Amine-containing adsorbents have been extensively investigated for post-combustion carbon dioxide capture due to their ability to chemisorb low-concentration carbon dioxide from a wet flue gas. However, earlier studies have focused primarily on the carbon dioxide uptake of adsorbents, and have not demonstrated effective adsorbent regeneration and long-term stability under such conditions. Here, we report the versatile and scalable synthesis of a functionalized-polyethyleneimine (PEI)/silica adsorbent which simultaneously exhibits a large working capacity (2.2 mmol g−1) and long-term stability in a practical temperature swing adsorption process (regeneration under 100% carbon dioxide at 120 °C), enabling the separation of concentrated carbon dioxide. We demonstrate that the functionalization of PEI with 1,2-epoxybutane reduces the heat of adsorption and facilitates carbon dioxide desorption (>99%) during regeneration compared with unmodified PEI (76%). Moreover, the functionalization significantly improves long-term adsorbent stability over repeated temperature swing adsorption cycles due to the suppression of urea formation and oxidative amine degradation.

Amine-containing solids have been investigated as promising adsorbents for CO2 capture, but the low oxidative stability of amines has been the biggest hurdle for their practical applications. Here, we developed an extra-stable adsorbent by combining two strategies. First, poly(ethyleneimine) (PEI) was functionalized with 1,2-epoxybutane, which generates tethered 2-hydroxybutyl groups. Second, chelators were pre-supported onto a silica support to poison p.p.m.-level metal impurities (Fe and Cu) that catalyse amine oxidation. The combination of these strategies led to remarkable synergy, and the resultant adsorbent showed a minor loss of CO2 working capacity (8.5%) even after 30 days aging in O2-containing flue gas at 110 °C. This corresponds to a ~50 times slower deactivation rate than a conventional PEI/silica, which shows a complete loss of CO2 uptake capacity after the same treatment. The unprecedentedly high oxidative stability may represent an important breakthrough for the commercial implementation of these adsorbents.

Poly(ethyleneimine) (PEI)/silica has been widely studied as a solid adsorbent for post-combustion CO capture. In this work, a highly macroporous silica (MacS), synthesized by secondary sintering of fumed silica, is compared with various mesoporous silicas with different pore structures as a support for PEI. The silicas with large pore diameter and volume enabled high CO adsorption kinetics and capacity, because pore occlusion by the supported PEI was minimized. The steam stability of the silica structures increased with the silica wall thickness owing to suppressed framework ripening. The silicas with low steam stability showed rapid leaching of PEI, which indicated that the PEI squeezed out of the collapsed silica pores leached more readily. Consequently, MacS that had an extra-large pore volume (1.80 cm g ) and pore diameter (56.0 nm), and a thick wall (>10 nm), showed the most promising CO adsorption kinetics and capacity as well as steam stability.

Substantial efforts have been made to increase the CO working capacity of amine adsorbents for an efficient CO capture. However, the more important metric for assessing adsorbents is the regeneration heat required for capturing a fixed amount of CO. In this work, we synthesized polyethyleneimine (PEI)/SiO adsorbents functionalized with various epoxides. This provided adsorbents with six different amine structures showing various CO/HO adsorption properties. Our studies revealed that the CO working capacity was not a decisive factor in determining the regeneration heat required for CO capture. This is because the benefit of large CO working capacity was canceled out by the difficulty of CO desorption. Instead, the suppression of HO co-adsorption was critical for reducing the regeneration heat because substantial latent heat is required for HO desorption. Consequently, the PEI/SiO functionalized with 1,2-epoxybutane required a much lower regeneration heat (2.66 GJ tCO) than the conventional PEI/SiO (4.03 GJ tCO) because of suppressed HO co-adsorption as well as moderately high CO working capacity.

No abstract