Rare-earth elements (which include all lanthanides, scandium, and yttrium) play a key role in many fields including oil refining, metallurgy, electronics manufacturing, and other high-technology applications. Although the available lanthanide resources are enough for current levels of manufacturing, increased future demand for lanthanides will require new, efficient recovery methods to provide a sustainable supply. Membrane adsorbers are promising separation materials to recover lanthanides from high volumes of wastewater due to their tailorable surface chemistry, high binding capacity and high throughput. In this work, membrane adsorbers were synthesized by first using ultraviolet-initiated free radical polymerization to graft a poly(glycidyl methacrylate) (p-GMA) layer to the surface of polyethersulfone membranes. Then, the reactive epoxy groups of the grafted p-GMA were used for the covalent attachment of lysine molecules via a zinc perchlorate-catalyzed, epoxide ring-opening reaction at 35 • C. Changes in membrane surface chemistry throughout the functionalization process were monitored with Fourier Transform Infrared Spectroscopy. The degree of grafting for the p-GMA film was quantified gravimetrically and increased with increasing polymerization time. Equilibrium adsorption experiments were performed for single specie solutions of La 3+ , Ce 3+ , Nd 3+ , Na + , Ca 2+ , and Mg 2+ at pH 5.25 ± 0.25. Lysine-modified membranes showed negligible uptake of Na + , Ca 2+ , and Mg 2+. The maximum capacities modeled by the Langmuir isotherm for La 3+ and Ce 3+ were 6.11 ± 0.58 and 6.45 ± 1.29 mg/g adsorbent, respectively. Nd 3+ adsorbed to the membrane; however, the fit of the Langmuir model was not significant and it adsorbed to a lower extent than La 3+ and Ce 3+. Lower adsorption of the higher charge density species indicates that the primary binding mode is through the amine moieties of lysine and not the carboxylic acid. Dynamic adsorption experiments were conducted with 1 ppm La 3+ feed solutions at different flow rates using either a single membrane or three membranes in series. The dynamic binding capacity at 50% breakthrough was independent of flowrate within the tested range. The low-temperature membrane functionalization methodology presented in this work can be used to immobilize biomolecules with even higher specificity, like engineered peptides or proteins, on membrane surfaces.

Physical aging of glassy polymers leads to a decrease in permeability over time when they are used in membranes. This hinders the industrial application of high free volume polymers, such as the archetypal polymer of intrinsic microporosity PIM-1, for membrane gas separation. In thin film composite (TFC) membranes, aging is much more rapid than in thicker self-standing membranes, as rearrangement within the thin active layer is relatively fast. Liquid alcohol treatment, which swells the membrane, is often used in the laboratory to rejuvenate aged self-standing membranes, but this is not easily applied on an industrial scale and is not suitable to refresh TFC membranes because of the risk of membrane delamination. In this work, it is demonstrated that a simple method of storage in an atmosphere of methanol vapor effectively retards physical aging of PIM-1 TFC membranes. The same method can also be utilized to refresh aged PIM-1 TFC membranes, and one-week methanol vapor storage is sufficient to recover most of the original CO2 permeance.

No abstract