The Measurement of the 1-Octanol/Water Partition Coefficient (Log P) for Thirteen Acrylates

Abstract

The partition coefficient, which describes the distribution of a chemical between 1-octanol and water, is referred as "P" or "Log P". It finds wide application for the purposes of understanding the fate, distribution and potential toxicity of chemicals. This report describes the measurements of Log P for a set of thirteen acrylates. These were performed for the Speciality Acrylates & Methacrylates (SAM) Panel of the Chemical Manufacturers Association (CMA). The set was chosen by SAM to be representative of several classes of acrylates and agreed upon with EPA as appropriate compounds to test. They exhibit a variety of structural characteristics including branching, straight hydrocarbon chain moieties, hydrocarbon chain moieties containing a hetero atom (oxygen), the alcohol grouping as well as variation in the number acrylate groups per molecule. Since measurements of Log Pare very costly and time consuming, methods have been developed in order to predict values of Log P from molecular structure. In these experiments we compared our measurements with the values predicted by the CLOGP program (MedChem Software Release 3.51, Pomona College, CA); a model that is widely used. The data indicate that the predictions are not accurate for cases where the acrylates: contain the {-CH2-CH2-0-} structural unit (DEGA, PDDA, DEGDA, TREGDA & TTEGDA), contain "significant" branching (PETA, NPGDA & TMPTA), and are predicted to be very hydrophobic (DA, XDA & SA). Measurements of Log Ps for the very hydrophic chemicals are considered to be unreliable. The new experimental design developed here shows that this may be due to the influence of the concentration of octanol in the water in the presence of the acrylate on the measured Log P. Predictions for the structurally "simple" molecules (PA, HMGDA) are in excellent agreement with the measurements.