50

DMF, the kea values are smaller than those for benz- hydryl chloride by factors of 2.4-2.5, so that the p-C1 substituent effect is in line with the id...
0 downloads 0 Views 259KB Size
COMMUNICATIONS TO THE EDITOR

Nov. 20, 1964

solvents. Further, in MeN02, MeCN, and MezCO, the kze values for p-chlorobenzhydryl chloride are either nearly identical with or somewhat larger than those for the unsubstituted benzhydryl chloride. The available evidence suggests that the kze values are rate constants for S Nattackzc ~ on covalent RCl. A similar conclusion has been reached by Pocker5 in the case of bromide exchange of benzhydryl bromide in MeN02. The keo intercept values represent rate constants for formation of capturable intermediates extrapolated to zero salt concentration. In MeNOZ, MeCN, and DMF, the kea values are smaller than those for benzhydryl chloride by factors of 2.4-2.5, so that the p-C1 substituent effect is in line with the idea that keo is associated with ionization of RCl. For benzhydryl bromide in MeNOZ, Pocker5 observed the same intercept values from bromide exchange or reaction with triethylamine, and he also noted common ion depression of the amine reaction by bromide salt. On this basis, he concluded that the capturable intermediate is the dissociated carbonium ion. Analogously, the capturable intermediate for the kea values in all four aprotic solvents may well be a dissociated species6 In Table TI are summarized kao/kto ratios for the solvolyzing solvents and k,O/keo ratios for the aprotic ones. These values are lower limits for the ratio of

Solvent

AcOH (25') 80% M e C O (25")

MeNOz MeCS HCONMe2 MezCO

D

6 30 37 37 37 21

AT

75.0"

kan/kel'

the tendency for such nucleophilic solvent intervention is high for DMF* and low for CH3CN.

111

I

While an ion like I would be sufficiently long-lived to become a dissociated cation (free of the counter chloride ion) the nucleophilic intervention of D M F in the radiochloride exchange scheme must occur a t a stage of ionization-dissociation of RCl earlier than the dissociated R + ion. If D M F attacks a carbonium chloride ion pair, one can understand kea being sensitive to solvent nucleophilicity, while kao is not. ~

( 8 ) See, e..?., F. C . C h a n g and R. T. Blickenstoff, J . A m Chem. Soc., 80, 2906 (1960), for nucleophilic a t t a c k by D M F ; f o r related effects of MeNOn and MeZCO, see Y . Pocker, J. Chem. SOL., 240 (1958); H. Burton and G. W . H. Cheeseman, ibid., 832 (1953); R. A . Sneen and H. Weiner, J. A m . Chem. Soc., 8 6 , 2181 (1963).

CONTRIBUTION KO. 1744 OF CHEMISTRY DEPARTMENT OF CALIFORNIA UNIVERSITY LOS h G E L E S , CALIFORNIA 90024

ARTHURF. DIAZ S. WINSTEIN

RECEIVED AGCUST10, 1964

Radiochlorine Exchange and Racemization of 1-Phenylethyl Chloride and p-Methylbenzhydryl Chloride in Nitromethane'.2

TABLEI1 RATE COMPARISONS FOR RCl

5011

Sir : kd/kt'

38 2.6 44 304 4 >50

ionization rate to rate of chemical capture.2 They are also lower limits to the ratio of rate of ionization to rate of formation of dissociated carbonium ions. Very interesting is the large range of kao/keo values in the aprotic solvents. Most striking are the values in MeNOZ, MeCN, and DMF, three solvents with essentially identical dielectric constants. In D M F the ratio is as low as 4, while in CH3CN it is as high as 304. The most plausible single factor responsible for the large solvent specificity of the kUo/keo ratios in these three solvents is nucleophilic solvent intervention which makes keo a larger fraction of kao than it would otherwise be. For example, in the case of DMF, cation I would be an intermediate for the kea rate. Such solvent intervention is conceivable with the other aprotic solvents as well, salts of the I, 11, and I11 types being known.7 The present data suggest that ( 5 ) Y. Pocker, J. Chem. Soc., 3939, 3944 (1959). (6) With @-chlorobenzhydryl chloride and triethylamine in MeN02 a t 75O, t h e situation is quite complex. First-order r a t e constants for t h e amine reaction, followed by amine consumption or chloride ion formation, obey a n equation like (1) with t h e intercept t h e same as for chloride exchange. F u r t h e r , in t h e presence of sufficient EtaN, t h e exchange reaction with BurNCl has its intercept kea value suppressed t o ca. zero and its k i , slope is relatively unaffected. T h e product of t h e amine reaction is not t h e R N -EtaCl - salt, however. T h e l a t t e r , independently prepared, decomposes rapidly a t 75' with chloride ion disappearance, and subsequently chloride ion slowly reappears. (7) H. Meerwein, el ai., A n g e w . C h e m . , 67, 374 (1955); Be?., 89, 209, 2060 (1956).

In the presence of added pyridine, the initial firstorder rate coefficients of styrene, acid, and chloride ion production from 1-phenylethyl chloride a t 99.8" in solvent nitromethane are practically identical ( 2 . e . ) dehydrohalogenation) and independent of pyridine concentration: V D = klDIRCl] with k l D = 0.29 X set.-'. In the presence of added Et4NC136, V D = kID[RC1] kzD[RCl][Et4NC1] with k l D = 0.29 X 10-5 set.-' and kzD 2.0 X set.-' 1. mole-', but the radiochlorine exchange3 is a strictly bimolecular process: VE = k ~ [ R c 1 ] = KZ~[RCI][Et4NC136],with kzE = 6.1 X sec.-I 1. mole-l. For racemization we find Vu = k,[RCl] = kl"[RCl] kza[RC1][Et4NC136], where kl" = 1.2 X lop5 sec.-I and kza = 12.4 X set.-' 1. mole-I. It will be noticed that in the unimolecular region, [NEt4Cl]+ 0, the polarimetric rate coefficient, k l P , for the formation of racemic starting material and product is about four times larger than the rate coefficient for chemical capture, (klD f klE) klD.4 The route corresponding to kl" is associated with ion pair return after reorganization4a5;its free energy of activation is lower than that of

+

+

( 1 ) This work was supported in p a r t by a grant from t h e National Science Foundation. (2) Presented in p a r t a t t h e Nineteenth I U P A C Congress, London. 1963. (3) T h e rate of loss of tracer from Et4iVC136 contains a term due t o t h e dehydrohalogenation of 1-phenylethyl chloride (isotopic dilution) and a t e r m due t o substitution (isotopic exchange). To obtain t h e t r u e component of isotopic exchange we corrected for dehydrohalogenation. T h e values t h u s obtained have also been confirmed by determining t h e r a t e of incorporation of Cl36 into 1-phenylethyl chloride. (4) I n liquid S O X ,k i a / ( k i D kiE) = 9 : Y. Pocker, T r a n s . Faraday Soc., 66, 1266 (1959); cf. also Y. Pocker, "Nucleophilic Substitution a t d S a t u rated Carbon Atom in Nonhydroxylic Solvents," Vol. 1, Pergamon Press, New York, N. Y . , 1861, pp. 227, 228. ( 5 ) S. Winstein, J. S. Gall, M . Hojo, and S. Smith, J . A m . Chem. Soc., 8 2 , 1010 (1960); S. Winstein, M. Hojo, and S. S m i t h , Tetrahedron Lellers, No. 2 2 , 12 (1960); S. Winstein, A . Ledwith, and M. Hojo, ibid., No. 10, 341 (1961); A. F. Diaz and S. Winstein, J. A m . Chem. Soc., 86, 5010 (1964).

+

COMMUNICATIONS TO THE EDITOR

5012

the routes leading to dehydrohalogenation and to exchange (KID klE). The slope of the plot k , v s . [Et4NCl]is linear and is twice the slope of the plot k~ vs. [EtJVCI] leading to k2, = 2k2E. This signifies that each act of bimolecular radiochlorine exchange proceeds with inversion of configuration and suggests that k2E is the rate constant for sN2 attack by chloride ion on covalent 1-phenylethyl chloride.6 The kinetics of substitution of p-methylbenzhydryl chloride by Y = C1-, Na-, pyridine, and triethylamine in nitromethane follow eq. 1, containing concurrent first- and second-order terms. The kinetics of the

+

I'

=

k[RCl]

=

ki[RCI]

+ kz[RCl][Y]

.

.

(1)

,

first-order component, kl [RCl] (Table I ) , embody three features: (a) the initial rates of substitution are practically the same: klcl- = klNa- = klpYr = klNEts, (b) the ionic strength effects are similar and positive, and (c) in the nonsymmetric substitutions by N3-, pyridine, and NEt3, added EtlNCl (common ion) is able to depress the entire first-order component leaving the secondorder component largely unchanged. This behavior is analogous to the one found for benzhydryl bromide in nitromethane,' and more recently for p-chlorobenzhydryl chloride by Diaz and W i n ~ t e i n . ~ TABLE I YALUESOF

Temp., OC.

0 0 25 0 25 7 34.8 35.6 44.6 60 0

klE,

klN3-, klPYr, klKEt3, A N D k," (IS SEC.-') KUMBEROF TEMPERATURES

Vol. Sfi

We interpret klE as the rate of dissociation of p methylbenzhydryl chloride. However, kl" is much larger than klE and we associate the former with ion pair r e t ~ r n ,after ~ , ~reorganization. DEPARTMESTS OF CHEMISTRY USIVERSITY COLLEGE LOXDOS LONDOSW.C. 2, EXGLASD UNIVERSITY OF WASHIXGTOX SEATTLE, WASHINGTON

7 ; . POCKER

LY..4. MUELLER FRAXCESCO SASO GIORCIOToccrir

RECEIVEDAUGUST28, 1964

The Hydrogen Chloride Catalyzed Radiochlorine Exchange and Racemization of Aralkyl Chlorides in Nitromethane1z2 Sir: In polar nonsolvolytic media the stoichiometric composition of the activated complex is clearly defined, the possible structures of the transition state are more easily scrutinized, and short-lived intermediates interact reversibly with solvent molecules. The ionizing solvent nitromethane, although not entirely inert toward carbonium ions,4 is sufficiently unreactive toward aralkyl halides to permit a study of the mechanism of electrophilic catalysis by molecular HCl. The kinetics of racemization of 1-phenylethyl chloride in the presence of HC1 in nitromethane follow eq. 1. At 99.8", kla = 1.2 X set.-' and k2 = 1.5 X

AT A

V,

=

k,[RCl]

=

kl"[RCl]

+ kza[RC1][HC1] . . .

10skiE

lOSk1'8-

1o6kiPy'

105kiKEt3

lOjkia

0.019 0 474

0.021 0.46

0.020 0.47

0 020 0 45

0 54" 8 9"

0.95

0.93

2 25 10 0

2 26 10 0

2.20 9 6

2.28

19.0" 47.0"

Hydrogen chloride exerts a powerful effect on k1". Pyridine was added to eliminate catalysis by hydrogen chloride. Similar results were obtained when HC1 was rendered inert by adding tribenzyl amine or EtaSC1.

The kinetics of racemization of p-methylbenzhydryl chlorideS in the presence of [Et4NC136]< 0.043 M or [pyridine] < 0.056 -If follow eq. 2.9

V,

=

k,[RCl]

=

kl"[RCl]

1. sec.-l. Isotopic chlorine exchange between 1-phenylethyl chloride and hydrogen radiochloride was studied over a reagent concentration, [HCl] = 0.03-0.1 1 ,If, in nitromethane. The stoichiometric reaction in this region is one of substitution because d [olefinlidt = 0 ; the first-order rate coefficients, k ~ increase , linearly with the concentration of HCl (eq. 2 ) . At 99.8", I'E

=

kE[RCl]

=

klEIRC1]

(6) E . D Hughes, F. Juliusberger. A. D. Scott, B. Topley, and J. Weiss, J C h e m Soc , 1173 (1936). ( 7 ) Y . Pocker, ibid., 3939, 3914 (1959). (8) Details of the iresolution of 0-methylbenzhydrol and its conversion t o optically active chloride are in the P h D . thesis of William A . Mueller. tlniversity College, University of London, Sept., 1959. ($3) At temperatures below ROC and [NEtrCI] < 0.043 the bimolecular components. kZE [KCI][EtrNC136] and kza(RCI] [EtrNCl], are very small. They become significant a t higher [EtrNCll Values of kiE and ki" for 60.0' were obtained b y extrapolation t o [EtrNCI) = 0.

+

kzE[RCl][HC136] . .

.

(2)

klE = 0.29 X 10Y5 set.-' and kZE = 1.52 X 1. mole-' sec. - I . The value of klE is practically the same as that obtained for dehydr~halogenation~ in the presence of

(2)

The Arrhenius equation for the component of unimolecular radiochlorine exchange is: log klE = 8.2 18,600/(2.3X7') while that for racemization is: log kl" = 8.6 - 17,300,'(2.3R7'). The ratio of k l " I k l E decreases with increase in temperature.

(1)

SCHEME I [l-R-( C1HC136)-1

ia +

e [l-R+(CI3'HC1) -1

I-RC1 HC136 retention without exchange

ih +

l-RC136 HC1 retention with exchange

(1) This work was supported in part by a grant from the Kational Science Foundation. (2) Presented in part a t the Sineteenth I U P A C Congress, London, 1963. (3) Y. Pocker, "Nucleophilic Substitution a t a Satut-ated Carbon Atom in Xonhydroxylic Solvents," in "Progress in Reaction Kinetics," Pergamon Press, Vol. 1 , 1961, pp. 218, 227, 228. (4) H. Burton and G. W . H. Cheesman. J Chem. Soc , 832 (19331, 1'. Pocker, ibid., 240 (1968) (.5) Y . Pocker, W. A . Mueller, F. Naso. and G Tocchi, J . A m . Chem S o c . , 86, 5011 (1964).