Adsorption of Polycyclic Aromatic Hydrocarbons (PAHs) on Rhizopus Oryzae Cell Wall: Application of Cosolvent Models for Validating the Cell Wall-Water Partition Coefficient.

The cell wall-water partition coefficients (K_w) of polycyclic aromatic hydrocarbons (PAHs) on fungal cell wall were rarely studied. The K_w of naphthalene, fluorene, phenanthrene and pyrene were determined for Rhizopus oryzae cell wall by controlling the volume fraction methanol (f) ranging from 0.1 to 0.5. Partition coefficients of naphthalene, fluorene and phenanthrene in pure water systems were also determined. Five cosolvent models for the effect of methanol concentration on the partition coefficient (K_m) between cell wall and water/methanol mixture were employed for extrapolating the K_w in pure water. All extrapolated K_w except the mole fraction model (MF) were higher than the directly measured K_w. The mean percentage deviation (MPD) and root mean square error (RSE) indicated that the log-linear model (LL) and the Bayesian model were most robust. The K_w of pyrene were extrapolated by different cosolvent models and predicted using a quantitive structure property relationship (QSPR) model constructed with various cosolvent models extrapolated K_w values of naphthalene, fluorene and phenanthrene. Comparison of various K_w of pyrene generated from extrapolation and the QSPR model, combined with MPD and RSE, revealed that the performance of the LL and Bayesian models were the best among all five tested cosolvent models. This work suggests that the cell wall of Rhizopus oryzae plays an important role in the partitioning of PAHs in the mycelium during bioremediation because of the high K_w of fungal cell wall and that the Bayesian regression is a potential tool for cosolvent modeling.