Nature Energy published a study on 7 October 2026 describing all-organic mixed-matrix membranes built with single-crystal covalent organic frameworks (COFs) for post-combustion carbon capture — a materials route that, if it scales, targets the energy cost base of the capture stack rather than the solvent chemistry that dominates installed amine plants today.

The paper (DOI: 10.1038/s41560-026-02149-9), received 20 January 2026 and accepted 17 September 2026, comes from a 17-author team led by Tianjin University's School of Chemical Engineering and Technology. Co-first authors are Hanze Ma and Shilin Guo; corresponding authors are Guangwei He and Zhongyi Jiang.

What does the work actually claim?

The journal's reference list frames the engineering problem the paper attacks. Membrane separation competes with chemical absorption for post-combustion CO2 capture — a comparison Giordano, Roizard and Favre worked through in a 2018 life-cycle assessment in the International Journal of Greenhouse Gas Control, and one that Merkel, Lin, Wei and Baker framed as a membranes opportunity back in 2010 in the Journal of Membrane Science.

The binding constraint is the permeability–selectivity trade-off formalised in Robeson's upper bound (2008) and redefined for CO2/N2 by Comesana-Gandara and colleagues in 2019. Mixed-matrix membranes — polymer films loaded with porous fillers — are the standard attempt to break that trade-off, and the Tianjin group's differentiator is the filler: single-crystal COFs, an entirely organic crystalline phase, dispersed in an all-organic matrix.

The article body and its supplementary information carry the supporting data; the published abstract page discloses methodology, funding and authorship but not headline performance figures. Those sit behind the paywall with the source-data files.

Who is behind it?

The institutional footprint matters for scale-up watchers:

• Tianjin University — four linked labs: the Ministry of Education Key Laboratory for Green Chemical Technology, the State Key Laboratory of Synthetic Biology, the Haihe Laboratory of Sustainable Chemical Transformations and the Zhejiang Institute in Ningbo.
• École Polytechnique Fédérale de Lausanne — Kumar Varoon Agrawal's Laboratory of Advanced Separations in Sion, Switzerland, a group with a track record in zeolite nanosheet and graphene membranes for CO2 capture, including pyridinic-graphene membranes reported in Nature Energy (2024) and Nature Sustainability (2025).
• University of Southern California and Ningbo University and Tiangong University — co-authors on processing and materials characterisation.

Funding came from three National Natural Science Foundation of China grants (22278301, U22A20412, 22338011), the Energy Revolution S&T Program of the Yulin Innovation Institute of Clean Energy (YIICE E411030316) and Liaoning Binhai Laboratory (LBLD-2024-03).

The authors declare no competing interests. Peer review was handled by Nature Energy with Mariolino Carta as the named reviewer.

Why the filler chemistry matters

The single-crystal COF route distinguishes itself from the metal-organic framework (MOF) fillers used in most high-performing mixed-matrix membranes — the amine-functionalised MOF work of Ghalei et al. (Nature Energy, 2017) and Datta et al. (Science, 2022). An all-organic membrane removes metal content from the composite, which the authors position as relevant to processing and deployment.

Recent COF literature the paper draws on includes single-crystal 2D COFs (JACS, 2023), functional COF microspheres via dynamic linker exchange (Advanced Materials, 2024) and organic solvent reverse osmosis with COF membranes for C6 alkane isomer separation (JACS, 2025) — a separation family adjacent to the CO2/N2 split at issue here.

What happens next

Nothing in the publication supports a capacity, cost or timeline claim; this is a peer-reviewed materials result, not a bankable project. The milestone to track is whether the single-crystal COF membranes hold their separation performance in module-scale testing against the Robeson upper bound benchmarks the field uses — and whether the Yulin clean-energy programme that co-funded the work converts the result into a demonstration unit.