Sept., 1942
STRUCTURE OF THE DEHYDRATION PRODUCT OF ANHYDRACETONEBENZIL 2123
inden-1-one (IV, IVa), forms prisms on crystallization of reagent without evolution of gas, and regenerate the from benzene-petroleum ether. The melting point is parent ketone IV. The bromine is not removed by pyridine or potassium usually 178-179 ', but occasionally a low-melting form is obtained; this melts a t 14&146", but on admixture with acetate (e. g., there is no adjacent hydrogen atom in the the higher melting form, the observed melting point is alpha position to the carbonyl group) but alcoholic potash 178-179°.4 The yield was 60%. gave an intractable black oil. It is unaffected by hydroAnal. Calcd. for CasHsoO: C , 91.1; H, 5.9. Found: gen bromide, by acetyl chloride, or by excess bromine. C , 90.9; H , 6.1. Summary This ketone does not react with maleic anhydride nor The bimolecular product, resulting from the form an oxime. In the Grignard machine it evolves 0.67 action of acidic dehydrating agents upon anhymole of methane, indicating that it is enolized to a condracetonebenzil, is transformed into a carboxylic siderable extent. An acetate is secured in a 70% yield by refluxing an acetyl chloride solution for half an hour; it acid by the action of alkaline reagents. The separates from benzene-ligroin in prisms that melt a t 115'. carbonyl bridge disappears in the process, being Anal. Calcd. for CalH3202: C, 88.5; H , 5.8. Found: converted into the carboxyl group. C, 88.1; H, 6.0. The acid is decarboxylated and dehydrogenWhen treated quantitatively with methylmagnesium ated to give a new dienone, isomeric with one iodide, the acetate consumes two equivalents of reagent previously known. without evolution of gas; upon acidification the ketone IV The new dienone is also obtained by the action is obtained. The bromoketone, 3a,4,7,7a-tetrahydro-7a-bromo-2,3,5,of mineral acids on the one described earlier, and 6,7-pentaphenylinden-l-one (X) was secured by bromina- both are isomerized to the same known indanone tion in chloroform in the usual manner; it crystallized from by acids or heat. chloroform-alcohol solution in pale yellow rods; m. p. 218The new dienone gives products with phenyl219". It separated from benzene solutions in pointed rods magnesium bromide formed by both 1,2- and with solvent of crystallization'; the solid softened a t about 144 ", finally melting a t 234 '. 1,4-addition. The 1,4-addition product is an Anal. Calcd. for (A) CssH2sOBr: Br, 13.5; for (B) easily enolizable ketone. CasHlgOBr.CeH~:C, 80.5; H, 5.2; Br, 11.9. Found: The 1,2-addition product is a carbinol, isomeric (A) Br, 13.8, 13.2; (B) C, 80.4, 80.2; H, 5.1, 5.1; Br, with one obtained from the closely related inda11.9, 12.0. In the Grignard machine both forms consume one mole none. Both are dehydrated to the same hydro(4) Drs. Jelley and Titus of these Laboratories report that examination of these substances failed to show any significant differences of readily determinable optical properties.
[COMMUNICATION NO.
855 FROM
THE
carbon, which is also formed from the isomeric dienone. ROCHESTER,
NEWYORK
RECEIVED MAY5, 1942
KODAKRESEARCH LABORATORIES]
The Structure of the Bimolecular Product Formed by the Action of Acidic Dehydrating Agents on Anhydracetonebenzil BY C. F. H. ALLENAND J . W. GATES,JR. Several years ago' from an exarr,ination of the substance obtained by degrading the bimolecular product, formed by the action of acidic dehydrating agents upon anhydracetonebenzil I, a structure, 111, which seemed in best accord with the available evidence, was suggested for this substance and a mechanism proposed to account for its formation. Others interested in this field2v3 appear to have accepted these conclusions. The salient features were the dehydration of anhydracetonebenzil to diphenylcyclopentadienone 11,which (1) Allen and Spanagel, T m s JOURNAL, 66, 3773 (1933). (2) Dilthey, private communication. (:O Rurton and Shoppee, J . ChPm. .SOL , 201 (1934).
then underwent a diene synthesis with itself, forming the bimolecular product 111. The latter lost carbon monoxide on heating, to give an indenone that was rearranged to an isomeric indanone IV, which was then degraded, stepwise, by unambiguous reactions to o-terphenyl. Later on4 several of the intermediate degradation products were synthesized, so that the position of the various groups around the benzene ring was established. Only the pertinent formulas are given below. The two novel features were the presence of a phenyl group on a carbon atom common to two (1) Allen, Bell, 13cll and VanAllan, Tlirs
JOIIHNAI..
63,I j A (1940)
2 124
C.
F.H . - 1 ~ 1 . m .\ND J . W. GATES,JR.
Vol. 6 4
both concepts, and to conclude that the phenyl group is in the eo 1 co c position where it is found in the I ' /co --+ CeH,C, indanone, alpha to the carbonyl; \CH/ C/ c~H,c-CH~ C$H,&=CH the new structure is written in H OH 0 formula V. r I1 111 It differs from the older one first H proposed' only in an interchange CaH /\ CsH 4 'COCsH, GH -iCsH, of the angular phenyl group and f1 1 CeHo\,, / CoII,JC0OII c6Hj()L i i -H one hydrogen atom. This was ac1 CoH counted for by assuming a 1,3-shift 11- 0 of a phenyl group. Such a rerungs in 111, and its apparent 1,d-shift to the posi- arrangement was, indeed, proposed in the first tion in which it is found in the indanone Ii-. No paper, but a t a different point, namely, following similar instances have been recorded in the litera- the decarbonylation. The experimental conditure. The formation of thiophenol, on sulfur tions under which the bimolecular products are fusion of the dienone formed after the decarbonyl- formed, favor such allylic rearrangements. ation of 111, was interpreted in favor of the struc'I'he mechanism of the reaction is, as previously ture having the phenyl group on the bond com- outlined, a diene synthesis. Alternative interrnon to the two rings, by analogy with the behav- pretations have had to be excluded, because of the lor under similar conditions of Ruzicka's alkyl various experimental conditions under which the decal in^.^ Further work was accordingly under- bimolecular product can' be formed, and the fact taken, with the view of clearing up the uncer- that substituted anhydracetonebenzils can partainty. take only if there is a t least one hydrogen atom As a resultfiisomeric indenones were discovered left to be removed as water with the hydroxyl and the prevalence of allylic rearrangements in the group (this is, of course, essential to the producseries was noted. I t was found that sulfur dehy- tion of the intermediate cyclopentadienone). hogenation (a high temperature R CoH, CfiH,C-CK C'aH ,C=CK C ! I eactiori) gax.-e varying amounts CoHsC' 1 \Co f thiophenol in instances where c=-0 3 co I[ co , fCsH. CaH,C\, I ,C, X R there wa< nco rea5onable doubt C'hH,C C!HK ii
Il9.42)
STRUCTURE OF THE DEHYDRATION PRODUCT OF ANHYDRACETONEBENZIL 2125
Sept., 1942
loses carbon monoxide on heating, it is the keto group on the indene ring that has been attacked, i. e., the one addition has occurred here. But, since no exposed carbonyl group could persist unchanged, the carbonyl bridge must have been protected in some way from the action of the excess reagent. The obvious explanation is that it forms an enolate, H V I I , t h u s acC----cC& counting for the CsHsC/F]H\ 1 C-OMgX 1 IICCeHj one active hyC6HdC\ /c\c / drogen observed ; VI1 R 'OMgX acidification then regenerates the H original carbonyl CH I C6HsC'I-'C-CCbH,, group.
&
been prepared in which R = CH3, C6&, a-CloH;. All carbinols lost carbon monoxide and water, when heated, and gave hydrocarbons. The carbinols were studied in more detail; in the Grignard machine they showed two active hydrogens (one due to enolate and one to carbinol). The hydroxyl was replaced by halogens on treatment with acid halides; they gave isomeric acetates with acetic anhydride and acetyl chloride. The acetates (VIII: R = CHa, X = OCOCH3) showed one active hydrogen and two additions in the Grignard machine, and regenerated the carbinol, as would be expected. The bimolecular product was previously reported1 to give two stereoisomeric dioximes. Both of these have now been hydrolyzed, with production of the bimolecular product; this excludes the possibility that the starting material was a mixture. The bimolecular product was brominated to
excludes any of the open-chained isomeric structures that might be written, confirming the conclusion already mentioned above, concerning the formation of bimolecular products from substituted anhydracetonebenzils. I n the Grignard machine it consumes two moles of reagent without evolution of gas. The product is a di-carbinol IXa, for it evolves two equivalents of gas when treated with excess methylmagnesium iodide. These reactions afford confirmation of the assumption previously made in regard to the bimolecular product, that the production of gas is connected with the presence of an enolizable hydrogen; in the tribromoketone there is no alpha hydrogen, and the two carbonyl groups behave normally, giving two additions.
R
VI11
H
'I'he monochloroketone is isomeric with otic described as a result of chlorination of ;3,4-diphenylcyclopenten-3-one1 .* It was mentioned earlier that hydrocarboiib were formed by loss of water during decarbonylation. From the phenyl carbinol (VIII: R = C6H6, X = OH) the hydrocarbon obtained was identical with one secured from 2,3,3,(i-tetraphenylindenone X by the following sequence of reactions
A
C6H6 Br " _
0
X
XI
XI1 (R =
form a tribromoketone IX, gave a monochloroketone with phosphorus pentachloride, did not condense with aromatic aldehydes, and was destroyed by concentrated sulfuric acid. The tribromoketone is the result of complete replacement of all the available hydrogen atoms in the alpha positions to both carbonyl groups. Its formation
VI11 X
-CO =
OH)
I
_1
-Hzo
I t is obvious from inspection of the formulas that in one instance there has been a rearrangement. The substance in question is also isomeric with a closely related hydrocarbon described previously6; a discussion of these and other related compounds will be considered separately. (8) Burton and Shoppee, J Chein Auc , 20.-1(1934).
phenyl group is involved, or that the ani1 is hydrolyzed in part to its components, the aniline so formed then adding to a molecule of unchanged mil, with subsequent ring closure and dehydrogenation. In considering the first alternative, i t should be noted that -before dehydrogenation the substance xiThas an allylic system, in which rearrangements are facile and common. CoH,
H
--+
l
€I
XI
,!
,:
I , ,I
The Grignard reaction products (carbinols, VIII) were prepared in the usual manner with yields of i5-85%, There was nothing new in the experimental procedure5 employed when they were treated with acetyl chlo,/\ \, ride, acetic anhydride and a trace of sulfuric acid, \ jCsH, thionyl chloride, hydrogen bromide in glacial XIV acetic acid or phQsphoru5 pentachloride. The proper&\ of the various compounds are given in Table I.
--'J
".
I
XI11
Chlorination: The monochloroketone was secured by refluxing for one-half hour a benzene solution of 4.6 g. of the bimolecular Product and 6.2 g. of phosphorus Pent&chloride. It separates in prisms from benzene, mi. p. 215'. A n d . Calcd. for CsiHl,OpC1: C1, 7.1. Found: Cl,
H
The explanation involving a 1,S-shift of a phenyl group seems preferable in this case, for Peine showed that cinnamalaniline showed a remarkable stability to acids, there being but a very slight hydrolysis to the components. I t is much simpler than the necessary series of consecutive reactions required by the second alternative. We have found that cinnamalaniline is monomolecular, thus excluding interpretations inYolving polymeric forms.
h O P E K I I E S OF
Formula VI11
OH OCOCH3 OCOCHI C1' H r' OH OCOCH,
Experimental The bimolecular compound V and its two dioxirnes were prepared as reported previously.' Both diosimes (in. p. 176 and 229") were hydrolyzed by refluxing a suspension of 4 g. of the dioxime in 100 cc. of alcohol containing 25 cc. of concentrated hydrochloric acid for twenty-four hours; after cooling, the precipitate was filtered. The yield was 2.*5 g . ; the melting point, 202204", was not lowered Lvhen niised with the purc hitnolectilar product. Bromination: 3a,4,7,7a-Tetrahydro-2,3,S,B-tetraphenylI,T,ia-tribromo-4,;-inethanoindcn1 ,8-dione. IX; a mixture of 100 cc. of glacial acetic acid and 5 g. each of broinine and the bimolecular product was heated on the steam-bath for seventy-tkvo hours; after the usiial manipulative procedures, I .3 g. of tribronioketone was obtained. It cryqtallizes from benzene--ligroin in prisnis, m . p. 229.-
x
R
C1
OH Rr'
TABI.~., I RELATED TO
SUBSrANCES
\-ield, M. p , % OC.
262 202 180 219 191 226 235 216 295 233
THE
CARBINOLS
Analyses, % Calcd. C H T d H
8 i 5" 5 X 87 6 5 8 85 l h 5 i 85 3 5 x 8-5 1" 3 i 84 9 .i 9 (C1 7 1) ic1 7 a) (Br 14 7 ) (Br 14 3 ) 88 6' 5 5 88 6 .i3 86 3h 3 5 85 6 5 rj ic1 6 3) (C1 6 4) 89 2' 5 4 89 1 *? *-> I
(Br 12 2) (Br 12 2) Prisms, from xylene. I' Prisms, from benzene -petroIcuin ether. Rods, froin benzene -petroleum ether. Needles, from benzene- -petroleum ether. e Two active hydrogens, no addition in Grignard machine. I Action of ac-etyl chloride. Action of acetic anhydride and oiic drop of concentrated sulfuric acid. One active hydrogeii, t w o additions i n Grignarcl machine and recover carbinol. ' Action of phosphorus pentachloride. action of X 2 ~ o hydrogcn bromide i i i acetic acid. Action of thioriyl chloridc. "
''
l'
;1,3,5,ii-Tetraphenylirtdenoiic X was obtained in a 20c/', yield from the indenone, ni. p. l f j i 0 , lor in a i O % yield from its isomer, ni. p. 240°,6 by heating a t 220-280" for A n d . Calcd. for CS4H2104RrB: Rr, 34.2. Found: Br, fifteen minutes with sulfur; there is a strong odor of thio33.8. phenol in each instance. 2,3,5,6-Tetraphenyl- 1,8-dimethyl-1,7,ia- tribromo- 1,81,2,3,5,6-Pentaphenylindenol-l, XI, was obtained in a dihydroxy-3a,4,7,'Ta-tetrahydro-~,i-methanoindene, IXa, yield of 87% from 2,3,5,&tetraphenylindenone and was formed when the diketone IX reacted with methyl- phenylmagnesium bromide in the usual manner. It sepamagnesium iodide. I t crystallizes from benzene-ligroin in rates from benzene-petroleum ether in prisms, m. p. 220". rods, m. p. 278". A n d . Calcd. for C ~ P H ~ ~C, O 91.4: : H 5.t5. Fourid: -. C , L41.3; H , 3..i. (!I) l'einc. l i t , l?, 2117 (1884).
$>y)
-
