A technique is presented for the direct measurement of octanol-water partition coefficients by HPLC. The method involves running solutes in octanol-saturated water as the mobile phase against water-saturated octanol entrained on an inert support. Log P correlates linearly with log tc for a number of standards. The measurable range in log P (so far) is -0.3 to +3.7. A critical review of chromatographic methods in Hansch analysis is given.

Azine π‐values are discussed in terms of Δπ, which is the difference in π‐value from that expected for benzene. It is shown that Δπ is close to zero for alkyl and most halogen groups, but for polar substituents capable of hydrogen bonding may be as high as ca. 1.6. Except for peri‐positions, these Δπ‐values may be correlated by a set of equations specific for different types of substituent position and containing terms which separately parameterise proton donor and acceptor ability. The rationale behind this treatment is justified in terms of the nature of the octanol‐water partitioning process and the manner in which electronic effects are expected to operate, in this context and that of the individual molecule. These equations will correlate those compounds for which they are expected to be applicable to Δ logP ± 0.12. Other topics discussed include: reasons for deviations among “irregular” substituents; the special problems of peri‐positions; multi‐substitution; and some consequences of this analysis for other types of compound.

It is shown that πPh has a minimum theoretical value of ca. 1.7 in aromatic rings. This value is attained whenever phenyl is forced into a highly non‐planar conformation, as may happen through peri‐interactions or the heavy solvation of sterically hindering neighbouring groups. In the absence of such hindrance πPh will be greater, and can attain values as high as ca. 2.6 when there is no steric hindrance to coplanarity and two hetero‐atoms can be shielded. These and a number of intermediate situations are delineated; estimates of πPh are provided for each distinct case. The effect of phenyl substitution at sp3‐carbon is discussed in similar terms. It is shown that these variations make qualitative chemical sense and can be used for predictive purposes.

No abstract