In
the development of solid amine CO2 adsorbents, the
CO2 adsorption performance of amine-functionalized adsorbents,
with various novel porous supports or modification of the amine structure,
has been widely studied. However, a lack of systematic research limits
the industrial application of amine-functionalized CO2 adsorbents,
especially the adsorbents prepared from inexpensive and readily available
commercial porous supports. In this work, machine learning (ML) was
employed to explore how the CO2 adsorption performance
of amine-functionalized adsorbents is correlated with five factors:
amine loading, amine type, pore volume, pore size, and specific surface
area. We found that amine loading contributed the most to the effect
of CO2 adsorption capacity, followed by pore volume. Pore
size was the most important factor affecting amine efficiency, while
the cycle stability of the adsorbent was basically related to the
amine type, and the interaction effect between the influencing factors
was explored by ML. In addition, the CO2 adsorption capacities
of TEPA/KXY and PEI/KYX adsorbents were predicted by ML, and the results
of ML prediction were consistent with our experimental results. Furthermore,
we constructed a “five-in-one” comprehensive comparison
of the CO2 adsorption performance of 45TEPA/KYX and 60PEI/KYX
adsorbents through a radar diagram, and it was considered that the
45TEPA/KYX adsorbent had a better comprehensive CO2 adsorption
performance. Our study provides insights into the development and
optimization of solid amine CO2 adsorbents using commercial
porous supports.