Bottom-up computational design of shape-selective organic macrocycles for humid CO capture.

The capture of CO emissions using porous solids is challenging because polar water molecules bind more strongly in most materials than non-polar CO molecules. This is a challenge for both flue gas capture and for direct air capture alike. Here we develop a bottom-up computational screening workflow to calculate the binding energy of 27,446 diverse molecular fragments with both CO and water. Most molecules favour water binding, but bent, clip-like aromatic molecules exhibit potential for the desired reverse selectivity. This suggests that aromatic macrocycles with specific shapes can promote multiple weak π-π interactions with CO that surpass stronger but less numerous dipole-π interactions with water. We synthesize two water- and acid-stable molecular prisms with triangular and square geometries, as suggested by computation. Experiments confirm that the CO capture capacity of these prisms is unaffected by high relative humidity, surpassing the performance of benchmark commercial porous materials.