Part 3: Corrosion and prevention in post-combustion CO₂ capture systems

Abstract

This article is part 3 of the review series on ‘Recent progress and new development of post-combustion carbon capture technology using reactive solvents’. This review focuses on alkanolamine absorption during post-combustion CO₂ capture from coal-fired flue gas, looking at a range of absorbents, including those that are commonly used, as well as blended and new solvents. The effects on corrosion of blended and new absorbents and process parameters (e.g., amine concentration, CO₂ loading, oxygen concentration, SO₂ concentration and temperature) are reviewed. Also reviewed is the effect of corrosion on the formation of heat-stable salts in the absorbent stream, as well as the presence of impurities in flue gas. Finally, corrosion mechanisms are discussed and corrosion inhibition approaches are reviewed.

Figure 1.  Simplified process flow diagram of a typical amine treating process.

Fossil fuel power plants are a major source of CO₂ emissions, which gives rise to an enhanced GHG effect. To control emissions and reduce this effect, a CO₂ capture technique may be integrated into the power plant. In the case of low pressure coal-derived flue gas streams, the application of a chemical-absorption process using an aqueous alkanolamine solution has been found to be the most effective of such a CO₂ capture approach. However, corrosion is a major drawback, which prevents the amine process from achieving its highest possible efficiency. Corrosion is an electrochemical process involving the transfer of electrons within the construction material, resulting in its deterioration. The electrochemical process is a redox reaction, consisting of two chemical reactions known as oxidation and reduction [1,2]. When a metal is immersed in a given solution, redox occurs at the surface of the metal, causing the metal to corrode. For example, corrosion of zinc in an acid environment proceeds according to the following reactions:

Overall reaction:

Equation 1

Oxidation reaction:

Equation 2

Reduction reaction:

Equation 3

At equilibrium, the electrode potential (E) is referred to as reversible electrode potential (E_rev). The corrosion electrode potential (E_CORR) is equivalent to E_rev and used specifically for corrosion reactions. When a potential is applied to an equilibrium system, it causes the potential to shift and deviate from E_CORR to E. The potential deviation is known as polarization (η), which can be either positive or negative. When η is positive, the metal surface is driven toward the anodic side of the reactions and loses its electrons, which is termed as anodic polarization, while cathodic polarization occurs when η is negative.

When a number of half-reactions occur simultaneously on the metal surface, ‘mixed potential theory’ must be utilized to create polarization curves. The theory states that, at equilibrium, the net oxidation rate must be equal to the net reduction rate. That is, the sum of anodic oxidation currents must equal the sum of cathodic reduction currents and the net measurable current is zero.

Equation 4

Electrochemically, corrosion rate measurement is based on the determination of the oxidation current (also called corrosion current, i_CORR) at the corrosion potential. The corrosion current is calculated as follows:

Equation 5

Equation 6

Application of mixed potential theory allows the determination of the corrosion rate using a method known as Tafel plot.

Since the corrosion current is related directly to the corrosion rate, Tafel plot and potentiodynamic techniques are performed to experimentally determine i_CORR, from which the corrosion rate is calculated.

This paper reviews corrosion studies in a CO₂ capture plant. Effects of various factors including type of amine and its concentration, CO₂ loading, process parameters (e.g., temperature) and flue gas contaminants (e.g., O₂ and SO₂) on corrosion are presented. Corrosion prevention strategies using corrosion inhibitors are also reviewed.

Background of amine absorption

Alkanolamines as absorbents for acid-gas removal in gas-treating plants have been used in gas purification processes for more than 60 years. Commercial interest in gas purification has focused primarily on two amines, monoethanolamine (MEA) and diethanolamine (DEA), but can include other alkanolamines such as triethanolamine, diisopropanolamine, methyldiethanolamine (MDEA) and diglycolamine. Triethanolamine was first used in gas-treating plants, but was replaced by MEA, due to the latter’s low reactivity and relatively low stability. MDEA has been proven to be a selective solvent for removal of H₂S in the presence of CO₂[3]. MDEA is less basic and has greater capacity to react with acid gas due to the fact that it is used in higher concentrations than MEA. MEA is the most widely used solvent for CO₂ absorption, due to the fact that it has the highest alkalinity and highest volume of acid- gas removal at the fastest rate. The alkalinity of amines increases in the order of MDEA < DEA < MEA [4]. On the other hand, MEA also has some drawbacks based on its own physical characteristics, such as the high energy consumption required for CO₂ regeneration and high vapor pressure compared with other alkanolamines, which leads to considerable vaporization and solvent loss. In addition, amines are degraded upon contact with oxygen, following which the degradation products become corrosive agents and consequently cause equipment corrosion in the absorption plant [5,6]. The corrosion rates of amines increase in the order of MEA > DEA > MDEA [4].

Rooney and DuPart reviewed the role of oxygen in oxidative degradation of amines and corrosivity of these degradation products [7]. To reduce the corrosion rate, the solvent strength is kept at 20–30% amine by weight in water, resulting in relatively large equipment sizes and solvent regeneration costs. Also, there have been attempts to use blended solvents, such as pairing the slower rate of MDEA with the faster rate of DEA [3] or mixing MDEA with piperazine (PZ), a rate-promoting agent [4,8,101]. The process might also be modified to incorporate inhibitors that reduce solvent degradation and equipment corrosion. For instance, Bhat et al. reported the use of inorganic inhibitors in an aqueous solution for oil and gas applications [102]. The corrosion inhibitions combine the effects of anodic inhibitors (ammonium heptamolybdate or sodium orthovanadate), cathodic inhibitors (cerium chloride) and metal complexing ligand (trisodium citrate-2-hydrate). Khusnutdinov et al. disclosed 2-propyl-3-ethyl-8-oxychinoline-ZnCl₂ complex as a steel corrosion inhibitor [103]. The complex inhibitor was applied in petroleum production in mineralized media with high O₂ content. Abdrakhmanov et al. proposed an organic inhibitor, N-acetyl-2-(2,3-dihydroxycilopentenyl) aniline in mineralized water petroleum solutions including H₂S [104].

Post-combustion CO₂ capture from coal-fired flue gas

CO₂ comprises the largest fraction of GHG emitted into the atmosphere. In order to comply with anticipated future environmental regulations, CO₂ in flue gas will need to be captured prior to release into the atmosphere. CO₂ capture using chemical or physical absorption is a likely technology for efficiently reducing large volumes of CO₂ at emission sources to comply with these anticipated regulations. An effective technique is chemical absorption using aqueous alkanolamine solutions [9]. Amine absorption technology for gas processing has been readily adapted for capturing CO₂ from coal-fired power plants and could be for other worldwide large CO₂ emission sources, such as steel manufactories, cement production, chemical industries and oil refineries (Table 1). However, power plants using fossil fuels generate the most CO₂ emissions [201].

Natural gas feed composition is typically composed of 75% light hydrocarbons (C₁–C₅), 13–15% CO₂ and no O₂[10], while flue gas composition from coal combustion also includes O₂, N₂ and trace contaminants such as SO₂ and NO_x(Table 2)[11,12]. Fly ash is also present in flue gas streams and typically consists of inorganic oxides of SiO₂, Al₂O₃, Fe₂O₃, CaO, MgO, Na₂O, K₂O and P₂O₅, which have undesirable effects on the amine solution and must be removed before entering into the treating unit. Oxygen in the flue gas originates from air and excess combustion in the boiler. The presence of O₂ introduces oxidative degradation of alkanolamine solvents and causes the formation of heat-stable salts (HSS) [13,14,101].

As amine technology for acid-gas absorption has been adopted for capturing CO₂ from flue gas, most of the process configurations from natural gas processing are still used in CO₂ capture. However, modification and optimization of the processes are still needed to reduce energy consumption and decrease solvent losses and corrosion problems. Nielsen and Hansen [15] reviewed corrosive agents, including acid gases, NH₃, O₂, HSS and amine degradation products; they specifically looked at the mechanisms of corrosion of carbon steel in refinery amine systems. These variable parameters in amine absorption must be reviewed and become well understood for CO₂ capture from flue gas.

Successful precommercial demonstration of CO₂ capture from a coal-fired power plant using aqueous MEA absorption occurred at Boundary Dam in Canada [9]. However, the existing problems, especially corrosion from using coal fuel, may be more complicated than seen in natural gas systems. These SO₂, NO_x, CO and particulate matter, if passing through the filter into the amine solution, could participate in the amine reactions and subsequently induce corrosion. Thus, a pretreatment of the flue gas feed is required.

A generalized process diagram of CO₂ absorption using aqueous alkanolamine (MEA) solutions is illustrated in Figure 1. Flue gas from coal firing is first filtered to remove particulate and scrubbed to remove SO_x and NO_x. The flue gas is fed into the bottom of the absorber column, flowing upward against the aqueous amine solution, where it is introduced to the top of absorber column. CO₂ in the flue gas stream rapidly reacts with MEA and remains in the solution. This stream is then called a rich-amine solution. The free CO₂ flue gas leaves the solution at the top section of the absorber. The temperature of the rich amine solution, while leaving the absorber, is approximately 60°C. The rich-amine solution exits at the bottom section of the absorber, is heated to a higher temperature in the rich-lean heat exchanger, and then returned to the top section of the stripper. In the stripper, the rich-amine solution is heated by the hot vapor steam from the reboiler at the bottom of the column resulting in gas mixtures of CO₂, water and amine. These gas mixtures exit at the top section of the stripper, flow to the condenser and reflux accumulator to recover, and return both water and amine back to the stripper. The aqueous amine solution containing a low concentration of CO₂ (called lean-amine solution) at the stripper bottom, now at a solution temperature of approximately 120°C, is pumped and cooled down through the rich-lean heat exchanger before being reintroduced to the absorber to complete the process cycle.

It is suggested that the stripper be operated at temperatures below 120°C to avoid corrosion [16,17]. Types of corrosion, which can be both uniform and localized, include pitting, galvanic, erosion, stress cracking and inter-granular corrosion [4]. Kittel et al. presents the major results of corrosion testing from two pilot plants (coal-fired power station in Esbjerg [Denmark] and natural gas burner at the International Test Centre for CO₂ Capture in Regina [Canada], under MEA operation [18]. In addition, the highest corrosion rates were always found in the hottest parts of the unit, including the inlet and outlet of the stripper. Gao et al. also reported that the most serious corrosion occurs at the rich-liquid side outlet of the heat exchanger [19].

Corrosion

Corrosion in CO₂ absorption processes is one of the most severe operational problems in gas purification plants [6]. Carbon steel is usually used as plant construction material for absorption and stripping units to reduce investment cost. However, carbon steel is more vulnerable to corrosion than stainless steel. Nevertheless, it is possible that corrosion in amine plants might not be inevitable and could be minimized or controlled. Before any corrosion prevention or protection is performed, it is important to understand the system being operated and the effects of impurities and process parameters on corrosion.

▪ Types of amines

Different amines for CO₂ capture yield different degrees of system corrosion. The order of absorption efficiency of amines is MEA > DEA > MDEA, but the degree of corrosiveness is also MEA > DEA > MDEA. Corrosiveness of amines were tested under conditions in ranges of 1–5 kmol/m³ amine concentration, 30–80°C, 0–0.4 CO₂ loading, and 0–10% feed gas O₂ concentration. For primary amines, the sterically hindered amine of 2-amino-2-methyl-1-propanol (AMP) has slightly less CO₂ solubility than MEA, but the corrosion rate of AMP is much higher than that of MEA under the same conditions [20]. AMP’s greater corrosiveness is explained in terms of the preferable formation of HCO₃^- rather than carbamate, while MEA and DEA favor more stable carbamate formation in the solution. HCO₃^- is known to be one of corrosive agents responsible for corrosion in an amine system.

For blended amines, including MDEA–DEA, MEA–MDEA, and MEA–AMP, the corrosion rates of mixed amine systems at equal amine mixing ratio and CO₂ saturation appear to be a combination of the rates seen in single amine systems. The corrosion behavior of the mixed amine systems for MEA–MDEA and DEA–MDEA also exhibit cathodic and anodic current density values in between those produced in the individual single amines constituting the mixed systems. For MEA–AMP, the corrosion behavior shows only cathodic current. It has been explained that this was probably caused by differences in the amounts of CO₂ absorbed under the condition of CO₂ saturation [20]. MDEA in a blended system of MEA–MDEA–H₂O–CO₂ and, in the presence of O₂, is preferentially thermally degraded at high temperature and drastically reduces loss of MEA due to degradation. It also decreases the amount of non-environmentally benign degradation products. This allows MEA to be more available for CO₂ absorption [21]. Eustaquio-Rincón et al. conducted corrosion experiments using conditions of MDEA, DEA and their various ratio mixtures in a concentration range of 15–60% mass fraction, with and without acid gases, different amounts of H₂S and CO₂ at a pressure range of 276–5861 kPa and 120°C [22]. The corrosion rate of carbon steel reported after averaged test time of 350 h was determined by the weight loss method in aqueous solutions of individual and blended MDEA and DEA. For a given blended mass ratio, the corrosion rate decreases as the total amine concentration increases in relation to increased amounts of MDEA. The overall corrosion rate was less than 1 mils per year unit (mpy). In the presence of CO₂ loading less than 0.35 mol/mol amine and a mass ratio of MDEA to DEA of 3.5/1, the effect of CO₂ on corrosion was negligible.

New types of amines have been introduced for CO₂ capture in the absorption process, including hetercyclic PZ and its derivatives [23–26]; analog structures of PZ [27]; amino alcohols such as 4-diethylamino-2-butanol [28,29], 2-(isopropylamino)ethanol, 2-(isobutylamino)ethanol, 1-methyl-2-piperidineethanol, and 2-(isopropyl)diethanolamine [30,31]; and sterically hindered amines including AMP, with the intention of increasing absorption and cyclic capacities, which results in a lower circulation rate. This creates energy savings in the CO₂ regeneration step as compared with conventional amines [32].

The structure of PZ is a heterocyclic six-member ring containing two secondary amines in the structure. PZ has been found to have similar volatility to MEA but has almost double the CO₂ absorption rate and capacity, and is more resistant to oxidative and thermal degradation [25]. A CO₂ capture pilot plant using concentrated PZ has also been constructed [33]. The kinetic reaction of PZ and its derivatives with CO₂ identified using the stopped-flow technique are in the order of PZ > 2-methyl-PZ > 1-ethyl-PZ > N-(2-hydroxyethyl)-PZ > 1-methyl-PZ [24]. The degradation rates of 1-methyl-PZ and 2-methyl-PZ are faster than PZ under the same conditions at 150°C [27]. For amino alcohols, modification of different chemical structures, including changing an alkyl group relative to the position of an amino group in an alcohol structure, results in a high absorption rate and low heat of reaction when compared with AMP and MDEA [30]. When hydroxyl functions are placed relative to the amino position, some amino alcohols, such as 4-propylamino-2-butanol, show the highest CO₂ absorption, and 4-(ethyl-methyl-amino)-2-butanol shows the highest cyclic capacities among the amino alcohols studied in comparison to MEA [28]. These new solvents have a very high potential for application in CO₂ absorption processes. However, the physical characteristics, oxidative and thermal degradation, as well as corrosion behavior of these new solvents, must first be studied extensively. Thus, it will take some years for these solvents to be commercialized. Blending the new solvents with existing commercially available amines to improve their CO₂ absorption efficiency and reduce operating costs have been reported. However, studies of corrosion on this set of solvents have not been reported.

Several attempts have been made to formulate blended amines in order to overcome the limitations of single amine absorption [19]. The objectives are to increase CO₂ capture efficiency, increase CO₂ solubility in amine solution [34] and decrease energy requirement in the stripping process [35,36]. MEA–AMP can absorb CO₂ with a similar capacity to MEA–MDEA, but at a much higher rate and with higher mass-transfer coefficients [37]. Blended K₂CO₃-hindered cyclic amine [25], a K₂CO₃-promoting amine [38], have been claimed to have very low heat absorption, low volatility, and high resistance to thermal and reactive degradation. PZ has been found to rapidly react with CO₂ and forms carbamates, which could be used as an effective promoter in existing amine absorption. Extensive study of PZ blends includes MEA–PZ [39,40], MDEA–PZ [41,42], and K₂CO₃–PZ [23,43]. Blended PZ with MEA for CO₂ showed increased absorption 1.5–2.5 times greater than that of MEA alone. The degree of absorption capacity increases in order of MEA–PZ > AMP–PZ or MDEA–PZ and MEA alone [34]. However, PZ not only increases the CO₂ absorption capacity, but also increases the corrosion rate. The corrosion rate of carbon steel in the blended MEA–PZ solutions is more corrosive than MEA at the same total concentration and under the same operating conditions. The corrosion rate in the blended MEA–PZ system is more pronounced with increasing total amine concentration, and it also rises with an increase in the ratio of PZ to MEA. Corrosion rates of MEA–PZ system determined under conditions of 5–8.7 kmol/m³ total amine concentration with various mixing ratio, 0.2–0.63 CO₂ loading, 0–10.13 kPa feed gas O₂ partial pressure, and 80°C (test time was not given) was reported to increase in the presence of O₂, CO₂ loading and high temperature [39]. Table 3 also summarizes various corrosion studies conducted using different amines and test conditions.

▪ Heat stable salts

The main problem of solvent loss is due to oxidative and thermal degradation experienced in the absorption process. Amine solvents can be degraded. Degradation products such as carboxylic acids can react with amine to form alkanolamine salts or HSS, which cannot be removed under solvent regeneration conditions and remain in the amine absorption solution throughout the plant. Various kinds of HSS and their negative effects are often reported [21,44–47]. Tanthapanichakoon et al. investigated the effects of HSS (oxalate, formate, malonate, glycolate, succinate and acetate) of various concentration on corrosion of carbon steel and stainless steel using 5 kmol/m³ MEA, 0.20 mol/mol CO₂ and 80°C (test time was not provided) [48]. Degrees of corrosiveness varied depending on type and concentration of salt. For carbon steel under the same conditions when different HSS were added, oxalate was the most corrosive (corrosion rate increase 54–64% as compared with the solution without added HSS), followed by malonate (14.7–17.2%), formate (12.2%) and the others (<4%). For stainless steel, the most corrosive oxalate did not reduce the corrosion resistance of stainless steel or have any impact on the corrosion behavior. The corrosion rate of carbon steel in MEA–PZ blends at 80°C, with CO₂ loading and in the presence of 1% wt of HSS (acetate, formate, oxalate and thiosulfate), is faster and shows greater degrees of deterioration [39]. Anions of HSS were found to be an oxidizing agent in the corrosion process, which altered the corrosion mechanisms on both the anodic and cathodic sides. The order of corrosiveness was formate, followed by acetate, oxalate and thiosulfate; however, in the presence of O₂, acetate was more corrosive than formate.

▪ Process parameters

Most corrosion behavior in all amine systems is sensitive to the variations in type of amines, amine concentration, solution temperature, CO₂ loading, and amounts of oxygen, HSS and impurities. The effects of the process parameters are typically studied in any new equipment or system process or in testing any new solvent.

Effect of O₂

Coal-fired power plant flue gas is composed of CO₂, N₂, O₂, SO₂ and NO₂. Considerable efforts have focused on understanding O₂-induced degradation of MEA. Dissolved oxygen in liquid solvent experiences high temperature at various locations in the capture plant, including the heat exchanger, stripper and reboiler/reclaimer, and this leads to a higher corrosion rate, especially in these areas [47]. Bello and Idem proposed pathways for the formation of O₂ degradation products in MEA solution during CO₂ absorption [44]. O₂ was not initially present in the feed gas stream, but could be produced as a degradation product. Thus, an oxidative degradation environment could be created without the presence of O₂ in the flue gas feed. In the absence of O₂, corrosion reactions occur due to iron dissolution (anodic reaction) and reduction of oxidizers (H₂O and HCO₃^- in cathodic reactions), and black slime of FeCO₃ was also observed during the experiment [48]. Increasing the O₂ partial pressure increased O₂ solubility in the solution. In an MEA system with CO₂ loading, the corrosion rate increased with increasing O₂ concentration under the test conditions of 5–9 kmol/m³ MEA concentration, 0.2–0.55 CO₂ loading, 40–80°C, 0–10.13 kPa feed gas O₂ partial pressure and 0–2000 rpm solution velocity [49]. Similar effect of O₂ concentration was also reported using 1–7 kmol/m³ MEA, 0–100% feed gas O₂, 0–204 ppm SO₂, 0–0.5 CO₂ and 30–80°C [50,51]. Quantitative information (i.e., corrosion rate) usually dependent on operating parameters (e.g., amine type, temperature and feed gas concentrations) was also given for MEA–H₂O-CO₂-O₂-SO₂ system to represent the effects of specific variables (i.e., MEA, O₂, CO₂ and SO₂ concentrations and temperature) on corrosion rate of carbon steel [50]. The rate equation is also given as follows:

Equation 7

Chemical analysis of the amine solution by inductively coupled plasma–MS and capillary electrophoresis found an increase of dissolved iron but no change in carbonate/bicarbonate ions in the solution, and analysis of surface of corroding of specimen using a scanning electron microscope–energy dispersive spectrometer showed a relative decrease in Fe and increase in O and C on the specimen surface. The corrosion products proposed were Fe(OH)₂, Fe(OH)₃ and FeCO₃[50,51].

Effect of amine concentration

Most amines show similar behavior with respect to corrosion rate, in that it tends to increase with increasing amine concentration. The corrosion behavior shows a greater impact on current densities on the anodic side. Higher amine concentrations result in larger amounts of absorbed CO₂, carbonate/bicarbonate formation and dissolved iron being present in the solution. The increase in carbonate/bicarbonate anion concentration is responsible for the increased corrosiveness of the system [50,51]. It is noted for AMP and amine concentration above 3 kmol/m³, the corrosion rate of AMP gradually deceases as AMP concentration increases. This was explained as being due to the decrease of the hydrolysis of carbamate, resulting in less HCO₃^- being produced. There is a limitation of amine concentration of 20–30% amine to minimize the corrosion rate, and there have been research attempts to overcome this limitation. For instance, it was reported that there was no MEA degradation when 40% MEA was used with oxidative inhibitors [52].

Effect of CO₂

High CO₂ absorption capacity of amines is one of the most important parameters in decreasing operating costs of CO₂ capture. Increasing the CO₂ loading in most systems, however, leads to a greater amount of HCO₃^- and, consequently, a higher corrosion rate. Higher CO₂ loadings result in both higher anodic and cathodic current densities in electrochemical reactions and, thus, higher amounts of HCO₃^-. The solution pH also becomes more acidic, suggesting increased amounts of H⁺ or RNH₃⁺ might occur in the solution. Both HCO₃^- and H⁺ accelerate the corrosion rate.

Effect of temperature

Temperature is an essential term in the study of kinetic rate of reaction. It shifts many reaction equilibria, including CO₂ absorption, carbamate formation, hydrolysis of carbamate, pH of the solution and solubility of chemical species. Raising the solution temperature, therefore, accelerates corrosion reactions [49,50]. A corrosion study conducted at 30–80°C, 0.0–1.0 mol CO₂ loading and 1.0–4.0 kmol/m³ AMP concentration showed that for AMP systems operated at a high temperature and lean CO₂ loading, the corrosion rate was less than in MEA systems [53]. At elevated temperatures, CO₂ solubility of AMP was usually lower than that of MEA, thus a smaller CO₂ loading. A decrease in CO₂ loading in AMP then produced a smaller concentration of corroding species, particularly bicarbonate (HCO₃^-), protonated amine (RNH₃⁺), and proton (H⁺), thus corrosion rates were reduced. On the other hand, an increase of CO₂ loading of AMP solution increased the corrosion rate of the carbon-steel material. Based on the polarization curves at different CO₂ loadings, the curves exhibited the most change in current density on the cathodic side as compared with a slight change on the anode. This work demonstrated more activity in the cathodic region where reduction normally occurs. This observation led to the conclusion that CO₂ loading enhanced the rate of reduction, whereby it provided more oxidizing species (HCO₃^-, CO₃^2- and H⁺). The corrosion inhibitors used for CO₂ absorption in various amine solutions are summarized in Table 4. To our knowledge, no data are available in terms of comparison of temperature effect of different contaminants. Typically, any study is usually conducted to evaluate the effect of parameters of interest in a step-wise manner. It usually compares the effect of temperature on corrosion rates induced by different concentrations of the different species at the same time. An example can be found in a study, which showed the effect of different temperatures on various CO₂ loadings [53]. A similar study also applies to effect of O₂ on corrosion rates. In the case of SO₂, in which very limited studies are available, its effect was only evaluated at 80°C [50]. However, this study proposed a corrosion rate equation, previously given in Equation 7, which could indicate the effect of temperature on each corrosion species (i.e., MEA, SO₂, O₂ and CO₂).

Effect of SO₂

SO₂ in flue gas streams arises from coal combustion. Concentration of SO₂ is dependent on the quantity of sulfur contained in the original coal and the conditions of the combustion process. High SO₂ concentration (>10 ppm) has a marked depolarizing effect on cathodic reactions on polished metal surfaces and also creates low pH in the surface film. Consequently, it generates a low rate of sulfate production and appreciable amounts of tetravalent sulfur (dithionite and S₂O₄^-), which can be reduced to sulfide. Ferrous sulfide is a stable corrosion product.

Although it is present in small amounts, SO₂ is more acidic than CO₂; its regenerative removal requires the use of weaker bases than primary amines [54]. The strong basicity of primary amines results in an irreversible reaction with SO₂ that produces corrosive and HSS, reducing the CO₂ absorption rate and capacity of the absorbent. However, it has been reported that the absorption of CO₂ in industrial gas streams that contain a typical concentration of SO₂ might lead to a loss of only 2 mol of MEA/(mol of SO₂) [55]. Thus, very low concentrations of SO₂ (<10 ppm) might be desirable in order to avoid excessive loss of costly solvent [56]. So far, the information on the effects of SO₂ and NO_x on amine degradation and corrosion is scant. Uyanga and Idem have made a comprehensive study of the effects of SO₂ and O₂ on the degradation of MEA [57]. The effect of SO₂ as well as other process parameters (CO₂ loading, O₂ concentration and temperature) on degradation of MEA was also reported in [13]. The SO₂-O₂-N₂ gas mixture was in contact with aqueous MEA solutions at elevated temperatures. In the absence of CO₂, the rate of MEA degradation increased with raising O₂ concentration and further increased when SO₂ was added, continuing to climb as the SO₂ concentration rose. In the presence of CO₂, the effects of O₂ and SO₂ became negligible. Later, Kladkaew et al. investigated the effect of SO₂ (0–204 ppm) on corrosion of carbon steel in a similar system and simulated the conditions of absorption–regeneration sections [50,51]. The corrosion rate increased with increasing SO₂ due to the increase of hydronium ion formation from the reactions of SO₂ and H₂O, as well as SO₂, O₂ and H₂O. However, the corrosion rate was slight as compared with the effects of MEA concentration, CO₂ loading and operating temperature.

Gao et al. evaluated the influence of SO₂ on the corrosion of carbon steel and stainless steel, and degradation of blended amine solvent in a pilot plant [19]. More serious amine degradation and HSS formation occurred with increasing SO₂ concentration. In addition, SO₂ has been found to play an important role in formation of corrosion protection films.

Corrosion mechanism

Corrosion in amine-treating units is normally quantified using a corrosion rate measured by two major techniques: weight loss and electrochemical measurements. The weight loss technique uses a metal coupon of known weight immersed into the amine solution at predetermined conditions. The weight after a specific time is determined and used to calculate the corrosion rate during the test period. The electrochemical technique measures current densities of the anode and cathode, dictated by the strengths of oxidation (metal dissolution) and reduction (reduction of corroding species) when corrosion occurs. In any technique, the corrosion rate is normally expressed in mpy. In electrochemical measurement, electrochemical polarization consists of an anodic reaction (oxidation of metal or iron dissolution, Equation 8) and cathodic reactions (reduction of oxidizing agents, Equations 9 & 3). Although redox potentials are important, all of the works reviewed did not include the potentials for the reactions; all of the studies proposed a specific system of various amines and corrosion inhibitors. Objectives set by these works were mostly focused on showing corrosiveness, effects of the studied parameters, and how corrosion occurred in the systems. Redox potentials were not used to explain any reported results, thus they were not available from their studies.

Anodic reaction:

Equation 8

Cathodic reaction:

Equation 9

In an aqueous amine-CO₂ system, the reaction of amine (primary and secondary) with CO₂ yields reversible formations of carbamate and protonated amine as follows:

Formation of carbamate [58]:

Equation 10

In the aqueous amine solution, water as a bulk phase plays an important role in hydrolysis of carbamate.

Hydrolysis of carbamate [59]:

Equation 11

Hydrolysis of bicarbonate:

Equation 12

Under conditions in ranges of 1–5 kmol/m³ amine concentration, 30–80°C, 0–0.4 CO₂ loading, and 0–10% feed gas O₂ concentration, 3 kmol/m³ amine and 80°C, the corrosion rates of different amines are in order MEA > AMP > DEA > MDEA and for the blended amines (1:1) are MEA–AMP > MEA–MDEA > DEA–MDEA, which the corrosion rates are in between of the amine precursors. The polarization curves of the amines also show higher anodic and cathodic current densities in the same order. The different corrosion rates of the amines are due to the different CO₂ absorption capacity, 0.565, 0.554, 0.442 and 0.243 mol/molCO₂, for MEA, AMP, DEA and MDEA, respectively. For low CO₂ loading, ex 0.2 mol/mol CO₂ loading, 3 kmol/m³ amine and 80°C, the corrosion rates are AMP >> MEA ≥ DEA, corresponding to 30, 20 and 19 mpy, respectively. Higher corrosion of AMP is explained due to the presence of HCO₃^-, corrosive species. Carbamate compound formation of MEA and DEA is stable, while that of AMP is unstable and thus undergone hydrolysis to generate HCO₃^-[20].

For MEA–PZ system evaluated at 5–8.7 kmol/m³ total amine concentration with various mixing ratio, 0.2–0.63 CO₂ loading, 0–10.13 kPa feed gas O₂ partial pressure and 80°C (test time was not given), specifically at 0.2 CO₂ loading, 80°C, constant total amine 6.2 kmol/m³ (MEA 5 kmol/m³–PZ 1.2 kmol/m³ vs MEA 6.2 kmol/m³), the mixed system (21.79 mpy) is more corrosive than MEA (19.23 mpy). The corrosiveness of PZ is more pronounced when PZ is increased and the evidence is also seen in the increase of the cathodic current densities. Possibly, there was a change in reduction of oxidizing agents in the presence of PZ; for example, metal complex formation [39].

Corrosiveness of HSS can be viewed by polarization. There are also changes in anodic (βa) and cathodic (βc) Tafel slopes. Linear polarization resistance is decreased as HSS concentration increased. For oxalic acid, both anodic and cathodic current densities in the presence of oxalic acid are greater than that of the same system with no acid, implying the change in iron dissolution and chelation of iron. With increasing oxalic acid concentration, βa is decreased, while βc remains unchanged as compared with no acid system, and linear polarization resistance is also decreased. The polarization parameters are more pronounced as the acid concentration greater than 1000 ppm for oxalic and 10,000 ppm for formic and malonic acids [48].

Reviews of mechanisms are mostly relied on analysis of electrochemical polarization and its data. Postulation is based on electrochemistry and changes in polarization plot (potential related to standard electrode versus logarithm of current density) due to what are added and cause change in the environment. There is no confirmation of what definite chemical species are taking part in corrosion or inhibition. Most recent studies report, ¹³C-NMR was used to determine AMP–protonated AMP (AMP/AMPH⁺), AMP carbamate (AMPCOO^-) and HCO₃^-/CO₃^2- in 30 wt% aqueous solution of AMP with different amount of CO₂ at 25°C [60]. Yamada et al. also used ¹³C-NMR to demonstrate that the product ratios of carbamate to bicarbonate of MEA are greater than that of AMP [61].

In the amine–CO₂ aqueous system, CO₂ is bound in the form of carbamate, bicarbonate and carbonate. Increased CO₂ loading accelerates the formation of these chemical species.

Equation 13

An increase in CO₂ loading also shows a shift of cathodic current density towards a greater value. This also indicates an increase of bicarbonate concentration in the solution, thus enhancing the corrosion rate [50,62]. Bicarbonate as a primary oxidizer in an aqueous MEA solution was also identified as playing an important role in corrosion in a mechanistic corrosion model [62]. Corrosion occurs due to iron dissolution (anodic reaction) and reductions of oxidizers (H₂O and HCO₃^- in cathodic reactions). Insoluble black particles of FeCO₃(Equation 13) were also observed [49].

Oxidized iron can be stabilized in the solution and formed into complexes with HSS anions such as formate and acetate [46]:

Equation 14

Complex formation [63]:

Equation 15

where n = 1 to 6. The complexation with HSS anions will increase dissolution of FeCO₃ into the solution with increased HSS anion concentration. As demonstrated by Cummings et al., solubility of FeCO₃ was significantly increased in the presence of bicine due to the complex formation of iron-hydroxide-bicine [63]. Diamines as degradation products of primary and secondary amines have no effect on corrosion, but can be strong chelators of iron and promote corrosion by CO₂.

Similar to the effect of CO₂, an increase in amine concentration intensifies the formation of carbamate, bicarbonate and carbonate. Also, increased amine concentration has a more pronounced effect on the anodic side of reactions, which accelerates the corrosion rate, as seen from the increase of carbonate/bicarbonate in the solution. There is also an increase in carbamate formation, as well as protonated amine and carbonate/bicarbonate [64].

It is known that the corrosion rates of individual amines are in the order of AMP > MEA > DEA > MDEA. For sterically hindered amines, the stability of carbamate is greatly reduced when the number of carbon atoms between the amine and alcohol groups increases from two to three, which increases bicarbonate formation [65,66]. It was also noted that the corrosion rate of AMP is higher than that of MEA under the same operating conditions because of increasing bicarbonate. The existence of chemical species of AMP–CO₂ system confirmed by ¹³C-NMR consist of the amine/protonated amine AMP/AMPH⁺ and the carbamate (AMPCOO^-) [60,61]. The carbamate stability constant (on mol fraction basis) was calculated to be 0.47 at 25°C [60]. Reactions responsible for AMP–CO₂ absorption are given as follows [60]:

Equation 16

Equation 17

Equation 18

Equation 19

Increased dissolved O₂ in solution means there is an increased presence of oxidizer in the solution [1,67] and an increased tendency for iron to be oxidized.

The oxidation–reduction reactions of iron (Equation 8) and O₂(Equation 9) in the solution:

Equation 20

According to Fontana, Fe(OH)₂ identified as a reaction intermediate is usually oxidized further to form a more stable compound such as Fe(OH)₃[1]:

Equation 21

SO₂ is a stronger acid gas than CO₂, quickly solubilizing in aqueous solution and producing hydrogen ions [68].

Equation 22

Equation 23

Equation 24

HSO₃^-, SO₃^-, and SO₄^- can react with protonated amine and form inorganic HSS, which cannot be regenerated by heat under stripping conditions. Hydrogen ions decrease solution pH and also act as an oxidizer in the solution. However, since the reduction of H⁺ provided by SO₂ hydrolysis shown in Equations 22–24 and 3 can be controlled by the strength of the amine as a base, there will be less possibility for free hydrogen ions to react with iron:

In the solution, SO₂ can react directly with water and O₂[69].

Equation 25

Temperature also has a profound effect on corrosion rate in the SO₂-amine system. In the MEA–O₂-CO₂-SO₂ system, as temperature raised to 30, 50 and 80°C, the polarization curve shifted toward greater values of both anodic and cathodic current densities, which corresponded to corrosion rates of 13 mpy, 41 mpy and 161 mpy, respectively [50]. These corrosion rates were measured under conditions of 1–7 kmol/m³ MEA, 0–100% feed gas O₂, 0–204 ppm SO₂, and 0–0.5 CO₂, and 30–80°C; however, test time was not specified.

Corrosion inhibition

Equipment corrosion and solvent loss are the essential problems in amine absorption and represent the most expensive costs in operating an amine system. The causes of corrosion include HSS from oxidation of amine, products from degradation of amine, and contaminants in flue gas and preparation water. It is possible that corrosion can be controlled and prevented, depending upon its route, using such approaches as filtering contaminants in flue gas before feeding it into the absorber unit, use of high purity water in preparation of amine solution, or separating [45] or neutralizing [70] HSS and degradation products to controlled levels. Since the anodic and cathodic reactions occurring during corrosion are mutually dependent, it is possible to reduce the rates of either reaction to reduce the corrosion [1]. Corrosion inhibitors slow corrosion processes by interfering with either the anodic and cathodic reactions or both, and injection of corrosion inhibitors is inexpensive. Chemical inhibitors, however, could accumulate and build up concentration over a period of time, causing a change in solution physical properties. Such changes cause surface tension to decrease when corrosion inhibitor is added in the MEA solution, leading to a foaming problem in the CO₂ absorption process [71]. In addition, the effectiveness of chemical inhibitors might be affected by variations in the process parameters. As well as the inhibitor type and concentration, operating temperature can have an impact on different corrosion inhibitors. A study showed that sodium metavanadate (VND) inhibition performance was not affected significantly with temperature change while the opposite was true for undisclosed inhibitors F, G and H [72]. It is important, therefore, to understand corrosion behavior when inhibitors are used.

Many corrosion inhibitors have been developed, patented and commercialized, and can be categorized into two types. The inorganic category includes salts of arsenic, vanadium, copper, cobalt, molybdenum, antimony and stannous [105]. The organic category includes nitro-substituted aromatic acids and its salts and naphthoquinone [106]. Inorganic inhibitors are more favored in practice than organic compounds because of their superior inhibition performance. However, these inorganic corrosion inhibitors are not environmentally friendly; as such inhibitors contain toxic arsenic, antimony, and vanadium. Vanadium compounds, particularly VND, are the most extensively and successfully used in amine treating plants. However, VND has been known to be toxic, as indicated by a much lower lethal dose to rat (10 mg/kg LD₅₀-oral) when compared with MEA solvent [72]. In addition, it also has a detrimental effect to MEA by boosting up the solvent degradation rate during CO₂ absorption operation [73]. The attractive, less toxic and more environmentally friendly inorganic corrosion inhibitor sodium molybdate, as indicated by a higher LD₅₀-oral of 4,000 mg/kg [202] is widely used in cooling water systems [74]. This inhibitor shows high performance in inhibiting both uniform and localized corrosion and pitting of ferrous and nonferrous metal [75].

Corrosion inhibition of VND and 2-aminothiophenol organic inhibitor (a good inhibitor for hot K₂CO₃ system) at various concentrations were compared using National Association of Corrosion Engineers and American Standard for Testing and Materials complied static weight loss methods in a 5 kmol/m³ AMP saturated with mixture of 52% CO₂ and 48% air (equivalent to 10% O₂) under boiling conditions [76]. To evaluate performance of the inhibitors (i.e.,% protection), a base run without inhibitor was initially conducted to obtain uninhibited corrosion rate, which was later used to compare with inhibited corrosion rates obtained from runs with inhibitors. For the tested period (not provided in this work), system containing 75 ppm VND yielded only 0.1 mpy corrosion rate equivalent to 99.9% protection, as compared with uninhibited system. On the other hand, 2-aminothiophenol failed the screen test due to nonreproduced corrosion data. The authors suggested that 2-aminothiophenol spiked AMP solution could be degraded under heat, CO₂ and O₂ during a test, which was observed from successive change of solution color from light yellow to purple.

For CO₂ separation using aqueous solutions of MEA, the possibility of using three types of low-toxic organic corrosion inhibitors (amines, carboxylic acid and sulfoxide), instead of inorganic heavy metal inhibitors, was investigated and compared with the commercial inorganic inhibitor VND in carbon steel and MEA solution under CO₂ saturation [72]. Inhibitors A–F respectively identified as imidazole, PZ, hexamethyleneimine, cyclohexylamine, 2,4-lutidine (or 2,4-dimethylpyridine) and long-chained aliphatic amine are amines that have nitrogen functional groups with different molecular structures, that is, aromatic and long-chain aliphatic. Based on this study, no inhibitors contain changes in both aromatic and aliphatic structures. Either aromatic or aliphatic compounds were tested for corrosion inhibition performance. The role of compound structures (except electron density) was rarely discussed in this study. The inhibition efficiency, shown by percentage of corrosion protection, depended on the inhibitor concentration and system temperature. VND provided the highest percentage of protection of 97% followed carboxylic acid with 92%, sulfoxide and long-chain aliphatic amine respectively. Mechanisms of inhibitors F-H were proposed as providing enhancement of corrosion resistance by suppressing reduction of HCO₃^- at the cathodic side and adsorption of the inhibitor at the surface of metal. Increase of temperature from 40°C to 80°C had no significant effect on the inhibition performance of VND, in which percentage of protection still remained as high as 97%. However, temperature had a significant impact on the F–H inhibitors, in which it increased percentage of protection of all inhibitors when temperature rose from 40°C to 80°C. For example, percentage corrosion protection of system containing inhibitor G doubled when the temperature was raised from 40°C to 80°C.

The inhibition performance of less toxic copper carbonate (CuCO₃) with 1350 mg/kg LD₅₀[203], as compared with vanadium based inhibitors such as VND having LD₅₀ of only 10 mg/kg, was also investigated [77]. Under the tested conditions of 5 kmol/m³ MEA, 0.20 CO₂ loading, various CuCO₃ concentration, 0–10% feed gas O₂ concentration, 0–2000 rpm solution velocity and 40–80°C, the addition of copper carbonate decreased the corrosion rate of carbon steel but might induce pitting corrosion. The inhibition efficiency was found to be greater at 40°C than at 80°C. This was probably due to the nature of the passive film. The inhibition performance was found negligibly affected by solution velocity range (i.e., 0–2000 rpm) used in this study, as observed from no apparent change in both anodic passive current density and cathodic Tafel slopes. Solution velocity only accelerated cathodic reactions by inducing more mass transfer rate of corroding species available for the metal surface. To our knowledge, there are no reports in the reviewed papers in regards to structure/nature of passive films. In real systems, corrosion is often determined by placing metal coupons in different location within the process to determine their weight loss after specific times or using a corrosometer. Determination of corrosion in pilot plants using these techniques can be found in literature [18].

Jovancicevic et al. studied inhibition of N-based surfactant inhibitor of imidazolines and their amide/amine counterparts containing different hydrocarbon chain lengths [78]. The corrosion tests were all conducted in CO₂-saturated brine done at 1 atm, 150°C and 6000 rpm rotating speed of electrode using inhibitor concentration in a range of 3–50 ppm. The corrosion rate was measured at steady state after 3–4-h test periods. In this study, contribution of inhibitor structures helps understand inhibitor molecular arrangements (e.g., spherical/nonspherical micelles) which strongly affect their adsorption strength at the metal surface for corrosion protection. These inhibitors function by forming an ordered structure known as micelle (i.e., an aggregate composing of hydrophilic head regions in contact with surrounding solvent while hydrophobic tails pointing into micelle center) to adsorb as monolayer or bilayer (i.e., molecules are arranged into a two-layered sheet with all of their hydrophobic tails pointing toward the center of the sheet and the hydrophilic heads being exposed to surroundings, such as metal surface and solvent) on hydrophilic surfaces of metal surface to give corrosion protection. Bilayer film specifically formed by surfactant inhibitors analyzed in this work could be considered a type of added passive film. Passive film is generally known as a 2–3 nm thick film composed mainly of iron oxides, Fe(OH)₂, Fe₃O₄ and Fe₂O₃, which adhere to the metal surface. The film functions as a barrier inhibiting oxidation reaction on the anodic side, thus minimizing corrosion of metals. Some inhibitors, such as CuCO₃[77] and VND [72] have been found to promote formation of a passive film on the anodic side. The bilayer film also passivates metal by protecting it from being corroded. However, bilayer film is formed and arranged on the metal surface in a special manner. The molecules of the surfactant inhibitor assemble themselves to a two-layered sheet, where all of the hydrophobic tails point toward the center of the sheet and the hydrophilic heads adhere to the metal surface. The adsorption of surfactant inhibitor can protect the metal surface from contact with oxidizing species in the systems (e.g., H⁺ and HCO₃⁺ in MEA–CO₂–H₂O system), thus corrosion is minimized.

Contributions of the constituent parts of the hydrophobic hydrocarbon chain lengths and types of hydrophilic head groups (i.e., imidazoline, amide and amine) on the inhibitor structure were discussed. For imidazoline inhibitors, hydrocarbon chain length was found to affect the minimum effective concentration (MEC), a 2 mpy corrosion rate. A decrease in hydrocarbon chain lengths from C20 to C10 or to C8, for the most part, linearly increased the MEC. Corrosion inhibition was unnoticeable when the chain length reached C8. Therefore, higher chain length imidazolines used smaller concentrations than their lower-chained counterparts to achieve maximum corrosion protection or a corrosion free environment (2 mpy), thus making them more effective inhibitors. The study suggested that longer chain length provided a better metal surface adsorption, due to higher cohesive energy and stability of the thicker bilayer as opposed to the small spherical micelle of the shorter chained C8 inhibitors. Inhibitors with amine and amide head groups were found to work more effectively than imidazoline inhibitors of a comparable hydrocarbon chain length. It was reasoned that amine/amide inhibitors were able to form nonspherical micelles at, or nearer to, the metal surface than their imidazoline counterpart, resulting in the formation of more stable admicelle bilayers, thus creating better corrosion protection. Ramachandran and Jovancicevic used molecular modeling to determine the binding of imidazoline and amide on the iron oxide surface [79]. The conceptual model was based on adsorption and bilayer film formation on a metal surface.

Conclusion

Most corrosion studies in amine system have been carried out to evaluate effect of amine type and concentration, temperature, CO₂ loading and O₂ concentration in the feed gas. More process parameters used in CO₂ absorption using aqueous amine should also be determined for their effects on the corrosion, especially effect of other impurities in gas stream such as NO_x and inorganic fly ash, which can lead to corrosion. Major degradation products other than HSS such as NH₃ and amides are also important species, which should be included in corrosion studies.

Most corrosion failures in amine service are localized corrosion. Thus, a focus should be given to localized corrosion, since it occurs in specific areas and is often undetected until an unexpected or premature failure occurs. Different forms of localized corrosion, namely pitting, erosion, galvanic, stress corrosion and inter-granular should be identified, as well as proper prevention techniques (e.g., use of effective inhibitors and control of system’s potential to passive regions).

More corrosion inhibitors must be screened, and their performances and roles in preventing corrosion in amine system should be evaluated. Evaluation of blends of two or more potential chemicals should be done to evaluate possible synergistic effect to enhance corrosion inhibition. One should also keep in mind that developed inhibitors should not trigger additional problems in the amine plants, such as solvent degradation and foaming. In addition, long-term exposure test using those inhibitors should be carried out to demonstrate their stability and extended inhibiting power.

Advanced analytical tools have been introduced to assist the electrochemical method to determine corrosive species in amine–CO₂ systems. These tools, including micro-Raman, should be applied to investigate corrosion species and to study corrosion inhibition of different types of corrosion inhibitors at various process parameters. The challenge is to find a way to use currently available instruments to perform in situ analysis of corrosion species in the vicinity of the electrodes in an amine–CO₂ aqueous solution.

Future perspective

As demand for energy increases, there is a growing reliance on coal combustion for power generation. This is leading to ever-increasing CO₂ emissions from large, stationary sources. Post-combustion CO₂ capture using alkanolamine absorption is considered by many to be one of the major technologies for reducing these emissions. MEA will likely continue to be the most effective solvent in use over the next 5–10 years; during this time, new candidate solvents with their characteristic properties for CO₂ absorption have to be developed and commercialized. In this regard, there will be more blending of other solvents with MEA (formulation) to overcome corrosion and other limitations compared with the use of MEA alone. This will complicate the corrosion behavior in amine systems, as every process parameter can also have an influence on corrosion and on the behavior of the various components of the blended amine solution. The effects of impurities such as SO₂ and NO_x in flue gas on corrosion need also to be included in future research, as there are few studies on SO₂ and none on NO_x. As seen in this review, a few advanced analytical instruments have been introduced for use in the study of corrosion. There will be more uses developed for state-of-the-art, advanced analytical instruments in order to explicitly uncover the electrochemistry of corrosion in order to identify the chemical species generated and their paths to the corrosion. Eventually methods will emerge that will demonstrate good methods of corrosion prevention and corrosion that can be inhibited in a cost-effective manner.

Table 1.  Worldwide CO2 emissions from large sources as reported by the IPCC.

Process	Number of sources	Emissions (MtCO2/yr)
Fossil fuels
Power	4942	10,539
Cement production	1175	932
Refineries	638	798
Iron and steel industry	269	646
Petrochemical industry	470	379
Oil and gas processing	Not available	50
Other sources	90	33
Biomass
Bioethanol and bioenergy	303	91
Total	7887	13,468

Data from [201].

Table 2.  Typical concentrations of coal-fired power plant flue gases after SO2 scrubbing.

Composition	Concentration (mol %)
CO2	7–15
O2	2–12
N2	65–75
H2O	5–15
SO2	2–400 ppm
SO3	1–10 ppm
NOX	1–400 ppm
Particulates	0.1–0.5 grains/standard cubic foot

Reproduced with permission from [11].

Table 3.  Summary of various corrosion studies conducted with different amines and test conditions.

Amine	Temperature (°C)	CO2 concentration	O2 concentration	Method	Corrosiveness	Mechanism	Ref.
3 M various amines	80	Saturated	0–10%	Electrochemical	MEA > AMP > DEA > MDEA	CO2 solubility	[19]
3 M various mixed amines (1:1)		Saturated			MEA/AMP > MEA/MDEA > DEA/MDEA
3 M various amines		0.2 mol/mol amine			AMP > MEA ≥ DEA
MEA/PZ (5:1.2) vs 6.2 M MEA	80	0.2 mol/mol amine	N/A	Electrochemical	MEA/PZ > MEA	Corrosive agent: HCO3^-	[38]
MEA and AMP (1–7M)	100	100% 52%	48% air	Static weight loss	MEA > AMP	N/A	[75]
5 M MEA containing HSS: acetic, formic, glycolic, malonic, oxalic and succinic acids	80	0.2 mol/mol amine	NA	Electrochemical	Oxalic > Malonic > Formic > Succinic > Acetic > Glycolic	N/A	[48]
5 M MEA	80	0.4 mol/mol amine	0–100%	Electrochemical	Increase with increase of % O2 in feed gas2 concentration	N/A	[51]
5 M MEA with SO2 (0–204 ppm)					Slightly increase with increase of SO2 concentration
MDEA: DEA of 3.5:1 and 2:1 with various H2S concentrations	120	Various CO2 loadings	N/A	Weight loss	Rate not affected by 0–0.35 CO2 loading range but increased with an increase of H2S and amine concentrations	N/A	[21]

AMP: 2-amino-2-methyl-1-propanol; DEA: Diethanolamine; HSS: Heat-stable salts; MDEA: Methyldiethanolamine; MEA: Monoethanolamine; NA: Not applicable; PZ: Piperazine.

Table 4.  Various corrosion inhibitor studies of different amine systems.

Amine	Temperature (°C)	CO2 loading	Test	Corrosion inhibitor	Inhibition efficiency (%)	Ref.
MEA and AMP	100	Saturated	Weight loss	10–75 ppm NaVO3 2–16 ppm NaVO3	93–99.9 >90–93	[75]
3 M MEA	80	Saturated	Electrochemical	1000–5000 ppm A: Imidazole, B: piperaxzine, C: hexamethyleneimine, D: cyclohexylamine and E: 2,4-lutidine	<60 (C turn the solution to orange)	[76]
				100 ppm F: long-chain aliphatic amine 1000–5000 ppm G: carboxylic acid 1000 ppm H: sulfoxide	75–92 (G > H > F)
				1000 ppm NaVO3	97
	40 vs 80			F, G and H: VS NaVO3	F, G and H increased with a temperature increase while NaVO3 remain 97
5 M MEA	40 vs 80	0.2–0.55 mol/mol amine	Electrochemical	250 ppm CuCO3	>80	[77]

A: Imidazole; B: Piperaxzine; C: Hexamethyleneimine; D: Cyclohexylamine; E: 2,4-lutidine; F: Long-chain aliphatic amine; G: Carboxylic acid; H: Sulfoxide.

AMP: 2-amino-2-methyl-1-propanol; MEA: Monoethanolamine.

Corrosion

Electrochemical process involving the transfer of electrons in oxidation–reduction reaction within the construction material, resulting in its deterioration.

Corrosion inhibitors

Chemical additives added to a process to decrease corrosion rate of construction materials, such as metal or alloy.

Electrochemical technique

Technique used to determine corrosion rate and behavior of a system of interest, based on plot of its potential and current densities, generated on anode and cathode during the test.

Executive summary

	▪ Alkanolamine absorption has played a major role in acid-gas removal, both past and present.
	▪ There are many activities using amine absorption in capturing CO2 from power plant coal-fired flue gas.
	▪ Corrosion is one of the drawbacks of capturing CO2 using amine absorption: ▪ Blending of existing typical alkanolamines (monoethanolamine, diethanolamine, methyldiethanolamine and 2-amino-2-methyl-1-propanol), as well as new solvents (piperazine and sterically hindered amines), has been studied as a means to increase the efficiency and absorption capacity of absorbents, to decrease energy requirements in the stripping process and to reduce corrosion; ▪ Corrosion by heat-stable salt is reported in the typical alkanolamines systems. There are few studies of corrosion in new solvents systems; ▪ Each system parameter, that is, concentrations of amine, CO2, oxygen and impurities, as well as temperature, increases the corrosion rate of carbon-steel equipment under alkanolamine operating conditions.
	▪ Generation of corrosion mechanisms is based on information from advanced analysis, using 13C-NMR, inductively coupled plasma–MS, scanning electron microscope and energy-dispersive spectrometers, assisting the electrochemical methods to measure chemical species in identification of corrosion products.
	▪ Corrosion in alkanolamine systems cannot be avoided: mixed anodic and cathodic inhibitors are recommended to change oxidative–reductive reaction behavior of potential corrosive species.

Financial & competing interests disclosure

The authors would like to acknowledge the research support over recent years to the International Test Centre by the followings organizations: Natural Sciences and Engineering Research Council of Canada, Canada Foundation for Innovation, Saskatchewan Ministry of Energy & Resources, Western Economic Diversification, EnCana Energy Inc., EON Energy, RWE Corp, Saudi Aramco, Doosan Heavy Industries, HTC Purenergy Inc. Saskatchewan Power Corporation, StatOil Hydro (Norway), SaskFerco Inc., Sulzer Chemtech (Switzerland), Fluor Corporation (USA), the Canada Centre for Mineral and Energy Technology, Alberta Energy Research Institute and the Research Institute of Innovative Technology for the Earth. In addition, the authors would also like to acknowledge the recent research supports from Provincial Government of Hunan, Federal Government of China, as well as Hunan University to the Joint International Center for CO₂ Capture and Storage. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.