Relationship between olive oil:air, saline:air, and rat brain:air partition coefficients of organic solvents in vitro.

Partition coefficients of 28 volatile organic solvents (13 alkylbenzenes, 10 chlorinated hydrocarbons, and 5 ketones) in olive oil, saline, and rat brain tissue homogenates were measured by equilibration in a closed vial and subsequent gas-chromatographic analysis of headspace air. The values of oil and saline partition coefficients correlate well with previously reported data. Brain partition coefficients were fit to a bilinear equation of the form P(brain:air) = alpha(o)P(oil:air) + alpha(s)P(saline:air) + c. The regression coefficients accurately predicted previously reported rat brain partition coefficients of 19 solvents with distinct physicochemical properties within a factor of 2.5. The combined data set of presently determined and previously reported brain partition coefficients (n = 46) yields tissue-specific regression coefficients for solvent partitioning in rat brain of 0.028 for alpha(O'), 0.845 for alpha(S'), and 0.90 for the intercept, with coefficients of variation amounting to 11%, 4%, and 463%, respectively. The generalized empirical relationship predicts the brain partition coefficients within a factor of 2.5 accurately for 95% of the compounds. The ratios of rat brain concentrations calculated from predicted and measured P(brain:air) and P(blood:air) values were within a factor of 4 for 95% of the compounds. It was concluded that the enlargement of the empirical data set leads to more reliable predictions of rat brain partition coefficients, particularly for the lipophilic volatile organic compounds.