Experimental determination of partition coefficients

Alternative experimental methods comprise several modifications of the shake-flask procedure, such as the slow-stirring method, which is especially suitable for highly lipophilic compounds (Bruijn et aL, 1989 Brooke et aL, 1990), and various chromatographic techniques. 

The chromatographic techniques make use of the partitioning between the stationary lipophilic and the mobile hydrophilic phases. They differ from the equilibrium partitioning in flask experiments principally by using a dynamic process, which may be more similar to the distributions in environmental systems. The TLC-derived parameter (Biagi et aL, 1975) provides easy access 

Especially when calibrated with standard compounds relative to the classical shake-flask method, rp-HPLC is the method of choice for the experimental determination of partition coefficients. The liability to large variation in measured log for the highly lipophilic compounds (log 5) subjects the experimental data to major uncertainties and for such molecules calculation is generally preferable (Taylor, 1990). Reported log P 1 (e.g. for some chlorinated dioxins up to 11), corresponding to concentration differences between the aqueous and the 1-octanol phase of about 10 and more are not exact values, but reflect the accuracy of the experimental analytical techniques used they may be subject to major uncertainties of 2 log units. The respective data have to be treated with great caution in QS AR studies, because at best they indicate the order of magnitude of the compounds lipophilicity. 

---

Long P. E. (1978). Experimental determination of partition coefficients for Rb, Sr and Ba between alkali feldspar and silicate liquid. Geochim. Cosmochim. Acta, 42 833-846. 

Long P.E., 1978, Experimental determination of partition coefficients for Rb, Sr and Ba between alkali feldspar and silicate liquid. Ceochim, Comochim. Acts, 42, 833-846. Lugmair G.W. and Marti K., 1978, Lunar initial Nd/ Nd differential evolution of the lunar crust and mantle. Earth Plartet. Set. Lett., 39, 349—3S7,... 

Jenner GA, Foley SF, Jackson SE, Green TH, Fryer BJ, Longerich HP (1994) Determination of partition coefficients for trace elements in high pressnre-temperature experimental ran products by laser ablation microprobe-inductively conpled plasma-mass spectrometry (LAM-ICP-MS). Geochim Cosmochim Acta 58 5099-5103... 

To further illustrate the possible experimental ranges of partition coefficients, examples of experimentally determined partition coefficient ranges are given below. In Table 4-10 data for BHT partitioned between polypropylene and different food and food simulants at 40 °C and in Table 4-11 for d-limonene partitioned between LDPE and various foods and food simulants are shown. The partition coefficient values in these tables are representative examples of the ranges one can expect with polyolefins which are the most common food contact plastics. 

To establish extraction efficiencies for a substance A one first needs to know, or experimentally determine, the partition coefficient, Kp. This constant relates the concentration of solute A in the extraction solvent to the concentration of the solute in the extracted phase, at equilibrium (Eq. 10.35). 

Partition coefficients are relatively simple to measure, at least in principle. However, the devil is in the details and certain aspects demand sufficient attention that rapid throughput methodologies have not yet been successfully developed to cover a broad range of partitioning. Several recent reviews of experimental methods provide an abundance of practical information on the accurate determination of partition coefficients (22, 23). Indeed, some of the motivation to develop reliable predictions of partition coefficient lies in the fact that measurement is often time-consuming and challenging (24). 

Both dynamic melting and equilibrium transport melting require that the porosity when two nuclides are fractionated from one another is similar to the size of the larger of the partition coefficients for the two nuclides. Given the low values of the experimental determinations of Du and Dxh, the porosities required to explain the observational data in these models are generally less than 0.5% and often times closer to 0.1%. Such low porosity estimates have been criticized based on physical grounds given the low estimated mantle permeability derived from the extent of melt connection observed in experiments (Paul 2001). 

Covers, H., Ruepert, C., Stevens, T.]., and van Leeuwen, C.]. Experimental determination and prediction of partition coefficients of thioureas and their toxicity to Photobacterium phosphoreum, Chemosphere, 15(4) 383-393, 1986. Govindarajan, M. and Sabarathinam, P. Effect of some inorganic salts on the ternary liquid-liquid equilibria of the water + 4-methyl-2-pentanone + propanoic or butanoic acid at 35 °C, / Chem. Eng. Data, 42(2) 402-408, 1997. 

Rothwell. J.A.. Day. A.J.. and Morgan. M.R.A. Experimental determination of octanol-water partition coefficients of quercetin and related flavonoids. J. Agric. Food Chem., 53(ll) 4355-4360. 2005. 

Raheim A. and Green D. H. (1974). Experimental determination of the temperature and pressure dependence of the Ee-Mg partition coefficient for coexisting garnet and clinopyro-xene. Contrib. Mineral Petrol, 48 179-203. 

Govers, H., et al., Experimental Determination and Prediction of Partition Coefficients of Thioureas and Their Toxicity to Photobacterium phosphoreum. Chemosphere, 1986 22, 383-393. 

While the variability has been assessed for LAS, little is known for other surfactants. However, the variability of the log Kow-values of LAS indicates that Kow measurements yield conditional partition ratios rather than well defined partition coefficients. On thermodynamic grounds, the partition ratio for ionic surfactants is a parameter depending very much on the electrolyte composition of the aqueous medium and therefore cannot be viewed as a partition coefficient in the sense of Henry s law (Schwarzenbach et al. 1993). In addition, experimental determination of Kow of surfactants could yield erroneous results, due to the emulsifying action of the surfactants in the octanol-water system (Morall et al., 1997). 

In the vast majority of experiments carried out to determine carbonate partition coefficients, the trace component concentration has been allowed to vary substantially during the precipitation process. This experimental approach results in constantly changing solution and solid compositions. The approach that has... 

Govers, H., Ruepert, C., Stevens, T.J., van Leeuwen, C.J. (1986) Experimental determination and prediction of partition coefficients of thioureas and their toxicity to photobacterium phosphoreum. Chemosphere 15, 383-393. 

Mysen, B. O. (1978) Experimental determination of nickel partition coefficients... 

Effect of Tween 80 on TFMS tt Values. Table VI lists octanol/water partitioning data for the 15 substituted TFMS compounds used in this study. Logarithms of the partition coefficients obtained both in the absence (log Px) and presence (log Px) of 0.1% (w/v) Tween 80 are given. (Procedures for determining the partition coefficients Px and Px have been outlined in the Experimental section.) tt, tt, and w" values in Table VI were calculated according to the following relationships ... 

There are different experimental methods for determining the gas-liquid partition coefficients leading to the determination of activity coefficients at infinite dilution y . The most frequently used methods are dynamic and static headspace methods. 

The octanol-water partition coefficient for surfactants can not be determined using the shake-flask or slow stirring method because of the formation of emulsions. In addition, the surfactant molecules will exist in the water phase almost exclusively as ions, whereas they will have to pair with a counter-ion in order to be dissolved in octanol. Therefore, experimental determination of K w does not characterize the partition of ionic surfactants (Tolls, 1998). On the other hand, it has been shown that the bioconcentration of anionic and non-ionic surfactants increases with increasing lipophilicity (Tolls, 1998). Tolls (1998) showed that for some surfactants, an estimated log Kow value using LOGKOW could represent the bioaccumulation potential however, for other surfactants some correction to the estimated log Kow value using the method of Roberts (1989) was required. These results illustrate that the quality of the relationship between log Kow estimates and bioconcentration depends on the class and specific type of surfactants involved. Therefore, the classification of the bioconcentration potential based on log Kow values should be used with caution. 

---