Creation of Well-Defined "Mid-Sized" Micropores in Carbon Molecular Sieve Membranes.

Carbon molecular sieve (CMS) membranes are candidates for the separation of organic molecules due to their stability, ability to be scaled at practical form factors, and the avoidance of expensive supports or complex multi-step fabrication processes. A critical challenge is the creation of "mid-range" (e.g., 5-9 Å) microstructures that allow for facile permeation of organic solvents and selection between similarly-sized guest molecules. Here, we create these microstructures via the pyrolysis of a microporous polymer (PIM-1) under low concentrations of hydrogen gas. The introduction of H2 inhibits aromatization of the decomposing polymer and ultimately results in the creation of a well-defined bimodal pore network that exhibits an ultramicropore size of 5.1 Å. The H2 assisted CMS dense membranes show a dramatic increase in p-xylene ideal permeability (≈15 times), with little loss in p-xylene/o-xylene selectivity (18.8 vs. 25.0) when compared to PIM-1 membranes pyrolyzed under a pure argon atmosphere. This approach is successfully extended to hollow fiber membranes operating in organic solvent reverse osmosis mode, highlighting the potential of this approach to be translated from the laboratory to the field.

Data-driven predictions of complex organic mixture permeation in polymer membranes.

Membrane-based organic solvent separations are rapidly emerging as a promising class of technologies for enhancing the energy efficiency of existing separation and purification systems. Polymeric membranes have shown promise in the fractionation or splitting of complex mixtures of organic molecules such as crude oil. Determining the separation performance of a polymer membrane when challenged with a complex mixture has thus far occurred in an ad hoc manner, and methods to predict the performance based on mixture composition and polymer chemistry are unavailable. Here, we combine physics-informed machine learning algorithms (ML) and mass transport simulations to create an integrated predictive model for the separation of complex mixtures containing up to 400 components via any arbitrary linear polymer membrane. We experimentally demonstrate the effectiveness of the model by predicting the separation of two crude oils within 6-7% of the measurements. Integration of ML predictors of diffusion and sorption properties of molecules with transport simulators enables for the rapid screening of polymer membranes prior to physical experimentation for the separation of complex liquid mixtures.