Modeling the biodegradability and physicochemical properties of polycyclic aromatic hydrocarbons

Abstract

The biodegradability and physicochemical properties of unsubstituted and methylated polycyclic aromatic hydrocarbons (PAHs) were investigated. The focus was on the development of models expressing the influence of molecular structure and properties on observed behavior. Linear free energy relationships (LFERs) were developed for the estimation of aqueous solubilities, octanol/water partition coefficients, and vapor pressures as functions of chromatographic retention time. LFERs were tested in the estimation of physicochemical properties for twenty methylated naphthalenes containing up to four methyl substituents. It was determined that LFERs can accurately estimate physicochemical properties for methylated naphthalenes. Twenty unsubstituted and methylated PAHs containing up to four aromatic rings were biodegraded individually by Sphingomonas paucimobilis strain EPA505, and Monod-type kinetic coefficients were estimated for each PAH using the integral method. Estimated extant kinetic parameters included the maximal specific biodegradation rate, the affinity coefficient, and the inhibition coefficient. The generic Andrews model adequately simulated kinetic data. The ability of PAHs to serve as sole energy and carbon sources was also evaluated. Quantitative structure-biodegradability relationships (QSBRs) were developed based on the estimates of the kinetic and growth parameters. A genetic algorithm was used for QSBR development. Statistical analysis and validation demonstrated the predictive value of the QSBRs. Spatial and topological molecular descriptors were essential in explaining biodegradability. Mechanistic interpretation of the kinetic data and the QSBRs provided evidence that simple or facilitated diffusion through the cell membranes is the rate-determining step in PAH biodegradation by strain EPA505. A kinetic experiment was conducted to investigate biodegradation of PAH mixtures by strain EPA505. The investigation focused on 2-methylphenanthrene, fluoranthene, and pyrene, and their mixtures. Integrated material balance equations describing different interaction types were fitted to the depletion data and evaluated on a statistical and probabilistic basis. Mixture degradation was most adequately described by a pure competitive interaction model with mutual substrate exclusivity, a fully predictive model utilizing parameters estimated in the sole-PAH experiments only. The models developed in this research provide insight into how molecular structure and properties influence physicochemical properties and biodegradability of PAHs. The models have considerable predictive value and could reduce the need for laboratory testing.