THESTRUCTURE OF GOS~YPOL
Sept., 1937
Analyses for sulfur were either too low or too high for cyclic polymers. Also, a cyclic polymer would not possess a variable melting point since it would represent one molecular species. These facts indicate rather clearly that the above polymers are linear. The length of the chain and consequently the physical properties were determined by the conditions under which they were formed. If the terminal valence were satisfied by non-sulfur groups, the culfur content was found to run low, and vice versa. Vacuum distillations of the polymers in the anthracene ?cries led to results entirely different from those obtljned io the atqhalirlene and benzene aeries. The pdymers, m. p. 366 sad W oupon , being hated abwc their melting pbif1t5, gave tars from which nothing detkite was isghted except 9-methylanthracene. Much hydrogen s a d e was
evofved.
[COFITRIBUTION FROM THB C H E M I W
1723
Summary 1. The 2,4-dinitrophenylhydrazonesof anthracene-aldehyde-9and 8-ethoxynaphthaldehydewere characterized. 2. A monomeric thioaldehyde was isolated and its tendency to polymerize was studied. 3 . An excellent synthesis for naphthyl substituted ethylenes was obtained. 4. Several new polymerized thioaldehydes were prepared. 5. The &ect of group size uport ring closure was discussed. CHAPELHILL,N. C. RECEIVB~ JWNE 23, €937
LABOWTORY OF THE UNIVERSITY OF
fLLINOIS]
The Stiucture of Gossypol. 1 BY K.N. CAMPBELL, R. C. MORRIS'AND ROGERADAMS Gosoypo1,2 the yellow pigment present in by limiting the study to the prepatation of those cotfunseed, was first isolated by Marchlew~ki.~derivatives of definite, unquestioned purity, Although it has been investigated in several could important contributions to the knowledge lab~ratories,~ no structural formula has as yet of the chemistry of gossypol be accomplished. been suggested for it. The principal difficulty thus far encountered in the study of gossypol has been due, primarily, to the lack of methods of preparation of more than aT very limited number of pure crystalline derivatives or degmdation produds suitable for stntctural d d d . Gossyp~lis highly reactive chemically but is sensitive to oxidation, partrdarly in alkaline salution, and exhibits an +al tendency under the usual experimental conditions to ykld amorpholls or ill-defined e r y s t d b wmpobds or mixtures extrwrdhwily resistant to pluribcation. he preEiminaq resu~tsin this investigation were sirnifav and led to the conclttsisn t&t only (21 A pmtW d a rtrrsia snbdtsd ia POrQa lalimrrrt tb Degree d Date# d PMlosOphy. (2) T b mthara are Webted to the M e y e Coftbn Oil wrp, nrbsidiary of The Proctor pnd Gamble Company, for a gift d ttra m t s fwkich the gossypol ' 1 ~ extracted. s taqYirrmcnta for
Mpobkrpld i ,W d t . cbcm., m, 84 (Isso). Wktters, J . Ch. 245 (1917): Rhwmtz &&E, 3. Abt Reresnh, S6,285 (I028); (bfCatruth dnd Winters, ibid., 12, 83 (1918); (c) Gallup, J . BioZ. Chem., 77, 437 (1928); (d) €!=ruth, Tms JOURNAL, 40, 647 (1918); (e) Clark, J . B i d . Chcm., 76, 725 (19271, (f) i b d , 76, 229 (1928); ( 9 ) +brd , 77, 81 (1928), (h) ibrd., 78, 159 (1928), (i) THISJOURNAL, 51, 1475, 1479 (1929): Oil &' Fat Industries, 6, July (1929), ( j ) Karrer and Tobier, H d v . Chim. Acta. 15, 1204 (1932), (k) Schmid and Margulies, Monatsh., 66, 391 (1934); (1) Grunbaumbwna and Marchlewski, Biochcm. Z . , 286, 295 (1936); (m) Podolskaja, ihid , 284, 401 (193fij; Frttchrmisrhr Uwisrhllil , 42, 96 (1935).
1sy
(4') (al ChdVtk d
m.
The extraction of gossypol, its purification, the explanation of its various melting points and its color reactions are described in this preliminary paper. Many researches have been reported on the methods of isolation. The procedure found most satisfactory is that described by C s r r r ~ t h , ~ ~ who extracted the oil from the ground seed with petroleum ether and then extracted the gossypol with ether. The gossypol was isolated as gosypol-acetk acid. A more rapid method of extraction described by ClatkBaka wtts investigated, but dtle to thc large a'mount of oil p-sent the separation of goss y p o l - w - k acid was found to be extremely slow. The method of Carruthwas adopted in prefemwc. The gosqpcrl-gatic acid was converted into gossypOl by clissulving in ether, adding water and evaporalhg tfte: ether from the mixture. A brownish p h c t is #us obtained which is difficult to get perfesffy pure by clpst.a&ation and which, even after purificaton, gives a reddish solution when dissolved in ether. The recorded observations that gossypol is oxidized readily and is not reduced by sulfur dioxide led us to a slight modification of the procedure which proved (.i) Clark, Oil
i" Far I t i d m t r i f ~6, ,
13 (1929)
1724
K. N. CAMPBELL, R. C. MORRISAND ROGERADAMS
to be highly advantageous. A small amount of sodium hydrosulfitewasmerely dissolved in the water used fordecomposingthegossypol-aceticacid. The pmduct thus obtained is crystallized readily in the form of brilliant yellow needles, m. p. 184'. which dissolve in ether to give a yellow solution. Several melting points have been reported for
Plate I.-Gossypol.
m. I). IN'.
Vol. 59
gossypol, 184°," 199O." and 214°,4e but no explanation of the wide variance has been offered. A careful study has revealed that three crystalline forms of this substance exist, each with a characteristic melting point, and that intermnversion of the forms may be accomplished by use of the appropriate solvents.
Plate 2.-Gossypol.
m. 1). lg!)'
Plate 3.-Gossypol. m . p. 214". Plate 4.--Gossypol-acetic acid. Piales I-.i.-X-Ray diffraction pattemsof crystals. The X-Ray diffraction patterns were registered using ihc mono. chromatic pinhole technique. the metalix iron target tube operating a t 15 milliamperes under a potential of 24 kilovolts. the sample thickness being adjusted to abrorb approximately two-thirds of the incident intensity. The plate to specimen distance was 5.0 em. and the exposure time three hours. (See opposite page for Plate 5).
THESTRUCTURE OF &SSYPOL
Sept., I937
1725
When pure gossypol, m. p. IW', is recrystallized from chloroform, it undergoes a change in crystalline form and the product melts at 199': crystallization from ether causes a reversion of the 199' to the 184'. The form melting a t 199' may he converted into that melting a t 214" by recrystallization from high-boiling ligroin (b. p. C i l l O O ) . The reverse change is effected by rccrystallization from chloroform. The form melting at 184' may be changed into the one melting a t 214' by the use of high-boiling ligroin (b. p. (iO-llOo). Ether reverses this change. 'l'heie experiments are summarized in t h following diagram: Gossypol. m. p. 184'
1
1
Ether? Ligroin Ligroin J. M. p. 211' M. p. 15%' CHCh
ISthert I CHCh
1 J.
1
X-Ray diffraction patterns were made of these various forms and showed unequivocally that the products are crystalline (Figs. 1, 2, 3 , 4, 5). Examination under a polarizing microscope showed that the various modifications have different optical properties and different crystalline forms (Figs. A red isomeric form of gossypol has been isolated recently by Podolskaja.'" This material melts a t 1S4l%' and gives the reactions of ordinary yellow gossypol to which it is readily converted by means of solvents such as alcohol or ether. Yellow gossypnl in chloroform gives a red solution when treated with dry hydrogen chloride, but this mlor is not due to red gossypol, since yellow gossypol can be recovered readily by precipitation with ligroin. Merely grinding the red form in a mortar converts it to the yellow. Cfissypol reacts readily with acetic acid to give gossypol-acetic acid, which is a very convenient intermediate in the extraction and purification process. Analogous derivatives are formed with other organic acids. Crystalline compounds composed of one molecule of gossypol and one molecule of the acids, formic, propionic, and butyric, have been prepared. I t is not impossible that the highly-colored products formed when gossypol is treated with mineral acids are of a similar character. Gossypol has been reported hy several inves14) The difiraietiom patterns m d photomsphr were kindly msde lor us by Dr. J . N. Mrgudieh 01 the Division oi Analyticst Chemistry the woive.rity iiiiaaoiS.
or
l'lale 5 - Ked g
