Adsorption and desorption kinetics of dry sorbents for CO2 capture

Abstract

This study focused on commercially available activated carbons' adsorption and desorption kinetics using a laboratory-scale fixed-bed reactor. The three activated carbons are CQ650, derived from coconut shells and activated with a combination of steam and KOH impregnation; CQ30P, derived from coal, also using a combination of steam and KOH impregnation. The third is CQ006, derived from coal and activated using steam and an acid wash. Ten continuous adsorption and desorption cycles were performed at 30 to 70 ℃ with 10 ℃ increments and two CO2 concentrations of 5 and 15 vol.%. The characterisation data of the samples show that CQ006 has the highest fixed carbon content while CQ650 has the highest BET surface area of 517.1±4.5 m2/g and the highest Dubinin-Radushkevich micropore surface area at 735.0±27 m2/g. A scanning electron microscopy analysis revealed that the activated carbon samples have well-developed pore structures over the entire surface.

All the investigated sorbents performed well under cyclic operation with no significant difference in adsorption capacity between cycle 1 and cycle 10. The most significant difference is the completion time, with CQ650 taking 8600 seconds at 60 ℃ and 15 vol.% CO2 feed concentration, while the quickest sorbent, CQ006, took 5900 seconds under the same conditions. CQ650 havethe highest adsorption quantity overall at 5 and 15 vol.% CO2 feed concentration with 0.96 mmol/g and 1.67 mmol/g, respectively. CQ30P and CQ006 have near-identical adsorption capacities of 0.85 mmol/g and 0.83 mmol/g at 5 vol.% CO2 feed and 1.03 mmol/g for both at 15 vol.% CO2 feed concentration.

The increased temperatures decreased the sorbents' saturation times and adsorption capacities, while the feed concentration also significantly affected the CO2 quantity adsorbed and the saturation times of the experiments. CQ006 has the best cyclic desorption efficiency, with the lowest efficiency achieved at 93.2%; CQ650 is second with the lowest cyclic efficiency at 90.1%, and CQ30P at 86.7%. The same trend is seen when looking at the overall desorption efficiency across the 5 temperatures (30, 40, 50, 60 and 70 ℃), where CQ006 averages at 98.3% and 99.1% for the 5 and 15 vol.% CO2 feed concentrations, 96.5% and 95.1% for CQ650 and 93.4% and 92% for CQ30P at 5 and 15 vol.% CO2 feed concentrations, respectively.

Three kinetic models were tested, with Avrami being the most suitable with a quality of fit of (99.1%), followed by the pseudo-first-order (94.7%) and the pseudo-second-order kinetic model (87.6%). The desorption activation energy is higher than the adsorption activation energy for all conditions, indicating that desorption (reverse adsorption reaction) requires more energy to remove the molecules from the surface of the sorbent than the energy released by binding to the surface of the sorbent. The thermodynamic results indicate that the adsorption mechanism is physical adsorption with a change in enthalpy of (-10 to -20 kJ/mol). The negative change in enthalpy indicates that the reaction is exothermic, aligning with other reports from the open literature. The entropy in all conditions is also negative (-0.00198 kJ/mol to -0.00660 kJ/mol), implying that the molecules are slightly more orderly when adsorbed. A negative change in the Gibbs free energy (-18.37 kJ/mol to -22.57 kJ/mol) implies that the reactions occur spontaneously.

The breakthrough analysis using the Avrami equation for adsorption and a modified Avrami equation for desorption provided satisfactory results. The Avrami is well suited to predict adsorption breakthrough behaviour, with the lowest QOF% achieved being 96.4%. In contrast, the modified Avrami equation does not adequately predict the desorption breakthrough behaviour of the sorbents, with the QOF% ranging from 63.2% to 87.9%.