May 5 , 1938
INDUCTIVE EFFECTSOF ALKYLGROUPS
2279
action according t o t h e method of Buck and Ide.13 T h e T h e salts, separated by centrifugation, when decompose: solid oxime from 0.5 g. of ketone was placed in 8 ml. of 10% with water and acid produced 2.57 g. (34.5%), b.p. 85-90 sodium hydroxide solution and 0.69 ml. of benzenesulfonyl (2 mm. ), of a-hydroxyisobutyrophenone. 3-sec-Butoxy-3-phenpI-2-butanonewas converted t o the chloride added in two equal portions with shaking. T h e solid disappeared leaving an oil and when the odor of bensemicarbazone in 69y0 yield, m.p. 184-185”. A n d . Calcd. for C l s H ~ l N 3 0 ~C : , 64.95; H, 8.36; N , zenesulfonyl chloride disappeared (approximately 30 min15.15; 0, 11.54. Found: C, 65.08; H, 8.54; K , 15.12; utes) the basic solution was extracted twice with 15-1111. portions of ether. T h e combined ether extracts xwre dried 0 , 11.62. over magnesium sulfate, filtered and evaporated t o (dryness. T h e oxime was made from 0.2 g. (0.9 mm.) of ketone. T h e residual oil was converted to the 2,4-dinitrophenylT h e yield was 0.26 g. (72%) of crude material, m.p. 118- hydrazone derivative and 0.47 g. (68%), m.p. 240-244’, of 123’. Recrystallization from hexane gave an analytical product was obtained. T h e material was recrystallized sample, m.p. 123-125’. A mixture melting point from chloroform, m.p. 244-245’. Anal. Calcd. for C20H&406: C, 59.99; H, 6.04; N, with authentic acetophenone 2,4-dinitrophenylhydrazone 13.99; 0, 19.98. Found: C, 59.92; H, 5.96; N, 13.73; was not depressed. 0, 19.68. Acknowledgment.-The authors wish to thank Beckmann Rearrangement of 3-sec-Butoxy-3-phenyl-Zbutanone Oxime.-A mixture of 0.5 g. (2.3 millimoles) of Dr. J , M. Vandenbelt and associates of Parke, 3-butoxy-3-phenylbutanone, 0.4 g. (5.75 mmoles) of hy- Davis and Co. for the spectral data published here droxylamine hydrochloride and 5 ml. of pyridine was allowed and C. E. Childs and associates of the same comt o stand overnight a t room temperature then heated for two hours on the steam-bath. T h e pyridine was then removed pany for the analytical determinations. via the vacuum pump and the residue treated with water. T h e oxime was subjected t o the second-order Beckmann re- DETROIT2, MICHIGAN
[CONTRIBUTION FROM THE
CHEMICAL LABORATORIES O F PURDUE UNIVERSITY]
The Inductive Effects of Alkyl Groups as Determined by Desilylation Reactions BY ROBERT A. BENKESER, RICHARD A. HICKNERAND DONALD I. HOKE RECEIVEDOCTOBER7, 1957 I t has been found that the rates of desilylation of trimethyl- and triethylphenylsilanes substituted in the m- position with alkyl groups increase in the order H < M e < E t < i-Pr < t-Bu. This is contrary to the decrease in rate for this series which is observed in the solvolysis of m-alkylphenyldimethylcarbinyl chlorides. It is rationalized that, a t least in this instance, the desilylation reactions more closely approximate conditions realized in true aromatic substitutions. The carbinyl chloride series is apparently complicated in the present instance by unusual hyperconjugative or steric effects.
I t is generally accepted that the inductive effect counterbalanced by steric hindrance to solvation of of alkyl groups increases moderately in the order the ring ortho to the alkyl substituents as suggested Me < Et < i-Pr < t-Bu.l Likewise these same by Schubert and co-workers.4 groups are thought to contribute hyperconjugative effects which tend to increase in the series t-Bu < i-Pr < E t < M e . 2 a , b Recently Professor Brown and co-workers8 have \\ demonstrated that the electrical effects of alkyl A CHtH’ groups can be studied conveniently by measuring the solvolysis rates of various alkyl substituted Recently5 it was demonstrated in this Laboratory phenyldimethylcarbinyl chlorides in 90% aqueous that the removal of a trimethylsilyl group (clesilylacetone. Their results generally bear out the se- ation) from an aromatic ring by acid is a reliable quence shown above for the hyperconjugative and convenient tool for studying electrical effects series. Rather unexpectedly these authors noted, a t a particular position in a benzene ring. however, that the rates of solvolysis of the m-substituted alkylphenyldimethylcarbinyl chlorides decrease in the order Me > E t > i-Pr > t-Bu (Fig. l.) H+ RgSi+ This is just opposite to the gradual increase in rate \ one would predict for this series from a consideraR’ ‘Rl tion of the inductive effects of the alkyl substituents. They3 explain this anomaly in terms of a It was decided therefore, to study the rate of re hyperconjugative effect in the o-position which is moval of the trimethylsilyl group from various relayed to the reaction site by a n inductive or field m-substituted alkylphenyltrimethylsilanes to deeffect. termine whether the same unexpected decreases in It should be noted that these results also can be rate would be obtained as were found in the carexplained in terms of a “normal inductive effect” binyl chloride series. As an added check on the method, the rates of cleavage of the triethylsilyl (1) C. K. Ingold, “Structure and hlechanism in Organic Chemist r y , ” Cornell University Press, I t h a c a , N. Y., 1953. series were also determined.
aYR8+
0+
(2) (a) E. D. Hughes, C. K . Ingold a n d N. T a h e r , J . Chem. Soc., (4) W. A Sweeney and W. hI. Schubert, i b i d , 76, 4625 (1954); 949 (1940); (b) J. W. Baker, “Hyperconjugation,” Oxford Univer79, 910 (1957); see also J . Org. Chem , 21, 119 (1956). sity Press, New York, S . Y., 1952. ( 5 ) R. A Benkeser a n d H. R. Krysiak, THISJOURNAL, 76, 6353 (3) H. C . Brown, J. D. Brady, M. Grayson and W. H. Bonner, THIS JOURNAL, 79, 1897 (1957). (1954); see also C. Eaborn, J. Chem. SOL,4858 (1956).
R. A. BENKESER, R. A. HICKNERAND D. I. HOKE
2280
Vol. 80
TABLE I
HZ-ALKYLPHENYLTRIMETHYLSILASES B.P.,
Compound (K = Mersi-)
R-Benzene m-R-Toluene m-R-Ethylbenzene
112OD
d 904
OC.
1.4910b 1,4924" 1,4924
.., . ....
169* 190" 206-207
Carbon,o
...
% ...
...
...
.\fRCZ
3lm.
Hydrogen,1'
Yield,
...
65 55 50
%
...
7c
59.88 73.90 9.96 (60.36) (74.08) (10. 17) ,8597 94 14 64.62 74.80 az-R-i-Propylbetizene 1,4894 10.58 75 (74.89) (10.48) (04,03) m-R-t-Butylbenzene 1.4884 ,8617 52 6 69.05 75.47 10.66 55 (68.66) 175.63) (10.75) a The value in parentheses is the calculated value. Rep2rted b.p. 169', n z n1.4908; ~ R . -4. Benkeser and P. E. BrumReported b.p. 189 , n z o1.4922; ~ H . A. Clark, A . F. Gordon, C. W. Young and field, THISJOURNAL, 73, 4770 (1951). M. J. Hunter, THISJOURKAL, 73, 3798 (1951). 0.8645
TABLE I1 m-ALKYLPHENYLTRIETHYLSILANES Compound = EtaSi-)
(R
B.p., )l?OD
dZQ4
"C.
nz-R-Toluene m-R-Ethylbenzene
1.5022h
....
1.5015
0.8833
123 125
nz-R-i-Propylbenzene
1.4972
,8902
131
nt-R-t-Butylbenzene
1,4956
,8747
145
a The value in t h e parentheses is the calculated value. THISJOCRSAL, 76, 904 (1954).
The desilylation reactions were accomplished in a cleaving medium of glacial acetic acid which was 2.35 molar in hydrogen chloride and 7.23 molar in
*
yo Yield, Yo 24 76.12 10.95 20 73.58 (76.28) (10.97) (73.29) 9 78.03 76.80 11.02 15 (78.02) (11.18) (76.84) 77.27 11.02 26 11 82.93 (77.33) (11.36j (82.55) Reported ? Z ~ O D 1.5030; R . A. Benkeser and H . Landesman, Arm.
.lJRn
Carbon,a %;
11 9
...
...
IIydrogen,a
...
rate of cleavage incveases as the inductive effect (+I) of the a1kyl group increases. It is not unexpected that the rates of cleavage of the triethylsilyl group are generally slower than
1.0
0.8 i 04
+m 0.6 0.4 I
C
H ble Et i-Pr t-Bu Fig. 1.-Effect of alkyl substituents upon the rates of solvolysis of the alkylphenyldimethylcarbinyl chlorides in 90% aqueous acetone a t 25".
water a t 25". The rates were followed by volume expansion employing di1atometers.j The results are listed in Tables I11 and I V and are plotted in Fig. 2. Results and Discussion It is apparent from Fig. 2 that, contrary to the results obtained in the solvolysis of the tertiary carbinyl chlorides, the rates of cleavage of the trialkylsilyl groups increase in a fashion which is completely predictable in terms of the inductive effects of the substituent alkyl group. Thus, the
0.2
H
31e
Et
i-Pr
1-Bu
Fig. 2. -Effect of alkyl substituents upon the rates of acid-catalyzed cleavage of trialkylsilylalk~lbenzc~~es by 2.35 M hydrochloric acid in glacial acetic acid containing 7.23 JI water at 95': top curve, MeaSi-series; lower curve, E taSi-series.
for the trimethylsilyl group. From a steric stanclpoint it would be predicted that both the proton attack on the carbon atom holding the silicon, as well as the attack of water molecules on the silicon
M a y 5 , 1958
INDUCTIVE
TABLE 111 RATECONSTANTS AND HALF-LIVES FOR
EFFECTSOF ALKYLGROUPS mula A).
THE
2281
On the other hand, the electronic de-
ACID-CATALYZEDmands of the desilylation system may never attain
CLEAVAGE OF SOME m-TRIMETHYLSILYLALKYLBENZENES I N A
this intensity, and hence hyperconjugative effects
of the alkyl groups are not called into play. I n GLACIAL ACETICACID SOLUTION (2.35 MOLAR IN HYDROGEN these cases the reaction rate would be governed CHLORIDE AND 7.23 M o L m I N WATERAT 25’)
only by the inductive effect of the alkyl group. The observed results also may be explained in 3 84 181 H 3.80,3.85,3.86 terms of a buttressing effectg A large group sub8.31 83 CH3 8.48,8.10,8.35 stituted on the benzene ring probably distorts the 8 54 81 CaH6 8.56,8.40,8.67 C-H bonds of the o-hydrogens. Another large 9.08 71 i-C3H7 9.07,9.36,8.83 group meta to the first would increase the distortion 11.0,10.6 10.8 64 t-C,Ho of the C-H bond between the two large groups. Thus, the larger the m-substituents the greater TABLE I\‘ would be the strain due to this buttressing effect. Such an effect would account for the difference in RATECOXSTANTS AND HALF-LIVES FOR THE ACID-CATALYZED slope of the trimethylsilyl- and triethylsilylalkylCLEAVAGE OF SOMEm-TRIETHYLSILYLALKYLBENZENES IN A GLACIAL ACETICACID SOLUTIOX (2.35 MOLAR I N HYDROGENbenzenes illustrated in Fig. 2 since the relief of strain in the triethyl series would be greater than CHLORIDE AND 7.23 MOLAR IN WATERAT 25”) EtsSiCaHdR(m) R a t e constant Half-life, for the trimethyl case and hence the rate of cleavR group X 108, min.-’ Average min. age would be more rapid in the former. This same 1.66,1.62 1.64 422 H effect would tend to hinder the formation of a co3.87 179 CHa 3.82,3.91 planar system in the solvolysis of the aryldimethyl4.69 148 CZH5 4.73,4.65 carbinyl chlorides. Coplanarity of this system is 5.18,5.09 5.14 135 i-CsH7 required for resonance stabilization of the carbon5.84,5.95 5.90 117 t-CtHo ium ion. The attainment of this requirement would be increasingly more difficult as the size of atom6 would be rendered slower by the bulkier the group in the m-position increased, and thus the ethyl groups. Further evidence bearing on this rates of solvolysis would decrease in the order Me point will be offered in a forthcoming paper. > E t > i-Pr > t-Bu as observed by Brown and coIt is interesting to speculate on the reason for the w o r k e r ~ . ~ , ~ opposite trends shown in Figs. 1 and 2. This is Another possible explanation for the opposite especially true since both desilylation and car- trends of these two reactions is that solva.tion is binyl chloride solvolysis have been put forth3y6 as more important in the solvolysis of the cilrbinyl reliable methods for determining electrical effects chlorides than in the desilylation reactions. Acin a ring position. cording to the qualitative theory of solvent effects Considerable work has been accumulated to date of Hughes and Ingoldlo an increase in the ion solwhich shows that hyperconjugative effects of alkyl vating power of the medium will accelerate the forgroups become most apparent in electron deficient mation, and inhibit the destruction of charges. systems, ie., those systems which involve carbon- Thus a reaction incapable of forming a charged ium ions. Thus Norris and Banta7 observed that species should be little affected by the solvent. If the rate of solvolysis of p-methylbenzhydryl chlo- one envisions the desilylation reaction to be a conride is eight times faster than that of the m-isomer. certed process, a charge such as that present in a Likewise p-methylphenyldimethylcarbinyl chlo- carbonium ion is not developed. In the case of the ride is solvolyzed thirteen times faster than the carbinyl chlorides solvation becomes very imporm-isomer. Obviously the p-methyl group in these tant in the formation and stabilization of the carcases is very effective in stabilizing the carbonium bonium ion. Any factor which interferes with this ion transition state through hyperconjugation. solvation probably would cause a decrease in the On the other hand, the hyperconjugative effects rate of reaction. Conceivably an increase in the size of the groups in the m-position could hinder R E =~ = C H ~ H the solvation of the carbonium ion and thus give are scarcely detectable in systems not electronically rise to the “bulk effect’’ suggested by Price and deficient. Thus the ionization constants for m- Lincoln. An additional factor which should be given conand p-methylanilines and benzoic acids are not sideration is the nature of the solvent. Shiner and greatly different.8 The mechanism of both the carbinyl chloride sol- Verbanic12have shown recently that the solvolysis volysis and the desilylation reaction involves the of m-alkylbenzhydryl chlorides shows a marked degeneration of a positive charge. However, i t is pendence upon the composition of the solvent. A conceivable that the electronic demands of the car- change in solvent from ‘‘S070’’acetone to “90%” bonium ion in the former case ( S N ~ mechanism) alcohol resulted in a reversal of reactivity of the may be so great as to call into play the hypercon- m-methyl and m-t-butyl compound. This effect jugative effects of the alkyl groups, even though the may be due, as suggested by Shiner and Verbanic, latter are in a m-position to the reaction site (for- to solvent enhancement of hyperconjugation in the MeaSiCsH‘R(m)R group
R a t e constant X 108, min.-’
Average
Half-life, min.
+
(6) See C. E a b o r n , J. Chem. Soc., 3148 (1953), where a proposed mechanism for these acid cleavages is p u t forth. ( 7 ) J. F. S o r r i s and C. B a n t a , THISJOURNAL, SO, 1804 (1928) (8) p K a , 25’: In-toluic acid, 4 . 2 8 ; $-toluic acid, 4 . 3 5 ; p K b , 2.5’. ?ti-methylaniline. 9.30; p-methylaniline, 0.70.
(9) Professor H. C. Brown, in a private communication, :iuggested this explanation t o us. (10) E. D. Hughes a n d C . K. Ingold, J. Chcm. Soc., 244 (1035). (11) C. C. Price a n d D. C. Lincoln, THISJ O C R N A L , 73,5836 (1951). (12) V. J. Shiner a n d C. J . Verbanic, ibid.,79, 369 (10.57).
2282
R.A. BENKESER,R. A. HICKXER AND D. I. HOKE
less hindered groups or to steric inhibition of solvation in the transition state. I t would not seem unlikely that other instances will be found in which there are discrepancies in the results obtained from two such vastly different approaches to electrical effects in an aromatic system. In the present instance, a t least, the desilylation reaction more closely approximates conditions which maintain during aromatic substitution than do the solvolysis reactions. Experimental Triethylphenylsilane was prepared according t o the diof 51YO was obtained, b.p. rections of B ~ g d e n l ~a ~yield ; 105' (9 mm.), W ~ O D 1.5024 (reportedI3b T Z ~ O D 1.5024). Triethylchlorosilane .-Ethylmagnesium bromide was prepared from 109 g. ( 1 mole) of ethyl bromide and 26 g. (1.1 g. atoms) of magnesium turnings in 250 ml. of anhydrous ether under an atmosphere of dry nitrogen. The mixture was filtered and added, with stirring, t o a solution of 157 g. ( 1 mole) of diethyldichlorosilane (redistilled, b. 128') in 250 ml. of anhydrous ether. When Color Test 114 was negative, the solid which had formed was filtered and washed with anhydrous ether. The filtrate was distilled through a U-idmer column. There was obtained 89 f . (59%) of material boiling continuously from 137 t o 149 . All the fractions in this range had the same refractive index ( ~ N D1.4340) (reported'j n Z o1.4320). ~ m-Bromoacetophenone was prepared according t o the directions of Pearson and Pope'c; yield 79y0, b.p. 96-97" (4mm.), W ~ 1.5759 D (reported16 T Z ~ 1.5740). D m-Bromoethylbenzene was prepared according t o the directions of Pope and Bogert"; yield 61Yc, b.p. 91" (20 nim.) (reported" b.p. 8&90D (20 mm.). m-Chlorophenyldimethylcarbinol.--m-Chlorophenylinagnesium bromide was prepared from 7.3 g. (0.3 g. atom) of magnesium turnings in 50 ml. of ether and 57.5 g. (0.3 rnolel of 1-bromo-3-chlorobenzene (Eastman Kodak Co.) in 150 ml. of anhydrous ether. dcetone (17.4 g., 0.3 mole) was added at such a rate as t o maintain gentle reflux. The mixture was stirred for an additional 15 minutes a t room temperature. It was then hydrolyzed by pouring onto chipped ice. The ether layer was separated, dried with magnesium sulfate, and distilled through a Yigreux column. There was obtained 34.8 g. (68%) of carbinol boiling at 86-88' (2.5 mm.), n z o 1.5380. ~ Anal. Calcd. for CSH~lC1O:C, 63.40; H , 6.46. Found: C, 63.68; H, 6.49. a-Methyl-m-chlorostyrene .-nz-Chlorophenyldimethylcarbinol (51 g., 0.3 mole), 2 g. of fused potassium bisulfate and 0.5 g. of hydroquinone were heated in a Claisen flask fitted for distillation. The product distilled from the reaction mixture at 125-130' (95 mm.) over about a 40minute period. The distillate was dissolved in ether and the ether solution washed twice with 570 sodium hydroxide and twice with water. It was then dried over anhydrous magnesium sulfate and distilled through a 12-inch Vigreux column t o give 29 g. (7270) of material boiling at 60-62" (4 mm.), n Z o1.5536, ~ and 6.4 g. of crude starting material. Anal. Calcd. for C S H ~ C I :C, 70.83; H , 5.95. Found: C, 70.94; H , 6.35. nt-Chloroisopropy1benzene.--;1 Parr hydrogenator was charged with 35.8 g. (0.24 mole) of a-methyl-m-chlorostyrene, 50 ml. of absolute ethanol and 0.3 g. of 10yp platinuin(13) (a) A. Bygden, Bel,., 4 6 , i l l (1!)12); (b) H. Gilman. R. K. lnxham a n d A. G. S m i t h , J Org. C h e m . , 1 8 , 1743 (1953). (14) H. Gilman a n d F. Schultz, THISJ O U R N A L . 4'7, 2002 (1925). (15) J. W. Jenkins and H. W. Post, J . Org. Chem., 1 6 , 55F (1950). (IC!) D. E. Pearson and H. W. Pope, i b i d . , 2 1 , 381 (1956). (17) G . XT'. Pope and RI. T.Bogert, ibid., 2 , 27G (1937).
ITol.
so
on-charcoal. T h e theoretical amount of hydrogen waj absorbed in about two hours from a n initial pressure of 55 p.s.i. The catalyst was filtered off and the ethanol distilled. Distillation of the product through a 12-inch Vigreux column gave 21.8 g. (70%) of product boiling at 66-68' (8 mm.), 1.5136. Anal. Calcd. for CgHllC1: C, 69.92; H , 7.17. Found: C , 69.62; H , 7.35. wt-Chloro-t-buty1benzene.-p-t-Butylacetanilide was prcpared according to the directions of Herstein.Is This was converted t o m-chloro-t-butylbenzene by the method of Capon and Chapman.1g The over-all yield based on acetanilide was 20%. The product, when distilled through a Todd continuous wire spiral column, boiled at 88" (11 mm.), %*OD 1.5127 (%*OD 1.5125 after distillation through a Podbielniak column20). m-Trialkylsilylalkylbenzenes were prepared by the following general method. -4solution of 0.1 mole of nt-chloro- or nt-brornoalkylbenzene and 0.1 mole of trimethylchlorosilane or triethplchlorosilane in 50 ml. of anhydrous ether was added over a 15-minute period t o a well-stirred suspension of 6.5 g. (0.28 g. atom) of sodium sand in 50 ml. of anhydrous ether under dry nitrogen. The mixture was stirred and refluxed until most of the sodium had reacted. The mixtures were various shades of blue a t this point. Hydrolysis was accomplished by cooling the mixture in an icebath and adding 100 ml. of cold water dropwise. The ether layer was separated, dried with Drierite, and distilled through a Todd continuous wire spiral column. The data for the preparation of the m-trimeth~-lsilS-lalk!.lbenzenes are summarized in Table I. T h e d a t a for the preparation of the nz-triethylsilylalkylbenzenes are summarized in Table 11. Purification of Materials.-Glacial acetic acid was purified by distillation from a mixture of glacial acetic acid, the calculated amount of acetic anhydride and a small amount of potassium permanganate. The hydrochloric acid was the Baker and AdamsoIi C.P. product, 37-38? HCl, used without further purification since the listed impurities were negligible. Thermostat and dilatometers in this study were described in a previous paper from this L a b ~ r a t o r y . ~ General Kinetic Procedure.--A stock solution of aqueous hydrochloric acid in glacial acetic acid was made t o such a concentration that a 15-ml. aliquot when diluted to 50 ml. would give a cleaving medium of the desired concentration. For the rate determinations an aliquot of stock solution was diluted to about 35 nil. and allowed t o stand overnight. A l tzero time this solution was added t o the silane (0.3 g.0.5 g.) contained in a 50-ml. volumetric flask suspended in the constant temperature bath. The volume was made up t o the mark with glacial acetic acid. The flask was removed from the bath, stoppered, and shaken vigorously. The solution was transferred rapidly by suction to the dilatometer. lieadiiigs of meniscus height were recorded with the aid of a cathetometer which could be read t o 0.01 cm. X11 ruiis exhibited an expansion in volume on cleavage, the increase in meniscus height amounting t o 1-2 cm. Kate constants (pseudo first order) were determined graphically from a plot of constant -I-log ( 1 , - It) against time, 1, in minutes ( 1 is the meniscus height). The slope of this plot is --k/2.303 where k is the pseudo first order-rate constant. The rate constants and half-lives are given in Tables I11 and IY.
Acknowledgment.-The authors are grateful to the Iiational Science Foundation whose financial assistance made this research possible. LAFAYETTE, ISD (18) B. Herstein, U. S . P a t e n t 9,092,972 (1936). (19) B. Capon and S B. C h a p m a n , J . C i i e ~ z .So
