Modeling Bisolute Adsorption of Aromatic Compounds Based on Adsorbed Solution Theories

Abstract

A large number of organic contaminants are commonly found in industrial and municipal wastewaters. For proper unit design to remove contaminant mixtures by adsorption, multicomponent adsorption equilibrium models are necessary. The present work examined the applicability of Ideal Adsorbed Solution Theory (IAST), a prevailing thermodynamic model, and its derivatives, i.e., Segregated IAST (SIAST) and Real Adsorbed Solution Theory (RAST), to bisolute adsorption of organic compounds onto a hyper-cross-linked polystyrene resin, MN200. Both IAST and SIAST were found to be less accurate in fitting the experimental bisolute adsorption isotherms than RAST. RAST incorporated with an empirical four-parameter equation developed in this work can fit the adsorbed phase activity coefficients, γ, better than RAST combined with the Wilson equation or the Nonrandom two-liquid (NRTL) model. Moreover, two polyparameter linear free energy relationships were developed for the adsorption of a number of solutes at low concentrations in the presence of a major contaminant (4-methylphenol or nitrobenzene). Results show that these relationships have a great potential in predicting γ of solutes when the adsorbed amounts are dominated by a major contaminant. To the best of our knowledge, this is the first study predicting γ for bisolute adsorption based on molecular descriptors. Overall, our findings have proved a major step forward to accurately modeling multisolute adsorption equilibrium.

How Reliable Is the Ideal Adsorbed Solution Theory for the Estimation of Mixture Separation Selectivities in Microporous Crystalline Adsorbents?

“…Other examples of the failure of the IAST, along with alternative approaches to RAST modeling of non-idealities, are available in the literature. ,− …”

Coupling a Feedforward Network (FN) Model to Real Adsorbed Solution Theory (RAST) to Improve Prediction of Bisolute Adsorption on Resins

“…The single isotherms were fitted with different adsorption isotherm models, including Langmuir, Freundlich, and the quadratic equation used in our previous research. 40 The best-fitting model was selected as the working isotherm, and the best-fitting parameters were then used to calculate the adsorbed amounts in bisolute adsorption. The AST equations (Text S2 and eqs S5−S9) were solved in Matlab 2019a (details of the code and input data setup are given in Texts S2 and S3).…”

“…In our previous research, 40 pp-LFERs helped build MLR models for the activity coefficients in bisolute adsorption under infinite dilute conditions and provided predictions for several bisolute mixtures. The results indicated that pp-LFERs are helpful in capturing interactions between different adsorbed solutes.…”