The evolution of MOF discovery for CO2 capture: From high-throughput screening to AI design and automated laboratories

Y Qiu and L Wang and C Liu and X Zhang and Y Tian and Z Zhou, MATERIALS TODAY, 91, 103-123 (2025).

DOI: 10.1016/j.mattod.2025.10.017

As a predominant greenhouse gas, CO2 has emerged as a critical environmental challenge on a global scale. Metal-organic frameworks (MOFs), with their large surface area, high pore volume, and tunable structure, exhibit extraordinary potential in CO2 capture. Despite advances in CO2 capture, lack of systematic understanding of inherent mechanisms governing framework flexibility and kinetic behavior in the guest-host system has largely impeded the practical applications. Although most existing work is still at the preliminary stage of brute- force material screening with limited integration of machine learning and theoretical calculations to solve these long-standing problems, AI- driven automated experimentation integrated with intelligent design framework has brought promising solutions, making it possible to bridge the gap between structural properties and separation performance. This review summarizes recent advances in machine learning, theoretical methods, and automated laboratories for CO2 capture with MOFs, highlighting the paradigm shift from theoretical computation to AI- driven automated experimentation.

Return to Publications page