Ternary Nonelectrolyte Systems
only uncertainty concerns the identity of some of the products. It should be noted that, if any of the unidentified products gave some color in the Davis-Morris test, the value of G(-Ade) would be even higher than that which has been quoted above. This possibility cannot be absolutely excluded, but it is very unlikely in view of what is known about the radiolysis process and the unidentified products.
Acknowledgment. The work described above was aided by a Biomedical Sciences Support Grant from the National Institutes of Health to the Research Foundation, Oklahoma State University, and by NSF Grant MPS 75-18967 to L.M.R.
References and Notes (1) R. N. Rice, G. Gorin, and L. M. Raff, J. fhys. Chem., 79, 2717 (1975). (2) G. Gorin, N. Ohno, and L. M. Raff, J. fhys. Chem., 80, 112 (1976). (3) D. T. Kanazir, frog. Nucl. Acid Res. Mol. Biol., 9, 117 (1969).
307 (4) G. Scholes, J. F. Ward, and J. Weiss, J. Mol, BiOl., 2, 379 (1960). (5) J. J. Conlay, Nature(London), 197, 555 (1963). (6) J. J. van Hemmen and J. F. Bieichrodt, Radiat. Res., 46, 444 (1971). (7) J. F. Ward in “Advances in Radiation Research”, Voi. 5,J. T. Lett and H. Adler, Ed., Academic Press, New York, N.Y., 1975, p 207. (8) H. J. Rhaese, Biochim. Biophys. Acta, 166, 311 (1968). (9) J. R. Davis and R. N. Morris, Anal. Biochem., 5, 64 (1963). (10) R. Uliana and P. V. Creac’h, Bull. SOC. Chim. Fr., 2904 (1969). (11) L. F. Cavaiieri and A. Bendich, J Am. Chem. Soc., 72, 2587 (1950). (12) “Radiation Dosimetry: X-Rays and Gamma Rays with Maximum Photon Energies Between 0.6 and 60 MeV” (ICRU Report No. 14), International Commision on Radiation Units and Measurements, Washington, D.C., 1969. (13) H. Stephen and T. Stephen, “Solubilities of Inorganic and Organic Compounds”, Vol. 1, Pergamon Press, Oxford, 1963, part 1, pp 87-88; W. F. Llnke, Ed., “Solubilities”, Vol. 11, American Chemical Society, Washington, D.C., 1965, p 1228. (14) C. A. Mannan, M.S. Thesis, Oklahoma State University, 1972. (15) L. Josefsson, Biochim. Biophys. Acta, 72, 133 (1963); P. Grippo, M. Iaccarino, M. Rossi, and E. Scarano, ibid., 95, 1 (1965). (16) M. D. Cohen and E. Fischer, J. Chem. Soc., 3044 (1962). (17) E,g., M. Poiverelli and R. Teouie, C. R. Acad S d faris Ser. C, 277, 747 (1973).
A Chromatographic Investigation of Ternary Nonelectrolyte Systems Jon F. Parcher” and Theodore N. Westlake Chemistry Department, University of Mississippi, University, Mississippi 38677 (Received May 17, 1976; Revised Manuscript Received November 8, 1976) Publication costs assisted by the University of Mississippi and the National Science Foundation
The “chromatographic”partition coefficients of eleven solutes at infinite dilution were measured as a function of composition in four binary solvents at 45 “C. The binary liquid phases were formed by utilizing a condensable component in the carrier gas. The molecular sizes of the components were disparate and the systems were used to test several nonelectrolytesolution theories. It is shown that Purnell’s “microscopic partitioning theory” was inadequate for most of the systems studied and that the Flory-Huggins theory provides a much better, although imperfect, description of the particular systems investigated.
In the last decade, gas-liquid chromatography has proven to be an accurate and convenient technique for measuring vapor-liquid equilibria data for nonelectrolyte systems. This is especially true in the low concentration ranges (infinite dilution) normally inaccessible by static vapor pressure measurements. The primary data usually obtained are the “chromatographic” partition coefficient and the Raoult’s law activity coefficient of the volatile solute in the stationary liquid phase. Binary liquid phases have also been investigated by several authors. There are two main purposes for this type of investigation. One is the attempt to control the selectivity of a chromatographic column by using mixed liquid phases and the other is an attempt to measure the equilibrium constants for the weak type of complexes, such as charge transfer, r bonding, or H bonding complexes, between the solute and one component of a binary liquid phase. The latter type of investigation has invoked a great deal of controversy, especially involving the comparison of complex formation constant data from chromatography, UV-visible spectrophotometry, and nuclear magnetic resonance spectrometry. Pilgrim and Kellerl reviewed the general area of mixed liquid phases in 1973 and pointed out the multiplicity of empirical and theoretical equations which “fit” the partition coefficient data. Dal Nogare and Juvet2first suggested a simple relation between the “chromatographic” partition coefficient of a
solute, i, in a binary solvent, A + S, and the composition of the liquid phase. Several author^^-^ have used the equation
where (tRi)j is the retention time of solute i in liquid phase j and Wj is the weight of component j in the mixed liquid phase. This equation was shown to be accurate for over 100 chromatographic systems by plotting (tRi)A+Svs. the weight percentage of one of the liquid phase components. Waksmundzki, Soczewinski, and S u p r y n o w i ~ zpro~~~ posed an equation of the form
In (ki)AtS = @A In (IZi)A+ $S In (2) where (kJj is the partition ratio (moles of i in the liquid phase per mole of i in the gas phase) of solute, i, in solvent j, and 0, is the volume fraction of component j in the liquid phase. The partition ratio is equal to the ratio of the retention time of a solute to the retention time of air or another inert solute. Thus, eq 2 predicts a linear relation between In tRi and the volume fraction of one of the liquid phase components. This equation was shown to be accurate for 17 chromatographic systems, although the authors laters modified eq 2 to include a term to account The Journal of Physical Chemistry, Vol. 81, No. 4, 1977
J. F. Parcher and T. N. Westlake
308
for the excess free energy of the solvent mixture. A third form of equation was derived by ReznikovgJo from the Van Laar equation for the activity coefficient as a function of solvent composition. This author used a form of relative partition coefficient and utilized the following expression In
( K R E L ) A + S = @‘A
TABLE I: Physical Characteristics of the Chromatographic Columns
In (KREL)A + @’s In (KREL)S
o/o Liquid Coat- phase ing vol, ml
Liquid phase
Solid support
1.Dinonylphthalate 2. Squalane 3. Squalane
Gas Chrom Z (60/80)
11.5
2.36
Gas Chrom Q (60/80) Gas Chrom Z (60/80)
13.4 10.7
2.59 2.28
(3) where ( 7 ~ )is s the activity coefficient of A in S and uj is the molar volume of component j and the subscript STD is used to denote the standard. 8’.is an “effective” volume fraction. If ui and uSTD are small relative to u,, the correlation term of eq 3 can be neglected and Reznikov showed that this simplified version of eq 3 is valid for 75 chromatographic system. Recently, P~rne11ll-l~ investigated a large number of systems and found that a simple equation was adequate to describe the vast majority of the systems involving two nonvolatile liquid phases over the entire range of composition
(4) where (KR?)~ is the chromatographic partition coefficient of solute i at infinite dilution in solvent j. Purnell postulated a model to explain this type of behavior which involved the idea that the two liquid phases “exist in their macroscopic solutions as microscopically immiscible groups of like molecule^."^^ Martire15 has criticized Purnell’s “Microscopic Partitioning Theory” and proposed an alternate model for complexing s y ~ t e m s . ’ ~Martire , ~ ~ suggested that the association constants measured chromatographically were actually a combination of terms measuring both chemical (complexing)and physical (solution) effects. This resulted in an equation of the form
(5) Vi
QI
=-(us
- UA)
+
+ a2
V A ( X ~ ) S- V A ( X ~ ) A
US @2= -vivA(XA)S (f)j is the infinite dilution activity coefficient of the solute i in solvent j; C A is the molar concentration of the moderator or additive, A, and (xJj is the Flory-Huggins interaction parameter for component i in solvent j. The partition coefficients are related to the activity coefficients and complex formation constant, K1, by the equation
In systems where strong chemical interactions or complex formation is not favored the partition coefficients can be related directly to a1 and cyz. Under certain conditions, Martire’s equations (eq 5 and 6) will reduce to an equation of the same form as Purnell’s equation (eq 4). If ( C q C A + (U2cA2)
