Detecting Reactive Products in Carbon Capture Polymers with Chemical Shift Anisotropy and Machine Learning

Aminopolymers are attractive sorbents for CO2 direct
air capture applications due to their high density of amine groups,
which can readily react with atmospheric levels of CO2 to
form chemisorbed species. The identity of these chemisorbed species
and the functional groups that form upon oxidative degradation depends
on both material properties and processing conditions, forming a variety
of carbonyl-type sites such as ammonium carbamates, bicarbonates,
carbonates, carbamic acids, ureas, and amides. 13C solid-state
nuclear magnetic resonance (NMR) is often used to help elucidate the
identity of these reacted species, but it is challenging due to the
narrow chemical shift range of carbonyl sites. Herein, we demonstrate
the application of a two-dimensional (2D) chemical shift anisotropy
(CSA) recoupling pulse sequence (ROCSA) to obtain CSA tensor values
at each isotropic chemical shift, overcoming limitations of isotropic
peak resolution. CSA tensor values describe the local chemical environment
and can readily differentiate between the chemisorbed and degradation
products. To aid identification, we also developed a k-nearest neighbor (kNN) classification model to distinguish the functional
groups via their CSA tensor parameters. This methodology was demonstrated
on poly(ethylenimine) in γ-Al2O3 exposed
to CO2 and showed that the chemisorbed products are ammonium
carbamate and a mixed carbamate–carbamic acid species. The
sample was analyzed again after desorption at 100 °C inducing
mild degradation, and the remaining products were strongly bound carbamate
and urea species. The combination of 2D CSA measurements coupled with
a kNN classification model enhances the ability to accurately identify
chemisorbed or degradation products in complex carbon capture materials.