(206f) Investigating Aerosol Formation in Flue Gas from Post Combustion CO2 Capture Plants Using Molecular Dynamics Simulation

Monoethanolamine (MEA) has a long history of being used as a solvent for acidic gas scrubbing. Through decades it has proven its economical and industrial viability. Nevertheless, solvent loss is observed during CO₂ capture process which becomes significant in large scale. During the absorption process, where MEA and flue gas mixture meet causing supersaturation of the gas, tiny particles named as particulate matter (PM) form as a separate phase. These particles can carry over some MEA out from the column. Few attempts have been done to predict MEA loss through simulation using commercial chemical engineering software such as Aspen Plus. However, due to the limitations of such commercial softwares a true picture of the particle formation through nucleation process is not captured. Molecular dynamics simulation can be instrumental to achieve such a goal.

In this work, molecular dynamics simulation was carried out using GROMACS software in a strive to understand the underlying mechanisms of formation of particles in CO₂ capture process. Five different systems were considered. The former three consist of only MEA and CO₂, while the latter two also contain air components, water, and sulfur dioxide. Analysis of the simulation included radial distribution function (RDF), cluster size and formation study, and energy examination. Initially, appropriate forcefield is selected by computing densities and self-diffusion coefficients of MEA and CO₂ and comparing themwith their known value. As a result, literature data has been chosen, which yielded the following data for MEA and CO₂ density and self-diffusion coefficient: 1094.50 ± 36.27 kg/m³ and (0.261 ± 0.005)×10^-5 cm²/s, 1.624 ± 0.10 kg/m³ and 15550.96 ± 436.89×10^-5 cm²/s, respectively. Graphical representation illustrates formation of molecule clusters (PM) in all the systems. These clusters contained all the components, but were mainly consisted of CO₂, water and MEA molecules. Furthermore, existing particles in the flue gas, which are represented by SO₂ particles are also engulfed into the clusters. RDF analysis demonstrated distinct peaks between 4 and 5 Ã… for all systems in both MEA...CO₂ and CO₂...CO₂ correlations. In cluster size analysis of MD ensemble, number of clusters decreased suddenly indicating quick formation of clusters within 2 ns. Notably, in case of presence of air components and water, cluster sizes were more stable. Energy analysis revealed that MEA – CO₂ interaction is suppressed when air components are introduced to the system. It increased from -76.7094 kJ/mol to -15.0158 kJ/mol, which means inhibition of attraction between MEA and CO₂. CO₂ – CO₂ potential dropped from -23851.0607 kJ/mol to -12113.3779 kJ/mol. Most likely, this is because CO₂ molecules are also attracted to the other components such as nitrogen and oxygen.