(196a) Multiscale Inverse Design of High-Performance Ionic Liquid Solvents for High-GWP Refrigerant Separation

Hydrofluorocarbons (HFCs) are predominantly used as refrigerants, but have high greenhouse gas potential (GWP). Currently, there are a million kilograms of HFCs that are being used worldwide, which need to be recycled and reused. Recycling these HFCs is important to prevent the release of unwanted high-GWP gases into the environment. However, due to the azeotropic/near azeotropic nature of some of the HFC mixtures, sustainable separation of HFCs using conventional distillation is nearly impossible. To that end, ionic liquids (IL) have shown great potential as a solvent in extractive distillation (ED)-based separation of HFCs [1,2]. The choice of IL plays an important role in ED-based process performance and the attainable separation quality [3]. Most ILs are liquid at room temperature which can be synthesized by combining anions and cations. Theoretically, there are millions of hypothetical ILs that can be constructed, which gives them the name “designer solvents”. To expedite the search for the optimal IL for separating refrigerant blends, it is imperative to construct a multiscale process synthesis, optimization, and material screening framework.

In this work, we propose a novel “inverse design” approach for materials discovery and illustrate the efficacy of this approach in finding the most desired ILs for ED-based separation of HFCs. Our proposed approach is able to decouple the multiscale simultaneous materials-process screening problem into two separate subproblems using feasibility maps [4]. These feasibility maps are obtained by solving the full-scale process synthesis problems for numerous hypothetical ILs, thereby allowing parallel computing and fast identification of the property domains of interest [5]. We extract the optimal domain of thermodynamic and physical properties of ILs based on our recent work where we employ SPICE_ED for ED-based intensified separation schemes of refrigerant blends [3,6]. Next, we obtain a reliable quantitative structure-property relationship (QSPR) of ILs by developing an artificial neural network (ANN). To overcome limited data availability for ANN training, we use COSMO-RS-based simulations to predict the thermodynamic and physical properties that include Henry’s constant, density, and specific heat capacity, among others. After establishing confidence in the derived QSPR by performing cross-validation, we transform the trained ANN model into an discrete optimization problem whose solution provides new IL structures with desired properties that match with the feasibility maps. We continue generating new structures until the ANN-predicted properties match with COSMO-RS predicted properties within sufficient accuracy. We apply the design framework to identify new ILs for separating R-410A (50 wt% mixture of R-32 and R-125). We anticipate that such a dynamically updated closed-loop inverse design procedure could be an exciting pathway for identifying novel molecular structures for separation applications beyond extractive distillations.

Keywords: Inverse design, Machine Learning, Hydrofluorocarbons, Ionic Liquids, Extractive Distillation, Process Synthesis.

References:

[1] Shiflett, Mark B.; Yokozeki, A. Solubility and diffusivity of hydrofluorocarbons in room-temperature ionic liquids. AIChE Journal (2006), 52 (3), 1205-1219.

[2] Baca, K.R., Olsen, G.M., Matamoros Valenciano, L., Bennett, M.G., Haggard, D.M., Befort, B.J., Garciadiego, A., Dowling, A.W., Maginn, E.J. and Shiflett, M.B., 2021. Phase Equilibria and Diffusivities of HFC-32 and HFC-125 in Ionic Liquids for the Separation of R-410A. ACS Sustainable Chemistry & Engineering, 10(2), pp.816-830.

[3] Monjur, M.S., Iftakher, A. and Hasan, M. M. F., 2022. Separation Process Synthesis for High-GWP Refrigerant Mixtures: Extractive Distillation using Ionic Liquids. Industrial & Engineering Chemistry Research, 61(12), pp.4390–4406.

[4] Iyer, S. S.; Hasan, M. M. F. Mapping the Material Property Space for Feasible Process Operation: Application to Combined Natural Gas Separation and Storage. Industrial & Engineering Chemistry Research, 2019, 58(24), pp.10455-10465.

[5] Monjur, M.S., Iftakher, A. and Hasan, M. M. F., Ionic Liquid-based Energy Efficient Separation Pathways for High-GWP Refrigerant Mixtures: Solvent Screening and Process Intensification. Under Review.

[6] Monjur, M.S., Iftakher, A. and Hasan, M. M. F., Sustainable Process Intensification of Refrigerant Mixture Separation and Management: A Multiscale Material Screening and Process Design Approach. Proceedings of the 14th International Symposium on Process Systems Engineering – PSE 2021+, Accepted.