this EET, such as Desulfopila corrodens, are called electrogenic and, therefore, the MIC caused by these SRB is called electrical microbiologically influenced corrosion (EMIC), which is also known as type I MIC within anaerobic conditions (Enning and Garrelfs 2014; Jia et al. 2019; Zhang et al. 2015). According to DET on the one hand, the SRB attach to the iron surface, thereby making direct contact via which they will directly transport and take away the electrons (Kato 2016; Philips et al. 2016). This can be typically realised in different ways. The first one is the use of redox protein-based structures to connect the SRB cell with the iron surface such as membrane-bound c-type cytochrome (Cyc1 and Cyc2) and rusticyanin (Mehanna et al. 2009; Kato 2016; Philips et al. 2016). A second possibility is the use of conductive filamentous apparatus, i.e. outer membrane extensions and pili, which are also denoted as nanowires (Kato 2016; Philips et al. 2016). These nanowires are used as the electrical connections needed for the DET (Reguera et al. 2005).

According to MET, on the other hand, soluble and diffusible redox species are used, also called electron mediators or electron shuttles, which can carry the released electrons from the iron surface to the SRB cell. The mediator is reduced by accepting the released electrons from the iron surface and will subsequently diffuse to the SRB cell, where it is oxidised to donate the electrons needed for the sulphate reduction reaction. Subsequently, they can diffuse back to the iron surface, ready for a new cycle (Kato 2016; Philips et al. 2016). The SRB self-secreted inorganic or organic molecules such as phenazine compounds (e.g. pyocyanin and phenazine-1-carboxamide) and flavine derivates (e.g. riboflavin and flavin adenine dinucleotide) can serve as mediators or artificial mediators can be added, such as methyl viologen (MV), anthraquinone-2,6-disulfonate (AQDS) and ferrocene derivates (Jia et al. 2017; Kato 2016; Philips et al. 2016; Zhang et al. 2015). Hydrogenase-positive SRB can use molecular hydrogen (H₂) as an electron mediator. Consequently, the classical CDT mechanism actually is an electron transfer mechanism using H₂ as the electron mediator, which implies that hydrogenase-negative SRB can only use the DET mechanism or other electron mediators than H₂ (Xu and Gu 2014; Zhang et al. 2015). Furthermore, precipitated iron sulphide corrosion products are believed to also mediate the electron transfer due to their electroconductive properties (Enning et al. 2012; Guan et al. 2016). Both the DET and MET mechanisms are schematically illustrated in Figure 10 (Gu et al. 2019; Li et al. 2018).

Figure 10:

Schematic representation of EET (extracellular electron transfer): DET (direct electron transport) and MET (mediated electron transport), based on Gu et al. (2019) and Li et al. (2018).

Besides the described mechanisms, other mechanisms are corrosive metabolic products or CMIC (H₂S and volatile phosphorous compound (1983)), Fe-binding exopolymers (1995), three stages or Romero mechanism (2005), anodic depolarisation theory (ADT) (1984) and sulphide and hydrogen-induced SCC (1995) (Kakooei et al. 2012; Lv and Du 2018). Comparing EMIC and CMIC for example, the cathodic reaction is, respectively, stimulated by the consumption of cathodic electrons or by secreted corrosive metabolic products (Kato 2016). Furthermore, the reduction of the oxidant in the latter, which attacks the metal, happens extracellularly without the need for biocatalysis (Li et al. 2018). The already mentioned H₂S corrosion serves as a nice illustration for this case.

4.1 General interaction between cathodic protection and microbial life

An applied potential of, e.g. −0.85 V with respect to Saturated Calomel Electrode (V_SCE) is in general sufficient to protect offshore carbon steel structures within aerated seawater against corrosion. However, if microbial life is involved such as SRB within anaerobic environments, the potential needed for CP should be lower, i.e. around −0.95 V_SCE and optimally around −1.03 V_SCE (Guan et al. 2016). Therefore, the interaction between CP and microbial life should be considered for offshore structures (Lunarska et al. 2007). Permeh et al. (2020) performed field condition tests, to evaluate the CP behaviour in the presence of SRB and crevices and, therefore, to elucidate localised corrosion associated with macro and microfouling on submerged steel bridge piles under CP in marine environment. Moreover, laboratory tests were conducted through potentiostatic polarisation tests in SRB-inoculated solution, to evaluate the cathodic reactions occurring under compact and porous crevices, which represented the real marine fouling. Both field conditions tests and laboratory tests were consistent with each other and showed that localised corrosion can still occur under crevice environments during CP, enhanced by the presence of SRB. This shows that even when CP is applied, its efficiency is affected by the presence of SRB, marine fouling and coverage by calcareous deposits. These are typical surface irregularities, which result in non-uniform polarisation conditions and thus the occurrence of localised corrosion.

Gordon et al. (1981) showed that the attachment of two different marine bacteria strains, named B27–5 and B27-13, on copper was increased by CP. However, Dhar et al. (1982) illustrated that the attachment of the marine bacteria Vibrio anguilarum (V. anguilarum) and Pseudonoma atlantica (P. atlantica) on tin oxide and titanium cathodes decreased during CP within both artificial and natural seawater. A decreased settlement and attachment during CP were also observed by Edyvean et al. (1992) for aerobic bacteria on 304 stainless steel and 50 D structural steel.

There is in general a mutual interaction between CP, in particular calcareous deposits, and microbial life. On the one hand, CP will alter the solution chemistry and conditions at the metal surface and, therefore, also the settlement and attachment of the microorganisms. CP will typically affect the growth and reproduction of the microorganisms (He et al. 2023). The consumption of oxygen by CP at the polarised surface (Equation (1)) may result in a decreased reproduction of aerobic bacteria (Edyvean et al. 1992). The production of hydrogen during CP (Equation (2)) is favourable for the growth of hydrogenase-positive SRB, which consume this hydrogen (Guezennec 1994). This hydrogen consumption combined with the precipitation of iron sulphide corrosion products, being more cathodic than steel, will result in an increased current demand for CP to maintain a certain level of protection (Guezennec 1994). Esquivel et al. (2011) concluded that a potential of – 0.950 V with respect to copper/copper-sulphate electrode (V_CSE) was not enough to control MIC of XL 52 steel by SRB and some localised corrosion was observed. This was attributed to both the hydrogen consumption by the SRB and the deposition of iron sulphide corrosion products causing a galvanic corrosion effect. Microorganisms and their biofilm will also interfere with the occurring electrochemical reactions, i.e. oxygen reduction (Equation (1)) and hydrogen evolution (Equation (2)), and, therefore, will also change the current demand for CP (De Gómez Saravia et al. 1997). A biofilm on the metal surface can for example, similar to calcareous deposits, act as a barrier against oxygen diffusion (Guezennec 1994). Secreted metabolic products, such as extracellular enzymes, exopolymer substances and acids, will also alter the solution chemistry at the metal surface and can, therefore, affect the deposition and nature of calcareous deposits (Edyvean et al. 1992; Guezennec 1994). Acid-producing bacteria (APB) that secrete acetic acid for example can dissolve protective calcareous deposits on stainless steel and, therefore, can cause an increase in the current density demand (De Gómez Saravia et al. 1997).

Guezennec (1994) reported that the production of volatile fatty acids such as acetic and butyric acids by the slime-producing bacterium V. natriegens prevented Mg deposition under the form of brucite due to a pH decrease and, therefore, only allowed aragonite to form as a calcareous deposit. Furthermore, if a Mg-rich deposit is already present, a pH decrease can induce its dissolution and, therefore, the bacteria may act as a kind of nuclei for the breakdown of the calcareous deposit. Tian et al. (2019) investigated the combined effect of CP and sulphur-based species on the formation of calcareous deposits on low carbon bainite E690 steel within deaerated ASW with a pH of four and a varying amount of thiosulphate (S₂O₃ ²⁻) concentration. This S₂O₃ ²⁻ tried to simulate SRB as a H₂S producer within an acidic solution. They observed that at 0.01 M S₂O₃ ²⁻, aragonite precipitation was hindered at potentials of −850, −1,050 and −1,200 mV_SCE. This could be attributed to three reasons: i) Other cathodic reactions were dominating, such as reduction reactions of sulphur-based species, thereby not producing OH⁻ needed for the precipitation of the calcareous deposit; ii) HS⁻ was formed, which had a more favourable adsorption on the steel surface compared to CO₃ ²⁻ and OH⁻; and iii) there was an enhanced hydrogen evolution within the solution with S₂O₃ ²⁻, for which H₂ bubbles inhibited the available deposit growth sites and increased the crystal growth time and, therefore, impeded aragonite nucleation. The formation of brucite was also suppressed via this mechanism; however, at a potential of −1,200 mV_SCE, its deposition became possible and brucite was detected by both scanning electron microscopy (SEM) and X-ray diffraction (XRD). Furthermore, Dexter and Lin (1992) investigated the influence of a pre-existing biofilm of marine bacteria on the formation of calcareous deposits under CP of stainless steel within filtered seawater. They concluded that at all used current densities, i.e. 20, 50, and 100 μA/cm², this pre-existing biofilm created more uniform and better distributed calcareous deposits by providing an increase in nucleation sites compared to the situation without biofilm. However, the combined behaviour of the biofilm and calcareous deposits varied depending on the applied current density. At 100 μA/cm², i.e. high current density, the combination of biofilm and calcareous deposits acted as a barrier. The interfacial pH was high enough to deactivate the bacteria’s metabolism, enhancing CP efficiency by actually making the calcareous deposit a more efficient barrier and making the potential more active compared to the situation without biofilm. On the other hand, at 20 μA/cm², the interfacial pH was not high enough to deactivate the bacteria and, therefore, allowed cathodic depolarisation and ennoblement to occur as dominant effects. In this way, the oxygen reduction was enhanced and the CP efficiency decreased, as such increasing the loss of protection compared to the situation without biofilm. It is clear that the bacteria’s metabolism can lead to changes in the Ca/Mg ratio, stability and protectiveness of the calcareous deposit and, therefore, the required current demand for CP (Guezennec 1994).

There are several theories that could explain the improved attachment or improved detachment of bacteria when CP is applied. An overview is provided in the next subsections.

4.1.1 Electrostatic-chemical theory

A first theory that explains the influence of CP on the attachment of bacteria is the electrostatic-chemical theory (Javaherdashti 2017). Zobell (1943) and Marshall et al. (1971) described the formation of microbial surface colonies as a two-step sequence:

1. Reversible sorption: this describes an initial instantaneous attraction to the metal surface. The bacterial cells are somehow weakly attracted by the charge of the metal surface and other forces. However, they have still a Brownian motion.

2. Irreversible sorption: this describes a stable adhesion to the metal surface caused by the production of EPS after several hours.

The reversible sorption phase describes electrostatic interactions. CP will cause the metal surface to be negatively charged, by the creation of a negatively charged OH⁻ diffusion layer. This repels negatively charged bacterial cells upon their approach (Edyvean et al. 1992; Lv and Du 2018). The fact that the cells have a negative charge should be attributed to the cell walls made up of biopolymers, e.g. peptidoglycan, which contain a wide range distribution of negatively charged groups, such as ionised phosphoryl and carboxylate substituents (Wilson et al. 2001). During the irreversible sorption, e.g. negatively charged acidic polysaccharide glycocalyx within the EPS is produced, which can also cause the electrostatic repelling from the negatively charged metal surface during CP (Edyvean et al. 1992). As illustrated in Figure 11, bacterium 1 undergoes a higher repulsive force due to the negatively charged surface compared to bacterium 2. However, if a bacterium, such as bacterium 3, is very close to the negatively charged metal surface, i.e. within a distance of 0.4 nm or less, attractive forces are created, which provide the attachment of the bacterium on the surface (Javaherdashti 2017).

Figure 11:

Interaction of bacteria with a negatively charged metal surface, based on Javaherdashti (2017).

Poortinga et al. (2001) for example found that applying an impressed cathodic current resulted in the desorption of bacterial strains on an indium tin oxide electrode because of both electrostatic repulsion and electrophoretic force, i.e. the force proportional to the electrophoretic mobility of bacteria and the strength of the electric field. (Hong et al. 2008) noticed the same findings when a cathodic current of 15 μA/cm² was applied. Furthermore, Busalmen and de Sánchez (2001) concluded that below potentials of −0.2 V and −0.5 V with respect to Ag/AgCl reference electrode (V_Ag/AgCl), the bacterial adhesion of Pseudomonas fluorescens on a gold film was inhibited. However, it should be noted that the surface charge of the bacterial cells is a function of the interfacial pH, because the latter will have an influence on the protonation degree of the functional and ionogenic groups of the cell wall or glycocalyx polysaccharide within the EPS, as mentioned above (Edyvean et al. 1992). De Gómez Saravia et al. (1997) reported that the isoelectric point of many aerobic bacteria corresponds to a pH value within the acidic region, while that of SRB corresponds to a pH value within the alkaline region. Therefore, at a neutral and an alkaline pH, the aerobic bacteria would have a negative charge and, therefore, will be repelled by a negatively charged metal surface, while the SRB are not.

Electrostatic repulsion is obviously the electrostatic part of the electrostatic-chemical theory. However, there is also a chemical part coupled with this theory. As was already mentioned before, CP will result in an increase of the interfacial pH, due to reduction reactions producing OH⁻ ions, as given by Equation (1) and Equation (2) (Javaherdashti 2017). For an anaerobic environment, such as for SRB, Equation (2) is of course dominating. This alkalisation actually is a negative factor for the bacteria and several studies were also done considering its influence. Pérez et al. (1994) studied the influence of a potential of −0.82 V_SCE and −1.10 V_SCE on the biofouling within a metal-anticorrosive paint-seawater system of two species, i.e. Balanus amphitrite and Polydora ligni. A decreased attachment of the former with more negative potential was observed due to an increased interfacial pH. However, the attachment of the latter was not affected. The bacteria that generally would adhere to the metal surface, despite the electrostatic repulsion, will nevertheless die because of the increased interfacial pH, if its value is of course sufficiently high (Javaherdashti 2017).

4.1.3 Chemical bridge theory and supply of excessive electrons

Another theory, which does not consider electrostatic forces, is the chemical bridge theory (Javaherdashti 2017). It is generally accepted that bacteria may attach to negatively charged surfaces by the use of divalent cations such as Ca²⁺ and Mg²⁺. They will use these cations as ‘bridges’, actually bypassing the influence of electrostatic repulsion caused by the negatively charged metal surface (Edyvean et al. 1992; Javaherdashti 2017). However, as already mentioned, because of an increased pH during CP caused by Equation (1) and Equation (2), calcareous deposits such as CaCO₃ and Mg(OH)₂ can precipitate as described by Equations (3)–(5). Due to their precipitation, there is a decrease of Ca²⁺ and Mg²⁺ at the metal interface. Because of this decrease, the bacteria do not have these ions anymore to attach to the surface via ‘bridge’ formation (Edyvean et al. 1992). Eashwar et al. (2009) observed a reduction of biofilm density on a stainless steel cathode polarised at −0.70 V_SCE within coastal Indian seawater with mild steel and zinc anodes. This was attributed to the deposition of calcareous deposits, mainly CaCO₃ and Mg(OH)₂.

Liu and Cheng (2017) studied the influence of CP and SRB, i.e. D. desulfuricans, on the corrosion behaviour of X70 pipeline steel within extracted soil solution. They observed that the application of CP did not affect the planktonic bacterial cells; however, when the applied potential was more negative, there was an increase in the number of sessile cells and thus in the bacterial attachment. This was caused by the fact that the bacteria could use the excessive electrons, supplied by the power source of CP, as an electron donor instead of steel. A potential of −850 mV_CSE was not sufficient to protect the steel from uniform corrosion, compared to −1,000 mV_CSE. However, the shielding effect of the biofilm when the cells were adhered at −1,000 mV_CSE caused fluctuations in the potential. This means that if locally the potential, and thus the electron supply, is not sufficient, the SRB can use the steel again as an electron donor. As a consequence, pitting corrosion could occur at −1,000 mV_CSE. The situation is illustrated in Figure 12. The use of excessive electrons provided by CP is an argument for an improved bacterial attachment (Lv and Du 2018).

Figure 12:

Use of electrons supplied by CP as electron donor (a) and possible pitting corrosion because of shielding (b). Reprinted from Journal of Alloys and Compounds, 729, Liu and Cheng, The influence of cathodic protection potential on the biofilm formation and corrosion behaviour of an X70 steel pipeline in sulphate reducing bacteria media, 180–188, Copyright (2017), with permission from Elsevier.

The corrosion behaviour of L360 N steel by SRB, i.e. Desulfovibrio, under CP was reported by He et al. (2023). On the one hand, at a potential of −0.85 V_SCE, an increased corrosion rate and anodic dissolution (AD), even higher than at open circuit potential (OCP) conditions, were observed due to the acceleration of the SRB growth and their metabolism by the promotion of the hydrogen proton (H⁺) reduction into adsorbed hydrogen, being consumed by the SRB, due to an increased electron supply. The local corrosion rate, under the form of pits, was accelerated and deeper pits were created compared to OCP. On the other hand, at −0.95 V_SCE, the SRB growth and metabolism were inhibited due to a higher alkalisation on the metal surface, creating conditions unsuitable for SRB adhesion. In this case, the corrosion was slowed down. Moreover, due to the higher alkalisation and thus higher interfacial pH, CaCO₃ calcareous deposits could precipitate, giving less Ca²⁺ available as possible ‘bridges’ for the SRB adsorption on the metal. In this way, CaCO₃ formed a protective film and any form of local corrosion could not be observed in this case. Lastly, at −1.05 V_SCE, the SRB could receive a large number of ‘free’ electrons by the supply of excessive electrons and, therefore, the SRB growth and metabolism were promoted again. However, calcareous deposits were also able to form, thus giving a lower corrosion rate compared to the situation at −0.85 V_SCE. The local corrosion rate under the form of pits was in this case low and fewer deep pits were created compared to OCP and −0.85 V_SCE. A similar study was done by Guan et al. (2016) on high strength EQ 70 steel and by using D. caledoniens at the same potentials. At −0.85 V_SCE, an increased corrosion rate was observed compared to OCP conditions. This was attributed to a promoted SRB metabolic activity. Corrosion cracks were observed on the surface consisting of a biofilm with SRB and corrosion products. After the removal of the surface products, bigger and deeper corrosion pits were observed compared to OCP conditions. At −0.95 V_SCE, the corrosion rate was decreased compared to −0.85 V_SCE; however, this potential was still insufficient to prevent corrosive attack by the SRB metabolism and corrosion cracks were again observed. After the removal of the products, there were again corrosion pits, however, more rounded compared to −0.85 V_SCE. Lastly, at −1.05 V_SCE, the specimen was effectively protected and calcareous deposits were observed on the surface, suggesting a restraining on the SRB metabolic activity. After the removal of the products, only a few pits were observed. It was suggested that, on the one hand, at −0.85 V_SCE, the SRB metabolism was promoted by the DET mechanism, as was discussed before. It is believed that the SRB could obtain electrons for their metabolism, i.e. the sulphate reduction (Equation (18)), more easily by the cathodic polarised electrode via the use of protein-based structures or conductive nanowires. The situation is illustrated schematically in Figure 13. MET is also mentioned via the use of hydrogen and electron mediators (med) in general. At −1.05 V_SCE, the calcareous deposits and alkalisation were not suitable for the SRB metabolism and growth, and the deposits also hindered the DET mechanism.

Figure 13:

Electron transfer between a negatively polarised surface and SRB via MET and DET. Reprinted from Guan et al. (2016), CC BY 4.0 license: https://creativecommons.org/licenses/by/4.0/, not adapted.

Lv et al. (2022) examined the corrosion behaviour of X65 steel by Pseudomonas, i.e. an IOB species, under CP within a simulated seawater solution at two different potentials, i.e. −850 mV_SCE and −1,050 mV_SCE. They concluded that at −1,050 mV_SCE, the protective effect against IOB-induced corrosion, in the form of pitting, was larger compared to −850 mV_SCE and OCP conditions. This was attributed to both an increased electrostatic repulsion between the steel surface and bacteria, an increase in pH value, a decrease in the amount of dissolved oxygen and the precipitation of calcareous deposits, i.e. CaCO₃ and Mg(OH)₂, the latter forming a protective barrier against bacterial attachment. Contrary, at −850 mV_SCE, these effects were still too low, and even more, an increased production of EPS, as a response to ‘damage’ caused by cathodic polarisation, could increase the adhesion of bacteria cells and could resist the electrostatic repulsion. Therefore, localised corrosion and the formation of corrosion products could still occur. Even more, at −1,050 mV_SCE, the bacteria still showed tolerance, and due to the thickening of the calcareous deposit layer, the electrostatic repulsion weakened and, therefore, led to a heterogeneous biofilm because of re-attachment of IOB, which could facilitate pitting. Therefore, this also clearly shows the competition between a MIC-promoting effect of the bacteria’s metabolism and biofilm formation on the one hand and a MIC-inhibitory effect of CP on the other hand. Furthermore, it was stated that CP affects less the growth and reproduction of planktonic bacteria, because they are further away from the protected surface compared to sessile bacteria within the biofilm attached to the metal surface. It is clear from the abovementioned studies that the influence of CP on bacteria will be different for different species and their metabolism, which affects the distribution of the biofilm, corrosion products and calcareous deposits and, therefore, the occurrence of localised corrosion.

4.2 The influence of cathodic protection on hydrogen uptake

As was already mentioned above, cathodic overprotection will result in an increased generation of hydrogen at the steel surface (Gurrappa et al. 2015; Lazzari and Pedeferri 2018). In general, if CP is applied such that the potential is more negative than the hydrogen reduction potential of the metal, the Volmer reaction occurs, which is given by Equation (22) for alkaline or neutral solutions and Equation (23) for acidic solutions with M the metal’s free surface and MH_ads the hydrogen atom adsorbed at the metal’s surface. It is clear that adsorbed hydrogen can be generated by the electrochemical reduction of water (H₂O) or a hydrated hydrogen ion (H₃O⁺) within alkaline/neutral solutions and acidic solutions (Dafft et al. 1979; Lasia and Grégoire 1995), respectively. For seawater, which is buffered by chemical species with an average pH of 8.2 and displays a pH variation between 7.5 and 8.3, an alkaline situation can be expected (Orozco-Cruz et al. 2023).

After the adsorption of hydrogen by the Volmer reaction, there are three possibilities. First, the adsorbed hydrogen atom (MH_ads) can recombine with another adsorbed hydrogen atom (MH_ads) through the Tafel reaction, given by Equation (24). This is a chemical recombination or desorption (Xu et al. 2021).

Second, recombination or desorption can also occur electrochemically, through the Heyrovsky reaction. In an alkaline/neutral solution, as represented by Equation (25) involving water (H₂O), while for an acidic solution, it involves a hydrated hydrogen ion (H₃O⁺), as represented by Equation (26) (Dafft et al. 1979; Lasia and Grégoire 1995).

It should be clear that in both cases, i.e. the chemical recombination by Tafel and electrochemical recombination by Heyrovsky, hydrogen gas molecules (H₂) are produced. If the Volmer reaction is followed by the Tafel reaction, it is called the Volmer–Tafel reaction pathway. If the Volmer reaction is followed by the Heyrovsky reaction, it is called the Volmer–Heyrovsky reaction pathway (Popov et al. 2018). Third, the adsorbed hydrogen atom (MH_ads) can enter the metal and absorbs into the metal’s inner surface (MH_abs), as shown by Equation (27). This absorbed hydrogen can subsequently diffuse in the bulk of the metal (Sun and Frank Cheng 2022).

Atomic hydrogen will diffuse via an interstitial lattice diffusion, driven by Fick’s law according to a concentration gradient (Sun et al. 2019; Sun and Frank Cheng 2022). The diffusing atomic hydrogen may also accumulate in hydrogen traps, with both reversible or weak and irreversible or deep traps possible, such as grain boundaries, dislocations, vacancies and precipitates (Koyama et al. 2017). Moreover, when atomic hydrogen is present within a microvoid or an inclusion, e.g. MnS, it may recombine with another hydrogen atom to form molecular hydrogen (H₂). This can result in either hydrogen-induced cracking (HIC) or hydrogen surface blistering, which can occur in the absence of an external load (Lazzari and Pedeferri 2018; Popov et al. 2018). However, in the presence of stress hydrogen-assisted cracking (HAC) can occur (Cabrini et al. 2011; T. Zhang et al. 2018b).

In general, an optimum balance is needed between a low level of corrosion, which means no underprotection, and an allowable risk of HE, thus no high overprotection (Batt et al. 2002). T. Zhang et al. (2018b) investigated the influence of CP on the hydrogen permeation, through the Devanathan-Stachurski cell (Devanathan and Stachurski 1962), of three different pipeline steels, i.e. X70, X80 and X100, within simulated seawater, according to ASTM standard D1141 (An American National Standard 2003). The results showed that the hydrogen permeation current density and subsurface hydrogen concentration increased with a more negative potential. Furthermore, fracture morphologies of the specimen after slow strain rate testing (SSRT) and polarisation curves at slow and fast sweep rates showed that HE fulfilled a more important role at the crack tip compared to anodic dissolution (AD) when a more negative potential was applied. This was also clear from the transition from a ductile fracture to a more brittle fracture with more negative potential.

Besides their role in the efficiency of CP, calcareous deposits influence the hydrogen uptake of steel. It is, however, still an open debate whether their role is promoting or retarding. It is logical to reason that because of the formation of calcareous deposits, less effective steel surface area is available for the water reduction reaction (Equation (2)) and, therefore, results in a lower current density and lower hydrogen uptake. However, by providing a barrier to the diffusion of dissolved oxygen, the increase in interfacial pH is limited and, therefore, they may promote hydrogen uptake. Moreover, slowing down the recombination reaction or diffusion of molecular hydrogen (H₂) away from the steel surface is also given as argument for an increased hydrogen ingress by calcareous deposits (Hartt 2006; Hinds and Turnbull 2005).

L. Zhang et al. (2018a) investigated the influence of Ca²⁺ on the formation of calcareous deposits and on the hydrogen permeation of X80 high strength pipeline steel. This was done by performing permeation tests and thermal desorption tests within 0.1 M NaNO₃ and 0.1 M Ca(NO₃)₂ solutions under galvanostatic charging. They concluded that the permeation rate of hydrogen atoms, sub-surface hydrogen concentration and hydrogen content were larger in the Ca(NO₃)₂ solution compared to the NaNO₃ solution. This was attributed to the formation of Ca(OH)₂ deposits, which would enhance the efficiency of hydrogen absorption according to Lillard et al. (2000) by lowering the rate constant for hydrogen desorption, hindering the escape of hydrogen bubbles (H₂) and increasing the rate constant for hydrogen absorption. Therefore, it was believed that Ca-rich deposits had an acceleration effect on hydrogen entry and hydrogen permeation. Ma et al. (2021) examined the hydrogen permeation under galvanostatic charging of X65 steel within simulated Canadian groundwater (NS4) solution and compared it with the hydrogen permeation behaviour within a modified NS4 solution without Ca²⁺ and Mg²⁺. They concluded that the hydrogen permeation current and the hydrogen concentration on the charging side were higher in the NS4 solution compared to the modified one. This was attributed to the formation of a dense thin Mg(OH)₂ layer with scattered CaCO₃ cubic and snowflake-like crystals on it, which were assumed to inhibit the hydrogen recombination reaction and to promote the hydrogen absorption within the steel. As a consequence, the risk of HE under CP could be increased. Gao et al. (2016) proposed that calcium deposits on 16Mn steel in seawater environment inhibited the diffusion of molecular hydrogen to the solution, resulting in an increase of the concentration of hydrogen at the metal surface and, therefore, inhibited the Tafel reaction (Equation (24)) directly and the Volmer reaction (Equation (22)) indirectly.

Contrary, Ou and Kuo Wu (1997) published opposite results. They performed hydrogen permeation tests on cold rolled milled steel at 25 °C at different cathodic current densities under galvanostatic charging within different solutions: solution A: synthetic seawater; solution B: synthetic seawater without CaCl₂, NaHCO₃ and MgCl₂; and solution C: synthetic seawater without CaCl₂ and NaHCO₃. They concluded that the hydrogen permeation rate at a certain cathodic current density was the lowest for solution C, intermediate for solution A and the highest for solution B. Solution B did not form calcareous deposits, while the calcareous deposits in solution A were CaCO₃ and Mg(OH)₂, and loose Mg(OH)₂ deposits for solution C. Both the thick Ca-rich deposit layer for solution A and a thin Mg(OH)₂ loose deposit for solution C were considered to act as hydrogen diffusion barriers, with the latter giving higher barrier efficiency and lower hydrogen absorption. These calcareous deposits were, therefore, beneficial for the resistance against HE. Lucas and Robinson (1986) observed a decrease in the hydrogen flux for carbon-manganese steel within ASW despite successive lowering of the potential. This was believed to be caused by the formation of Ca-rich and Mg-rich deposits, which were protective and impermeable to hydrogen. Further, Hinds and Turnbull (2005) reported that the formation of calcareous deposits, being CaCO₃ in the form of aragonite needles, decreased the hydrogen uptake of super martensitic stainless steel within ASW at −1 V_SCE compared to permeation tests within deaerated 3.5 % NaCl solution. Smith and Paul (2015) also observed a reduction of the hydrogen ingress for high-strength low-alloy steel under a potential of −1.1 V_SCE within synthetic seawater compared to 3.5 % NaCl solution. This was attributed to the decrease of exposed surface needed for the water reduction reaction, caused by brucite formation.

Simoni et al. (2017) compared the hydrogen permeation behaviour for API 5CT P110 steel at potentials of −1 V_SCE and −1.5 V_SCE in ASW solution, ASW solution without Ca²⁺ and Mg²⁺, and 3.5 % NaCl solution. A competition was proposed between a surface effect of the calcareous deposits, caused by the blocking of available adsorption sites for the hydrogen reduction reaction, and hydrogen overpotential, being larger if the local pH is lower. At −1 V_SCE, on the one hand, a lower hydrogen uptake within the ASW solution was observed compared to the two other solutions, i.e. ASW without Ca²⁺ and Mg²⁺, and 3.5 % NaCl. This was attributed to the compensation of the higher hydrogen overpotential, because of buffer reactions within ASW, by the surface effect caused by double-layered deposits consisting of a brucite inner layer and an aragonite outer layer. At −1.5 V_SCE, on the other hand, a higher hydrogen uptake within the ASW solution was observed compared to the other solutions. This was attributed to the dominating difference in hydrogen overpotential, making the surface effect insignificant. Moreover, the deposits consisted now mainly of porous brucite, being more detached due to the destructive influence of intensive hydrogen bubbling and, therefore, it made the surface effect even less crucial.

Xing et al. (2021) investigated the hydrogen permeation behaviour of X80 steel under impressed cathodic current density within the range of 10–125 mA/cm² within a simulated soil environment. The sub-surface hydrogen concentration increased and subsequently decreased under an increased current density, with a maximum at 75 mA/cm². When the current density was below 75 mA/cm², the deposit mainly consisted of porous CaCO₃, and at 75 mA/cm², a squamous structure with a lot of gap interfaces was formed, promoting the hydrogen permeation. When the current density was above 75 mA/cm², the deposit consisted of a dense Mg(OH)₂ film, inhibiting the hydrogen permeation. Furthermore, the hydrogen permeation was also galvanostatically evaluated within three different solutions: simulated soil solution, simulated soil solution without CaCl₂ and NaHCO₃ and simulated soil solution without CaCl₂, NaHCO₃ and MgCl₂. The sub-surface hydrogen concentration was the highest within the simulated soil solution with Ca²⁺ and Mg²⁺, and the lowest within the simulated soil solution without Ca²⁺. An intermediate situation was observed for the solution without Ca²⁺ and Mg²⁺, where deposits could not form on the steel surface. For the simulated soil solution with Ca²⁺ and Mg²⁺, this finding was attributed to porous CaCO₃ crystal deposit, where the hydrogen atoms were trapped within cavities under the crystals or within the gaps between the crystals. The Volmer reaction (Equation (22)) could easily occur within these cavities and gaps. Moreover, the Tafel recombination reaction (Equation (24)) was inhibited by a decrease in its rate constant, because it was difficult for the molecular hydrogen (H₂), due to its larger volume compared to hydrogen atoms, to diffuse out of the cavities or gaps. Therefore, the coverage of the hydrogen atoms on the cavity or gap interface was increased, which promoted the absorption rate constant and reduced the desorption rate constant. For the simulated soil solution without Ca²⁺, the observation was attributed to amorphous and dense Mg(OH)₂. This deposit reduced the hydrogen adsorption sites and, therefore, decreased the hydrogen uptake. In this situation, the desorption or recombination step was according to the Heyrovsky reaction (Equation (25)). The different discussed situations are schematically illustrated in Figure 14.

Figure 14:

Influence of the calcareous deposits on the hydrogen uptake for (a) simulated soil solution without Ca²⁺ and Mg²⁺, (b) simulated soil solution with Ca²⁺ and Mg²⁺ and (c) simulated soil solution without Ca²⁺. Reprinted from International Journal of Hydrogen Energy, 46, Xing et al., Effect of surface calcareous deposits on hydrogen uptake of X80 steel under strong cathodic current, 4,555–4,566, Copyright (2021), with permission from Elsevier.

Furthermore, Xu et al. (2021) examined the role of previously precipitated calcareous deposits, formed by the application of potentials ranging from −0.8 V_SCE to −1.2 V_SCE, on the subsequent hydrogen permeation behaviour of AISI 4135 steel under galvanostatic conditions in synthetic seawater. Therefore, the direct influence of the calcareous deposits on the hydrogen permeation could be elucidated. It was concluded that the calcareous deposits facilitated hydrogen permeation compared to a bare steel surface. This was attributed to the inhibition role of the deposits on the hydrogen recombination, both the Tafel reaction (Equation (24)) and Heyrovsky reaction (Equation (25)), and, therefore, they hindered the escape of hydrogen molecules from the steel surface. However, it was noted that the hydrogen permeation rate was related in a proportional way to the integrity and the compactness of the calcareous deposit. The highest difference in hydrogen permeation efficiency between bare steel and steel with deposits was obtained for the deposit formed at −1.0 V_SCE, because it had a very good integrity and compactness, being double-layered with an intact brucite inner layer and cluster-shaped aragonite outer layer. A decrease in the efficiency difference at the more negative potentials −1.1 V_SCE and −1.2 V_SCE was attributed to a weakening of the integrity and compactness. At both potentials, the brucite layer was more porous, showing a honeycomb-like structure with clear cracks. Furthermore, the aragonite at −1.2 V_SCE consisted of smaller and sparsely distributed bouquet-shaped deposits. At −1.2 V_SCE, hydrogen evolution under the deposit layer could also occur, making a calcareous deposit of low integrity due to induced stress and causing flaky or blistered and non-adherent deposits. This realised more pathways for the hydrogen molecules to escape and oxygen molecules to diffuse to the steel surface, reducing the importance of the Volmer reaction (Equation (22)) among other possible cathodic reactions. The situation is schematically illustrated in Figure 15.

Figure 15:

Schematic of the influence of calcareous deposit on the hydrogen permeation and the relation with integrity and compactness. Reprinted from International Journal of Hydrogen Energy, 46, Xu et al., Investigation of the calcareous deposits formation controlled by interfacial pH and its effect on the hydrogen entry into AISI 4135 steel in seawater, 5,824–5,841, Copyright (2021), with permission from Elsevier.

It is clear that the integrity of the calcareous deposit is indeed a very important factor. Piri and Arefinia (2018) studied the influence of potential, bicarbonate (HCO₃ ⁻) concentration and temperature on the CaCO₃ deposition within an artificial solution of 0.20 g/L NaHCO₃ and 1.16 g/L CaCl₂. The variation in HCO₃ ⁻ concentration and temperature were studied at a potential of −1.0 V_SCE. By chronoamperometry curves, i.e. current as a function of immersion time, they concluded that a shift towards more negative potential, increase of HCO₃ ⁻ concentration and increase of temperature all individually increased both the residual current density and calcareous deposit scaling time. They attributed this to a dominant effect of hydrogen evolution for all three cases. Created hydrogen bubbles caused detachment and spalling of the calcareous deposit and led to an increased porosity and a reduced compactness of the scale. The influence of hydrogen bubbles on calcareous deposits was further examined by Salvago et al. (2003), by testing different metals under cathodic polarisation within both artificial and natural seawater. By PDP curves, different situations could be observed: a) The hydrogen evolution via macroscopically visible bubbles, b) the adherence of calcareous deposits to the hydrogen gas bubbles and suspension of deposit particles within the solution and c) the detachment of the calcareous deposit due to the hydrogen evolution. These events occurred at more negative potentials and cathodic current densities, respectively.

It is clear from the abovementioned studies that the findings are contradictory, and, therefore, the role of calcareous deposits in hydrogen uptake remains unclear and the underlying mechanism remains unresolved. As stated by Tavassolian et al. (2025), the origin of the contradictory results can be first of all attributed to the different factors influencing the deposition, morphology and protectiveness of the calcareous deposits. These were mainly the solution composition (basically related to the presence of only Ca²⁺, only Mg²⁺, both of them or none of them), the applied potential and the applied current density in the abovementioned literature studies. Such different conditions can alter the hydrogen permeation behaviour to a large extent. It is clear that the separate effect of the calcareous deposit constituents, i.e. predominantly CaCO₃ and Mg(OH)₂, on hydrogen uptake should be more isolated, to minimise the influence of other ions within the electrolyte or other constituents within the deposit. Therefore, simplified solutions should be used for a more controlled assessment of the influence on hydrogen uptake and the underlying mechanism. A second reason is related to the duration of the performed experiments, being often (too) short. The duration influences both the compactness, number of crystals and thickness of the calcareous deposits, thereby affecting also the hydrogen uptake. As concluded from experiments by Rousseau et al. (2010) in natural seawater under CP, for a maximum of 12 months, densification of the deposit’s microstructure was observed as the immersion time was longer. During a short-term experiment, hydrogen saturation of the metal may occur faster than the deposition of a full-covered calcareous deposit, thereby affecting the results. Therefore, after a longer immersion time, the calcareous deposit may act as a protective coating and can block hydrogen uptake.

Tavassolian et al. (2025) already examined the role of calcite on the hydrogen permeation of S235JR steel by performing CP within a simplified artificial seawater solution consisting of NaCl, CaCl₂ and NaHCO₃, therefore, isolating the effect of CaCO₃, under the form of calcite, on hydrogen uptake. On the one hand, it was shown that both short-term immersion and lower CP current density facilitated hydrogen uptake, because of fine and porous calcite formation. On the other hand, long-term immersion, i.e. by using preformed deposit samples, and higher CP current density reduced hydrogen uptake, because the calcite acted as a protective coating. The results were confirmed by both SEM and in situ EIS analysis. The latter serves as a very powerful technique since it provides information regarding the coverage, compactness and morphology of the deposits, and it gives a correlation with the observed hydrogen permeation behaviour.

To evaluate further the individual influence of Mg(OH)₂ and the combined influence of both CaCO₃ and Mg(OH)₂ on hydrogen uptake, a similar approach with simplified solutions and investigation of parameters, such as immersion time and current density, individually, should be used. Besides these parameters, other parameters or experimental conditions such as applied potential, temperature and applied stress could be examined, since they are all relevant for e.g. offshore applications. All these parameters can affect the composition, morphology, compactness or porosity and thickness of the deposits, which all affect the subsequent hydrogen uptake.

4.3 The influence of sulphate-reducing bacteria on hydrogen uptake

Besides cathodic overprotection, corrosion, caused by SRB in particular, is the main source of hydrogen uptake of steel within a seawater environment (Robinson et al. 1998). It is stated that SRB could aggravate hydrogen uptake and, therefore, HAC of jack-up offshore installations (Mathisen 2024). As already mentioned, H₂S corrosion can occur in the presence of SRB because of their metabolism based on the sulphate reduction reaction. During the corrosion process, atomic hydrogen can be generated at the steel surface, as it involves several cathodic reactions consisting of a series of depolarisation processes involving both H₂S, HS⁻ and H⁺ (Wen et al. 2018; Zhang et al. 2012). As was already clear, this atomic hydrogen can be partially absorbed within the steel (Carneiro et al. 2003). HIC and sulphide stress cracking (SSC) are typical problems encountered for pipelines within H₂S-containing environments (Huang et al. 2017). SSC is generally defined as ‘the cracking of a metal under the combined action of tensile stress and corrosion in the presence of water and hydrogen sulphide’ (NACE 2002; Zhou et al. 2020). It is generally accepted that it is caused by HE, in particular for high-strength steels (Berkowitz and Heubaum 1984; King 2017). Several factors influence the hydrogen uptake within H₂S-containing environments.

First of all, H₂S and other related sulfidic species such as S²⁻ and HS⁻ act as hydrogen recombination poisons, by inhibiting the transformation of hydrogen atoms into hydrogen molecules being possible via the Tafel or Heyrovsky reaction. Therefore, hydrogen absorption and diffusion are promoted, which can result in hydrogen-related failures (Huang et al. 2017; Lazzari and Pedeferri 2018). Furthermore, H₂S has a catalytic action and, therefore, accelerates the proton discharge, i.e. Volmer reaction, within the cathodic reaction mechanism (Banakhevych et al. 2009).

Second, iron sulphide corrosion products (Fe_xS_y) will also influence hydrogen uptake. However, it is still unclear whether their role is accelerating or inhibiting. This is closely related to their protectiveness and compactness influenced by the layer thickness, morphology, porosity and composition (Karimi et al. 2023a), which in turn depend on several factors such as temperature (Zheng et al. 2013), H₂S concentration (Tang et al. 2010), pH (Huang et al. 2017), flow (Anyanwu 2019) and applied stress (Ge et al. 2020). Huang et al. (2017) reported that pre-formed Fe_xS_y films had an increased blocking effect on hydrogen permeation with an increase in H₂S concentration and a decrease in pH value. The highest blocking effect was given by a mixed corrosion product consisting of both mackinawite and crystalline FeS. Hao et al. (2011) concluded that a mackinawite-dominated scale and a pyrrhotite-dominated scale promoted and reduced the hydrogen steady-state permeation flux, respectively. However, despite their possible role as a diffusion barrier, Fe_xS_y products are typical corrosion products that are well conductive, as mentioned before, and, therefore, provide active surface sites for the hydrogen reduction reaction or Volmer reaction (Kim et al. 2012). Furthermore, Fe_xS_y products possessing ion-selectivity are also reported as determining factors for hydrogen permeation (Wan et al. 2023; Zhao et al. 2005). It should be noted that the dual role of these Fe_xS_y corrosion products on the hydrogen uptake during H₂S corrosion is very analogous to the dual role of calcareous deposits on the hydrogen uptake during CP.

As is clear, SRB will influence the hydrogen permeation process. The hydrogen permeation of high strength steel was in situ monitored by Y. Xu et al. (2022a) for 3 months in different marine corrosion zones, i.e. marine-atmospheric zone, splash zone, tidal zone and immersion zone, to get insight into the hydrogen permeation behaviour within a marine environment. The results showed a positive correlation between the corrosion rate of the steel, dependent on the marine zone, and hydrogen permeation current density. Within the immersion zone, a high hydrogen permeation current density was attributed to iron sulphide corrosion product produced by the metabolic activity of SRB. Zhu et al. (2007) investigated the influence of SRB on the hydrogen permeation of API X56 steel by comparing the permeation within a sterile sea mud solution and a SRB-inoculated sea mud solution. They observed that the hydrogen permeation current density within SRB-inoculated sea mud solution was three times larger compared to the sterile sea mud solution. H₂S, HS⁻ and S²⁻ were the main products of the SRB metabolism affecting the hydrogen permeation, by catalysing the Volmer reaction (Equation (23)), but inhibiting the Tafel reaction (Equation (24)). Therefore, the hydrogen concentration at the steel surface was increased and hydrogen permeation thus accelerated. Furthermore, the cathodic depolarisation reaction (Equation (11)) was stated as another reason for the improved hydrogen permeation by SRB. Lunarska et al. (2007) observed an increase in the hydrogen permeation and diffusible hydrogen content of hot rolled steel at a certain potential within a SRB-inoculated seawater solution compared to a sterile seawater solution. This was attributed to both generated S²⁻, acting as a promotor for hydrogen entry, and to the increase in polarisation current due to the formation of less protective calcareous deposits with a patched appearance. These less protective deposits should result from their decreased adhesion attributed to biofilm formation.

D. Wang et al. (2017a) found a correlation between the hydrogen permeation behaviour, the occurrence of hydrogen blisters on the surface and the SRB activity of X80 steel within Yingtan simulated soil solution. On the one hand, hydrogen permeation was investigated at OCP conditions within both sterile solution and SRB-inoculated solution with the bacteria in their stable growth phase. The steady-state hydrogen permeation current was three times higher within the SRB-inoculated solution compared to the sterile solution. Furthermore, clear hydrogen blisters were observed on the steel surface after immersion within the SRB-inoculated solution, while less frequent and less serious blisters were observed after immersion within the sterile solution. On the other hand, the hydrogen permeation and blistering were investigated at OCP conditions during specific phases of the SRB growth curve: at 4 days, i.e. in their logarithmic growth phase, at 8 days, i.e. in their stable growth phase, and at 14 days, i.e. in their residual phase. The typical growth curve of the SRB is illustrated in Figure 16. The highest hydrogen permeation current density and most severe blistering, under the form of dense and uniform blisters, were observed for the situation of 8 days, because the most active SRB cells were then present and the most sulphide was produced by their metabolism. The lowest hydrogen permeation current density and least hydrogen blistering were observed for 14 days, because most SRB cells had apoptosis, i.e. cell death, and a low amount of sulphide was present within the solution. An intermediate situation was observed for 4 days, for which the hydrogen permeation current density and severity of hydrogen blistering were intermediate between 8 days and 14 days. It should be mentioned, however, that during all three discussed situations regarding the SRB growth curve, the hydrogen permeation current density and blistering were still higher compared to the sterile solution. The increased hydrogen permeation within SRB-inoculated solution was again attributed to the catalysing effect of sulphide species (H₂S, HS⁻ and S²⁻) on the Volmer reaction and inhibition of the Tafel reaction by their poisoning effect. Furthermore, the cathodic depolarisation by the sulphate reduction reaction was again an argument.

Figure 16:

Typical growth curve for SRB. Reprinted from International Journal of Hydrogen Energy, 42, D. Wang et al., The effect of sulphate-reducing bacteria on hydrogen permeation of X80 steel under cathodic protection potential, 27, 206–27213, Copyright (2017), with permission from Elsevier.

Bueno et al. (2014) simulated the influence of H₂S produced by SRB on API 5 L grade X60 steel by adding thiosulphate (S₂O₃ ²⁻) to NS4 synthetic soil solution. Ductility loss was observed through SSRT if more S₂O₃ ²⁻ was added to the standard NS4 solution, which was attributed to HE. Furthermore, a higher degree of hydrogen permeation was observed when the S₂O₃ ²⁻ was added. Therefore, a catalytic effect of H₂S on the Volmer reaction and hydrogen concentration within the steel was again concluded.

Further, a competition between the hydrogen promotion effect by H₂S and the hydrogen blocking effect by mackinawite was observed by Tian et al. (2018). The influence of S₂O₃ ²⁻ concentration (10⁻⁴, 10⁻³ and 10⁻² M), simulating again H₂S production by SRB, was investigated within acidic deaerated ASW on the hydrogen permeation behaviour of E690 steel under OCP conditions. It was concluded that the hydrogen permeation was higher at 10⁻⁴ M compared to 10⁻³ M and 10⁻² M because within the latter environments formed mackinawite corrosion product inhibited the hydrogen uptake. According to the SCC behaviour, HE fulfilled the dominating role at the lower concentration, while severe pitting corrosion or anodic dissolution (AD) was dominating at the higher concentrations of S₂O₃ ²⁻.

The formed biofilm can also have an influence on the hydrogen evolution and hydrogen absorption reaction (Lunarska et al. 2007). However, it is reported that it has a dual role. On the one hand, it can disrupt a present protective oxide layer or passive film and, therefore, it allows hydrogen to have direct access to the metal surface (Cai et al. 2025). On the other hand, it is believed that the biofilm itself can block the hydrogen uptake (De Romero et al. 2005; Li et al. 2022; Lunarska et al. 2007).

Within the abovementioned research studies, sulphide metabolites and Fe_xS_y corrosion products are always the only reported species regarding hydrogen uptake within a SRB-inoculated solution. However, other metabolites can also be secreted by SRB, which influence hydrogen uptake. Besides sulphide metabolites produced by the sulphate reduction reaction, there are of course also produced metabolites by the oxidation reaction. As mentioned in Section 3.1, several electron donors, besides iron according to the biocatalytic sulphate reduction theory (Section 3.2.3), can be used by SRB. Such a donor, which is typically an organic species, e.g. pyruvate, lactate, acetate, etc., can be incompletely oxidised with acetate (CH₃COO⁻) as the end product or completely oxidised with CO₂ as the end product. Moreover, upon the oxidation of these organic species, hydrogen is generated, as can be seen for example in Equation (16) for lactate, which could promote hydrogen permeation (Zhang et al. 2024).

On the one hand, acetate as a product may affect the hydrogen permeation process. Duvall Deffo Ayagou et al. (2019) for example showed that acetate could increase the hydrogen permeation, because it competed with H₂S to react with Fe²⁺. In this way, the formation of protective iron sulphide was delayed and disrupted, and iron acetate corrosion product (Fe(CH₃COO)₂) formed, which had higher solubility in an aqueous medium and, therefore, was less protective. This decreased protectivity could facilitate hydrogen entry. Moreover, acetate may also be used as an electron donor itself, and its oxidation by SRB may result in the formation of HCO₃ ⁻ (Z. Zhang et al. 2022b), which will affect hydrogen permeation as will be mentioned below.

On the other hand, CO₂ as a product can also affect the hydrogen permeation process, caused by its corrosive action, which is also known as sweet corrosion. When CO₂ is dissolved in an aqueous solution, it can form carbonic acid (H₂CO₃). In analogy to H₂S, during the corrosion process, atomic hydrogen can be generated at the steel surface, as it involves several cathodic reactions consisting of depolarisation processes involving both H₂CO₃, HCO₃ ⁻ and H⁺ (S. Wang et al. 2017b). However, hydrogen uptake is promoted to a lesser extent compared to H₂S because of two reasons. The first one is the fact that in a CO₂ environment, hydrogen generation through dissociation reactions includes one more step, i.e. the dissolution of the CO₂ in the aqueous medium to form H₂CO₃, while H₂S can dissociate directly. The second one is related to the extra role of H₂S, HS⁻ and S²⁻ acting as hydrogen recombination poisons (Karimi et al. 2023b). Moreover, upon CO₂ corrosion of typical mild steel with high carbon content and a ferritic-pearlitic structure, corrosion products such as iron carbide or cementite (Fe₃C) and iron carbonate (FeCO₃) are able to form (Kahyarian et al. 2017). The corrosion process usually initiates with selective ferritic dissolution, giving rise to a thick porous Fe₃C network. This is attributed to Fe₃C being conductive and, therefore, it acts as a cathodic site compared to ferrite, resulting in galvanic coupling. When its surface area increases, the dissolution of the ferrite can accelerate and, therefore, the corrosion process is accelerated. However, FeCO₃ can precipitate in between the Fe₃C particles and, therefore, these will act as a kind of anchor points for the former one, which can form a protective layer against corrosion and can affect hydrogen uptake, possibly blocking it (Karimi et al. 2023a). This precipitation is affected by the morphology, size, distribution and fraction of Fe₃C. Moreover, these parameters also affect the hydrogen uptake of Fe₃C, because they affect its hydrogen diffusivity and solubility, as reported by several researchers (Karimi et al. 2023b; Ogawa et al. 2022; Ramunni et al. 2006; Skilbred et al. 2022). In the presence of SRB, FeCO₃ was reported by Alabbas et al. (2013).

Da Silva et al. (2019) studied the hydrogen permeation of both pre-corroded Fe₃C-rich API 5 L X65 steel and pre-corroded FeCO₃-rich API 5 L X65 steel, formed within CO₂-saturated 3.5 wt% NaCl solution with different temperature, pH and immersion time for both. The former had five as much hydrogen, while the latter had twice as much hydrogen, compared to a wet-ground sample without a corrosion layer. On the one hand, the promoting effect of Fe₃C was attributed to its promotion role in the hydrogen reduction reaction, its porosity and negligible hydrogen trapping. On the other hand, the promoting effect of FeCO₃ was attributed to its non-barrier role due to the presence of voids along its surface, being beneficial sites for hydrogen saturation and permeation, despite its ability to weaken cathodic reactions. In practice, the combined influence of both H₂S/iron sulphide product and CO₂/iron carbonate product, which in general act competitively in the hydrogen permeation process, is studied by several researchers, mostly within the context of oil and gas industry where both corrosive species are present (Bai et al. 2016; Karimi et al. 2023b; Plennevaux et al. 2013; Skilbred et al. 2022; S. Wang et al. 2017b; Zhou et al. 2020). For example, Zhou et al. (2020) performed hydrogen permeation experiments for X80 pipeline steel in both CO₂, H₂S and CO₂/H₂S containing solutions and stated that the permeation was influenced by the mutual coupling of both the extent of hydrogen promotion of CO₂ and H₂S, the promotion or hindrance by corrosion products, i.e. being porous or dense, and hydrogen evolution of the corrosion reaction. However, the presence of both iron sulphide and iron carbonate products is also relevant for SRB in an environment where CO₂ and HCO₃ ⁻ are present in solution, among others within a marine environment (Enning et al. 2012) or oil field (Liu et al. 2018). Competitive adsorption between sulphur compounds (e.g. HS⁻ and S²⁻) and bicarbonates (HCO₃ ⁻) or carbonates (CO₃ ²⁻) is generally stated (Plennevaux et al. 2013).

Further, as is clear from the SRB corrosion mechanism of the cathodic depolarisation theory (CDT) (Section 3.2.1), Fe(OH)₂ corrosion product is also able to form besides iron sulphide corrosion product. It is generally stated that iron-oxygen (Fe–O) species such as FeOH_ads, i.e. an oxy-hydride species, and FeO or Fe₃O₄ products have a promoting role in hydrogen entry. This can be attributed both to the acceleration of the hydrogen evolution reaction (HER) or the promotion of hydrogen adsorption, i.e. the Volmer reaction. The former can be linked in analogy to the electrolytic catalyst activity of transition metal oxides, e.g. IrO₂ and RuO₂, towards the HER, while the latter can be linked to a blocking role on the recombination reaction, thereby acting as a recombination poison. More detailed information on the related mechanisms can be found in (Flis-Kabulska et al. 2007; Heyden et al. 2005; Kodintsev and Trasatti 1994).

Furthermore, phosphine (PH₃) is stated to be a possible metabolic product, based on the corrosion mechanism of volatile phosphorus compound (Iverson 2001), which is also mentioned in Section 3.2. This phosphine has been reported to enhance the hydrogen permeation rate (Lillard et al. 2000). Furthermore, upon reaction with iron, iron phosphide corrosion product can form, such as Fe₃P (Schreibersite), which is known to have a catalytic effect on HER (Wu et al. 2017; Y. Xu et al. 2013b).

4.4 Synergism of sulphate-reducing bacteria and cathodic protection on hydrogen uptake

Since both CP and SRB separately affect the hydrogen uptake in steel, often negatively, a synergistic effect can be expected. A synergistic effect can be defined as (Sun et al. 2019): ‘an effect of different multiple factors acting together being stronger than the sum of either of these factors acting alone’. The improved hydrogen uptake by SRB under CP can be seen as a two-step process. As a first step, adsorbed atomic hydrogen is created on the steel surface under CP according to the Volmer reaction, being more intensive to overprotection conditions. As a second step, sulphide metabolites secreted by SRB inhibit the recombination of atomic hydrogen and, therefore, increase the hydrogen surface coverage and stimulate the hydrogen absorption within the steel (Robinson et al. 1998).

Wu et al. (2015b) studied the influence of SRB on the near-neutral SCC susceptibility and permeable hydrogen concentration of X80 pipeline steel under different potentials during CP. SSRT was performed for both a sterile soil solution and SRB-inoculated soil solution under the following potentials: OCP (−725 mV_SCE), −976 mV_SCE and −1,176 mV_SCE. Moreover, the SRB-inoculated soil solution was tested under different pre-incubation times. These authors found that after a pre-incubation time of 3 days, i.e. when the SRB were in their cellular exponential growth phase, the potential of −976 mV_SCE and the SRB had a synergistic effect on the SCC occurrence. However, the synergistic effect decreased when the potential was −1,176 mV_SCE. It was suggested that both AD, enhanced by the formed S²⁻, and HE were the responsible mechanisms for the SCC. HE was supposed to be caused and enhanced on the one hand by the poisoning effect of secreted sulphide metabolites (S²⁻, HS⁻ and H₂S) on the chemical Tafel reaction (Equation (24)) and electrochemical Heyrovsky reaction (Equation (26)), while they promoted the Volmer reaction (Equation (23)). On the other hand, a more negative potential generated more hydrogen atoms at the steel surface, which promoted also the Volmer reaction and the hydrogen entering the steel and, therefore, enhanced HE. The Volmer reaction was also promoted by a pH decrease, caused by the SRB, which consequently increased the amount of protons on the steel surface. Furthermore, SCC could be enhanced by the SRB due to localised pitting, resulting in stress concentrations near the bottom of the pit, which could be attributed to DET and sulphur compounds. The formation of highly electroconductive and porous FeS acting as an enhancer of electron transfer could be a reason for this DET (Sun et al. 2018; Xie et al. 2021). Moreover, pitting could also be caused by the heterogeneity of the biofilm and the EPS, resulting in the formation of concentration cells as already mentioned (Xie et al. 2021). The decreased synergistic effect at −1,176 mV_SCE could be attributed to alkalisation at this potential, which prevented the growth of the SRB. Fracture surface analysis indicated that the SRB-inoculated solutions caused more quasi-cleavage fracture at a certain potential compared to sterile solutions, which had a more ductile dimpled fracture. Moreover, the ratio of quasi-cleavage over dimpled fracture area increased at −976 mV_SCE compared to OCP, while at −1,176 mV_SCE, pure brittle rupture or cleavage fracture was observed. A direct relationship between plasticity loss and permeable hydrogen concentration was suggested.

Li et al. (2022) investigated both the hydrogen permeation and SCC behaviour of X100 pipeline steel in a simulated marine mud environment, being both sterile and inoculated with SRB. A synergistic effect was again observed for the applied CP and SRB on both SCC and hydrogen permeation and specifically on the diffusible hydrogen concentration, hydrogen diffusion flux and SCC sensitivity. However, there was no synergistic effect on the hydrogen diffusion coefficient. The effect weakened again when the potential was too negative, due to alkalisation and subsequent reduced bioactivity of the SRB. Similar reasons for the synergistic effect were quoted as in the work of Wu et al. (2015b). Moreover, it was stated in their research that SRB accelerated cathodic polarisation, providing more corrosion pits and, therefore, led to an improved SCC susceptibility.

The combined effect of CP and SRB was also investigated by other studies, where anodic dissolution (AD) or pitting and HE were the dominating and competing mechanisms (Sun et al. 2018; Xie et al. 2021). In general, corrosion pits are the typical sites where the nucleation of microcracks and SCC starts because of stress concentrations or localised acidification (Sun et al. 2018; Xie et al. 2021). The crack tip can serve as an anode, while the crack wall can serve as a cathode with a larger area, which results in a kind of galvanic corrosion cell between the crack tip and the crack wall, which promotes the crack propagation (Sun et al. 2018).

Further, the synergistic effect of CP and S₂O₃ ²⁻, simulating SRB as a H₂S producer, was investigated by Tian et al. (2019). A competition was concluded between the hydrogen promotion ability of sulphur species, i.e. H₂S and HS⁻, and an increase in the interfacial pH. The increase in S₂O₃ ²⁻ concentration, on the one hand, facilitated the hydrogen permeation in two ways: an increased production of H⁺ ions by increased H₂S and/or HS⁻ at the metal–solution interface and the poisoning effect of these sulphide products on the hydrogen recombination step, leaving more hydrogen within the atomic state. The increase in the interfacial pH at more negative potentials, on the other hand, led to a decrease in the hydrogen permeation due to the competitive adsorption of brucite, which had a blocking effect on the hydrogen entry. Moreover, the local alkalisation weakened the availability of H⁺, causing a decrease in H₂S proportion and a decrease in the possible adsorption sites of hydrogen due to competitive OH⁻ adsorption. The maximum hydrogen flux and the highest synergistic effect were observed for a S₂O₃ ²⁻ concentration of 0.01 M and a potential of −850 mV_SCE.

The influence of both the presence of S²⁻ and CP on the HE of 2,205 duplex stainless steel in deaerated acidic artificial seawater (AASW) with a pH of 6.5 was examined by Zucchi et al. (2006) by the use of SSRT. Different concentrations of S²⁻, i.e. 1, 10 and 30 ppm, and different potentials, i.e. −0.9 V_SCE, −1.0 V_SCE and −1.2 V_SCE, were used. When no S²⁻ was present, the HE susceptibility increased at −0.9 V_SCE and −1.0 V_SCE compared to OCP within both AASW and air. However, at −1.2 V_SCE, a calcareous deposit, consisting mainly of brucite, was formed, which could reduce the HE. Alternatively, when S²⁻ was present, the HE susceptibility increased and the effect was larger at higher S²⁻ concentrations and more negative potentials. At −1.2 V_SCE, the promoting action of S²⁻ on HE surpassed the blocking effect of the brucite, which resulted in the highest HE susceptibility.

Wu et al. (2015a) examined the simultaneous effect of present superficial stress, introduced through mechanical abrading, the presence of SRB and CP on the hydrogen permeation behaviour of X80 steel within a simulated soil solution. A synergistic effect between superficial stress and the presence of SRB on the apparent hydrogen concentration and hydrogen permeation flux was concluded at a CP potential of −1,176 mV_SCE. This could be attributed to more aggregation of SRB on an abraded surface, because of its higher surface roughness and free energy, and coupled sulphide metabolite production improving the hydrogen uptake. Moreover, larger surface area, surface roughness and free energy were assumed to increase the activity of iron sulphide product, which increased the active surface sites for hydrogen reduction and thus absorption, due to its catalytic action. In general, it is known that the presence of stress influences the hydrogen permeation behaviour, as is clear from several studies (Van den Eeckhout et al. 2020; Zhao et al. 2016; Z. Xu et al. 2022b). This is very relevant, since, e.g. submarine pipelines are exposed to both present complex load, e.g. within sea mud, and the presence of anaerobic SRB. The combined influence of tensile stress, within both the elastic and plastic regions of the examined steel, and the presence of SRB and/or iron sulphide layer on the hydrogen permeation behaviour has been examined (Ge et al. 2020; Kim et al. 2012; Sun et al. 2019). Further details on the exact mechanism governing the effects of both stress and SRB are considered to be outside of scope.