Ulrich Wiesner, the Spencer T. Olin Professor of Materials Science and Engineering at Cornell University, along with co-authors Lily Zahl and Fernando A. Escobedo, demonstrated a new ultrafiltration (UF) membrane fabrication technique that can sort molecules by chemical affinity. Published on November 13th nature communicationsthe Cornell engineering team utilized chemically distinct block copolymer micelles (self-assembled polymer spheres) and applied machine learning segmentation to identify patterns within the membrane’s porous structure. This approach enables control of pore surface chemistry and provides a route to create UF membranes with customizable selectivity for complex mixtures such as antibodies, previously limited to size-based separations, potentially revolutionizing industrial filtration processes.
Chemically diverse membranes made from block copolymers
Researchers at Cornell University have developed a new approach to ultrafiltration (UF) membrane creation that moves beyond size-based separation to chemical affinity. By blending different block copolymer micelles (nanoscale polymer spheres), we created porous films with chemically diverse pore surfaces. This is achieved through controlled self-assembly, exploiting neutral and repulsive interactions between micelles. Initial results demonstrate that up to three different block copolymers can be incorporated and pore chemistry can be tailored to specific molecular separations. This is an important advancement for biopharmaceutical manufacturing and beyond.
The key to this innovation lies in identifying the micelle distribution within the separation layer of the membrane. Due to the limitations of imaging, the research team used machine learning to analyze the subtle pore patterns revealed by scanning electron microscopy. This made it possible to map the position of each type of copolymer. Complementary molecular simulations using coarse-grained modeling further revealed the self-assembly rules governing the micellar organization, despite the system’s complexity and being far from equilibrium.
This work builds on previous work on self-assembly of block copolymers previously commercialized by Terapore Technologies. The potential impact is significant. Existing UF manufacturing processes can be adapted by simply changing the “recipe” of the block copolymer used. This promises a paradigm shift in filtration, opening possibilities for affinity separations, smart coatings, and sensitive biosensors. This research was supported by the National Science Foundation and utilized Cornell University’s Materials Research Facility.
Machine learning identifies chemical patterns in micelles
Researchers at Cornell University have demonstrated a route to ultrafiltration (UF) membranes that can separate molecules based not only on size but also on chemical affinity. Published in nature communicationsthe team blended block copolymer micelles (nanoscale polymer spheres) during the fabrication of the membrane. Importantly, machine learning was employed to analyze the obtained scanning electron microscopy images of the pore structure. This enabled the identification of subtle pattern differences that reveal where each micelle type assembles, a task that is not possible with imaging alone, and reveals control over pore chemistry.
This innovation addresses a key limitation of current UF technology, which struggles to distinguish between molecules that are similar in size but have different chemical structures, which is important in biopharmaceutical manufacturing. The researchers controlled the self-assembly and pore chemistry of the micelles by combining up to three different block copolymers. Molecular simulations using a coarse-grained model corroborated the experimental results and explained the micellar organization. This approach built on previous research at Cornell University and led to the startup Terapore Technologies.
This method provides a cost-effective way to obtain chemically diverse membrane surfaces without expensive post-fabrication treatments. Essentially, manufacturers can modify the “recipe” of existing UF processes to achieve affinity-based separations. Beyond filtration, the ability to program pore surface chemistry unlocks the potential for smart coatings and biosensors. This NSF-supported work represents a paradigm shift, moving UF beyond simple size exclusion to functional separation based on molecular recognition.
A new approach revolutionizes ultrafiltration technology
A new ultrafiltration (UF) technology developed at Cornell University promises a revolution in membrane technology, moving beyond size-based separation to chemical affinity. The researchers were able to create porous membranes by mixing block copolymer micelles (nanoscale polymer spheres) during manufacturing. This innovative approach allows the creation of pores with diverse chemical properties, allowing the separation of molecules of the same size and weight but different chemical structures. This is a long-standing challenge in fields such as biopharmaceutical manufacturing.
The key to this breakthrough lies in the control of micelle self-assembly. By combining up to three different block copolymers, the research team demonstrated how neutral and repulsive interactions determine the arrangement of different chemical structures within membrane surface pores. Determining these arrangements proved difficult, requiring hundreds of scanning electron microscopy images and machine learning algorithms to map the micelles’ locations. Molecular simulations using coarse-grained models further reveal the governing rules of this complex self-assembly process.
This method offers significant advantages over current post-manufacturing chemical modification techniques that are prohibitively expensive for industrial scale-up. By leveraging an existing scalable block copolymer process already leveraged by startup Terapore Technologies, companies can create membranes with chemically diverse pores by simply adjusting the “recipe.” This is expected to lead to a paradigm shift in UF operation, opening new possibilities for filtration, smart coatings, and sensitive biosensors.
