Development and Optimization of the Coating Method for Multichannel Carbon Molecular Sieve Membranes

The transition toward a hydrogen-based economy has increased the need for efficient, scalable, and cost-effective gas purification technologies. Carbon molecular sieve membranes (CMSMs) represent a highly promising alternative to conventional separation methods. However, their practical application is limited by fabrication complexity, scale-up limitations, and the cost of ceramic supports. Multichannel geometries can improve scalability by increasing the active surface area per unit volume and enhancing mechanical stability. However, this configuration introduces additional challenges in depositing a homogeneous, defect-free, and selective carbon layer on the internal surface of the channels. This work was carried out within the framework of the European HERMES project and focuses on the development and optimization of coating strategies for the preparation of multichannel supported CMSMs. The Novolac-based coating solution was deposited onto the ceramic supports, followed by drying and carbonization, to obtain the final selective carbon layer. Different coating approaches were investigated, including multilayer deposition, simulated and conventional dip-coating, and a funnel-assisted coating configuration. Particular attention was also paid to the drying step as it may play a key role in the final membrane performance. The funnel-assisted coating method in a 3-cycle configuration combined with an extended drying time led to promising results. In particular, membrane 27_CM_100_C exhibited clear molecular sieving behaviour, reaching H2/N2 and H2/CH4 ideal selectivities of 56.97 and 107.80, respectively. However, this performance was not consistently reproduced under identical conditions, suggesting that the overall fabrication process is highly sensitive to small variations in each stage, thus highlighting the need for a more systematic optimization of the entire fabrication procedure, to improve reproducibility and support the future scale-up of multichannel CMSMs.