945
SOTES
May, 1962
NH4+and Kf (9.3 ml./mole), but it corresponds to the difference expected for NH4C1 and KC1 if the excess concentration in the bound layer on the polyelectrolyte is about 1.5 N , a reasonable value on geometrical grounds. The study of the variation of P with the nature of the counterion, therefore, gives us a means of examining the question of the specificity of ionbinding. The method is more general than those of Raman spectroscopy or nuclear magnetic resonance. With th_e available techniques of density measurements V can be determined to better than 0.1g;;0.4 It should be- possible, from the difference of the measured AV and the value a t infinite dilution, to get some indication of the strength of the electrostatic interactions, as in the examples cited above, in terms of an effective concentration, or some more detailed theoretical picture of ion-binding. An extension of the method, applicable particularly to association c_olloidal electrolytes, is to compare the change in V from below the c.m.c. to above it for the case of different counterions. For this precise data below the c.m.c. are needed as well. From the available data, we can compare laurylsulfonic acrid a t 2507 with sodium laurylsubmate a t 2 3 O . l : ’ The extra oxygen in the latter is not expected to cause much difference. On- the formation of micelles from the single ions, V increases by 10.0 =t 2.0 ml./mole for sodium laurylsulfate and 11.3 st 1.0 ml./mole for laurylsulfonic acid. This again shows that the sodium and the hydrogen forms behave in about the same manner, and that the charge interactions are mainly electrostatic. (13) L. M. Kushner, B. D. Duncan, and J. I. Hoffman, J. Research Nat2. BUT.Standards, 49, 85 (1952).
-
FLUOlUNE N.M.R. SPECTROSCOPY. VII1. COUPLING CONSTANTS IN NORMAL AND ISOTOPIC C3F8 BY GEORGEVANDYKETIERS Contrabutton No. 626 f r o m the Central Research Dept., Mznnesota Mzning and M f g . Co. St. Paul I D , Mtnn. Rerezved September 86. 1961
Several recent papers’-’ have given indicatioii of the strong dependence of i’tuorine shielding value^'-^*^-^ and coupling constants3-7 upon the molecular conforniation or stereostructure, 1-6 and upon the nature of substituent groups.3PJ However, to place such observations upon a quantitative basis, it would be necessary to secure extremely detailed information about molecular geometry in fluorocarbon derivatives. Owing to the conformational flexibility of most of these comis, quite p o u n d ~ ’ * ~it, ~ ~ * ~ unlikely that precise bond angles and dihedrsl angles will be secured. Two (1) G. V. D. Tiers. Proc. Chem. Soc., 389 (1960). 12) G. V. D. Tiers. J . Phys. Chem., in press (Part IX). (3) J. Feeney and L. H. Sutchffe, Trans. Faraday SOC.,56, 1559 (1960). (4) L. M. Crapo end C. H. Sederholrn, J . Chem. Phys., 5 3 , 1583 (1960). ( 5 ) G. V. D. Tiers, ibid , in press (Part V). (6) E. Pitcher, A. D. Buckingham. and F. G. A. Stone, Spectrochzm. Acta, in press. (7) G. V. D. Tiers, J. Phys. Chem., 66, 764 (1962).
courses remain: first, the investigation of rigid molecules having a more or less well defined geometry‘m2J; and secondly, the study of certain symmetrical molecules (or groups) for which there is an obvious simplification of the conformational problems. In this paper the second course is adopted, the molecule studied being perfluoropropane, C3Fs. Owing to the symmetry, the coupling constant between CF, groups can only be seen for the isotopic isomer C21?5C’3F,3,which is present in natural abundance; this powerful method, due to Sheppard,s is particularly valuable for fluorine compound~.~ Experimental The n.m.r. equipment and techniques have been described .3,9Jo The perfluoropropane was a purified sample available in these laboratories. It was studied as a 34% (by liquid volume) solution in CCllF in order to reduce the hazard due to high pressure. Fluorine shielding v a l ~ e s +*, ,~ measured a t 28.9’ on this and on a 12% solution t’ien were extrapolated to infinite dilutiong; results, given in Table I, are the averages of six to eight separate determinations, and error values given are standard deviations in the averaged values.
TABLE I FLUORXNE N.M.R.SHIELDINQ VALUES FOR PERFLUOROPROPANE
Concn. of CsFs,
% 34
+* (CF3)
Std. dev.
+82.861 &0.006 4-82.833 f .003 0 f82.8Ba i .005 Extrapolated value; a t infinite
12 (1
+* (CF2)
Std. dev.
4-131.181 10.009 +131.200 i. .003 4-131.210“ ,008 dilution in CC13F, +* =
4. In Table I1 are listed the results of the coupling constant measurements, and also the associated “C13 isotope effect” shifts, both “direct”11 and “distant.”’O As in all previous cases,6JoJ1 the heavier isotope produces shifts corresponding to higher shielding. In all cases it was permissible to use first-order spin-spin analysis due to the large disparity in magnitudes of the coupling conetants. This was fortunate, as the spectra are very weak, the C13being in natural abundance; finer details in the patterns would have been very difficult t o observe reproducibly. For J ( F - 0 3 ) a 0.18 e./ sec. second-order correction was a p ~ l i e d . ~
TABLE 11 FLUORINE N.M.R. COUPLINGCOWPANTSAND ISOTOPE EFFECTSFOR THE CFa GROUPIN PERFLUOROPROPANE Coupling systema
J,
c./sec.
Ptd. dev.b
A+, p.p.m.
0
Std. dev.b
No. of meas.
F-C13 285.74d zkO.13 +0.135 10.005 20 F-C-C’a 40.34 =IC.10 +0.020 1 0 . 0 0 2 18 F-C-C-F 0.70 i. .01 . . . . . ..... 32 F-C’“C-C-F 7.31 i. .07 . .... . . . . . 25 Coupling is between the first- and last-mentioned nuclei in each system. A 4 = 4PlCl3 Std. dev. of the average. isomer) - 4(C13 isomer). Corrected according to ref. 5 ; ML Ma: = 285.93 c./sec.
+
Discussion It is noteworthy from Table I that there is little concentration dependence of shielding values for (8) N. Shcppard and J. J. Turner, Proc. Roy Soc. (London), A252, 506 (1959); Mol. Phys., 3, 158 (1960). (9) G. Filipovich and G. V. D. Tiers, J . Phys. Chem., 6 5 , 761 (1959). (IO) G. V. D. Tiers, J . Phys. Soc. Japan, 15, 354 (1960). (11) P. C. Lauteibur, private communication; paper in course of preparation.
946
P\TOTEB
C3F8, and indeed for nearly all purposes @* (10% to 20%) might be substituted for @ with negligible error, as indicated previously. l B 9 Interestingly, the weak solvent dependences for CF3 and CF2 groups are in opposite directions. While shielding of the CF3 group is slightly higher than is found in longer-chain compound^,^ the CF2 @-value, 131.210, is very substantially increased in C3Fs. For n-perfluoroalkanes the average values are 122.5 @ and 126.64 @ for midchain CF2 and for CF2 adjacent to CF312; in the present case proximity to two CF3 groups brings about a further, almost linear, increase in shielding. The n.m.r. isotope shifts due to attached and to “distant” CL3correspond reasonably well with previous values for CFa groups.6J0 Coupling constants between fluorine and C13 also have magnitudes close to those previously reported5.l0; a further treatment of this subject11 will demonstrate the range of variation observed in J(F-C13) in a variety of compounds. No information could be obtained as to the relative sign of J(F-C13) and J(F-C-C13). The coupling constant between dissimilar fluorines in the perfluoroethyl group is now known to be somewhat dependent upon structure. Thus, in CzFd J(FCCF) is 4.6 c./sec.,6 and in several (perfluoroethy1)-metal compounds it ranges from 1.4 to 1.8 c./sece6; however, when the C2Fs group is attached to carbonyl, the coupling varies from 1.4813 to 0.7 c./sec.B The value measured here for C3F8, 0.70 c./sec., corresponds well with the latter cases, provided there is no inversion in sign. The principal outcome of the carbon-13 study is the observation of the coupling constant between the otherwise equivalent CF3 groups in C3Fs. Such “distant” coupling has often been measured and it between CF3 and CF, groups so has been suggested6 that there is a correlation between high coupling constants, J(CF~-C-CFZX), and low shielding values for the CFzX group, and vice versa. However, in the present case (X = F) a “low” coupling constant, 7.3 c./sec., accompanies a low shielding value; the suggested correlation6 is thus seen to be based upon too limited a choice of data. Acknowledgment.-I thank R. B. Calkins for excellent operation of the n.m.r. spectrometer, and Jane E. H. Tiers for careful measurement of the spectra.
Vol. 66
use of the profile of a sessile drop, as in the work of Kingery, et aZ.,3 but this is difficult to apply to the interface between two liquids, although good results have been obtained by the use of radiograp h ~ .An ~ alternate method has been devised, which invoIves only the measurement of the height of a column of liquid at the moment that its meniscus breaks away from a sharp-edged circular hole in a flat plate. It may be called the “Meniscus Break-through Method.” I n principle it IS identical with the maximum bubble pressure method, but it uses the liquids under study as manometric fluids. The method can be used at room temperature, but is especially advantageous in the difficultly-accessible interiors of furnaces or cryostats, since the observation of the sudden instability of the interface is independent of the measurement of height, and can be done telescopically in a vertical direction. The method also has the advantage that the interface is easily renewed and is, therefore, fresh and relatively uncontaminated. The orifice plate, sealed to the end of a tube of appropriate length, is either raised or lowered through the interface (depending on whether the contact angle of the lower liquid against the orifice material is less or greater than 90’) until the liquid under higher pressure breaks through. TO a first approximation the interface energy
where p1 and p2 denote the densities of the upper and lower liquids, respectively; g is the gravitational constant; and h, the critical hydrostatic head, is experimentally the difference between the height, Ho, of the plate at the point where the interface first touches the orifice plate and the point, H , where it breaks through (both of which are easily identified by watching the meniscus from above with a telescope), corrected for the thickness of the disk and for the change in external liquid level caused by the displacement of the orifice tube. The correction for non-sphericity of the meniscus due to the gravitational pressure gradient in the liquids is done by the Sugdeii method, as in the conventional maximum bubble pressure method. I n practice it has been found convenient to mount the orifice tube on a rod attached to the carriage of a vertical cathetometer which also carries a horizontal telescope (sighted, via a prism, on the meniscus) and a suitable light, all moving together. (12) G . V. D. Tiers, paper t o be submitted shortly. The rod enters a vertical furnace containing the (13) C. A. Reilly, J . Chem. Pl~ys.,86, 604 (1956). crucible with the experimental materials through a small hole in the lid, which also is arranged to ,4 SIMPLE METHOD OF MEASURING admit thermocouples and an inert gas. LIQUID INTERFACIAL TENSIONS, If large amounts of both liquids are available, ESPECIALLY AT HIGH TEMPERATURES, it is simplest to allow the upper liquid to maintain WITH MEASUREMENTS OF THE SURFACE its own level within and without the tube support(1) Massachusetts Instltute of Technology, Cambridge, hIassaTENSIOX OF TELLURIUM BYCYRILSTANLEY SWITH’AND DONALD P. SPITZER
Institute for the Study of Metals, Unmerszty o f Chtcago, Chtcago,Illinois Received Octobei 86,1961
The common methods of measuring surface tensions at elevated temperatures have been reviewed by Ilozakevitch.* Perhaps the best method is the
chusetts. (2) P. Kosakevitch, “Suifaoe Tension” in “Physioochem~calMeasurements a t High Temperatures,” Edited by J. 0’51.Bockris, J. L. White, and J. D. Mackenzie, Academic Press, New York, N. Y., 1959, pp. 208-224. (3) W. D. Kingery and M. Humenik, Jr,, J . Phys. Chem., 67, 359 (1953). (4) P. Xosaker-itch, S. Chatel, G. C h a i n , and51. Sage, RBu. MBtallurgie, 62, No. 2, 139 (1955).