Evaluating coal tar-water partitioning coefficient estimation methods and solute-solvent molecular interactions in...

Area of Science:

• Environmental Chemistry
• Physical Chemistry
• Chemical Engineering

Background:

• Understanding solute partitioning in complex mixtures like coal tar is crucial for environmental risk assessment and remediation.
• Existing models for predicting partitioning coefficients (KCT/w) vary in accuracy and applicability to diverse industrial waste streams.

Purpose of the Study:

• To determine equilibrium partitioning coefficients (KCT/w) for various solutes in an industrial coal tar.
• To evaluate the performance of different modeling approaches (Raoult's law, SPLFER, LSER, SPARC, COSMOtherm) in predicting KCT/w.
• To assess the influence of coal tar composition on solute partitioning behavior.

Main Methods:

• Batch systems were used to measure KCT/w for 41 polar and nonpolar solutes.
• Experimental data were combined with literature values for a total of 69 KCT/w data points.
• Data were analyzed using Raoult's law, single parameter linear free energy relationship (SPLFER), linear solvation energy relationships (LSERs), SPARC, and COSMOtherm models.

Main Results:

• Raoult's law and SPLFER models showed good agreement with experimental log KCT/w values (RMSE of 0.31 and 0.33).
• LSER models provided comparable estimations (RMSE=0.29) and indicated significant hydrogen-bond acceptor properties of the coal tar.
• SPARC and COSMOtherm offered fair predictions (RMSE ≈ 0.64) without empirical parameters, using naphthalene as a surrogate.
• Nonpolar solute partitioning was minimally affected by minor tar components, but hydrogen-bonding solute partitioning varied significantly with polar component content.

Conclusions:

• Raoult's law and SPLFER are effective for predicting partitioning of nonpolar solutes (e.g., PAHs, alkylbenzenes) in certain coal tars.
• The accuracy of simple models for H-bonding solutes depends heavily on the specific composition of the coal tar.
• More sophisticated models or composition-specific adjustments are needed for accurate prediction of hydrogen-bonding solute partitioning in diverse coal tar matrices.