| Summary: | Concerns about the impact of anthropogenic CO2 emissions on global climate change are necessitating the development of low-carbon technologies. Despite its energy requirements, carbon capture and sequestration (CCS) provides a promising approach to limiting carbon dioxide emissions, in addition to renewable energy and nuclear power. Due to the depth of technological experience and retrofitting capabilities associated with the process, post- combustion CO2 capture based on absorption in alkanolamine solutions is presently the most feasible technology available for mitigating the soaring levels of atmospheric CO2 and achieving the ambitious targets set by the Intergovernmental Panel for Climate Change (IPCC). Whilst the process of CO2 capture by alkanolamines involves a technologically mature process, it is still faced with numerous challenges, such as high energy requirements and operating costs, as well as operational difficulties. Monoethanolamine (MEA), the benchmark amine for the process, frequently becomes contaminated with degradation products and, when CO2-loaded, is corrosive to the carbon steel process equipment. Hence, this thesis aimed to determine the oxidation and reduction reaction kinetics and mechanisms of iron in aqueous MEA-CO2 systems as function of experimental variables, aiming to predict iron corrosion rates under process conditions. The behaviour of iron in aqueous MEA solutions was characterized by voltammetry with a rotating disc electrode (RDE) and an electrochemical quartz crystal microbalance (EQCM), as functions of temperature (25-80oC), CO2 loading (0-0.6 mol CO2 (mol amine)-1), pH (8.10- 12.55), MEA concentration (5-60 wt%) and oxygen concentration. Electrode potential-pH and activity-pH diagrams of iron-water-CO2 systems were used to assist with reaction assignments. The passive electrochemical behaviour of Fe in MEA at pH ca. 12 switched to active dissolution on loading the MEA with CO2, causing the pH to decrease to ca. 8. Analysis of the resulting kinetic data enabled corrosion rates to be predicted as functions of the experimental variables. Based on the proposed corrosion mechanisms from the voltammetric results, a mechanistic model was developed for the uniform corrosion of iron in CO2-loaded MEA systems, taking into account the CO2 absorption equilibria reactions and the electrochemical reactions at the iron | solution interface. Equilibrium concentrations of the amine species (RNH2, RNH3+, RNHCOO-), carbon(IV) species (HCO3-, CO32-) and hydrogen ions (H+) were calculated with a Kent-Eisenberg type model. The electrochemical reactions incorporated in the model were the anodic dissolution of iron and the cathodic reduction of H+, direct water reduction and the reduction of oxygen. The corrosion model was developed by simulating polarization curves based on the species concentrations and the transport limited current densities of the iron RDE defined by the Levich equation. In order to measure dissolved iron concentrations under typical CO2 absorption conditions, an electrochemical flow reactor was designed and fabricated from PTFE with an iron/steel anode, platinised titanium cathode and a cation-permeable Nafion membrane. Inductively coupled plasma optical emission spectrophotometry (ICP-OES) was used to determine FeII concentrations in the amine solution after potentiostatic electrolyses as a function of solution flow-rate, temperature, CO2 loading, pH, MEA concentration, oxygen content, steel type and amine type, enabling partial current densities leading to iron dissolution to be deconvoluted from measured current densities. Results from the voltammetric data from the RDE were used to propose a mechanism for the oxidation and reduction reactions occurring at the iron surface. During positive-going potential sweeps, the voltammograms were characterised by 3 anodic peaks corresponding to the anodic dissolution of Fe to FeII at low potentials, leading to iron carbonate formation if its solubility product was exceeded, and a passive region from the formation of adsorbed products such as iron(III) (hydr-)oxide at higher potentials. The cathodic reactions at the iron surface included reductions of H+, H2O and O2. Based on the thermodynamic predictions, the anodic dissolution of iron was assigned to the formation of a soluble FeII species. Using a flow reactor operated in batch recycle mode, constant potential electrolyses at a potential -0.6 V, within the anodic dissolution potential range, resulted in a dissolution charge yield of less than unity, which varied depending on the experimental conditions, implying the formation of insoluble iron species such as Fe(OH)2 or FeCO3. The mass transport behaviour of the flow reactor was characterised as a function of solution flow rate using the transport controlled reduction of hexacyanoferrate(III) ions at a platinised titanium electrode, resulting in a mass transport correlation for the reactor. A value of the diffusion coefficient for FeII from the literature enabled prediction of the conditions required for FeCO3 and Fe(OH)2 formation, based on measured fluxes and dissolved FeII concentrations in solution, from which the rate of FeCO3 formation could also be estimated. Kinetic analysis of the data from the RDE, EQCM, flow reactor and the corrosion model resulted in a complementary set of results. The corrosion behaviour of iron in aqueous MEA- CO2 solutions was sensitive to changes in the operating parameters; iron dissolution rates were enhanced by increasing temperature, CO2 loading, solution velocity and oxygen content. However, at high temperatures, high CO2 loading and high FeII concentration and lower solution velocity provided favourable conditions under which a protective FeCO3 layer could precipitate. Dissolution rates increased with concentration at lower concentrations of MEA (5- 40 wt%) and decreased with increasing concentration at higher concentrations (40-60 wt%), due to viscosity effects. Significantly lower corrosion rates were measured in the other commercially available solvents tested namely methyldiethanolamine (MDEA), 2-amino-2- methyl-1propanol (AMP) and aminoethylpiperazine (AEP). Carbon steel exhibited similar dissolution rates to those of iron, whereas those for stainless steel were significantly lower. From the comprehensive results on the oxidation and reduction kinetics of iron in benchmark MEA-CO2 systems, the effect of MEA concentrations on predicted iron corrosion rates brings into question the effectiveness of the concentration of MEA used in amine scrubbing being limited to 30 wt%, specifically to limit corrosion related issues. The results also indicated that MEA was the most 'aggressive' in corrosion behaviour, as other amines such as MDEA, AMP and AEP provided more promising alternatives, based on both their predicted corrosion behaviour and their reported efficiencies in the absorption process.
|
|---|
