2628
CHEVESVALL LING
AND
chromatography; e < 100 in the region 220-300 m p ; intense carbonyl absorption at 5.78 p ; n Z 41.4311. ~ Anal. Calcd. for ClaHlsO?: C, T0.54; H, 10.66. Found: C , 70.51; H , 10.79. The last elution peak shown bj- each of the above product mixtures in gas chromatography (Fig. 1) corresponded in retention time t o t h a t of authentic 2-inethosyisophorone. A mixture of the pair of methoxyisophorones obtained in the present reactions was warmed on a steam-bath for 30 minutes with concentrated hydrochloric acid. This treatment converted the lower-boiling inethoxyisophorone component into 3,~,5-trimethylpheiiol, m.p. 108.5' (lit.3 n1.p. 108.5"), b u t left 2-methoxvisophorone largely unchanged. .Isample of 2-methoxyisophorone recovered from the hydrochloric acid treatment gave a single elution peak in gas Chromatography, and gave a 2,1-dinitrophenglhydrazonederivative, m . p . 190-191.5' dec. after recrystallization from ethanol. The structure of this product was confirmed by hydrolysisll to 2-hydroxyisophorone, IU . p . 91-93" (lit .I1 m.p. 92-93 O j .
[CONTRIBUTIOS FROM
THE
IIEPARTME'iT
OF
MICHAEL NAIMAX
Vol. 84
From the product mixture uf run C, a sample of 6-methoxy isophorone was obtained by distillation, b . p . 101-103" (12.5 m m . ) , T I ~ ~1.4735, D Xma, 236 m p (ethanol). Gas chroof 2matography revealed the presence of a few mole inethoxyisophorone as the only impurity. .-lnai. Calcd. for C H ~ O C ~ H I :C~, ~i :l . 3 9 ; H , 9.59; OCH3, 18.15. Found: C, 71.55; H , 9.88; OCH:,, 18.94. The 2,4-dinitrophenylhydrazone derivative of 6-methoxyisophorone was recrystallized from ethanol; 111. p . 129.5130.5O. . 4 n d . Calcd. for ClsHl,,S,Oj: C , 55.16; H , %5.78; S , 16.08. Found: C, 55.08; H , 6.08; S,16.17. In control experiments, known mixtures of the methoxyisophorone products arid of the methyl trimethylcyclopenteneCarbixylate products ( i f the above reactions were treated with inethanolic potassium hydroxide for 3 hours a t room temperature. Products recovered from the alkaline solutions were examined by means of gas chromatograph>-, and in neither mixture was any change in composition found to have resulted from the treatment with alkali.
CHEMISTRY, COLUMBIA U S I ~ E R S I T Y S E, \ Y YORK 27,
Organic Reactions under High Pressure. VI. Rearrangements
s.I*.]
Some Claisen and Cope
B Y CHEVES L ~ X L L I N GA N D MICHAELX A I M A N ~ RECEIVED J A N U A R Y 27, 1962 Measurements are reported on the effect of pressures up t o 6000 kg.jcm.2on the rates o f rearrangements of allyl phenyl ether ( I ) a t 160' and ethyl (1-ethylpropeny1)-allylcyanoacetate( 1 7 1 1 1 ) a t 119', each in several solvents. Both reactions are pressure-accelerated, A V *Is being generally -6 t o - 7 cc./mole between 1000-6000 kg./cm.2 and essentially solvent independent. Extrapolations t o atmospheric pressure are less certain, but indicate Al'*'s of - 10 t o - 15 cc./mole, again with no obvious solvent dependence. These results are interpreted as consistent with rather tightly bonded cyclic transition states in which new bond formation has proceeded farther than bond breaking, rather than with transition states which are essentially ion-pairs, or in which homolytic bond cleavage has been extensive.
,Ilthough there is a general agreement3 t h a t the mechanism of the Claisen rearrangement is represented by a sequence such as
I
0
11
?c:H= c H
@H~
OH &CH~CH=CH~
on reaction rates, and of the and stereochemical course' and geometric requirem e n t ~lo ~of the rearrangement. .I similar situation arises in connection with the Cope rearrangement of 1,4-dienes, usually activated by cyano, carbethoxy or similar substituents on carbon :?.l' Here the reaction has been shown to be intramolecular" and to possess a large negative entropy of activation,I3so that it may be considered a close analog of the Claisen rearrangement, occurring in an all-carbon chain system. ' I
11)
I11
IV
the detailed structure of the transition state I1 remains ill-defined. Several recent studies have attempted to establish more precisely the degree of bond breaking and bond forming, amount of ionic character and exact geometry which is involved through investigation of the effects of solvent^,^,^ I I ) Taken from t h e P h . D . nihsertation of Llichael S a i m a n , 1961 Support of this research b y a grant from t h e Petroleum Research Fund of the Amprican Chemical Society is gratefully acknowledged 1 2 ) Culumbia University FEIIow,1959-1960 (:i) For recent reviews, c t . ( a ) D . J . C r a m and .\IS. . Sewman "Steric Effects i n Organic Chemistry," John U'iley and Sons. I n c . . h-ew York. N . Y . , 1956, pp. 255-303; (b) E . S. Gould "Mechanism and Structure i n Organic Chemistry." Henrv Holt and Co.. I n c , Kew York, X . T.. 1939, pp. 611-649 (4) W. N. White, D. G . U'ynn, R. Schlitt, C Girard and W Fife, J . A m . Chem. Soc., 80, 3271 (1558). 1.5) H. I,. Goering and R R Jacobson. i b i d . , 80, 3277 (19.38). (6) U'. S . White and 1%'. K . Fife, i b i d 83, :3346 (1961); W. S . White, C . D. Slater and N' K . Fife, J Ore ( ' h e m , a?, 627 '1961)
It seemed to us that a study of the effect of high pressures of the rates of some typical Claisen and Cope rearrangements would provide further information on the nature of the transition states involved, and our results are described here. Since our work was completed, Brower14 has reported a study on the effect of pressure on three Claisen rearrangements leading to conclusions which are in qualitative agreement with our own. ( 7 ) E. S . Marvel a n d J. 1, Stephenson. ibid., 2S, 676 (1960). (8) A W. Burgstahler, J . A f n . Chem. Soc.. 8Z,1681 (1960). (9) L . D. Huestis a n d L J. Andrews, i b i d . , 83, 1963 (1961). (101 1%'. X. White and B. E. Norcross. ihid., 83, 1Y68, 326.5 11561). (11) A C . C o p e and E. M H a r d y , ibii:.. 62, 441 11940). (12) n. E . White and A . C C u p ibid , 66. 1999 (194X). (13) E. G.Foster, .?, C Cope and I. Daniel\, i b i d , 69, 18Y.'< (19471 (14) K . K Rrower, ibid.. 83,4:-lene glycol Ethanol 1-Hexanol Ethyl ether %-Butyl bromide Pentane Decane
(I In I'/dP X IOc,
!atm.-')
35 32 111 90 149 140 200
144
(18) S. D. Hammann a n d 13. Strausr. 7'vuiac I:oiailay S u i , 51, ID84 (lY35). (1Y) K . G. Gibsun, It. \V. I'awcctt and A I . W. i'crriri, I>rt,: ROJ, SUC.( L o n d o n ) , 8160, 223 (lY35). (20) Bridgeman's d a t a : fur summary cf, ref. 17, p'p. 21'9-226.
CLAISEN.im COPEREARRAXGENENTS
July 3, 19G2
solvent compressibility or variations in atmospheric pressure rates. Scheme 2 is our understanding of the transition state proposed by White for the Claisen rearrangem e n t 6 I n the extreme form it seems incompatible with our results, since other homolytic bond scissions of peroxides?' and azo compoundsz2show positive AV*'s of 5-13 cc./mole. Scheme 3, in which bond-forming precedes bond breaking, we consider in the best accord with our results. Here AV*s should be negative, and their upper values can be estimated from volumes differences between starting materials and cyclic products approximating the possible transition states. Molar volumes of some model systems are listed in Table IV. For the Claisen rearrangement the difference in molar volume of allyl phenyl ether and 1-chromene is 13.8 cc./mole a t 15', and would be somewhat larger a t 160'. For two models of the Cope rearrangements (hexadienes cyclohexadienes) differences are 16-1s cc./mole. As we have seen, different methods of extrapolation to atmospheric pressure yield somewhat different values of AV'*s from our data, hut the general range is smaller than the above differences, as would be expected for transition states in which two bonds are somewhat stretched. TABLEIV ( C c . ) OF MODELCOMPOUNDS MOLAR1-OLUMES ,411yl phenyl ether 135.Q Z 3 1-Chromene 120.124 4-Methylheptadiene-l,5 180,926 2,6-Dimethylcyclohexadiene-1,3 132. j2& 2,5-Dimethylhexadiene-l,8 146 425 1,4-Dimethylhexadiene-1,3 130.025
While our results thus favor cyclic transition states with marked bonding between fragments of the rearranging molecules for the two systems studied (which is certainly plausible, since the driving forces of the reactions must be chiefly the formation of new bonds), i t must be kept in mind t h a t the three schemes discussed represent limiting models of the possible transition states. Accordingly, contributions of other structures sufficient to account for phenomena such as the effects of substituents on reaction rates are certainly not ruled out. Indeed, in our present state of knowledge, i t is difficult to make clean cut distinctions between largely "u-bonded" transition states and species which are very tight ion pairs or charge transfer complexes, or even to say to what extent these formulations are operationally differentiable.
Experimental
2631
bromide in acetone in the presence of potassium carbonate; b.p. 8.5' (19 mm.), 1.5196. Reference samples of 2-allylphenol were prepared by rearrangement of t h e ether by refluxing at 180" and extracting with alkali; b.p. 72' (3 m m . ) n Z 61.5448. ~ Ethyl (1-Ethylpropeny1)-allylcyanoacetate, (VIII) was prepared in two steps from diethyl ketone, ethyl cyanoacet a t e and allyl bromide. T h e ketone and ester were condensed in the presence of acetic acid and ammonium acetate essentially a s described in reference 26. T h e resulting ethyl (1-ethylpropy1idene)-cyanoacetate, b.p. 122-125' (12 mm.) was alkylated with allyl bromide in the presence of sodium ethoxide, essentially as described by Cope and H a n ~ o c kfor ~ ~a series of similar compounds, to give VIII, b.p. 34-88' (0.7 mm.) n Z 5 D1.4618. A sample of VI11 was rearranged t o ethyl (1-ethyl-2-methyl-4-penteny1idene)cyanoacetate (IX) by heating 7 hr. a t l i 0 ' ; b.p. 147-148' (12 mm.), n 2 5 ~1.4775. Apparatus and Technique.-All experiments were carried o u t in collapsible Teflon vessels in the high pressure apparatus used previously.28 Aliquots of standard solutions were used in all runs, and reaction times were long enough, B-8 hr., so t h a t t h e short time required for samples to come to thermal equilibrium could be neglected. At the end of each experiment, samples were removed from the pressure system and analyzed a s described below. Analyses.-Early runs involving allyl phenyl ether were analyzed by acetylating the 2-allylphenol produced with a standard solution of acetic anhydride-pyridine, and backtitrating unreacted acetic anhydride with S a O H .29 Subsequently i t proved simpler to analyze samples by determining the optical density of aliquots, dissolved in 0 . 1 N NcOH in 50% ethanol, at 291 mp using a Beckman DU spectrophotometer. Allyl phenyl ether is transparent at this wavelength, while the 2-allylphenoxy anion absorbs strongly e = 3638. TABLEV REARRANGEMEKT OF A L L Y L PHENYL ETHERI S 1-OCTASOL AT 160" P,
kg./cm.'
1 1510 2130 2880 3620 4200 5030 5980
Time, X 101
Reacn.,
SCC.
25.2 25.2 25.25 25.2 25.3 25.2 24.75 25.2
70
6.33 10.3 11.8 14.3 16.7 17.1 19.3 23.7
k, sec. -1
x
108
2.59 4.31 4.97 6.12 7.23 7.44 8.66 10.75
TABLEV I REARRANCEMEXT OF VI11 IS ~-OCTANOL AT 119' (ALL R u s s 18,900 SEC.) P,kg./cm.l
1 1840 2210 2660 3430 3940 5100 5910
Keacn.,
19.3 34.4 36.6 38.0 42.7 43.8 50.5 56.7
%
k, set.-' X 1.13 2.23 2.41 2.53 2.94 3.05 3.72 4.42
1oj
Materials .--Solvents were commercial materials, distilled before use. Since most of our analyses of products involved absorption spectrometry, their optical absorption a t t h e wave lengths involved was checked before use. Allyl phenyl ether was prepared by treating phenol with allyl
The conversion of VI11 to IX was followed by determining the optical density a t 238 mp of aliquots dissolved in absolute ethanol. Although both materials absorb significantly at this wave length (for V I I I , e = 1045, for I X , e = 9789) they differ sufficiently for analysis. I n all analyses, three aliquots were made up independently, measured, and (21) C . Walling and J. Pellon, J . A m . Chem. Soc., 79, 4786 (IQZ7); the average taken.
C. Walling and G . Metzger. ibid., 81, 5363 (1959); A. E , Sicholson and R . G . W. Sorrish, Disc.Faraday SOL.,22, 97 (1956). ( 2 2 ) A. H . Ewald, i b i d . , 22, 138 (1956). (23) I. Heilbron, "Dictionary of Organic Compounds," Vol. I, Eyre and Spottiswoode, London, 1943. (24) P. .Maitte, A n n . chem., 9 , 431 (19.54). ( 2 5 ) G . Egloff, "Physical Constants of Hydrocarbons," Reinhold Publishing Corp., Piew York, N. Y.,1939, Vol. I and 11.
(26) "Organic Syntheses," Coll. Vol. 111, John Wiley and Sons, Inc., New York, S . Y . , 1955, p. 399. (27) A. C . Cope and E. M. Hancock, J. A m . Chem. S O L . 6, 0 , 2903 (1938). (28) C. Walling and J. Pellon, ibid., 79, 4776, 4786 (1957). (29) "Organic Analysis," Interscience Publishers, l n c . , Kew York, h7.Y., 1953, p. 24.
Treatment of Data.---Since this study ab primarily a survey of reactions under a variety o f conditions, rate constants were calculated from single experiments (e.g., one point) for each pressure in each solvent, a procedure which appeared justified, since the first-order kinetics of Claisen and Cope rearrangements a t atmospheric pressure are well established, and since the plots of rate I C \ . pressure (Figs. 1 : i r i c l 2 I give rens(inably smooth ciirvrs. Original dntn for
[COSTRIRUTIOS FROM
THE ISSTITUTE FOR
two seta o l runs are listed in Tablei \ ' and \ . I . Tlic value^ of A V#'s listed in Tables I and I1 were obtained fro111 the slope of lines obtained by the method of least squares, and experimental errors given are standard deviations. S o experimental errors are assigned t o the Benson-Berson extrapolations, which involve plots of (log k / ' k a ) / P *LIS. Pa.5z3, since points often scattered badly and extrapolation seemed t o involve a good deal of subjective judgment.
CHEMICAL I < E S E A R C H ,
K Y O T O 1 - S I V E R S I T Y , \-OSHII)A,
KYOTO, J A P A S
~
The Kinetics of the Reactions of P-Acetoxyethylmercuric, 6-Acetoxypropylmercuric and P-Acetoxy-P-phenylethylmercuricAcetate with Aromatic Substances BY KATSUHIKO ICHIKAWA, KOICHIFUJITA AND O S A ~ITOH J RECEIVED SEPTEMBER -5, 1961 The reaction rates of 8-acetosyeth~lInercuric,8-acetoxypropylmercuric and 8-acetos!--8-phen?.leth~-l~nercuric acetate with aromatic substances have been studied in acetic acid-perchloric acid-water systems. The results showed that the reactivity order of various aromatics is the same with those in the usual electrophilic aromatic substitutions. The relative rates of the three mercurials mentioned above with anisole were 1.0, 1.:3 and 710, respectively, under the same conditions. Froin these results, the mechanism is discussed as a n electrophilic aromatic substitution. As a preliminary experiment for the rate study above, the stabilities nf the rnercurials in acetic acid containing perchloric acid have been studied and the stoichiometry, rates and mechanism of the decornpositirms are reported.
In previous papers,' i t has been shown that $acetoxyalkylmercuric acetates (obtained by the addition of mercuric acetate to olefins in acetic acid) react with aromatics by eq. 1 The kinetics and
apparently through the formation of the addition compounds, i.e., P-acetoxy mercuric compounds. The stoichiometry of this reaction, however, is not yet established, because these oxidations have been carried out in the presence of excess mercuric Hacetate usually. \Vith the addition compounds of .\cO-C--C-HgO.lc + .\rH --+ , lower olefins, such as 11, no detailed results have been reported. Therefore, the product analysis .\r -C-C-HgO.\c + HO.\c (1) and kinetics have been studied. -It room temperature, the solution (50 i d . ) mechanism of this reaction between d-acetoxy- which contained 4,S.l millimoles of I1 and 1.03 ethylmercuric acetate and anisole have been re- molar perchloric acid in SOYc acetic acid gave S.0 ported also.? I t is desired, however, t h a t the ml. (at N.T.P.) of gaseous product (3.1 millimoles kinetic studies are extended to other mercurials as propylene). This gas was identified to be and aromatics more extensively as a new type of propylene by vapor phase chromatograph. The aromatic substitution. In this paper, the results analysis of the liquid mixture showed that 5.1 of the rate studies of the reactions of @-acetoxy- millimoles of mercurous acetate was formed. ethylmercuric ( I ) $-acetoxypropylmercuric ( I I) -1lthough the quantitative analysis of the liquid and +acetoxy-J-phenylethylniercuric acetate (I1I ) organic products was difficult, propylene glycol (prepared by the addition of mercuric acetate to diacetate and its hydrolyzed product glycol were ethylene, propylene and styrene in acetic acid) found to be formed which were expected from the with various aromatic substances are reported. results in terpene chemistry mentioned above Unfortunately, these mercurials are not always (no acrolein mas detected). These results show stable enough for the kinetic measurements of re- that the reaction of I1 in the presence of perchloric action 1 in the presence of perchloric acid which is acid can be expressed by ey. 2. The results deone of the typical catalysts Ais a preliminary Hstudy, therefore, the decompositions h a r e been 2CHaCH--CH? -+. studied. HO.\c 0.1~HgOAc Stabilities of I, I1 and I11 in Acetic Acid ContainCH{CH--~CH?+ Hg?(O.\c), + CH:CI-ICH: (21 ing Perchloric Acid.-In 80% acetic acid containing perchloric acid up to 2 molar, I is stable indefinitely. o.\c o.\c .It room temperature, no appreciable decomposiscribed above were obtained in the range of less tion was observed even after 5 days. V3th I1 and 111, however. a reaction to form mercurous than about 20cq reaction. .It higher reaction salt was observed, while their acetic acid solutions percentages, the situation is much more complicontaining no perchloric acid are much more stable. cated. The mercurous acetate produced may disMany examples of this type of reaction can be proportionate to g i w metallic mercury and merfound in terpene chemistry, since the oxidation of curic acetate which reacts with propylene to give terpene with mercuric acetate to form the cor- I 1 again. The reaction equation in this case responding diacetate and mercurous salt proceeds might be expressed by the equation I
(1) K . Ichikawa, S Fukushima, H . Ouchi and hl. Tsuchida. J A m C h e m . S o r , 8 0 , 600.5 (19.58); 81, 3401 (1959). ( 2 ) K . Ichikawa. K Fujita and H Ouchi. i b i d , 81, 5316 (1959)
(3) T h e formation of acrolein by the reaction of propylene with mercuric sulfate in aqueous sulfuric acid (du P o n t , 5.S. Patent 2.197.258> appears t o belong t o a different t y p e o f reaction
