DESC0024774

GlycoSurf, Inc. will build on past successes with metal ion extraction technologies by developing a new membrane-based process for critical minerals and materials (CMMs) recovery from underutilized sources. GlycoSurfĺs rhamnolipids are known to be highly selective chelators for rare-earth elements (REEs); and they have shown to be great leads for commercial extractants of these elements from a variety of sources using ion-floatation or hydrophobic attachment to solid supports. To extend the utility of these materials for elemental separations, we propose to utilize these molecules, and analogs thereof, in membrane extraction technologies. Current membrane technologies (e.g., liquid surfactant membrane, supported liquid membrane, hollow fiber supported liquid membrane, etc.) all broadly use a hydrophobic membrane to partition aqueous and organic solutions. GlycoSurf proposes to create a new class of materials using glycolipid impregnated electrospun polymer inclusion membrane (EPIM). The use of rhamnolipid-based extractants allows for broad design of molecules to tune the hydrophobicity, size, charge, and selective chelation to be applied in a number of these technologies. GlycoSurf and Norther Arizona University (NAU) have collaborated in the past to produce electrospun membranes that incorporate rhamnolipids into the membrane structure. This project will advance these activities at NAU to include higher concentrations of extractant impregnation with new membrane material sets that are more acid stable. EPIMs are a novel class of membranes that are advancing the separation of individual rare earths, by utilizing high concentrations of extractants as carriers across the membrane. However, these extractants are designed for solvent extraction and not membranes. Therefore, the use of machine learning will enable GlycoSurf to design more selective extractants for use in electrospun membranes. GlycoSurfĺs synthesis platform, combined with the U. Tennessee Knoxvilleĺs (UTK) modeling expertise and burgeoning AI technologies, will allow for facile tunable selectivity and deployment of a new and novel approach to selective extraction. UTK will perform computational modeling and design related to separations of REEs, which include quantum chemical calculations, HPC- driven efforts, and AI applications through the REACKT project and topological ligand exploration. Precisely, REACKT (Rare Earth Actinides Constants & logK Thermodynamics) is an AI model designed to predict REE and actinides reaction characteristics. REACKT can focus on selective binding for targeted REE separations based on molecular descriptions in liquid separations. This will aid evaluations of ligand-based separations that can inform interactions with GlycoSurfĺs tailored chelators (involved in the membrane-based separations, ligand-based separations, or methods that include alkyl variations). In addition, binding reactions driving REE separations will be modeled with electronic structure methods as shown in previous work.