The reaction‐field concept is examined quantitatively using nuclear magnetic spectroscopy by determining the magnetic shielding constants of three protons within the same molecule as a function of dielectric constant: the results are based on the fact that the magnetic shielding constant of a bonded proton depends linearly on the effective electric field along the bond axis. It is shown that (1) the reaction field must take on different values at different points in the molecule, and (2) the calculated reaction field's dependence on ε for a dipole in a spherical or spheroidal cavity does not adequately describe the experimental reaction field's dependence on ε.

1.
L.
Onsager
,
J. Am. Chem. Soc.
58
,
1486
(
1936
);
A. J.
Dekker
,
Physica
12
,
209
(
1946
);
T. H.
Scholte
,
Physica
15
,
437
,
450
(
1949
).
2.
A. D.
Buckingham
,
T.
Schaefer
, and
W. G.
Schneider
,
J. Chem. Phys.
32
,
1227
(
1960
);
A. D.
Buckingham
,
Can. J. Chem.
38
,
300
(
1960
).
3.
P.
Diehl
and
R.
Freeman
,
Mol. Phys.
4
,
39
(
1961
).
4.
See Paper I in this series and references cited therein.
5.
J. I.
Musher
,
J. Chem. Phys.
37
,
34
(
1962
).
6.
For a proton in a C‐H bond reasonable values of k are k = −3×10−12esu while for a proton in a N‐H bond k = −5.0×10−12esu. Recently obtained evidence for a fluorine nucleus in a CF bond shows k to depend critically on the other substituents on the carbon atom. In particular, for F in a CF(H)(C2) group k∼2.5×10−10esu while for F in a CF2(C2) group k∼1.95×10−10esu. Except for the k for a proton in a CH bond all these values have been obtained from studying the effects of intramolecular electric fields.5,7
7.
J. I. Musher, unpublished arguments based on the experimental data of
J. P.
Kokko
,
L.
Mandell
, and
J. H.
Goldstein
,
J. Am. Chem. Soc.
84
,
1042
(
1962
) for CH and NH bonds;
on that of
F. A.
Bovey
et al.,
J. Chem. Phys.
40
,
3099
(
1964
),
and
J.
Homer
and
L. F.
Thomas
,
Trans. Faraday Soc.
59
,
2431
(
1963
) for CF(H)(C2) bonds;
and on that of G. V. D. Tiers Proc. Chem. Soc. 1960, 389
and
J. D.
Roberts
,
Angew. Chem.
75
,
20
(
1963
) (and unpublished data for CF2(C2) bonds.) The unpublished axial‐equatorial chemical shift difference for 1,1‐difluorocyclohexane of Roberts and E. Glazer is 874 cps, presumably at 56.4 Mc/sec.
8.
(a) This innovation was first used by
F.
Hruska
,
E.
Bock
, and
T.
Schaefer
,
Can. J. Chem.
41
,
3034
(
1963
). These authors used mixtures of dioxane and water which we feel are in principle very susceptible to hydrogen bonding and therefore might give ambiguous results. In fact Hruska et al. found strict linear behavior for their molecules in these mixtures, although the intermolecular differences in chemical shift that they studied were very small, i.e., 1.4 and 1.8 cps.
(b) One would hope, therefore, to minimize differences between self‐association of solvent molecules and nondipolar association with solute molecules, so that solvent effects discussed here are due only to variations of ε. The same holds for interactions between the solvent and the TMS standard. The curves for the cyclohexane‐cyclohexanone mixtures are not given here since they are similar to those of Fig. 1. Unfortunately, insufficient solubility in cyclohexane prevented the further extension to lower ε.
9.
D. Decroocq, Bull. Soc. Chim. (France) 127 (1964).
10.
P. Laszlo and J. I. Musher, Bull. Soc. Chim. (France) “Constantes de couplage et structure en résonance magnétique nucléaire II.” (to be published).
11.
J. I.
Musher
,
J. Chem. Phys.
34
,
594
(
1961
).
12.
R.
Freeman
and
D. H.
Whiffen
,
Mol. Phys.
4
,
321
(
1961
);
R. Freeman, ibid., p. 385.
13.
R.
Freeman
and
D. H.
Whiffen
,
Proc. Phys. Soc. (London)
79
,
794
(
1962
);
R.
Freeman
and
W. A.
Anderson
,
J. Chem. Phys.
37
,
85
(
1962
).
14.
N. L.
Allinger
,
M.
DaRooge
,
M. A.
Miller
, and
B.
Waegell
,
J. Org. Chem.
28
,
780
(
1963
).
15.
B. Waegell and G. Ourisson, Bull. Soc. Chim. (France), 496 (1963).
16.
Using the procedure of
J. I.
Musher
,
J. Chem. Phys.
40
,
2399
(
1964
).
17.
The curves shown in Fig. 2 and the several significant deviations from the linear plots of Ref. 3 make an attempt to confirm this assertion as important and not trivial.
18.
That is, that the protons in both cases are shifted by the same interaction, presumably that with the reaction field, but the ε dependence of this reaction field is not that expected.
This content is only available via PDF.
You do not currently have access to this content.