Lumped approach for sorption of organic contaminant mixtures to simulated aquifer sorbents

Abstract

In this study, the feasibility and accuracy of using a lumped approach to describe the sorption of 12 neutral organic compounds (NOCs), 6 polar NOCs (acetone, 2-butanone, 2-hexanone, phenol, p-cresol, and 2,4-dimethyl phenol) and 6 nonpolar NOCs (1,2,4-trichlorobenzene, 1,4-dichlorobenzene, chlorobenzene, m-xylene, toluene, benzene), in mixtures to simulated aquifer sorbents with fractions of organic carbon (foc) ranging from 0.005 to 0.221 % were evaluated. The lumped approach consists of grouping one or more components in mixtures of NOCs into a fewer number of pseudocompounds that exhibit similar behavior with respect to a process being considered (e.g., sorption). Batch equilibrium sorption tests (BESTs) were performed to measure the sorption behaviors of the 12 NOCs existing as both single NOCs and in mixtures of NOCs in aqueous solution with each sorbent. The results of the BESTs were analyzed using the Freundlich sorption equation, and the lumped approach was evaluated on the basis of experimentally derived lumping criteria (i.e., Freundlich sorption parameters). Also, the use of the ideal adsorbed solution theory (IAST) with the determined pseudocompounds for predicting the sorption of the 12 NOCs within complex mixtures to simulated aquifer sorbents was investigated. The lumped approach for sorption of mixtures of NOCs to simulated aquifer sorbents was based on cluster analysis from statistics. Each pseudocompound contained components of relatively similar sorption capacities and nonlinearity, and approximated the sorption behavior of the NOC mixtures to within a 95 % confidence level of the Freundlich sorption isotherms for the pseudocompounds. The sorption behavior of the 12 NOCs in mixtures to the hydrophilic mineral surfaces was found to be described adequately on the basis of only four pseudocompounds, with each pseudocompound composed of from 2 to 4 NOCs. However, the number and composition of the pseudocompounds required to describe the sorption of the 12 NOCs in mixtures to humic acid-mineral complexes varied as a function of the foc due to the greater effect of foc on sorption for NOCs with greater hydrophobicity. For a given sorbent, the lumping schemes for the nonpolar NOCs in mixtures were consistent, i.e., independent of the compositions of the NOC mixtures, whereas those for the polar NOCs in mixtures were inconsistent, due to the complex mutual competition for hydrophilic specific sorption sites among the polar NOCs. In addition to the experimentally derived lumping criteria (i.e., Freundlich sorption parameters), the potential for a priori theoretical lumping (i.e., prior to experimental evaluation) based on the quantitative structure-activity relationships of the NOCs was evaluated. The evaluation indicated that the lumping of the NOCs into pseudocompounds for sorption can be described according to the hydrophobicity (ywsat) and the organic carbon partitioning coefficient (Koc) of the NOCs in mixtures. Thus, the pseudocompounds could be determined directly a priori (i.e., without the need for experimentation) simply from the properties of the individual NOCs in mixtures. Finally, based on comparisons of predicted sorption behavior using IAST with pseudocompounds to the measured sorption behavior, incorporating pseudocompounds into IAST to predict sorption of NOCs in complex mixtures was found to be feasible except in the case where the underlying assumptions of IAST (i.e., same access to sorption sites and ideal sorbed phase) became invalid. Nonetheless, further study is required to improve IAST predictions using pseudocompounds by incorporating terms to account for the differential availability of sorption sites and nonideal competition.