ENOLIZATION OF CYCLOALKANONES AND CYCLOALKYL PHENYL KETONES
Aug. 5, 19G2 [CONTRIBUTION FROM
THE
DEPARTMENTS OF CHEMISTRY OF THEOHIOSTATEUNIVERSITY, COLUMBUS, O., AND UNIVERSITY OF CINCINNATI, CINCINNATI, 0.1
2905 THE
The Effects of Ring Size on the Rates of Acid- and Base-catalyzed Enolization of Homologous Cycloalkanones and Cycloalkyl Phenyl Ketones BY HAROLD S H E C H T EAND R ~ MICHAEL ~ J. C O L L I SAND , ~ ~ RAYMOND D E S S Y YUZI , ~ ~ O K U Z U M I L b * CA N D ALANC H E N ' ~ RECEIVED MARCH2, 1962 T h c rcactivity for cnolization of cycloalkanones in hydrochloric a ~ i d - 9 0 7acetic ~ acid as a function of ring size is G > S > t h a t for cnolization by deuteroxide in dimethylformamide-deuterium oxide is 4 > 5 > 6 > 7. The reactivity order for acid-catalyzed enolizaLion of cycloalkanones is interpreted primarily on the basis of differences in steric requirements in conversion of the ketones t o transition states having endocyclic unsaturated character which is highly developed ; t h a t for base-catalyzed enolization is based on the s-character of the carbon orbital directed toward enolizable hydrogen as a function of ring size. The reactivity for acid-catalyzed enolization of cycloalkyl phenyl ketones with respect t o ring size of t h e cycloalkyl group is 5 > 4 > 6 > 3 ; for base-catalyzed enolization: 3 2 4 > 5 > 6. The relative rates of enolization of the cycloalkyl phenyl ketones by acid are generally correlated by the steric and hybridization requirements for conversion of a tetragonal atom t o near trigonal in saturated monocyclic systems. The sequence for cycloalkyl phenyl ketones in t h e presence of bases is attributed t o factors similar t o that considercd for base-catalyzed enolization of cycloalkanones.
5
> 7 > 4;
The present study is concerned with the effects of ring size on the rates, kinetic parameters and mechanisms of enolization of homologous cycloalkanones and cycloalkyl phenyl ketones in acids and in bases. Model dialkyl ketones and alkyl phenyl ketones have also been investigated for purposes of comparison with related cyclic derivatives. Acids and bases catalyze enolization of ketones by similar mec:hanisms. In the presence of strong acids, a ketone is converted reversibly to its conjugate acid (eq. 1); loss of an alpha proton occurs by reaction with a nucleophile in the rate-determining step to give the corresponding enol (equation 2 ) .z Enolization by strong bases2n3involves reaction of a base and a ketone with assistance of the
1 " "i B--H ~
I
weak electrophiles present to yield the enolate ion (eq. 3). The transition states for enolization in the presence of acids (I) and bases (11) differ with respect to charge, charge distribution and to the strengths of bonds developed by the nucleophiles and electrophiles. (1) (a) T h e study of acid-catalyzed enolization was effected by authors l a with suppurt from t h e Graduate School of T h e Ohio State University, E. I. d u Pont de Nemours and Co., and the Office of Ordnance Research (Contract No. DA-33-019-ORD-3218). (h) T h e investigations of base-catalyzed enolization were conducted b y authors l b with support by t h e Petroleum Research Fund ( P R F 417A). (c) Petroleum Research Fund Predoctoral Fellow, 19.59-1961. (2) (a) Catalysis of enolization of ~-phenylisocaprophenoneby acetic acidZb+ does not involve simultaneous addition of t h e acid t o t h e carbonyl Ernup and nucleophilic attack a t t h e a-position during t h e ratedetermining step, h u t rather a pruton transfer in a rapid equilibrium process and subsequent nucleophilic reaction h y aretate iun; ( h ) C. C.. Swain, A. J. UiMilo and J. 1'. Curdner, J . A m . Chem. SOL.,8 0 , 5083 (19.58); (c) C. G. Swain, E. C. Stivers, J. F. Keuwer, Jr., and L. J. Schaad, ibid., 80. 5885 (1958). (3) A. Lapworth, J. Chem. SOC.,86, 30 (1904).
The over-all rates of acid- and base-catalyzed enolization may be determined using methods which involve halogenation to give a-haloketones, isotopic exchange of a-hydrogen or racemization of enolizable asymmetric center^.^ As a result of isotope effects on rates of e n o l i ~ a t i o n , ~the ~,~,~~ stereochemistry of ketonization of enols5h and correlations of reaction rates and equilibria on the basis that an enol is a high energy intermediate relative to ketone,jc it is apparent that the transition state I for acid-catalyzed enolization is very close to enol in geometry. Related data and logic also indicate that the structure of the transition state I1 for base-catalyzed enolization has certain enolate c h a r a ~ t e r . ~ ~ Enolization of acetophenones in the presence of acids is accelerated by electron-donating and retarded by electron-withdrawing substituents in m- or p-positions.6a The effects of electron-donating and electron-withdrawing groups on the rates of base-catalyzed enolization of acetophenones6b are opposite to that for the acid-catalyzed reactions. Halogenation of alkyl methyl ketones and cycloalkyl methyl ketones in acidic media generally occurs in alkyl and cycloalkyl rather than in methyl group^^^,^; base-catalyzed halogenation (4) H. h1. Dawson and M.S. Leslie, ibid., 96, 1860 (1909); P. D. Bartlett, J. A m . Chem. Soc., 56, 967 (1934); C. K. Ingold and C . L. Wilson, J . Chem. Soc., 773 (1934); C . K. Ingold, S. K. Hsu and C. L. Wilson, i b i d . , 78 (1938); W. D. Walters and K . F. Bonhoeffer, Z . physik. Chem., 8182, 265 (1938). ( 5 ) (a) W. D. Emmons and M. F. Hawthorne, J . A m . Chem. Soc., 78, 6593 (1956); (b) H. E. Zimmerman, ibid., 7 8 , 1168 (1956); (c) 0. S. IIammond, i b i d . , 7 7 , 331 (1055). (6) (a) D. P. E v a n s . V. G. Morgan a n d 11. B. Watson, J. Chem. SOC.,1167 (1936); (b) V. 0. hlorgan and 11. B. Watson, i b i d . , 1173 (19:L-,). ( 7 ) ( a ) 11. h l . E. Cardwell aiid A. E. 11. Kilner, i b i d . , 2430 (l!J51); (b) enolization of such derivatives is of t h e Saytzeff type in t h a t t h e more substituted enol is formed t h e more rapidly; (c) H. M. E. Cardwell, ibid., 2442 (1951); (d) t h e position of attack of alkyl methyl
of such ketones takes place preferentially in the methyl g r o ~ p s . ~Such ~ , ~information along with limited kinetic data8 have led to the conclusion that acid-catalyzed enolization of ketones is significantly controlled by hyperconjugative factors,ia whereas the base-catalyzed process depends primarily on inductive effect^.^^^^ Recently it has been suggested that steric factors are important in acid- and base-catalyzed enolization of cyclopentyl phenyl ketone and cy-clohexyl phenyl ketone5a and in acid-catalyzed enolizationg of cyclopentanone and cyclohexanone. Experimental Acid-Catalyzed Bromination. Reagents.-Cyclobutatlulle ( Aldrich Chemical Co .) was purified by precision rectification; its purity was attested by vapor phase chromatography. Cyclopentanone, cyclooctauone and di-n-propyl ketone were obtained oia decomposition of their highly purified semicarbazoues with aqueous oxalic acid, steam distillation and fractionation. Cyclohexanone and cyclooctanone were purified by decomposition of their bisulfite addition compounds followed by fractional distillation. The cycloalkyl phenyl ketones were prepared by described niethodslo and purified by rectification. Initial purification of cyclopropyl phenyl ketone was effected through its semicarbazone; cyclohexyl phenyl ketone was obtained in highest purity by recrystallization from petroleum ether a t -70" and subsequent vacuum sublimation. The acetic acid used throughout the acid-catalyzed enolizations was refluxed over excess chromium trioxide, distilled, and rectified in a bubbleplate column. Acid-catalyzed Bromination Rates .-The rates of acidcatalyzed bromination of the various ketones were determined upon adaptation of the method of Evans.8 I n general the rates of bromination were determilied under pseudo-first order conditions at 14, 20 and 29.9'; bromination of acetone, di-n-propyl ketone and cyclobutduone were investigated as pseudo-zero order processes (the ketone was used in 20-fold excess)." The extent to which a reaction was followed depended oii the ketone and the temperature. In general, reaction was followed much nearer to completion with the more reactive ketones. Thus, a t 30" cyclopentyl phenyl ketone was followed until 75tq of the ketone had reacted; bromination of cyclohexyl phenyl ketone was followed only to 30y0completion. Bromination of the cycloalkanones is complicated by the facts t h a t their rates of enolization are considerably greater than t h a t of the cycloalkyl phenyl ketones, and the cycloalkanones h a r e more than one enolizable hydrogen. The catalyst concentration (0.0502 J 1 HCI) used for the cycloalkanones was thus considerably smaller than t h a t for the cycloalkyl phenyl ketones (0.53 M HCl); autocatalysis of bromination of the c>-cloalkanones by hydrogen bromide formed during reaction was found important under certaiu conditions. Thus when the concentration of cyclopentanone was 0.07 31,the reaction obeyed a first-order rate law over the entire period studied (39% mono-enolization); when the concentration of the ketone was 0.125 X , the reaction only followed a first-order rate law until monobromination was ,a '>-,1-, :, complete. In general tiiono-enolizatiori of cycloalkanones was followed up to 30-50[,{ completion. The rate of bromination of a ketone is first order vvith rcspect to ketone but independent of bromine concentration. The rate coilstarit was evaluated [in use of eq. 4 in which [ketotic], and [ketone] refer to crinccntratious of ketonc a t t = 0 ant1 tiinc t , respcctivcly. The rate constaiit W:IS deterrnincd graphically from the slqie of the line (multiplied by k c t u i ~ c sl,y b a i e i is
t h a t i i i I1din;iiiii c l i m i n ~ t i ~( ~ j f t (i[ i i i i tcrnary aininoniiim conipoiin,ls; ( e ) 1) 1' Evan\ ani1 J . .I (;or(Ion, J . C h e i n . S O L , 1434 (1938). ( 8 ) D. P. Evans, ibid.. i83 I Y Y G ) . ( 9 ) 0. 13. Wheeler a n d J . I, hlateos, J . Ore. C'hrvz , 22, 1153 (19.57). ( I O ) R . P S1arirll.i .ind R. K . K a u l , r , . I , . A m i A i i r i .So!.., 74, 521 ~ l ! l . 7 2 ) , C I1 'lilfortl ancl SI. C;, \-;in C'impei;, 11-, tCi,/., 76, 2 1 3 1 (1974). ( 11) T h e acid-catalyzecl hrornination of acetone ha5 l x e n previ(,ubly ntudied ;intier pseudo-zerc Co > C, 2 C, > Cs.'br I t is not necessary that the rates of acid-catalyzed enolization of cycloalkanones parallel their ketoenol cquilihria. I t has also Ireen found t h a t t h c enol content of a ketone is markedly dependent on t h e solvent environment.Md T h e present kinetic results in a single solvent are thus not strictly comparable t o the equilibrium d a t a for enolization of t h e pure cycloalkanones since in t h e latter system each ketone functions as its own internal solvent, arid thus i h e p d a r arid steric environment for enolization of each pure ketone is different. (b) G . Schwarzenbach and C. Wittmer, Helv. Chim. A d a , SO, 656, 669 (1947). (c) A. Gero, J . Otg. Chein., 26, 3156 (1961). (d) A . Gero, ibid., 19, 1960 (1954). (17) T h e relative rates of enolization of t h e Cs through CS cych~. alkaniines primarily reflect differences in the entropies of activation.
increase resulting from bond angle deformations in the small-ring transition state thus appears to be of greater significance than any effects such as removal of eclipsing, l8 solvation or entropy factors which may lower the energy of the transition state. The greater rate of acid-catalyzed enolization of cyclohexanone than cyclopentanone parallels their enol contents in waterLBband is of theoretical significance in that in this system a 6-membered ring accepts a double bond more readily than does its 5-membered homolog. l 9 The specific ordering of enolization of cycloheptanone and cyclooctanone presumably is related to the differences in conformational and solvation requirements of these exocyclic derivatives and their rate-controlling endocyclic transition states. It is of interest to compare the present rates of enolization of the C4-8-cycloalkanones in 90% acetic acid and the equilibria for enolization of cyclopentanone and cyclohexanone in waterIBb with thermodynamic data for the equilibria between methylenecycloalkanes and l-methylcycloalkenes in glacial acetic acid a t 25' (Table II).20 TABLE I1 EQUILIBRIA BETWEES METHYLENECYCLOALKASES AND 1-METHYLCYCLOALKENES IS ACETICACID AT 25'22 Ring size
Ksndo/cio
AFO"
5.66 -1.05 5 1144 -4.17 6 240 -3.24 c I 74.4 -2.55 8 598 -3.79 a Kcal./mole Cal./inole. 'I;. 4
AH05
ASjh
-0.95 -3.9 -2.4 -2.3
0.0 +0.9 f2.8 +0.8
Since there is a correlation between the rates of enolization of the C4-, Ci- and Cs-cycloalkanones and the free energies of isomerization of the corresponding methylenecycloalkanes in similar sol(18) Development of t h e transition state m a y allow remuval of as much a s a pair of non-bonded interactions. (19) (a) H. C. Brown, J. H. Bren'ster a n d H. Shechter, J . A m C h e m Soc., 76, 467 (1954); (b) H. C. Brown, J . 01.8. Chem., 22, 439 (1957). (20) (a) A. C . Cope, U. Ambros. E. Ciganek. C. F. Howell and Z. Jacura, J . Atn. Chem. Sac., 82, 17SO (1960); (b) E. Gil-Av and J. Herling, Tetvuhedvon L c l l e v s , 1, 27 (1961); (c) R B. Turner in "Kekuli. Symposium on Theoretical Organic Chemistry," Rutterwr~rthScientific Publications, London. lU;!l, p. 79
?!)US
13. SHECHTER, h.1. J. COLLIS,R.DESSY,U. OKUZUMI AND A. CHEN
vents, i t appears that the dominant structural features involved in enolization of these ketones and isomerization of these methylenecycloalkanes are similar. Neither the equilibria nor the rates of enolization of cyclopentanone and cyclohexanone are of the same relative order (C,> Cb), however, as the equilibria for isomerization of methylenecyclopentane and methylenecyclohexane to their corresponding I-methylcycloalkenes (Table 11, C5 > C,).20a,21 Since cyclopentanone and cyclohexanone served in part as the models for the generalizationlgthat conversion of an exo-unsaturated 6-membered ring to its endocyclic isomer is favored when compared with corresponding 5-membered homologs, the lack of correlation with the exo-endo equilibria of methylenecyclopentane and methylenecyclohexane is of intere~t.~Z~23 Prediction of the abilities of methylenecyclohexane and methylenecyclopentane to undergo endo isomerization was based primarily on enolization data and thermochemical (enthalpy) correlations for cyclohexanone and cyclopentanone. l 9 In extrapolating the conclusions with respect to the cyclic ketones to methylenecyclohexane and methylenecyclopentane it was presumed that because of eclipsing introduced between the exomethylene hydrogens and the two equatorial hydrogen atoms in the a-positions of the cyclohexane ring there should be even greater relative tendency for the 6-membered ring becoming endocyclic. The reasons for the relative equilibria of isomerization of methylenecyclopentane and methylenecyclohexane and the lack of correlation with the enolization of cyclopentanone and cyclohexanone are not yet ~ n d e r s t o o d . ~It~ is possible that the failure in correlation of abilities of these 5- and 6-membered derivatives to accommodate double bonds2j stems primarily from the greater difference in solvation energies for isomerization of the cycloalkanones than for the methylenecycloalkanes, 26 or indeed that the non(21) T h e differences in free energy ( A A F ' ) for enolization of cyclohexanone and cyclopentanone a t 2 5 O in t h e absence of a n external solvent and in water are -1.52 and -0 85 kcal./mole; t h a t for isomerization of methylenecyclohexane t o 1-methylcyclohexene and of methylenecyclopentane t o 1-methylcyclopentene a t 25' is 0 93 kcal./ mole. (22) R . B. Turner, J . A m . Chern. Soc., 80, 1424 (1958). (28) E. Gil-Av and J. Shahtai, Ciieniistry & I n d u s t r y , 1630 (1959) have shown t h a t t h e exo-endo equilibria for ethylidenecyclopentane and ethylidenecyclohexane. respectively, agree with t h e previous gcncralization concerning t h e abilities of 6- and 6-membered rings in accepting unsaturation. 19 (24) (a) T h e f a c t t h a t substituted 6-membered rings accept endo double bonds more readily than do analogous &membered systems appears t o result primarily from greater cis steric interaction of t h e alkylidene substituents and a,a'-methyIene hydrogen atoms in t h e (b) H. Shechter. rings of 6- t h a n of &membered derivatives.2','J,"b D. E. Ley and E. B. Robertson, Jr., J . A m . Chern. Soc., 78, 4984 (1956). (26) I t has been previously proposed's t h a t t h e correlation failure stems from differences in abilities of carbonyl and methylene groups t o acrept angle distortion in 5-membered rings or t h a t cyclopentanone is relativcly stabilized by hyperconjugation since its geometry is presuniahly more lavorable for such hybridization than is cyclohexanone and t h e electrical demand of t h e carbonyl function is greater t h a n t h a t < > fthe methylene groop. (26) hletliylenecyclopentane, methylenecycluliexane a n d tlieir corresponding 1-methylcycloalkenes are much less polar and thus less subject t o solvation in polar media t h a n are t h e 5- and 6-membered ketones and tlicir corresponding enols. On t h e supposition t h a t t h e
VOl. s4
bonded electrons on carbonyl oxygen have greater steric effects than previously envisioned. The kinetics of bromination of homologous cycloalkyl phenyl ketones have been determined in 90% acetic acid-0.53 M hydrochloric acid. The first-order rate constants and the kinetic parameters for these reactions are included in Table I. The reactivity order with respect to ring size of the cycloalkyl groups in enolization of these ketones is 5 > 4 > G > 3.27 The greater rates of reaction of cyclopentyl phenyl ketone than of cyclohexyl phenyl ketone indicate the importance of steric factors in these enolizations. The results are in agreement with previous correlations of steric effects in that 5-membered alicyclic rings have greater abilities for accepting trigonal atoms than do their 6-membered homologs.28 The data also support the postulate that the structure of the rate-controlling transition state in the reactions is close to that of enol (I11 and IV) .5a.c The result of particular interest is that the rates of acid-catalyzed enolization of cyclobutyl phenyl ketone are considerably greater than those of cyclohexyl phenyl ketone and only slightly less than that of cyclopentyl phenyl ketone. It is t o be expected that because of angle strain a 4membered ring which is tetragonally substituted will resist conversion to its trigonally related enol, V.28a The relatively rapid rate of enolization of cyclobutyl phenyl ketone may result from
0 I11
0 IV
the smaller cis steric effects in the transition state related to V than in those related to I11 and IV, and thus there is relatively large conjugative interaction of the benzal and cyclobutylidene groups with the developing carbon-carbon double bond. Additional features which may be expressed in the reactivity of cyclobutyl phcnyl ketone are removal of 1,2-non-bonded interactions on enolization and interactions of a solvent are greater for t h e ketones t h a n for t h e enols whereas those for t h e methylenecycloalkanes and l-methylcycloalkenes are more similar, it is possible t h a t cyclopentanone will be solvent stabilized relative t o cyclohexanone because of its greater dipole moment, rigidity'and more favorable stereochemistry. T h e relatively greater ground state energy of cyclohexanone t h a n of cyclopentanone in such a n environment m a y thus he expressed in t h e enolization equilibria and t h e rates of enolization. (27) First-order rate constants reported for bromination of cyclopentyl phenyl ketone and cyclohexyl phenyl ketone in aqueous acetic acid-hydrochloric acid (0.0722 M ) a t 26.7O are 1.3 X 10-6 and 9.0 X 1 0 7 sec.-l, respectivc1y.e These results are consistent with those of t h e present study. (28) (3)H. C. Brown. R . S. Fletcher and R. R. Johannesen, J . A m . Chenz. Soc., 73, 212 (1951); I I . C. Brown and M. Borkowski, i b i d . , 74, 1894 (1952); (b) G. S. Hammond in M. S. Newman, "Steric Effects in Organic Chemistry," John Wiley and Sons, Inc., New York, N. Y., 1956, p. 444.
ENOLIZATION OF CYCLOALKANONES AND CYCLOALKYL PIIEWLKc*rams
Aug. 5, 1962
2900
TABLE 111 BASE-CATALYZED DEUTERIUM EXCHANGE RATES O F O H
CYCLOALKANOXES AND CYCLOALKYL PHENYL KETONES I N
O
D
lo7
k16Q0
DMF
DiMF ll I /I I -c-c+-c-c-
1M k,oo X 107’=d kL’Qo X 107
Ketone
Cyclobutanone 140 C yclopentanone 35 Cyclohexanone Cycloheptanone Dipropyl ketone Diisopropyl ketone Cyclopropyl phenyl ketone Cyclobutyl phenyl ketone Cyclopentyl phenyl ketone Cyclohexyl phenyl ketone Isobutyrop henone Propiophenone Acetophenone First-order rate constants in src.-’.
300
lM‘5hl kl‘QO X 107 &PoX
2300 680 100 19
8 0.6 10 8.4
6
H
O
37
19 5.9 1.2
lo7
320 77 38 49 31 9.9 3.0 2.8
kPo X
138
165
lo7
Rel. rateq 40°
Ea, kcal.
290 85 12 2.4 1.0
12.3 15.9 12.2 14 6
20 1T 5.7 1.D 1.4
15.7 16.8 13.1
2.9 44 0.5 14 0.7 9.5 1700 490 2200 The rate constants were corrected for the statistical factor where
the relatively minimal steric restriction to solvation during reaction.29 The relative inability of cyclopropyl phenyl ketone to undergo enolization is interpretable in terms of the large strain and rehybridizational energies involved in formation of a transition state similar t o that of enol V I in which a double bond is attached to a 3-membered ring.30.a1 Base-catalyzed Enolization.-Table 111 summarizes the first-order rate constants and the kinetic parameters for base-catalyzed isotopic exchange of a-hydrogen in the series of homologous cycloalkanones and cycloalkyl phenyl ketones (eq. 5) in deuterium oxide-triethylamine-dimethylformamide. It has been shown that in the system O
170
X
D
D20, EtsN (29) First-order r a t e constants (sec. -1) reported8 for bromination of t h e following ketones in 76% acetic acid-0.5 M hydrochloric acid a t 25’ are: acetophenone, 0.241; propiophenone, 0.104; butyrophenone, 0.0721; valerophenone. 0.0851; isobutyrophenone, 0.0213; and isovalerophenone, 0.0320. Present interpretation of these results indicates t h a t steric effects (cis and solvation) in enolization are of importance along with possible hyperconjugative a n d inductive factors previously described.7’Va (30) T h e authors wish here t o emphasize t h a t t h e rate constant presently reported for acid-catalyzed bromination of cyclopropyl phenyl ketone is a maximum value. T h e reaction product isolated contained more t h a n one equivalent of bromine a n d the structures of t h e various components have not been determined. (31) Acid-catalyzed bromination of 7-ketocholestane yields 6bromo-7-ketocholestane rather t h a n 8-bromo-7-ketocholestar1e. T h e reaction is of interest in t h a t the 6,7-enol which leads t o product is derived by enolization of secondary rather t h a n tertiary hydrogen and thus is of the Hofmann rather t h a n the Saytzeff type. T h e result has been discussed previously9 on the basis t h a t an equilibrium exists between the two enols which favors the A7-isomer; however, attack of bromine on t h e once-formed A7-enol is impeded b y t h e angular methyl groups on (2-10a n d C-13a n d hence t h e ultimate reaction involves t h e 6,7-enol. T h e authors wish here t o suggest a more conventional^*' alternate mechanism in which, because of steric factors, t h e 6,7enol is formed much more rapidly t h a n t h e A7-isomer; bromination of t h e 6,7-enol in a step which is not rate-determining gives 6-bromo-7ketocholestane. Decision as t o t h e possible mechanism of acid-catalyzed bromination of 7-ketocholestane can be made upon determining t h e rates of deuteration and bromination of the ketone under ideutical conclitions.
mole-1
15.5
17.0 14.2 13.1 11.4 necessary.
AS*, e.u.
-31 -26 -42 -37 -36
-35 -35 -48 -39 -49 -42 -39
triethylamine-deuterium oxide the actual base involved in proton abstraction is deuteroxide ion (OD-)32; thus steric factors involving the base are kept low in the transition state for enolization. The following correlations and conclusions may be drawn: (a) in cyclic ketones the rate of exchange decreases as the ring size increases (4 > 5 > 6 > 7) approaching as an approximate limit the openchain analog dipropyl ketone; (b) in cycloalkyl phenyl ketones the rate of exchange decreases as the ring size increases (3 > 4 > 5 > 6) approaching as an approximate limit the open-chain analog i~obutyrophenone~~; and therefore (c) the order of reactivity with respect to ring size for basecatalyzed enolization of cyclic ketones and cycloalkyl phenyl ketones is decidedly different from that for the acid-catalyzed processes. The ratios of rate constants kh/ks and ks/kB (where k , refers to the rate constant for the 12 carbon ring size) are about the same in both series of ketones and suggest a common origin. With respect to base-catalyzed enolization,2cSwain, et al., have more clearly delineated the details of the structure of the transition state I1 and concluded that there is little bond making between B and H in the transition state, but that the bond between H and C is nearly broken. Particularly for hydroxide ion catalysis, as in the present cases, “distances are long in the transition state,” and “hydroxide ion supplies more driving force through coulombic and Pauli repulsions against the electron pair in the old bond than through covalent bond formation with hydrogen.” This would also seem to suggest that the sequences of enoli(32) Evidence for this conclusion is the f a c t t h a t E.-values for N a O D and “EtaN” catalysis in t h e ketone series are identical, and no steric effect is observable in kVn”/k“E’aN’’ ratios for a series of ketones with increasing steric requirements. Rough calculations of the catalytic constants for EtO? a n d OD - species support this; cf. ref. 13. (33) Ultraviolet and gas chromatographic analysis of t h e exchange mixtures indicated t h a t there were no products derived by dehydration of possible intermediates obtained by aldol condensation. Because the kinetic plots mere well-behaved, formation of protium oxide by a n y process competitive with deuterium exchange could not hnve been nppreciahle, e.f., non-rlehyrlratrd 311101 condensation
2910
WILLIAMP. JENCKS
AND
zation rates for cyclic and cycloalkyl phenyl ketones found in the present study are not a reflection of steric hindrance involving the base. Application of the concepts involved in the stabilities of emo and endo double bonds in cyclic systemsig and the factors involved in the ability of a carbocyclic ring to have one of its members undergo a coordination number change of four to three fail to correlate the present data. One possible explanation that seems to agree with the experimental results centers around the position of the transition state along the reaction coordinate, and is concerned with the effect of hybridization of the ring carbons on the acidity of the hydrogens attached. Cardwell and Kilner‘‘ suggest that much more unsaturation develops, through partial carbonium ion formation, in the transition state for acidcatalyzed enolization than in the base-catalyzed reaction. It is therefore suggested that the transition state for base-catalyzed enolization occurs early in the process, and has its energy determined to a large extent by factors affecting the ketone, unlike that for acid-catalyzed enolization, which occurs much later and has its energy determined to a larger extent by steric and hybridization factors in the enol. Coulson and Moffitt34 have calculated the hybridization coefficients for a series of cycloalkanes, acetylene and ethane, with the results : cycloC3H6, 1.51; cyclo-C4H8, 1.67; cyclo-C6Hlo, 1.73; C2H6, 1.73; C2H4, 1.41; C2H2, 1.36. This coefficient reflects the s-character of the carbon orbital directed toward hydrogen and thus in part expresses the ionic character of the C-H bond. This effect is, of course, evident in the carbonyl stretching frequencies for cycloalkanones which increase as one goes from cyclohexanone to cyclob~tanone.~ The ~ effect is also reflected in the basicities of these ketones as measured by hydrogen bondingY6%or by ultraviolet spectral shifts,36b 7>6>5>4. ( 3 4 ) C . A . Coulson and W. 13 A l o f l i t t , P h i / . .\log., 40, 1 (19.49). (35) R. Bellamy “Infra-red Spectra of Complex Molecules,” Juliti Wiley and Sons. Inc., New York. S . Y . , lY5U.
MARYGILCHRIST
VOl. 84
If one assumes a transition state that resembles factors (ketones) more than the product (enol), or one in which the a-carbon atom is predominantly sp3-hybridized, one would conclude, combining Swain’s postulated mechanism for enolization which involves long distances, and Pauli repulsion a t the a-C-H unit, and the relative acidities of hydrogens attached to succeedingly smaller carbocyclic rings, that a sequence of exchange rates (3) > 4 > 5 > 6 > (7) is consistent. It is interesting that the cycloalkariones exchange more rapidly than the cycloalkyl phenyl ketones. It is to be noted that in the changes in E a and AS*, a facet other than just “acidity” of the cyhydrogen is to be detected. Presumably the fluctuations are to be attributed to solvation and steric effects. However, since strict compartmentalizing of solvent and steric effects into the energy parameter does not occur, it is a t present not profitable to dissect the data further. The reported equivalent acidity of cyclopropyl phenyl sulfone37and isopropyl phenyl sulfone is not relevant here since the comparison was made on an equilibrium (thermodynamic) basis rather than a kinetic one. Cyclopentyl phenyl sulfone and cyclohexyl phenyl sulfone exchange their a-H’s for D a t more nearly equal rates38; kb,’k6 1.5. In the analogous ketones k5/’k6= 6. IX-ork in the sulfone area, which is currently in process, we hope will shed some light on the differences in a-H activation by carbonyl and sulfone groupings. The recent work by indicating the markedly different steric courses followed by these substrates in D/H exchange reactions, suggests a distinct difference. Acknowledgments.-R.E.D. wishes to thank Professors H. H. Jaff6 and H . Zimmerman for several valuable discussions. (36) (a) A I . Tamres and S. Searles, J . Ain. Chein. .Soc., 81, 2100 (1969); (b) H . J. Campbell and J, T. Edwards, C Q ~ ZJ . Chem., 38, 2109 (1960). (37) 13. Zimmerman and I3 S. Thyagarajan, J . Am. C l i e n t . Soc., 82, 2305 (1960). (38) J. \Veinstock, J. I,, Bernardi and R. G . Pearson, ibid., 80, 4‘361 (1958). (39) U. J. Cram. L). A . Scott and W, P. Sielscn, ibid., 83, 3696 (1961).
The Nucleophilic Reactivity of Alcoholate Anions Toward p-Nitrophenyl Acetate BY I ~ I L L IP. AM JENCKS
AND
MARYGILCI~KIST
RECEIVED MARCH8, 1962 Tlic tiuclco~iliilicreactivity o f a series uf alcoiidatc anions toward the ester, p - i ~ i t r i i ~ ~ lacetate, ~ c ~ ~ y lItas been d e ~ c r m i r ~ e d by iiicasureirieiit o f reaction rates in dilute, buffered solutions or by i i i ~ a ~ ~ r c n i eofi i product t ratios after base-catalyzed solvolysis of the ester in alcohol-\wter iriistures. Strongly basic oxygen anions shnw only a slight increase it1 reactivity with incrcasing basicity, and thus deviate from the BrZiisted relationship of slope 0.76-0.80 observed with less basic oxygen anions.
It was shown recently that, near neutral pH, the base-catalyzed reaction of p-nitrophenyl acetate with the hydroxyl group of N-acetylserinis llluch faster than the basc-catalyzed hydrolysis or cthanolysis of this ester.’ Uruicc, ct
al., found that the pKa of N-acetylserinamide is 13.6 and that the reactivity of its anion toward p-nitrophenyl acetate does not deviate signifi( 1 j 13 A t . Anderson, 13 II Cordcs and W. 1’. Jencks, J B z d . Chem. 236, 4; (lutii).
