Advancing membrane carbon capture in liquefied natural gas-fueled ships via upper bound theory and multi-objective optimization for CO2/O2 separation

The commitment of the shipping industry to reduce greenhouse gas emissions has spurred interest in onboard carbon capture solutions for liquefied natural gas (LNG)-fueled ships, which generate exhaust with low CO₂ and high O₂ contents. Membrane-based carbon capture is a promising alternative to bulky amine scrubbing. However, the critical challenge of CO₂/O₂ separation in O₂-rich exhaust has been largely overlooked in prior studies. Herein, we leveraged the upper bound theory to evaluate the performance of advanced membrane materials for a CO₂/O₂ gas pair, thus integrating this into a multi-objective optimization framework for system design. New CO₂/O₂ upper bound correlations were derived from established CO₂/N₂ and O₂/N₂ relationships to define several performance scenarios for next-generation membranes. Pareto optimization was then applied to identify the optimal trade-offs between the specific energy consumption (SEC) and membrane area. The results demonstrated a clear SEC–area trade-off: achieving an ultralow SEC required large membrane areas, whereas smaller membrane systems incurred higher SEC. Improving CO₂/O₂ selectivity emerged as a pivotal factor. In an upper bound analysis, membranes approaching to the current CO₂/O₂ performance limits achieved SEC values of approximately 3.5 GJ/ton LCO₂. Only a highly selective hypothetical membrane (CO₂/O₂ selectivity ≈ 20, roughly double of the current values) reduced the SEC to <3.15 GJ/ton. This performance surpassed that of the conventional amine-based capture systems and met stringent energy targets. These findings underscore the novelty of combining upper bound material insights with process optimization. We identified specific membrane performance targets required for efficient shipboard CO₂ capture to guide future membrane development for maritime carbon capture applications.