2
Prediction of Gas-Apolane Partition Coefficients (L
87) via the Solubility-Parameter-Based Method
M. Zare Baniasadi a, M. Mousavi a, A. Nasehzadeh b*
a Department of Chemistry, Shahid Bahonar University, Kerman 76175, Iran
b School of Chemical Engineering and Analytical Science, the University of Manchester, Manchester, UK, M60 1 QD
"maryam Zare" <[email protected]>
Introduction
The distribution of a solute between two phases has been an important subject for theoretical and experimental studies for many years [1-8]. The ratio of the concentration of solutes distributed between two phases is constant and is called partition coefficients.
There are some excellent reviews that gathered the most practical methods for determining partition coefficients [2,3]. Apolane is a branched alkane with chemical formula of C87H176. Weckwerth et al [4] have measured gas-apolane partition coefficient, L87, at 309.2 K and proposed it as a new non polar reference substance in order to be used in many linear salvation energy relationships (LSERs).
XYZ = ( XYZ
)0 + l(log L78 ) + sπ * + aα + bβ
2
+ dδ 2
XYZ can be any thermodynamics or kinetics quantity and the independent variables π 2 ,α 2 , β 2 and δ 2 represent the properties of the individual solutes.
The logL87 as a descriptor may also be used to estimate normal boiling temperature and critical properties of organic compounds, and is very useful in characterizing a wide variety of solute properties, including chromatographic retention and partition coefficient at liquid-gas and solid-gas interfaces and solute-solvent interactions in gas chromatography [4].
Serious difficulties arise when L87 of non-volatile compounds is to be measured. It would be very useful to have a method to predict the accurate gas-apolane partition coefficient of any solute to minimize the number of measurements required measuring logL87 experimentally. A theoretical method based on the Hildebrand–Scatchard solubility parameters of the solute and solvent are applied into a thermodynamic model
2
2
for the prediction of gas-apolane partition coefficients (L87) of different groups of polar and nonpolar compounds. It is demonstrated that reliable values of log L78 of non- volatile compounds can be predicted. The results obtained using this thermodynamic method is in good agreement with experimental[4] and with the results of the other existing methods.
The precision of the prediction of the log L87 criterion was tested on an independent experimental data set, obtaining a correlation coefficient of 0.99. Five characteristic constants of solute are presented which can be used to predict any desired value of gas- apolane partition coefficients.
Keywords: Gas-apolane partition coefficients, apolane, partition coefficient, Hildebrand-Scatchard solubility parameter, Ostwald solubility coefficient
Theory and Method and Results
A general characteristic of solubility-parameter-based models for the activity coefficient is that they have some qualitative molecular interaction explanation, without transition into a true statistical thermodynamic theory. The mixing process of solute and solvent is modeled as an exchange of molecules in similar lattices filled with just solute or solvent molecules. A number of partitioning theories [8] have been based on the concept of cohesive energy density or solubility parameter (δ) after its introduction by Hildebrand and Scatchard [9].
δ =[(∆Hv − RT)/V]1/ 2 where V , ∆H v , R and T
are the molar volume, molar heat of vaporization, universal gas constant and the absolute temperature, respectively.
This work is a continuation of our previous work [10,11] and we present a method for derivation of four characteristic parameters that can be used to predict logL87 of desired co mpound.
87 bi ∆H
i V87 Vi 2 2 (c2,i,s δ 87 + δ
i ) Vi Vi
log L
= −ai +
RT − log
103 RT − RT
δ i
+ δ 87 − 2δ i δ 87 c1,i,s
(δ +
c
δ ) − X i,s − log V +
V −1
87 2,i,s i s s
The method of calculation was applied to groups of homologous compounds that their values of logL78 are experimentally known.
The values of parameters which were obtained by a best fit non-linear regression
analysis are calculated. These constant characteristics values of different sets of
4
homologous were applied to predict the values of log L87 of desired compound. The consistency of our results with the literatures is of interest, and there is a good agreement between our predicted results and those obtained by the others, experimentally.
Figure 1. correlation between experimental and calculated values of logL87 at 298.15.
References
1. M. H. Abraham and P. L. Grellier, R. A. McGill, J. Chem. Soc. Perkin Trans.
2(1987)797
2. A. Leo, Chem. Rev. 1281 (1993) 93.
3. A. Leo, C. Hansch, D. Elkins, Chem. Rev. 525 (1971) 71.
4. J.D.Weckwerth and P.W.M.F.Vitha,and A.Nasehzadeh. Anal .Chem. 1998, 70, 3712.
5. J.W. Grate, B. M. Wise and M. H. Abraham .Anal .Chem.71 (1999)4544.
6. S. K.Poole,C. F. Poole J. Chromatogr.A.845(1999)381.
7. G.Defayes, D.F.Fritz, T.Gorner, G.Huber, C.De Reyff, E.sz.Kovats, Chromatogr .500 (1990)139.
8. P.Havelec, J. G. K. Sevcik, J. Phys. Chem. Ref. Data 25(1996)1483.
9. J. H. Hildebrand and R. L. Scott, the Solubility of non electrolytes, Reinhold, New York, 1950, pp. 119-134.
10. A. Nasehzadeha,*, E. Jamalizadeha, G.A. MansooribJournal of Molecular Structure (Theochem) 623 (2003) 135–143.
11. A. Nasehzadeha,*, E. Jamalizadeha, G.A. MansooribJournal of Molecular Structure (Theochem) 629 (2003) 117–126.
logL87(calc)
0 1 2 3 4 5 6 7
3 2 1 0
y = 1.0125x - 0.0399 R2 = 0.9900 7
6 5 4
logL87(exp)
