Abstract
<jats:p>Net-negative methanol production has long been constrained by an energy penalty. The thermal energy required to release captured CO2 (desorption) often offsets the net carbon benefit of the entire conversion cycle. We address this issue through a staged Reactive Capture and Conversion (RCC) process. This system bypasses the energy-intensive desorption step by directly hydrogenating the captured CO2 within the capture solvent. The captured carbon is converted to methanol through staged reactions, each carried out in its own optimal chemical environment. To identify optimal amines that mediate each stage, we screened 176 million compounds using a graph neural network (GNN)-based molecular screening framework. We identified piperidine as the key mediator, which facilitates the rate-limiting C–N bond cleavage and markedly raises the methanol yield. Our system achieves a single-pass methanol yield of 56%, which is more than three times the yield of conventional CO2 hydrogenation and double that of industrial coal-to-methanol synthesis. It delivers net-negative emissions while achieving a minimum selling price of $0.55/kg, which is competitive with fossil-based market prices. Comprehensive techno-economic and life-cycle assessments across 24 major countries further confirm that the proposed RCC platform remains robust under diverse supply scenarios. These findings demonstrate how molecular-level engineering can be translated into a globally deployable strategy for carbon-negative manufacturing.</jats:p>