Thermodynamic and mass transfer modeling of aqueous hindered amines for carbon dIoxide capture
MetadataShow full item record
With the detrimental effects of global climate change beginning to be felt, there is a growing consensus that something must be done. One part of the solution is carbon capture and storage using amine scrubbing to capture 90% of the CO2 from power plants burning coal and natural gas. To actualize this solution, process models are necessary. A process model requires an accurate thermodynamic and mass transfer model with physically meaningful parameters. While hindered amines are commercially used, the reason for their mass transfer rates is still an open question. These two needs are addressed in this work. To improve thermodynamic modeling, the physical significance of the electrolyte non-random two-liquid (eNRTL) regressed binary interaction parameters were examined. To improve mass transfer modeling, a response surface methodology (RSM) approach was used to give statistically significant regressed parameters. The mass transfer of two hindered amines, 2-amino-2-methyl-propan-1-ol (AMP) and 2-piperidineethanol (2PE) was studied to determine the role of carbamate. The absolute difference in eNRTL binary interaction parameters was found to moderately correlate with the pKa of the amine. An analogy method was developed to enable thermodynamic model creation for amines in the absence of some physical property data. The carbamate reaction plays a determining role in mass transfer of hindered amines. Based on Bronsted plots, 2PE appears to form carbamate using the same mechanism as unhindered, cyclic secondary amines, while AMP does not seem to use the same mechanism as unhindered, primary amines. The rate constant for bicarbonate formation for both amines is a factor of twelve faster than predicted from tertiary amine bicarbonate formation, indicating that neither seems to form bicarbonate using the tertiary amine mechanism. The six models constructed in this work enable process modeling and economic comparisons of solvents. Four binary interaction parameters were the most physically significant and should be regressed for future solvents. The high bicarbonate reaction rate of the hindered amines should be further investigated to determine if the mechanism is different or if this is model artifice, as either outcome will substantially improve mass transfer modeling for all amines.