The separation of complex liquid hydrocarbon mixtures was recently demonstrated using the glassy polymer SBAD-1, showing that small molecule fractionation is possible by such organic membrane materials. Here, we develop a framework that will enable workable predictions of permeate flux and composition in complex hydrocarbon liquids through intrinsically porous glassy polymers. The predictions are made by incorporating experimentally-derived unary sorption and diffusion parameters in a Maxwell-Stefan framework coupled with multicomponent sorption models and various distinct diffusion phenomena. Across the range of sorption and diffusion phenomena considered, both the conventional Flory-Huggins model and the proposed Langmuir + Flory-Huggins sorption model combined with a simple average guest diffusivity or a more complex free-volume theory-based transport resulted in the lowest prediction error for three chosen multicomponent separations. The proposed Maxwell-Stefan framework simply requires pure component transport parameters to allow a fast approximation of the separation of multicomponent liquid hydrocarbon feeds that can potentially be extended to more complex feeds such as crude oil fractions.

最近使用玻璃状聚合物 SBAD-1 证明了复杂液态烃混合物的分离，表明这种有机膜材料可以进行小分子分馏。在这里，我们开发了一个框架，该框架将能够通过本质上多孔的玻璃状聚合物对复杂烃液体中的渗透通量和成分进行可行的预测。这些预测是通过在 Maxwell-Stefan 框架中结合实验得出的一元吸附和扩散参数以及多组分吸附模型和各种不同的扩散现象来进行的。在考虑的吸附和扩散现象的范围内，传统的 Flory-Huggins 模型和提议的 Langmuir + Flory-Huggins 吸附模型结合简单的平均客体扩散率或更复杂的基于自由体积理论的传输，对三个选定的多组分分离产生了最低的预测误差。提议的 Maxwell-Stefan 框架只需要纯组分传输参数，以允许快速近似分离多组分液态烃原料，这些原料可能扩展到更复杂的原料，例如原油馏分。