Adsorption-Based Separation of Near-Azeotropic Mixtures—A Challenging Example for High-Throughput Development of Adsorbents

Abstract

Adsorption of gas mixtures is central to adsorption-based gas separations, and the number of adsorbate mixture/adsorbent systems that exist is staggering. Because examples of machine learning (ML) models predicting single-component adsorption of arbitrary molecules in large libraries of crystalline adsorbents have been developed, it is interesting to determine whether these models can accurately predict mixture adsorption. Here, we use molecular simulations to generate mixture adsorption data with a set of 12 near-azeotropic molecules in a diverse set of MOFs. These data provide a challenging example for any method to rapidly predict mixture adsorption in MOFs. We combine a previous ML single-component isotherm model with ideal adsorbed solution theory (IAST) to make predictions that can be compared directly with molecular simulation data for these adsorbed mixtures. This combination of ML and IAST illustrates the scope that is available with these methods, but the accuracy of the resulting predictions is disappointing. By examining the same examples with IAST based on minimal molecular simulation data for single-component isotherms, we show that having an accurate description of adsorption in the dilute loading limit is critical to being able to accurately predict mixture adsorption. This observation points to a useful direction for future work developing robust ML models of adsorption isotherms for diverse collections of molecules and adsorbents.

Efficient Models for Predicting Temperature-Dependent Henry’s Constants and Adsorption Selectivities for Diverse Collections of Molecules in Metal–Organic Frameworks

“…To enrich Tang's data set with more molecules, we considered another set of ∼450 molecules that can be described by the TraPPE force field (FF). 15,26 We restricted our attention to hydrocarbons and organic compounds with oxygen or nitrogen atoms up to C 7 species with a molecular weight of <100 g/mol, giving a set of 130 molecules.…”

“…13 Moreover, the selectivity in this limit is often a good approximation for the adsorption selectivity over a wide range of adsorption conditions. 15 The model introduced above therefore allows us to predict not only the Henry's constant for a wide range of molecules in MOFs but also the adsorption selectivity for mixtures of these molecules. As in the recent work of Gharagheizi et al, we restrict our interest to the challenging case of near-azeotropic mixtures, since these provide a strong test of any model that aims to make quantitative predictions about mixture adsorption.…”

“…15,16 By comparing ML-based predictions and direct molecular simulations for multiple mixtures of near-azeotropic molecules (that is, pairs of molecules with very similar boiling points), it was shown that accurate prediction of singlecomponent adsorption in the dilute loading limit is essential to making accurate mixture adsorption predictions. 15 Nearazeotropic molecular pairs are interesting from an applied point of view because they are much more likely than other molecular pairs to be difficult to separate with traditional methods such as distillation. They are also useful in a modeling context because the adsorption selectivity for near-azeotropic pairs in most adsorbents is likely to be moderate at best.…”

“…By using Henry’s constants obtained from molecular simulations, Tang et al showed that relatively accurate predictions of complete single-component isotherms were possible with simple models for 24 molecules adsorbing in 471 MOFs . Subsequently, ML models for single-component adsorption were trained by using this data set and applied to mixture adsorption by combining these isotherms with ideal adsorbed solution theory (IAST). , By comparing ML-based predictions and direct molecular simulations for multiple mixtures of near-azeotropic molecules (that is, pairs of molecules with very similar boiling points), it was shown that accurate prediction of single-component adsorption in the dilute loading limit is essential to making accurate mixture adsorption predictions . Near-azeotropic molecular pairs are interesting from an applied point of view because they are much more likely than other molecular pairs to be difficult to separate with traditional methods such as distillation.…”

Artificial Neural Network Model in Spatial Analysis of Geographic Information System

“…(2) e accuracy of the model is improved by using geological information to calculate the marriage, and the improved algorithm reduces e predictability of the data sequence improves the classification effect of the network. (3) Finally, through experiments and mutual comparisons, efficient classification can be achieved, reducing the number of interactions, simplifying adaptive calculations, simplifying algorithms, and improving classification accuracy and output data efficiency [5].…”