curs by an exchange mechanism which is effective only at distances ca. 10 Triplet-singlet energy transfer, occurring by the long-range dipolt-dipole mechanism, approaches unit probability at ca. 20 A. Diffusion of the triplet during its lifetime enhances both processes, but the probability of energy transfer increases more rapidly as a function of distance. Therefore, oxygen quenching is less significant in this type of system, and will interfere only by quenching the fluorescence of the acceptor. Uses. T + S energy transfer can be utilized in the analysis of compounds with high triplet yields and negligible fluorescence. The method is thus applicable to a wide range of aromatic carbonyls and azines, certain aromatic quinones and amines (13), and alkaloids such as codeine and morphine (14). A possible use of the method would be in metal ion analysis. Paramagnetic ions and heavy atom diamagnetic ions greatly enhance intersystem crossing rates. Complexation
A.
-
(13) S. P. McGlynn, T. Azumi, and M. Kinoshita, “Molecular
Spectroscopyof the Triplet State,” Prentice-Hall Inc., Englewood Cliffs, N. J., 1969, Chapter 6. (14) H. C. Hollifield and J. D. Winefordner, Talanta, 12, 860 (1965).
with an appropriate ligand will thus yield a complex with negligible fluorescence and high triplet yield, (15,16) making the complex amenable to energy-transfer analysis. With appropriate choice of acceptor (to give high +F and J(v)) and an acceptor concentration of >10-4M, a detectability limit of ca. lOPM can be achieved in energy-transfer analysis. The method thus lacks the trace analysis abilities of phosphorimetry. However, for routine analyses at >lO-5M analyte, the energy-transfer method is preferable because of its rapidity, ease of sampling, and relatively simple instrumentation. ACKNOWLEDGMENT
The authors thank Anthony Vaudo for his discussions of the theory of energy transfer in fluid solution. RECEIVED for review August 21, 1969. Accepted November 7, 1969. Work supported through funds provided by the United States Atomic Energy Commission under Contract AT(30-1)-905. (15) R. S. Becker and J. B. Allison, J . Phys. Chem., 2662 (1963). (16) P. Yuster and S. I. Weissman, J. Chem. Phys., 17, 1182 (1949).
Gas Chromatography and Thermodynamics of Divinylbenzene Separations on 4,4‘- Dihexoxyazoxybenzene Liquid CrystaI Walter L. Zielinski, Jr. and David H. Freeman Analytical Chemistry Division, National Bureau of Standards, Washington, D . C . 20234
Daniel E. Martire and Laurence C . Chow Chemistry Department, Georgetown Uniuersity, Washington, D . C. 20007
Gas-liquid chromatography in the nematic region of 4-4‘-dihexoxyazoxybenzene liquid crystal has shown base-line separations for the meta and para isomers of divinylbenzene and ethylvinylbenzene. Thermodynamic treatment of the data has illustrated that the meta isomers have a lower solubility (resulting in shorter retention times) than the respective para derivatives owing to a greater enthalpic requirement for solution in the rod-like ordered solvent. The para/ meta separation factors depict a greater rate of decrease with increasing column temperature in the isotropic liquid region because of the loss of order which is present in the anisotropic nematic region. Data for naphthalene present as a contaminant in divinylbenzene samples, were included in the treatment. Chromatographic extension to preparative scale appears feasible.
DIVINYLBENZENE (DVB) has extensive commercial use as a crosslinking monomer in the preparation of poly(styrene/ DVB) copolymers. Such copolymers serve as important precursors in the production of anionic and cationic exchange resins. I$ is unfortunate, however, that the commercial monomer contains both m- and p-DVB isomers, along with m- and p-ethylvinylbenzene (EVB), traces of naphthalene and as many as twenty additional impurities ( I , 2). A case has 176
0
been established for the use of highly purified DVB isomers by recent reports that synthetic conditions using such monomers are reproducible ( 3 ) and that qualitative ( 4 ) and quantitative (5)infrared studies on the type and degree of crosslinking in high-purity copolymers are tractable. The preparation of ion exchange materials from high-purity copolymers leads to the important premise that such materials should prove of value to future studies of their fundamental properties and to the analytical chemist in the final analysis, in terms of reproducible separations and predictions of chromatographic behavior. Gas chromatographic separations of meta- and para-DVB have been reported in the past on Carbowax 6000 and di-2-
(1) R. E. Hannah, M. L. Cook, and J. A. Blanchette, ANAL. CHEM., 39, 358 (1967). (2) R. H. Wiley and R. M. Dyer,J. Poly. Sci., 2,3153 (1964). (3) R. H. Wiley and T. K. Venkatachalam, ibid.,4, 1982 (1966). (4) D. H. Freeman, W. L. Zielinski, Jr., and W. F. Rittner, Proc. Intern. C o d on Ion Exchange - in the Process Industries.. July- 16-18. 1969, in prkss. (5) National Bureau of Standards, Washington, D. C., unpublished
data.
ANALYTICAL CHEMISTRY, VOL. 42, NO. 2, FEBRUARY 1970
~-
~~
Table I. Gas Chromatographic Retention Data5 Column T (OC)* 96.1 105.3 115.5 125.8 135.3 154.7 a
f0.6 f0.2 f 0.1 f0.1 f 0.4 f 0.1
Carrier flowc cma/min
m-EVB
pEVB
m-DVB
pDVB
Naphthalene
41.17 39.99 38.61 37.21 35.93 33.30
20.37 16.62 13.30 10.72 9.30 7.37
25.27 20.48 16.14 12.85 10.89 8.27
36.67 29.15 22.57 18.15 14.70 11.10
47.60 37.45 28.64 22.30 17.96 12.77
65.97 52.80 40.82 32.24 27.03 19.87
tx' (minutes)d
Mean values of 3 to 9 analyses.
* Column temperatures with standard deviations of the mean of individually monitored temperature measurements. c
Calculated helium flow rates from soap bubble flowmeter. Net retention time: fR' = [ I 8 (compound) - fR (CH,)].
ethylhexyl sebacate (1) and on Bentone-34 (6, 7). Complete base-line separation of these isomers by analytical gas chromatography has been closely approached by Wiley and his coworkers (2, 6), but separation factors greater than 1.1 have not been reported. It has been realized, on the other hand, that excellent separations of meta and para isomers are obtainable using liquid phases which exhibit an ordered molecular arrangement within a defined temperature interval (8); viz., the so-called liquid crystals. This work describes the chromatographic behavior and the related thermodynamic data for the meta and para isomers of DVB and EVB, and for naphthalene on a nematic liquid crystal stationary phase.
I
I
1
I
I
1
-
,?'
-EVE ---DVE A NEMATIC E ISOTROPIC
I
P X
EXPERIMENTAL
U
Chemicals. A sample of mixed DVB isomers (Shell Chemical Corp.) was obtained from Ionac Chemical Co. (Birmingham, N. J.) and diluted 1 :1 with reagent-grade benzene. The liquid crystal 4,4 '-dihexoxyazoxybenzene (4,4'-DHAB) was purchased from Frinton Labs. (South Vineland, N. J.) and was recrystallized from methylene chloride. The column packing was 9.407% by weight (determined by replicate ashing) of 4,4'-DHAB on 100-120 mesh acid-washed, dimethylchlorosilane-treated Chromosorb W (high performance), packed in a '/a in. 0.d. by 10-ft stainless steel column in a U-tube configuration, which was then coiled to a 31/*in. diameter. Apparatus and Procedure. Gas chromatography was carried out using a chromatograph equipped with a dual flame ionization detector. Carrier flow rates were measured with a soap bubble flowmeter. An iron-constantan thermocouple was placed in the column oven and the column temperature was monitored potentiometrically referenced to a 0 "C ice bath. Chromatograms were recorded on a 10-mV recorder. Sample injection volumes of 0.1 pl and electrometersettingsof 1 x 10-11A(for m-andp-EVB, and naphthalene) to 1.6 x 10-loA (for m- and p-DVB) full scale were used. The usual precautions were taken to ensure meaningful and accurate retention volumes.
RESULTS The transition temperatures of 4,4'-DHAB have been reported by differential-scanning calorimetry (9) to be 81.0 "C from solid to a nematic anisotropic phase, which becomes isotropic at 128.2 "C. Four column temperatures (96.1, (6) R. H. Wiley, G. DeVenuto, and T. K. Venkatachalam, J. Gus Chromatog., 5, 590 (1967). (7) Narl. Bur. Srds., Tech. Note, 459,19 (1968). (8) H. Kelker and E. von Schivizhoffen, "Advances in Chromatography,'' Vol. 6, J. C. Giddings and R. A. Keller, Eds., Marcel Dekker, Inc., New York, N. Y., 1968, p 247. (9) L. C. Chow and D. E. Martire, J. Phys. Chem., 73, 1127 (1969).
s s
Figure 1. Rate dependence of separation factors ( p isomer/rn-isomer) on column temperature in 4,4 '-DHAB nematic and isotropic phases 105.3, 115.5, and 125.8 "C) were employed within the nematic region, and two temperatures (135.3 and 154.7 "C) were used in the isotropic region. In a nematic phase model the liquid crystal molecules are arranged as nonlayered parallel rods, enabling para-substituted aromatics to align and interact with the liquid crystal phase more intimately than meta- and orthosubstituted derivatives. The retention time data corrected for the dead volume of the system are presented in Table I. The values represent means of replicate analyses, which were reproducible to within less than 1%. The sample size was purposely kept low to eliminate overloading effects and to ensure infinite dilution conditions at which the solutes obey Henry's law. It was observed that the retention ratios or separation factors [a! = V,' (para)/V,' (meta)] for the EVB and DVB isomers d e creased with increasing column temperature, and furthermore, that the rate of decrease was more pronounced in the isotropic region of the liquid crystal. (Figure 1). However, the rate
ANALYTICAL CHEMISTRY, VOL. 42, NO. 2, FEBRUARY 1970
177
Table 11. Computed Vapor Pressure Data. Temperature, “C 96.1 105.3 115.5 125.8 103.7 151.2 47.45 69.62 m-EVBb 103.5 149.3 70.19 p-EVBb 48.33 84.98 124.4 38.83 56.94 m-DVBC 78.50 114.9 36.54 52.91 p-DVBd 46.11 68.93 30.16 Naphthaleneb 20.04 a [Newtons per (meter)2] X b From Antoine constants of Ref. (11). c From Antoine constants of Ref. (12). d From computed Antoine constants from vapor pressure data of Ref. (13).
change in CY with temperature appeared to be similar for both EVB and DVB isomer pairs in the nematic and isotropic regions, respectively. The data in the isotropic region were intentionally limited (two temperatures) since the principal objective of the study lay in investigating DVB separations in the more selective nematic region. Hence, the following thermodynamic treatments apply specifically to solute behavior only in this region. The specific retention volumes (V,’) were computed from the corrected peak maxima retention times of Table I using the classical expression developed by Littlewood et al. (10). Solute vapor pressures (Table 11) were calculated directly from the Antoine equations (14). Where the Antoine constants were unavailable, they were calculated (11, 15) from available vapor pressure data. The solute activity coefficients at infinite dilution uncorrected for non-ideality in the vapor phase (ypm)were then calculated from the expression given by Martire (16). The activity coefficients were corrected to ideal vapor phase conditions (yf”) using (17)
Table III. Critical Constants Tc (K) Vc (cmS/mole) m-EVBa 676.36 462.72b pEVBa 681.16 456.08a m-DVBC 709.63 440.13 p-DVBC 693.30 443.90 Naphthalene 748. 36d 413.43b a Ref. (11). Calculated from de of Ref. (11). Calculated from Herzog, Ref. (20). Ref. (21).
solute in question. Where critical constants were unavailable (m-and p-DVB), they were estimated from the parachor equations of Herzog for aromatics (20) TClTt, = 2.640 - 0.4634 log P
(3)
and V,
=
2.92P1.2/Tca.3
(4) where P is the molecular parachor (m-DVB, 337.2; p-DVB, 337.65) and TOthe normal boiling point (K) of the solute in question. The parachor was calculated by subtracting four hydrogen atom values from and adding two terminal double bond values to the parachor values of m-and p-diethylbenzene. The critical constants computed in this manner for m-EVB showed an error of 0.14% and 0.55% for the Tc and V, values reported in the literature (11). The critical constants of the solutes are given in Table 111. The solute specific retention volumes and activity coefficients are presented in Table IV. The solute partial molar excess free energies at infinite dilution were calculated from
dem= RT In y
(5)
Also, from the Gibbs expression where p Z oand BZ2are the vapor pressure and second virial coefficient of the pure solute at temperature T(K). The virial coefficients were obtained from the corresponding states equation of McGlashan and Potter (18, 19) B,/Vc = 0.430 - 0.886(TC/T) O.694(Tc/T)*- 0.0375 (n
- 1) (Tc/T)4.5
(2)
where n is the effective carbon number, which is 10 for all solutes used in this study, and T, and V, are the critical temperature ( K ) and volume (cma/mole), respectively, of the (10) A. B. Littlewood, C. S.G. Phillips, and D. T. Price, J. Chem. SOC.,1480 (1955). (11) R. R. Dreisbach, “Physical Properties of Chemical Compounds,” Advances in Chemistry Series, No. 15, American Chemical Society, Washington, D. C., 1955; ibid., No. 22, 1959. (12) R. R. Dreisbach and R. A. Martin, Znd. Eng. Chem., 41, 2875 (1949). (13) D. R. Stull, ibid., 39, 517 (1947). (14) C. Antoine, Compt. Rend., 107,681 (1888). (15) G. W. Thomson, Chem. Rev., 38, l(1946). (16) D. E. Martire, “Gas Chromatography,” L. Fowler, Ed., Academic Press, New York, N. Y., 1963, p 33. (17) D. E. Martire and L. Z. Pollara, “Advances in Chromatography,’’ 1, J. C . Giddings and R. A. Keller, Ed., Marcel Dekker, Inc., New York, N. Y., 1966, p 335. (18) M. L. McGlashan and D. J. B. Potter, Proc. Roy. SOC.(London), A267, 478 (1962). (19) E. A. Guggenheim and C. J. Wormald, J. Chem. Phys., 42, 3775 (1965). 178
and Equation 5, one obtains the expression In yI” = (Rem/R)( l / T ) - Bem/R
(7)
where and 3,’” are the solute partial molar excess enthalpies and entropies, respectively, and may be obtained from a linear least squares calculation. The solute partial molar excess quantities are given in Table V. Since the enthalpy of mixing is principally governed by the potential energy change which accompanies the formation of a solution, we can appreciate from a simple lattice model that
where z is the number of effective nearest neighbor sites, No is Avogadro’s number, ZAE is the molecular interchange energy, and €11, w, and €12 are, respectively, the solventsolvent, solute-solute, and solvent-solute pair-wise potential energies of interaction (22). Since excess enthalpy of mixing values include e2 terms, and since the enthalpy of vaporiza-
(20) R. Herzog, Znd. Eng. Chem., 36,997 (1944). (21) A. P. Kudchadker, G. H. Alani, and B. J. Zwolinski, Chem. Rev., 68,659 (1968). (22) D. E. Martire, P. A. Blasco, P. F. Carone, L. C. Chow, and H. Vicini, J. Phys. Chem., 72,3489 (1968).
ANALYTICAL CHEMISTRY, VOL. 42, NO. 2, FEBRUARY 1970
~~~
Table IV. Solute Specific Retention Volumesa (V, ") and Activity CoeflBcientsb (ypm,ylm) in the 4,4'-DHAB Nematic Phase Temperature, "C
-
96.1
v,
rn-EVB p-EVB rn-DVB pDVB Naphthalene
665.7 825.9 1198.0 1556.0 2156.0
105.3
YPrn
71-
1.804 1.428 1.225 1.002 1.319 115.5 1.360 1.124 0.978 0.835 0.997
1.816 1.437 1.232 1.008 1.323
O
v L 7
"
523.2 644.7 917.7 1179.0 1662.0
404.1 1.375 m-EVB 490.4 1.136 p-EVB 685.7 0.988 rn-DVB 870.1 0.842 p-DVB 1240.0 1.003 Naphthalene cm3 helium per gram of 4,4'-DHAB liquid phase. b 7 ," uncorrected and 7," corrected, for non-ideality of the vapor phase.
Yfrn
YP-
1.565 1.260 1.091 0.914 1.137 125.8 1.201 1.015 0.862 0.760 0.876
,
313.8 376.1 531.3 652.8 943.7
1.578 1.270 1.099 0.920 1.142 1.218 1.029 0.874 0.768 0.883
0
Table V.
Solute Partial Molar Excess Free Energies (0,") of Mixing and Partial Molar Excess Enthalpy and Entropy (8,")of Mixing in 4,4'-DHAB Nematic Phase
a*-
sem
R," b
5
Temp., "C 96.1 105.3 Solute rn-EVB 1831 .O 1435.0 752.5 p-EVB 1133.0 640.1 297.5 rn-DVB -263.2 24.2 pDVB 418.2 Naphthalene 860.4 Joules per mole. Kilojoules per mole; 96.1 to 125.8 "C. e Joules per mole-deg.; 96.1 to 125.8 "C.
115.3
125.8
1029.0 411.8 -39.0 -556.8 9.1
655.0 95.4 -447.6 -873.3 -411.5
(R,") 0
16.45 =k 0.33 13.73 f 0.42 14.02 i 0.28 11.12 i 0.10 16.56 i 0.39
39.6 i 0 . 9 34.2 f 1 . 1 36.2 i 0 . 7 30.1 i 0 . 3 42.6 i 1.0
(1
tion of pure solute (AHu,2) is approximately equal to zN0ezz/2, we have from Equation 8 that AReoln.2
Rem - AHo,2 = zN,(t11/2 - €12)
(9)
where ARsoln,2 is the solute partial molar enthalpy of solution at infinite dilution. Finally, the ell term may be taken as constant for a given phase of 4,4'-DHAB, and a comparison of ARsoln,2 values for a series of solutes essentially permits a comparison of relative e,? terms. Heats of solution may be calculated by combining the differentiated Antoine equation with the Clausius-Clapeyron equation, giving
+
AHu,z = 2.30259*R*B*T2/(t C)*
(10)
where B and C are the Antoine constants, R the gas constant (8.3143 joules mole-' degree-'), and t the temperature in OC; and Equation 9. Likewise, the corresponding entropy term may be obtained from AS'soln,2
Bern
- AS,,2
Sarn - AH,,z/T
(1 1)
The reference state for the determination of AS'soln.2 was realistically chosen to be an ideal gaseous mixture in which the solute is at infinite dilution. It should be noted that a choice of the pure solute vapor as the reference state would render the entropy terms infinite and meaningless. The computed mean solute partial molar heats and entropies of solution from an ideal gas phase into the 4,4'-DHAB nematic phase are given in Table VI.
Table VI. Solute Partial Molar Enthalpy. (ARBoln.2m) and Entropy6 (AS'soln,2m) of Solution in 4,4'-DHAB Nematic Phasec ARsoin.~"
A$oin,2"
rn-EVB -31.36 -85.1 p-EVB -32.80 -87.2 rn-DVB -34.00 -89.0 pDVB -36.16 -93.2 Naphthalene -34.41 -90.4 Kilojoules per mole. Joules per mole-deg. Values are means of individual values calculated at each column temperature. Their absolute deviations from values calculated at the nominal mean temperature of 110 "Cwere all less than 0.5 %.
solutes at the column temperature, and of the selectivity of the liquid phase employed for the solutes. It is well known, and can be easily shown, that the separation factor (a)for 1) at a two solutes having similar vapor pressures (ratio given temperature, is closely approximated by the ratio of their activity coefficients; Le., a = Vo"(para)/Vo"(meta) = (ry" ~~")rneta/(~~y~~~")~,,,
(yjm)meta/(Y/m)psrs (12)
It follows from Equation 5 then, that In a = In(y,m)meta- ln(rym)psra = A,-,(Be")/RT
DISCUSSION
The separation of two solutes by gas-liquid chromatography is a function of the ratio of the vapor pressures of the
(13)
Since the vapor pressures are indeed close for the meta and para isomer pairs of EVB and DVB (Table 11) and actually contribute less than 2 to a,an index of the relative selectivity
ANALYTICAL CHEMISTRY, VOL. 42, NO. 2, FEBRUARY 1970
179
of 4,4’-DHAB between the isomeric species is given by Am-p (dem).Inspection of Table V reveals that such free energy differences for the meta and para isomers do exist and are appreciable. One may go further, however, in terms of understanding the mechanism of separation by examining the thermodynamic quantities which contribute to the excess free energy values obtained-uiz., the partial molar enthalpies and entropies of solution. Note that these values (Table VI) are more negative for the para isomers than for the meta isomers. More negative enthalpy values (more exothermic) are indicative of stronger solute-solvent interactions (greater e13, while more negative entropy values reveal a more ordered solution state. This behavior is consistent with the postulate that parasubstituted solutes, being more rod-like spatially, “fit” better into, and thereby interact more strongly with, the rod-like ordered solvent. In other words, in going from a completely disordered vapor state (similar for both meta and para isomers owing to the virial correction) to an ordered liquid state, the para isomer sacrifices more translational and rotational freedom, but, in return, its favorable geometry allows it to interact more strongly with the aligned liquid crystal m o l e cules. In the balance then, its entropy loss is overcome by its enthalpy gain, making it more soluble than its meta counterpart (which, although its motion is less restricted, has a greater enthalpic requirement for solution relative to an ideal solution). Finally, the longer retention associated with naphthalene relative to the previously eluted p D V B is due both to its lower vapor pressures (Table 11) and to its more
positive free energy values (Table V). Once again, as in the case of the meta isomers, solution for naphthalene is entropically favored but enthalpically disfavored because of its un-rod-like molecular structure. The liquid crystal solvent becomes far less selective in the isotropic region of temperatures (Figure 1). Although some residual short-range order is probably maintained in this region, the long range disorientation of the solvent molecules prevents any appreciable alignment (and, hence, selective r e tention) of the para isomers. The practical realization of complete separation of DVB isomers is of considerable interest to the isolation of quantities of pure isomers via preparative scale gas chromatography (6). The chromatographic data obtained in this study, taken together with considerations for liquid phase capacity, should prove valuable to approximate the requisite number of theoretical plates and appropriate conditions with which meaningful preparative scale separations of DVB isomers may be achieved on a nematic liquid crystal stationary phase. ACKNOWLEDGMENT
The authors thank Stanley P. Wasik of the Physical Chemistry Division, National Bureau of Standards, for stimulating discussions of the material in this study. RECEIVED for review September 15, 1969. Accepted November 18, 1969. Two of us (D.E.M. and L.C.C.) acknowledge grant support from the U. S. Army Research Office, Durham, N. C.
Identification of Polycyclic Naphthenic, Mono-, and Diaromatic Crude Oil Carboxylic Acids1 Wolfgang K. Seifert Chevron Oil Field Research C o o ,P.O.Box 1627, Richmond, Calif. 94802
Richard M. Teeter Chevron Research Co., Richmond,
Gal$ 94802
Many detailed structural features of substituted naphthenic, naphtheno-aromatic, and mono- and diaromatic carboxylic acids from a California crude oil are elucidated after isolation of fractions of derived hydrocarbons by repeated chromatography on alumina and the application of a variety of methods of high resolution molecular spectrometry. This comprehensive instrumental analytical approach results in compound class identification of many types of carboxylic acids which have not been reported previously in petroleum. The analytical methods involved are a combination of computer-averaged nuclear magnetic resonance, high resolution mass spectrometry, ultraviolet, fluorescence and infrared spectrometry, and gas chromatography combined with low resolution mass spectrometry. All acids of low hydrogen deficiency, postulated previously solely on the basis of high resolution mass measurements of acids and trihydroperfluoroalcohol esters, were confirmed by the combination of these methods of molecular spectrometry applied to hydrocarbons derived from the acids. The identified carboxylic acids represent at least 12% of all acids naturally occurring in this crude oil. A novel technique of structural analysis applicable to complex mixtures of hydrocarbons by high resolution mass spectrometry based on fragment ions combined with NMR is described and leads to detailed structural information within compound types. 180
THE importance of the identification of carboxylic acids naturally occurring in petroleum is that these acids are considered to be the precursors of petroleum hydrocarbons ( 1 , 2 ) and the knowledge of their structure is linked directly to the problem of the origin of petroleum (3-6) and of life on earth ( 7 , 8 ) . In spite of the enormous efforts spent on this problem during the last 100 years (9), and particularly during the past several years because of the advent of modem instrumental 1 Partial preliminary communication: W. K. Seifert and R. M. Teeter, Cbem. Znd. (London), 1464 (1969).
(1) J. E. Cooper and E. E. Bray, Geochim. Cosmochim. Acta, 27, 1113 (1963). (2) K. A. Kvenvolden, J. Am. Oil Chem. SOC.,44, 628 (1967). (3) H. M. Smith, ibid., 44, 680 (1967). (4) E. D. McCarthy and M. Calvin, Nature, 216,642 (1967). ( 5 ) W. Henderson, G. Eglinton, P. Simmonds, and J. E. Lovelock, ibid., 219, 1012(1968). (6) T. Maclean, G. Eglinton, K. Douraghi-Zadeh, R . G. Ackman, and S . N. Hooper, ibid., 218, 1019(1968). ( 7 ) R. Robinson, ibid., 212,1291 (1966). (8) G. Eglinton and M. Calvin, Sci. Am., 216, 32 (1967). (9) H. L. Lochte and E. R. Littman, “Petroleum Acids and Bases,” Chemical Publishing Co., Inc., New York, 1955.
ANALYTICAL CHEMISTRY, VOL. 42, NO. 2, FEBRUARY 1970
