July 20, 1953 NOTES 3601

saponifiable nitrile groups in stable aromatic a linear array parallel to the helix axis, and reso-
systems. nance structures have been proposed for the poly-
Experimental iodine chaim2 We may expect that this poly-
2,3,6,7-Tetramethylnaphthalene- 1,Cdinitrile (11).-The iodine chain has a structure similar to that of poly-
author is indebted t o Dr. P. E. Hoch for a sample of this enes. The corresponding two p-electrons (those
material which he prepared by a standard Knoevenagel having the same symmetry) of each iodine atom
reaction of biacetyl with 3,4-dimethyl-o-phenylenediaceto-
nitrile. This product, when decolorized with charcoal and may form a a-orbital along the polyiodine chain.
recrystallized from cyclohexane, formed felted yellow The iodine chain may have two such a-orbitals,
needles which appear greenish in reflected light as a result because each iodine atom possesses two pairs of
of a blue fluorescence. I t decomposed a t about 240' with- such p-electrons, and these a-electrons may be
out melting.
Anal. Calcd. for CiGHllN2: C, 82.0; H, 5.98; N,11.96.
responsible for the light absorption of the complex.
Found: C, 81.9; H , 5.88; N,11.75. We applied to our experimental results the relation
2,3,6,7-Tetramethylnaphthalene-l,4-dicarboxamide (111). which has been proposed by Kuhn3 for polyenes.
-A mixture of 250 mg. of the dinitrile 11, 4.0 g. of potas- Kuhn's relation is expressed by
sium hydroxide, 40 ml. of ethylene glycol and 5 ml. of water
was refluxed one hour. The solid I1 slowly dissolved to a
clear yellow solution, which, when it was poured onto ice
and filtered, gaxe 260 mg. (97% yield) of pale buff solid, where Xmax is the wave length of maximum absorp-
m.p. above 320 . tion, L , the resonating chain length, N , the number
2,3,6,7-Tetramethylnaphthalene(IV).-To a solution of
3.0 g. of phosphorus pentoxide in 7.0 g. of reagent 85% of a-electrons, h, Planck's universal constant,
phosphoric acid was added 250 mg. of the diamide 111. m, the mass of the electron, c, the velocity of light,
The mixture, contained in a small flask with reflux con- Vo,the perturbation parameter.
denser, was heated at 200' for one hour with occasional shak- The results obtained indicate that the Xmax
ing. White micro-platelets of product sublimed into the upper
part of the flask and into the condenser. The mixture was shifts toward the shorter wave lengths and the
cooled, diluted with water and extracted well with cyclo- resonating chain length becomes shorter with the
hexane. The cyclohexane solution was dried and chromato- increasing iodide concentrations. Our data are also
graphed on an 80 X 8 mm. column of alumina. The prod- in good agreement with those obtained by Rundle
uct, recovered from the effluent liquid by vacuum stripping,
weighed 145 mg. (7'9% yield). One recrystallization from and co-workers4 and Swanson,6 concerning the
methanol gave shining, fluffy platelets, m.p. 191.2-191.8'. relationship between starch chain length and Amax.
Anal. Calcd. for C14H16: C, 91.25; H , 8.75. Found: Potentiometric6 and spectrophotometric titrations*
C, 91.14; H , 8.80. suggested that iodide enters the helix with iodine.
-The TNF -derivative, when recrystallized from benzene, Our suggested explanation for the shift of Xmax with
gave scarlet microneedles, m.p. 176.0-176.6'. the change of iodide concentration is that iodide
Anal. Calcd. for C27H21N301: C, 65.0; H, 4.21; N, ions enter into the polyiodine chains and break the
8.42. Found: C, 65.5; H,4.62; N, 8.32.
When 20 mg. of the dinitrile I1 and 2.0 g. of polyphos- resonating chains into shorter lengths.
phoric acid were heated a t 200' for a short time, white Experimental
crystals sublimed into the upper part of the tube. The Amylose used was obtained from potato starch by hot
identity of these crystals with I V was confirmed by melting water extraction (at 60°), and it had the following charac-
point and mixed melting point. teristics: intrinsic viscosity was [a] = 0.80. Using this
A comparison of the ultraviolet absorption spectrum and value, the degree of polymerization was calculated to be
physical properties of this material with those of other tetra- D.P. = 480 glucose units.'
methylnaphthalenes has been made. Absorption spectra were measured with a G. E. Recording
~

( 5 ) W. L. Mosby, THISJOURNAL, 75, 3348 (1953).

Spectrophotometer at room temperature, using a cuvette
10 mm. thick. Amylose concentration used in this investi-
CENTRAL RESEARCHLABORATORY gation was 0.01 %. Iodide concentrations were varied from
GENERAL ANILINEAND FILMCORPORATION zero to 0.5 M at constant iodine concentration (13% of
EASTON,PENNSYLVANIA theoretical maximum complex formation). A parallel
investigation was conducted using ethanol instead of iodide.
Results and Discussion
On the Starch-Iodine Complex
In our study, N is equal to twice the numberoof
BY SBZABURO ONO, SHINPEI TSUCHIHASHI AND TAKASHIiodine atoms, n, and L is equal to 6.2(n/2 - 1) A.,
KUGE based on the average 1-1 distance of 3.1 A. in the
RECEIVED FEBRUARY 4, 1953 polyiodine chain.a It was postulated that in the
This paper presents a quantitative study of the complex free of iodide, 80 iodine molecules form one
relationship between iodide concentration and wave polyiodine chain, since D.P. of the amylose used
length of maximum absorption of amylose-iodine was 480 glucose units and 6 glucose units take up
complex. The authors attempt to explain the one iodine m o l e ~ u l e . At ~ ~ ~zero iodide concentra-
manner in which the iodide ion influences the tion, all terms in equation (1) are known except VO.
starch-iodine complex formation. Vois then calculated to be 3.09 X lo-'* erg (1.9
Rundle and his co-workers1 have found that the e.v.).
nature of the absorption spectra of the colored (2) R . S.Stein and R . E. Rundle, J. Chcm. Phys., 16, 195 (1948).
iodine in the starch-iodine complex is something (3) H.Kuhn, ibid., 17, 1198 (1949).
(4) R. R.Baldwin, R . S. Bear and R . E. Rundle, THISJOURNAL, 66,
quite different from that of iodine in any other 1 1 1 (1944).
medium. The structure of the amylose-iodine ( 5 ) M. A. Swanson, J . B i d . Chcm., 172, 825 (1948).
complex has been established as a helical amylose (6) F. L. Bate, D. French and R . E. Rundle, THISJOURNAL, 65, 142
chain in which the iodine molecules are arranged in (1943).
(7) A. L. Potter and W. 2. Hassid, ibid., 78, 593 11951).
(1) R . E. Rundle, J. F. Foster and R. R . Baldwin, THISJOURNAL, (8) C. D. West, J . Chcm. Phys., 16, 689 (1947).
66, 2116 (1944). (9) R . E.Rundle and D. French, THISJOURNAL, 66, 1707 (1943).
 3602 NOTES Vol. 75

Hach and RundlelO stated that it had not been having a nitro group ortho to the chlorine can be
determined experimentally whether polyiodine accomplished by heating the compound with
consists of chains with two different 1-1 distances, copper and benzoic acid.’
or whether all distances have become equal to 3.1 We wish to report a study of the application of
8. From the VOvalue determined above, 1.9 e.v., this reaction to nitro compounds containing
it seems likely that polyiodine consists of chains fluorine, bromine or iodine, and also the use of the
with two 1-1 distances, due to unequal contributions reaction for the preparation of 3,5-dinitrotoluene.
of two limiting forms A and B. In the earlier work it was found that a chlorine
. . .I-I.. . . .I-I. . . . .I-I.. (A) . meta or para to a nitro group was not removed by
-1.. . . .I-I.. . . . I - I . . . . .I- (B)
the treatment with copper and benzoic acid. It
has now been shown that a chlorine or bromine
TABLE
I ortho to a nitro group can be removed without
Concn. of KI, Concn. of EtOH, n (chain
NO. M M Xmax length) removal of a chlorine or bromine located meta to
1 ....... ........ 628 160 the same nitro group. Thus, the dehalogenation
2 1.2 x 10-4 ........ 625 134 of 2,5-dichloro- and 2,5-dibromonitrobenzene leads
3 6 . 1 x 10-4 ........ 621 108 to m-chloro- and m-bromonitrobenzene. The yields
4 5 . 0 x 10-3 ........ 613 77 in these cases are 37-38%.
5 3.0 X ........ 600 52 The monofluoronitrobenzenes did not dehalo-
6 2 . 1 x 10-1 ........ 572 30 genate under the conditions used. The 0-, m- and
7 5.0 X 10-1 ....... 557 22 p-isomers were all recovered without any forma-
8 ........ s.2 x 10-1 594 45 tion of nitrobenzene.
9 ........ 1 6 588 40 The results observed with the monobromonitro-
benzenes were comparable to those obtained with
Table I shows the values of Amax observed and the the chloronitrobenzenes. Only in the case of o-
corresponding chain lengths (n) which have been bromonitrobenzene did dehalogenation take place.
calculated with equation (1) and the VOvalue. While it might be expected that p-bromonitro-
Both the Amax and the estimated chain length benzene might be sufficiently more reactive than
change in a regular fashion with the iodide concen- the corresponding chlorine compound to permit
tration. Also, it appears that ethanol causes an dehalogenation, such was not the case. $-Bromo-
effect similar to that of iodide. nitrobenzene was recovered in 78% yield and there
Our data may not describe exactly the quantita- was no evidence that any nitrobenzene was formed.
tive relation between the chain length and the The results with the iodonitrobenzenes are quite
Am
,, because the length of the polyiodine chain a t different than those obtained with the other halo-
zero iodide concentration is estimated. Neverthe- nitrobenzenes. As would be expected, the ortho
less, our results indicated that Kuhn’s theory is isomer dehalogenated readily to give a 53% yield
applicable to the polyiodine chain. They also of nitrobenzene. It was not surprising to find
account for the behavior of iodine in shifting the that the presence of a nitro group para to the
absorption spectrum of the complex. iodine was suffcient to activate the iodine, so
Gilbert and Marriott” suggested that the amyl- that it was easily replaced by hydrogen upon treat-
ose complex is composed mostly of groups of the ment with copper and benzoic acid. The experi-
type (31~-21-),each probably a linear resonating ment with m-iodonitrobenzene indicates, however,
group which develops the pure blue color in the com- that the location of the nitro group, and probably
plex. According to our model, it may be that even its presence, is not an important factor in
iodide is chain terminating, even if it is in the reson- determining if iodine is replaced. The yield of
ating chain. nitrobenzene was 32y0 when m-iodonitrobenzene
I t is well known that the color of the complex was used.
disappears a t higher temperature. The authors The dehalogenation of 2-chloro-3,5-dinitrotoluene
intend to investigate this phenomenon spectro- provides a new synthesis of 3,5-dinitrotoluene,
photometrically. The presence of nitro groups located both ortho
We are indebted to Mr. Tamotsu Fukuda of the and para to the chlorine appears to enhance the
Osaka Industrial Research Institute for the meas- ease of dehalogenation and 3,s-dinitrotoluene is
urement of absorption spectra. obtained in 83% yield. This method of synthesis
(10) R . J. Hach and R . E. Rundle, THISJOURNAL, 1 3 , 4321 (1951). suffers from the low yield obtained in the prepara-
(11) G. A. Gilbert and J. V. hlarriott, Trans. Faraday SOC.,44, 84 tion of the 2-chloro-3,5-dinitrotoluene. The usual
(1948).
“mixed acid” nitration of o-chlorotoluene was not
LABORATORYOF BID-PHYSICAL
CHEMISTRY successful, even a t elevated temperatures. The
OF AGRICULTURE
FACULTY
UNIVERSITY
SANIWA use of concentrated sulfuric acid and sodium
S A K A I C I T Y , JAPAS nitrate was also unsuccessful. By using a mixture
_ _ _ _ _ . . ~ of nitric, sulfuric and fuming nitric acids, o-
Removal of Halogen from Aromatic Nitrohalo chlorotoluene may be dinitrated to give a 9%
Compounds yield of 2-chloro-3,5-dinitrotoluene.
BY WALTERT. SMITH,
JR., A N D LOUISCAMPANARO Experimental
RECEIVEDFEBRUARY 20, 1953 Dehalogenati0ns.-The dehalogenations and attempted
dehalogenations were carried out as illustrated below for o-
Earlier work has shown that the replacement of fluoronitrobenzene. The results are summarized in Table I.
chlorine by hydrogen in aromatic compounds (1) W. T.Smith, J r . , TWSJOURNAL, 71, 2855 (1949)