Article pubs.acs.org/IECR
FeCO3 Coating Process toward the Corrosion Protection of Carbon Steel in a Postcombustion CO2 Capture System Liangfu Zheng,† James Landon,† Naser S. Matin,† and Kunlei Liu*,†,‡ †
Center for Applied Energy Research, University of Kentucky, 2540 Research Park Drive, Lexington, Kentucky 40511, United States Department of Mechanical Engineering, University of Kentucky, Lexington, Kentucky 40506, United States
‡
S Supporting Information *
ABSTRACT: Corrosion mitigation was evaluated at 80 °C in a naturally aerated 30 wt % monoethanolamine (MEA) solution with a CO2 loading of 0.43 molCO2/molalkalinity for a FeCO3-coated A106 carbon steel sample fabricated from solutions of NaHCO3 and piperazine with CO2 loading. The results showed that spontaneous passivation of A106 carbon steel was promoted by this FeCO3 coating, which resulted in a Fe3O4 layer formation. FeCO3 was metastable in this aerated MEA solution and eventually would transform into Fe3O4. Overall, the FeCO3 coating reduced the corrosion rate by a factor of 70. In this case, a semi in situ process to form and/or recover a dense FeCO3 layer to mitigate corrosion in a postcombustion CO2 capture system is proposed. Because of its more environmentally friendly nature, easier process control, and faster FeCO3 coating synthesis, NaHCO3 is recommended as a FeCO3 fabrication solution with respect to a piperazine solution with CO2 loading.
1. INTRODUCTION Postcombustion CO2 capture with the application of aqueous alkanolamine-based solutions is a potential method to mitigate anthropogenic CO2 generation (a main greenhouse gas) from large-scale sources such as fossil-fueled power plants.1−3 However, one concerning issue for this process is corrosion of the wetted process units such as the absorber column, stripper column, pipelines, heat exchangers, and reboiler.2−5 Most alkanolamine-based solutions such as monoethanolamine (MEA) and diethanolamine (DEA) are highly corrosive after a certain amount of CO2 is captured under process conditions.2−4,6−15 Corrosion is exacerbated because of the presence of 4−8 vol % O2 and trace amounts of SOx and NOx in the flue gas during the absorption process.1−3,6−11,16 These issues will inevitably increase the total capital cost of postcombustion CO2 capture. To minimize corrosion issues and thus lower the total capital cost for CO2 capture, techniques such as protective coatings, corrosion-resistant materials, corrosion inhibitors, and alternative solvents with less corrosivity may be applicable.3,5,12−22 However, in the open literature, most of these techniques have secondary issues that must be addressed before their application. For example, an alumina coating was reported to be highly corrosion-resistant in a MEA solution with CO2 loading.14 However, the reported fabrication process, i.e., a three-step process with electrochemical nickel plating, a pack cementation aluminizing process at 615 °C, and a preoxidation process at 900 °C,14,23 is somewhat complicated and could potentially damage the mechanical strength of the underlying © 2016 American Chemical Society
carbon steel because of its limited heat treatment temperature (≤475 °C). Stainless steel was reported to have much higher resistance to corrosion in alkanolamine-based solutions after capturing CO2,8,9,15 but its price is about 4 times higher than that of carbon steel and pitting corrosion remains a concern for its application.24 For corrosion inhibitors, ecofriendly sulfurcontaining organic inhibitors such as 2-mercaptobenzimidazole were observed to have severe chemical stability issues under CO2 stripping conditions.15 Under this circumstance, inorganic inhibitors with much higher stability were considered. The reported effective corrosion inhibitors are normally heavy-metal compounds such as lead-, antimony-, bismuth-, arsenic-, vanadium-, and tin-based compounds,12,17−21 and their use is somewhat restricted because of their high toxicity to humans and the environment. These compounds could cause concern during disposal and subsequently increase the total cost. Although a less toxic compound of copper carbonate (CuCO3) was observed to perform well in an MEA-based solution as a corrosion inhibitor by Soosaiprakasam and Veawab,12 Cu2+ was reported to significantly accelerate amine oxidation,25 and this initial evaluation lacks data concerning long-term stripping conditions. Some alternative solvents such as piperazine (PZ) have been shown to possess low corrosivity because of its ability to promote formation of a dense and protective layer of Received: Revised: Accepted: Published: 3939
November 4, 2015 March 3, 2016 March 22, 2016 March 22, 2016 DOI: 10.1021/acs.iecr.5b04145 Ind. Eng. Chem. Res. 2016, 55, 3939−3948
Article
Industrial & Engineering Chemistry Research FeCO3 on the surface of carbon steel,13,16,22 although there remain other disadvantages to PZ systems such as solubility concerns.26,27 Formation of a dense layer of FeCO3 was reported to be highly effective to mitigate the corrosion of carbon steel in a CO2 environment.11,13,16,22,28 However, for solvents such as MEA, the major issue is their lack of ability to promote formation of this layer under optimum operating conditions for postcombustion CO2 capture.4,6−9,11−15 Therefore, this work seeks to build and recover a reliable dense layer of FeCO3 semi in situ that could be applied to carbon steel to enhance its corrosion resistance in postcombustion CO2 capture processes for solvents with severe corrosion issues; i.e., a flushing process with certain solutions where a dense layer of FeCO3 can be formed (FeCO3-forming solutions) such as NaHCO3 and/or PZ solutions with certain CO2 loading is proposed. This provides an additional choice for application of these solvents for the capture process without the addition of a corrosion inhibitor. In this work, corrosion of A106 in solutions of NaHCO3 and PZ with CO2 loading and A106 with and without a FeCO3 coating in a MEA solution with CO2 loading was conducted by electrochemical testing methods such as linear polarization resistance (LPR) testing and corrosion potential (Ecorr) measurements, along with solution analysis and surface characterization after corrosion.
reason for the selection of a rich solution with a representative CO2 loading of 0.43 C/N in this work. All solutions were exposed to air from the preparation process, storage, and experimental testing in a nonpressurized vessel open to air. No deaeration method was used in this work. Therefore, the solutions were naturally aerated. For a PZ solution with a CO2 loading of 0.43 C/N at ambient temperature, the total alkalinity and pH are 6.14 mol/kg and 8.97, respectively. For a MEA solution with a CO2 loading of 0.43 C/N at ambient temperature, the total alkalinity and pH are 4.64 mol/kg and 9.70, respectively. To simulate the operating condition of the crossover heat exchanger between the stripper and absorber, a temperature of 80 °C was chosen for the corrosion study in this work. The CO2 loading for amine solutions and the temperature for all solutions used in this work are 0.43 C/N and 80 °C, respectively, unless otherwise noted. 2.3. Electrochemical Corrosion Testing. An electrochemical corrosion unit with a standard three-electrode setup was used for electrochemical corrosion testing with a saturated calomel electrode (SCE) as the reference electrode and a graphite rod as the counter electrode (see the schematic in Figure S2 in the Supporting Information), as described elsewhere.22 LPR testing, with polarization from −10 to +10 mV versus Ecorr and a scan rate of 0.166 mV/s, was carried out for approximately 15 min after an initial potentiostatic reduction at −1.2 V vs SCE for 2 min to help remove airformed oxides on the A106 surface. This step was used to help eliminate the effect of surface oxides on the subsequent electrochemical corrosion measurements.29 In the case with the FeCO3 coating, this precathodization process was omitted in order not to affect the protective layer, and electrochemical testing was conducted after the system reached a steady state, i.e., with Ecorr being stable around −0.82 V vs SCE (with a fluctuation of