(1) Use of Johnson's formulas for sp3 levels in conjunction with existing experimental data makes it doubtful whether the 5S state of the carbon atom is only 1.6 volts above the ground state, as often assumed in the literature. (2) A formula is derived for the energy of the ``valence state'' of the C atom, which is a linear combination of sp3 levels wherein the spins are paired with those of attached atoms. This state is characteristic of tetravalent carbon compounds involving four electron pair bonds, and is shown to involve an increase of about 7 volts in the internal energy of the C atom over that in the ground state. (3) Because of this increment, the net or observed bonding energy is less than the gross or true inter‐atomic bonding energy. The gross energy per bond is probably greater in CH3 than in CH4 although the reverse is true of the net. (4) A critical comparison is given of the Slater‐Pauling theory of directed valence and the nondirectional Heitler‐Rumer |theory based on a 5S state of the C atom. The former leads to much firmer bonds. (5) The assumption of electron pairing, made in Part II, is shown to be well warranted in CH4, as it yields an energy value which is very nearly the same as the deepest root of Eyring and colleague's more exact cubic secular equation. (6) The energies of CH4 and 4CH are compared in the light of theory.

1.
Part I of the present series appeared in
J. Chem. Phys.
 
1
,
177
(
1933
);
Crossref
Search ADS
 
Part II,
J. Chem. Phys.
 
1
,
219
(
1933
).
Crossref
Search ADS
 
See Part I for references to the literature. We neglected to state there that the method of “molecular orbitals,” which is the essence of the Hund‐Mulliken procedure, was first introduced by
J. E.
Lennard‐Jones
, though only for diatomic molecules (
Trans. Faraday Soc.
25
,
668
(
1929
)). The present article connects particularly with Section 5 of Part II, based on the Heitler‐London method.
Google Scholar
Crossref
Search ADS
 
2.
J. C.
Slater
,
Phys. Rev.
34
,
1293
(
1929
).
Google Scholar
Crossref
Search ADS
 
3.
E. U.
Condon
 and
G. H.
Shortley
,
Phys. Rev.
37
,
1025
(
1931
).
Google Scholar
Crossref
Search ADS
 
4.
M. H.
Johnson
, Jr.,
Phys. Rev.
39
,
209
(
1932
): These formulas for sp3 may also be derived very simply by the Dirac vector model, as will be shown in a future paper by the writer in Phys. Rev. We have used this model to determine the additive constant in the energy, which was not specified in Johnson’s paper since he took D3 as the origin.
Google Scholar
Crossref
Search ADS
 
5.
W.
 
Heitler
 and
G.
 
Herzberg
,
Zeits. f. Physik
 
53
,
52
(
1929
);
Google Scholar
Crossref
Search ADS
 
L. A.
Turner
,
Proc. Nat. Acad.
15
,
526
(
1929
).
Google Scholar
Crossref
Search ADS
 
6.
In their tables, Bacher and Goudsmit use different origins of energy for the doublets and quartets of C+, but a common origin can immediately be secured by making use of the fact that absolute term values are known for both the triplets and singlets of C++.
7.
N. F.
Beardsley
,
Phys. Rev.
39
,
913
(
1932
).
Google Scholar
Crossref
Search ADS
 
8.
The expressions αK, α(O+23P), αP, αQ, on page 918 of Beardsley’s paper are respectively the same as F0(2s;2p), F0(2p;2p), 3F2, F0(2s;2s) in our notation.
9.
R. S.
Mulliken
,
J. Chem. Phys.
1
,
500
(
1933
).
Google Scholar
Crossref
Search ADS
 
10.
W.
Heitler
 and
G.
Rumer
,
Zeits. f. Physik
68
,
12
(
1931
).
Google Scholar
Crossref
Search ADS
 
11.
J. C.
Slater
,
Phys. Rev.
38
,
1109
(
1931
).
Google Scholar
Crossref
Search ADS
 
12.
L.
 
Pauling
 and
G. W.
 
Wheland
,
J. Chem. Phys.
 
1
,
362
(
1933
);
Google Scholar
Crossref
Search ADS
 
L.
Pauling
 and
L.
Sherman
,
J. Chem. Phys.
1
,
679
(
1933
).,
J. Chem. Phys.
Google Scholar
Crossref
Search ADS
 
13.
H.
 
Eyring
 and
M.
 
Polanyi
,
Zeits. f. Physik. Chemie
 
12
,
279
(
1931
);
Google Scholar
 
H.
Eyring
,
J. Am. Chem. Soc.
53
,
2537
(
1931
). Dr. A. Sherman informs the writer that it is better to take the exchange part as 86 rather than 90 percent of the Morse function, at least in calculations of activation energies. This modification is too small to be of consequence for our work, and so we use the round value 90.
Google Scholar
Crossref
Search ADS
 
14.
For a quintet state involving four electrons, the Pauli principle allows only one electron to each orbital state. There are two kinds of 2pπ states, viz., the left‐ and right‐handed varieties, or orthogonal linear combinations thereof.
15.
A. S.
Coolidge
,
Phys. Rev.
42
,
189
(
1932
).
Google Scholar
Crossref
Search ADS
 
16.
Eyring
,
Frost
, and
Turkevich
,
J. Chem. Phys.
1
,
777
(
1933
).
Google Scholar
Crossref
Search ADS
 
17.
F.
Seitz
 and
A.
Sherman
,
J. Chem. Phys.
1
,
11
(
1934
). The appropriate cubic secular determinant is given in this paper in a form in which W appears only down the principal diagonal and which hence facilitates the application of perturbation theory. In reference 16, a less convenient, non‐orthogonal form is given. The writer is indebted to the various authors mentioned in notes 16 and 17 for the opportunity of seeing their manuscripts in advance of publication.
Google Scholar
 
18.
Eq. (19) is proved by the same methods as given in Section 5 of Part II or in fine print in
J. H.
Van Vleck
 and
P. C.
Cross
,
J. Chem. Phys.
1
,
357
(
1933
). Concerning the need of the k term see note 13 of the latter reference.
Google Scholar
Crossref
Search ADS
 
19.
This numerical value is taken from
L.
Pauling
 and
J.
Sherman
,
J. Chem. Phys.
1
,
606
(
1933
). Mulliken bases his calculations on a lower value 15.8 volts. To facilitate comparison with Mulliken’s article, we have utilized 15.8 rather than 17.3 volts as the heat of formation of CH4 in all our estimates of the relative energies of CH3 and CH4 in Section 3. Use of 17.3 instead of 15.8 would not affect our conclusions in Section 3 that the gross energy per bond is greater in CH3 than CH4, and that our theory leads to an energy for the reaction CH4 = CH3+H of approximately the proper magnitude 5 volts. This is all, of course, provided the promotional energy to the valence state of carbon is 7 volts or more. If it is only about 4 volts, as is the case provided there is really a S5 state only 1.5 volts above normal, or provided we accept Beardsley’s numerical estimates of the F’s and G’s, the gross energies per bond are nearly equal in CH3 and CH4. In this case WV(C)∼4 volts, Eq. (20) demands Nσs = 0.4 rather than 1 volt.
Google Scholar
Crossref
Search ADS
 
20.
R. S.
Mulliken
,
Rev. Mod. Phys.
4
,
80
(
1932
).
Google Scholar
Crossref
Search ADS
 
21.
C.
Ireland
,
Phys. Rev.
43
,
329
(
1933
).
Google Scholar
Crossref
Search ADS
 
22.
H. J.
Woods
,
Trans. Faraday Soc.
28
,
877
(
1932
).
Google Scholar
Crossref
Search ADS
 
23.
If the promotional energy to the valence state of C is 7 volts, the gross energy per bond in CH4 is 6 volts, rather than 5 as stated by Woods.
This content is only available via PDF.
© 1934 American Institute of Physics.
1934
American Institute of Physics
You do not currently have access to this content.