Quantum chemical modeling of humic acid/air equilibrium partitioning of organic vapors

Classical approaches for predicting soil organic matter partition coefficients of organic compounds require a calibration with experimental partition data and, for good accuracy, experimental compound descriptors. In this study we evaluate the quantum chemical model COSMO-RS in its COSMOtherm implementation for the prediction of about 200 experimental Leonardite humic acid/air partition coefficients without calibration or experimental compound descriptors, but simply based on molecular structures. For this purpose a Leonardite Humic Acid model monomer limited to 31 carbon atoms was derived from (13)C NMR analysis, elemental analysis, and acidic function analysis provided in the literature. Altogether the COSMOtherm calculations showed a good performance and we conclude that it may become a very promising tool for the prediction of sorption in soil organic matter for compounds for which the molecular structure is the only reliable information available. COSMOtherm can be expected to be very robust with respect to new and complex compound structures because its calculations are based on a fundamental assessment of the underlying intermolecular forces. In contrast, other empirical models that are also based on the molecular structure of the sorbate have an application domain that is limited by their calibration data set that is often unknown to the user.