Lipophilicity means the tendency of the compound to partition between lipophilic organic phase (immisciple with water) and polar aqueous phase, and value of lipophilicity most commonly refers to logarithm of partition coefficient P (logP) between these two phases. Up to certain limit, compounds with higher lipophilicity have higher permeation across biological membranes (but lower aqueous solubility). Relationship between logP and permeation is however non-linear, permeation decreasing in both low and high end of logP values, and often a logP value of 5 is considered as a upper limit of desired lipophilicity.The partition coefficient is commonly expressed as partition of the compound in two-phase system of octanol and water (buffer), i.e. logP = log10 [in octanol]/[in water]. For ionizable compounds, the partition is changed as a function of pH, this relationship is called as distribution constant (logD), or sometimes called also as an apparent partition coefficient. Whereas logP describes partition of the unionized (neutral) form of the compound, logD describes the total partition of both ionized and unionized form. Thus, for acidic compounds, logD is decreased as a function of increased pH (acids are more ionized and thus more water soluble in higher pH), while for basic compounds logD-value is increased as a function of increased pH (bases are less ionized and thus less water soluble in increased pH). Often when characterizing the compound, logD is studied across the wide pH range. Usually logD at pH 7.4 is considered as an index of the behavior of the compound in plasma.
Traditionally logD (or logP) is measured using a “shake-flask” method, meaning incubating the compound in two-phase system under shaking, and samples collected from both phases after equilibration are analyzed with using analytical methods such as HPLC, LC/MS, or simply by spectrophotometer. Time required to reach the equilibrium is usually more than 3 hours, and for very accurate estimates even 24h equilibrium-times are used. Longer equilibrium times also set some criteria to chemical stability of the compound. Analysis by spectrophotometer is naturally more rapid than by using methods involving chromatography, but the latter ones are able to eliminate errors produced by impurities or degradation products. As concentration differences between two phases may be even 100 000-fold (logD=5), it also means that the analytical method has to have adequate sensitivity and linear range, and that even minor impurities present may have relatively large effect to the result, if having different partition than the actual study compound. The used octanol and buffer has to be presaturated with respect to each others, as otherwise the minor solubility of the phases to each others during the experiment may have large effect to the results (as small droplets of phases within each others may contain relatively high concentrations of the study compound with respect to the surrounding phase). Too high test concentrations should also be avoided, as high amounts of compound with respect to total test volume may cause some problems in a form of saturation of the phase with the higher solubility and precipitation problems. Typically concentrations below 1 mM (in total test volume) are used.
Sometimes also methods based on chromatographic retention times under validated conditions are used for high-throughput estimation of partition coefficient, but these methods are more suitable for comparison of compounds within certain chemical family, as the group of compounds used for validating the relationship between logD and retention time should have as closely similar structure with the study compounds as possible. Due to this reason the approach is rather for in-house studies in a drug discovery/development company having long experience with compounds from certain chemical series, than for outsourced studies for a contract laboratory.