Environ. Sci. Technol. 1985, 19, 628-632
Aqueous Solubility and Octan-1-01 to Water Partition Coefficients of Aliphatic Hydrocarbons Michael Coates, Des W. Connell," and Diane M. Barron
School of Australian Environmental Studies, Griffith University, Nathan, Old. 41 11 Australia The aqueous solubility (S) and octanol-water partition coefficients (P)of homologous series of n-, 2-methyl-, and 3-methylalkanes, as well as 1-alkenes, have been determined by extrapolation of known results, direct measurement, and high-pressure liquid chromatography (HPLC). Long-term equilibration experiments, used to reduce aggregate formation, indicated that n-dodecane and ntetradecane have S values in agreement with those obtained by extrapolation of the data on lower members. HPLC data from reverse-phase columns further validated the use of extrapolation. By use of published values for P and S for lower n-alkanes, the relationships between log P, log S, and N, were obtained. Cochromatography of n-alkanes with members of the other series then allowed these relationships to be determined for the 2- and 3methylalkanes and the 1-alkenes. The derived S values were in reasonable agreement with values from previous work and those obtained by extrapolation. The log P values have not been previously determined for these compounds. W
Introduction The hydrocarbons are important petroleum-derived contaminants in the world's oceans (1-3). Little work has been carried out on the hydrocarbons relating their physicochemical properties to their bioaccumulation by aquatic organisms. However, as a general rule, the bioaccumulation of persistent lipophilic compounds by aquatic organisms is positively correlated with their hydrophobicity, particularly as measured by the octan-1-01 to water partition coefficient, P (e.g., see ref 4-6). For a wide range of lipophilic compounds, P is inversely proportional to s,the aqueous solubility (7). Thus, to develop an understanding of the bioaccumulation of the hydrocarbons, data on their S and P values are required. McAuliffe (8-10) has determined the aqueous solubilities of C1-CI2 n-alkanes and C5, cg, and C, 1-alkenes and found linear relationships between log S and N , (carbon number) for each group. Later, Price (11) found a linear relationship between log S and N , for Cg-cB 3-methylalkanes. By extrapolation of McAuliffe's data @-lo), the theoretical values for the S values of the n-alkanes above C12can be obtained. However, S values obtained in this way and those of other workers (12,13) determined experimentally are in substantial disagreement. McAuliffe (10) has suggested that errors occur due to the incomplete removal of clusters or aggregates of hydrocarbon during the solubility experiments. In addition, solubilities of the longer chain length compounds are very low, making the concentrations difficult to measure. Recently, Hutchinson et al. (14) have reported log P values for cg, C,, Cl0, C12,and C14n-alkanes which exhibit a linear relationship between log P and N,. In recent years high-pressure liquid chromtography (HPLC) using reverse-phase columns has been shown to be a valuable method for estimation of P for lipophilic compounds from their retention times. Our objectives in this present work were to determine the S and P values for various homologous series of aliphatic hydrocarbons, particularly those members with chain lengths greater than 628
Environ. Sci. Technol., Vol. 19, No. 7, 1985
(210, utilizing the HPLC technique and other methods as appropriate. It was expected the HPLC technique would eliminate many of the difficulties due to the presence of aggregates or clusters that occur with more direct measurement techniques. Further, it was expected that the HPLC method would allow the determination of P values for the higher aliphatic hydrocarbons whose insolubility in water (log P > 6) prevents use of the classic shake-flask method. By using homologous series, we expected to be able to make use of any systematic relationships between physical properties and carbon number to confirm or estimate values.
Experimental Section The hydrocarbons were obtained from Analabs Co., North Haven, CT. Hexane solvent was twice distilled in an all-glass fractional distillation apparatus with an efficiency of about 10 theoretical plates using a reflux ratio of about 10 to 1. Water was deionized by mixed bed resins and glass distilled. Analytical-grade absolute ethanol was used for HPLC after filtration through a 0.45-pm Millipore filter for organic solvents. Aqueous Solubility Experiments. Short-term experiments were performed by shaking 150 pL of n-alkane with 1L of water for 16 h. The aqueous layer (200 mL) was then filtered through 0.45-pm Millipore filters with gentle suction (12, 13, 15) and extracted with hexane (3 X 20 mL). An internal standard consisting of 2 pg of the n-alkane one carbon longer than the n-alkane under investigation was added to the first extractant. The extract was then concentrated to