Solvent regression equations

It is known I0.43.53 ° -62> that the partition coefficients of the same solutes in various solvent biphasic systems are interrelated according to the so-called solvent regression equation ... 

Leo et al.53,62) stress that the intercept bj values of the solvent regression equations are clearly related to the extent to which water is dissolved in the organic phase of the partitioning system. They found 53,62) that the b, coefficient value is related to the above parameter as follows ... 

An analysis of the published partition coefficients measured in various solvent systems for homologous series of fatty acids, aliphatic alcohols, and amines was performed by Zaslavsky et al. 63) in order to clarify the physical meaning of the a. and bj coefficients in the solvent regression Equation (8). The partition coefficients for... 

The most comprehensive set of solvent regression equations is that given by Leo and Hansch [27] and repeated in the subsequent publication by Leo, Hansch, and Elkins (28). Their original equations were written with log KSw as the dependent variable but are restated here in the following form ... 

Thirty-one such equations are provided, allowing Kow to be calculated if a value of KSw for the solute is available with one or more of approximately twenty different solvents. A modified solvent regression equation, developed by Seiler [41J, is provided if the K3W value is for the cyclohexane/water system. Several solvent regression equations are also given by Rekker [39J, but most of them involve the use of special fragment constants or correction factors and are thus slightly more difficult to use. Rekker s equations are not included in this chapter. 

Solvent Regression Equations. The selection of the appropriate solvent regression equation sometimes depends upon the nature of the solute. Table 1-8 lists a number of solute classes in two basic groups A (hydrogen donors) and B (hydrogen acceptors). Table 1-9 provides values of a and b for the basic set of solvent regression equations (Eq. 1-7 to -37), all of which are of the form shown in Eq. 1-6. If the solute (the chemical for which Kow is to be calculated) is listed under Group A or B in Table 1-8 and if the solvent (associated with the available KSw value) is one of those listed in the first two sections of Table 1-9, then a choice between two equations must be made. For example, if a value of Ksw is available from the xylene/water system, one must choose between Eqs. 1-10 and 1-21. The choice depends on where the solute is listed in Table 1-8 — e.g., Eq. 1-10 would be used if the solute were an alcohol, and Eq. 1-21 would be used if it were an ether. 

Seiler [41] has proposed a modified solvent regression equation to cover the cyclohexane/water — octanol/water calculation ... 

If the solvent is one of those in Set C" in Table 1-9, select the appropriate solvent regression equation from this group. Substitute the given values of a and b, along with log KHW, in the generalized equation shown at the head of Table 1-9 (also shown as Eq. 1-6) and solve for log Kow-... 

Alternative approaches to the calculation of log P values from fragment contributions comprise solvent regression equations - extrapolations from... 

Marriott and Topsom have recently developed theoretical scales of substituent field and resonance parameters. The former correspond to the traditional inductive parameters but these authors are firm believers in the field model of the so-called inductive effect and use the symbol The theoretical substituent field effect scale is based on ab initio molecular orbital calculations of energies or electron populations of simple molecular systems. The results of the calculations are well correlated with Op values for a small number of substituents whose Op values on the various experimental scales (gas-phase, non-polar solvents, polar solvents) are concordant, and the regression equations are the basis for theoretical Op values of about 50 substituents. These include SOMe and S02Me at 0.37 and 0.60 respectively, which agree well with inherent best values in the literature of 0.36 and 0.58. However, it should be noted that a, for SOMe is given as 0.50 by Ehrenson and coworkers . 

A good illustration is provided [53] by fluorenone (FL) in a group of n = 19 solvents, which possesses both donor and acceptor groups (see structural scheme of FL in Fig. 6), the multiple regression equations were obtained as ... 

Scheme 7.1, relevant to an amphiprotonic solvent). Using Eq. (7.2), the multiple linear regression equation for the fluorescence maximum (expressed in 103 cm-1) is... 

The regression equations were established for data in 11 alcohols as solvents and were used to assess the peculiar behaviour of another 15 orf/zo-substituents in respect of conformational effect and intramolecular hydrogen-bonding143,145. Flere we are concerned with assessing the situation for o-N02. We first give as an example the regression for 2-methoxyethanol as solvent ... 

The application of ab initio molecular orbital theory to suitable model systems has led to theoretical scales of substituent parameters, which may be compared with the experimental scales. Calculations (3-21G or 4-31G level) of energies or electron populations were made by Marriott and Topsom in 1984164. The results are well correlated with op (i.e. 07) for a small number of substituents whose op values on the various experimental scales (gas phase, non-polar solvents, polar solvents) are concordant. The nitro group is considered to be one of these, with values 0.65 in the gas phase, 0.65 in non-polar solvents and 0.67 in polar solvents. The regression equations are the basis of theoretical op values for about fifty substituents. The nitro group is well behaved and the derived theoretical value of op is 0.66. 

This pH definition for non-aqueous and mixed solvent systems is practically the same as that for aqueous solutions (Section 6.2.1). Thus, if a pH standard is available for the solvent or mixed solvent under study, the glass electrode is calibrated with it and then the pH of the sample solution is measured. The pHRVs values for 0.05 mol kg-1 KHPh have been assigned to aqueous mixtures of eight organic solvents (see 5 for pHRVs at 25 °C). Although they are for discrete solvent compositions, the pHRVs in between those compositions can be obtained by use of a multilinear regression equation [14b],... 

In Charton s work pKa data for 4-X-substituted-bicyclo[2.2.2.]octane-l-carboxylic acids in 50% w/w EtOH-H20 are the basis for primary molecular skeleton, the substituent is somewhat remote from the acidic centre and the geometry for 4-X and 1-COOH closely resembles that for groups in a 1,4-disubstituted benzene)81. However, the pKa value for the vinyl-substituted acid in this solvent was not available, so the 07 value for the vinyl group was calculated by substituting the pKa value for vinylacetic acid in water at 25 °C (4.352) in the regression equation 10 ... 

A recent re-examination of the 19F NMR data for me/a-subsl.ihil.cd fluorobenzenes in hydrocarbon solvents by Hansch, Leo and Taft92 led to a slightly different regression equation. When this was applied to the vinyl group, o/ was found to be 0.07, in fair agreement with the reactivity-based value of 0.10. 

The multiparameter treatment of solvent effeets ean be criticized from at least three complementary points of view. First, the separation of solvent effects into various additive contributions is somewhat arbitrary, since different solute/solvent interaction mechanisms can cooperate in a non-independent way. Second, the choice of the best parameter for every type of solute/solvent interaction is critical because of the complexity of the corresponding empirieal solvent parameters, and because of their susceptibility to more than one of the multiple facets of solvent polarity. Third, in order to estabhsh a multiparameter regression equation in a statistically perfect way, so many experimental data points are usually necessary that there is often no room left for the prediction of solvent effects by extrapolation or interpolation. This helps to get a sound interpretation of the observed solvent effeet for the process under study, but simultaneously it limits the value of such multiparameter equations for the chemist in its daily laboratory work. 

---