Further discussion of the octanol/air partition coefficient Koa as a correlating parameter for gas/particle partitioning coefficients

The pure-compound liquid vapor pressure (p(L)(o)) has been used extensively as a log-log correlating parameter for gas/particle partitioning constants. When the partitioning to the aerosol particle phase involves partitioning into an organic material (om) phase, the success of p(L)(o) in this context relies on compound-to-compound constancy of the molecular activity coefficient zeta(om) in the om phase. While zeta(om) for a particular type of aerosol may indeed tend to remain constant within a given compound class, significant class-to-class variations in zeta(om) are certain. This paper discusses the theory underlying the advantages of using the octanol/air partition coefficient K-oa as a correlating parameter for gas/particle partitioning constants. In particular, it is far more likely that different compounds will exhibit similar values of the ratio zeta(oct)/zeta(om) than it is that they will exhibit similar values of zeta(om). The conclusion that the octanol-based partition coefficient K-oa has significant advantages over the pure-compound partition coefficient p(L)(o) when parameterizing gas/particle partitioning has a direct analog in the selection of the octanol/water partition coefficient K-ow as a correlating parameter for soil/water partition coefficients like the organic carbon/water partition coefficient K-oc. The nature of this analogy is discussed in detail. The acknowledged success of log K-ow in this regard presages wide application of log K-oa as a correlating parameter for gas/particle partition coefficients. (C) 1998 Elsevier Science Ltd. All rights reserved.