The direct air capture (DAC) of carbon dioxide is a potentially transformative technology for addressing climate change. This program aims at the realization of a next generation of cooperative adsorbents, consisting of polyamine molecules that function as porous molecular solids and reversibly react with CO2 to form higher dimensional ammonium carbamate networks with the right thermodynamics to extract it efficiently from air. Materials discovery is accelerated through computational methods that capitalize upon state-of-the-art machine learning techniques. The studies in this program will lead to powerful insights into structure-function relationships that will then feed into the machine learning models to identify potential new materials even better suited for DAC.