Abstract
Polymers of intrinsic microporosity (PIMs) have rigid and contorted backbones that result in inefficient packing and a correspondingly high fractional free volume, enabling these materials to overcome the permeability–selectivity trade-off inherent in membranes. However, as a result of their excess free volume, these polymers suffer from a phenomenon known as physical aging, which results in a decrease in free volume over time as the polymer chains relax toward their equilibrium state, thereby lowering a membrane's overall permeability. Therefore, it is important to understand this phenomenon in microporous materials, especially in industrially relevant form factors such as hollow fibers. In this work, high molecular weight PIM-1 was synthesized at a multi-decagram scale for use in hollow fiber spinning. The spun fibers then underwent solid-state functionalization to yield an amine-functionalized derivative of PIM-1 known as PIM-NH2, providing moieties for hydrogen bonding and modifications to polymer chain mobility. These chemical and structural features can influence physical aging. As a result of their thinner selective layers, both PIM fiber compositions experienced accelerated aging and greater gains in selectivity when compared to thicker dense films. For PIM-NH2 fibers, greater gains in selectivity were observed with aging, especially for CO2-based separations when compared to PIM-1 fibers due to the higher sorption affinity that amines have towards CO2. This work represents an in-depth aging study performed on hollow fiber membranes spun from microporous organic polymers, allowing for a better understanding of physical aging and the long-term performance of PIM-based materials in the hollow fiber geometry.
Original language | English |
---|---|
Article number | 123875 |
Journal | Journal of Membrane Science |
Volume | 722 |
Number of pages | 12 |
ISSN | 0376-7388 |
DOIs | |
Publication status | Published - 2025 |
Bibliographical note
Funding Information:This work was funded by ExxonMobil. The authors would like to thank Dr. Stephen DeWitt and Prof. Ryan Lively for helpful discussions early on during this work and Dr. Michael Ocheje for assistance with GPC.
Publisher Copyright:
© 2025 Elsevier B.V.