(353b) Methods to Derive the Thermodynamic Properties of CO2 Interaction with Amine Based Sorbents

Introduction

In order to effectively capture carbon dioxide (CO₂) from point sources such as the flue of pulverized coal power plants, we must first understand the effects that both pressure and temperature play on the thermodynamics and kinetics of the amine - CO₂ interaction.  Amine based sorbents for capture rely on the basic nitrogen site to interact with the acidic CO₂ forming a covalent bond which is subsequently broken under regeneration conditions.  Recent findings with carbon capture using gas phase adsorption experiments in a plug flow reactor highlight the benefits of this type of capture over membrane or absorption capture processes (Gray et al. 2006).  Gas phase capture and regeneration can be characterized as a thermodynamic shift in the Gibbs free energy from conditions resulting in spontaneous adsorption to conditions such that the adsorbed species is no longer thermodynamically stable on the surface.  In this work we use density functional theory calculations in conjunction with experimental analysis of a plug flow reactor to evaluate the thermodynamic properties of the amine ? CO₂ interaction.

Experimental Details

                A custom built tubular plug flow reactor (PFR) was used for all experimental results.  Gas concentrations were recorded by a mass spectrometer in tandem with a CO₂ analyzer to provide information including the volumetric concentrations of CO₂ adsorbed and desorbed from the surface during the capture and regeneration steps.  The amines under investigation involve bicyclic tertiary amidines including Diazabicyclononene (DBN) and Diazabicycloundecene (DBU) as well as Aminopropyltriethoxysilane (APTES) which were immobilized on the surface of both activated carbon and mesoporous silicas by either wet impregnation or co-condensation techniques.  Changes in the operating conditions during the adsorption and regeneration steps were modified in order to determine the effect that changes in the driving forces for adsorption played on the interaction.  A wide variety of initial conditions were selected to emulate the range of conditions which could be found in the flue of a pulverized coal power plant (30°C ? 60°C, 0.05 ? 0.15 pCO₂/p°, 0-7.5% water).  The final regeneration conditions, were selected such that all CO₂ was removed from the surface and generally consisted of an increase in reactor temperature (85°C) and a reduction in the partial pressure of CO₂ (pCO₂/p°) by flowing an inert gas through the reactor.  Water concentration was controlled by passing a percentage of both the reaction and inert streams through a temperature controlled humidity tank where the water concentration was determined from relative humidity calculations based on the ambient pressure of the room.  Extrinsic properties of the reactor setup, including the dispersion and space velocity, were determined from a pulse of argon introduced through a test loop in the setup. 

Results and Discussion

                Measurements from the mass spectrometer are analyzed in conjunction to the volumetric flow of gas exiting the reactor in order to obtain the number of moles of CO₂ interacting with the sorbents in each step.  From these results, we obtain the delta loading of the reaction on both a weight and molar basis by integration of our concentration measurements with respect to time.  We see in Figure 1 that changes in the concentration of water in the feed stream result to changes in the measured capacity of our sorbent and in the case of DBN, result in an optimal capture capacity between 2.5% and 7.4% water.

Figure 1.  Adsorption and desorption profiles of 1,5 Diazabicyclo [4.3.0] non-5-ene.

In addition to our experimental analysis we also conduct Density Functional Theory calculations on our model sorbent systems to better understand the interactions on an atomic level.  The energies of adsorption at 0 Kelvin are related to our experimental conditions through the use of a thermodynamic framework, allowing us compare the trends in energies between our experimental and computational work.  By modeling these energies using a single site, single adsorption model, we can obtain an estimate for the capacity we could expect from each sorbent at a given set of operating conditions.

Conclusion

                In the course of this work, we develop an underlying framework to evaluate the effectiveness of amine based sorbents supported on mesoporous silica and activated carbon to capture and release CO₂.  We find that changes in our operating conditions greatly impact the capture efficiency of each sorbent and that each amine has an individual set of operating conditions such that its capture capacity is maximized while minimizing the cost of regeneration.

References

1.       Gray, M.L. et al., 2006. Novel Tertiary Amine Solid Adsorbents Used for the Capture of Carbon Dioxide.