Science‐guided data analytics for selecting ionic liquid solvents for aromatic extraction

Abstract

AbstractIonic liquids (ILs) are promising solvents for the aromatic extraction process. An attractive characteristic is the existence of hundreds of ILs that exhibit different properties. To identify key properties of IL solvents for an energy‐optimum aromatic extraction, we use process simulation to generate the process datasets for multivariate data analytics with partial least squares, and use science‐guided fundamentals to develop an IL heat load variable (HLV). We consider 16 well‐studied ILs and correlate process steam duty and process variables affecting equipment size to the HLV for ethylene cracker feeds of low aromatic content. For such feeds in an IL aromatic extraction process, 11 of 16 ILs show energy advantage compared with sulfolane solvent with the lowest energy IL process requiring 57% of total energy required for an equivalent sulfolane process. Our results facilitate the IL solvent selection for pilot tests and subsequent commercialization of an IL aromatic extraction process.