MECHANISTIC QUANTITATIVE STRUCTURE-ACTIVITY RELATIONSHIP MODEL FOR THE PHOTOINDUCED TOXICITY OF POLYCYCLIC AROMATIC-HYDROCARBONS .2. AN EMPIRICAL-MODEL FOR THE TOXICITY OF 16 POLYCYCLIC AROMATIC-HYDROCARBONS TO THE DUCKWEED LEMNA-GIBBA L. G-3

Photoinduced toxicity of polycyclic aromatic hydrocarbons (PAHs) occur s via photosensitization reactions (e.g., generation of singlet-state oxygen) and by photomodification (photooxidation and/or photolysis) of the chemicals to more toxic species. The quantitative structure-activ ity relationship (QSAR) described in the companion paper predicted, in theory, that photosensitization and photomodification additively cont ribute to toxicity. To substantiate this QSAR modeling exercise it was necessary to show that toxicity can be described by empirically deriv ed parameters. The toxicity of 16 PAHs to the duckweed Lemna gibba was measured as inhibition of leaf production in simulated solar radiatio n (a light source with a spectrum similar to that of sunlight). A pred ictive model for toxicity was generated based on the theoretical model developed in the companion paper. The photophysical descriptors requi red of each PAH for modeling were efficiency of photon absorbance, rel ative uptake, quantum yield for triplet-state formation, and the rate of photomodification. The photomodification rates of the PAHs showed a moderate correlation to toxicity, whereas a derived photosensitizatio n factor (PSF: based on absorbance, triplet-state quantum yield, and u ptake) for each PAH showed only a weak, complex correlation to toxicit y. However, summing the rate of photomodification and the PSF resulted in a strong correlation to toxicity that had predictive value. When t he PSF and a derived photomodification factor (PMF; based on the photo modification rate and toxicity of the photomodified PAHs) were summed, an excellent explanatory model of toxicity was produced, substantiati ng the additive contributions of the two factors.