Isotope Effect Studies on Elimination Reactions. VIII. The Mechanism

0 downloads 0 Views 754KB Size Report
Elcb mechanism is favored by the presence of .... mechanism for the reaction will result in the loss ..... converted to the Grignard reagent which was treated with.

Isotope Effect Studies on Elimination Reactions. VIII. The Mechanism of the 1,3-Elimination Reaction of 3-Phenylpropyltrimethylammonium Iodide with Amide Ion in Liquid Ammonia1 K. C. WESTAWAY~ AND A. N. BOURNS

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

Department of Chemistry. McMaster University, Hamilton, Ontario Received January 28, 1972 Reaction of 3-phenylpropyltrimethylammonium iodide with potassium amide in liquid ammonia at - 55 "C was found to give concurrent 1,3-elimination forming phenylcyclopropane and 1,2-elimination forming 3-phenylpropene and cis- and trans-1-phenylpropene. Deuterium tracer studies on the salt labelled at C-3 established that neither a carbene nor an ylide are intermediates in the 1,3-elimination process. Isotopic exchange at C-3 was shown to accompany the reaction of the deuterated salt in ordinary ammonia, but it was not detected in the reaction of unlabelled salt in ammonia-d3. A nitrogen isotope effect, k14/k15, of 1.022 0.001 was found for the 1,3-elimination, while the corresponding hydrogen isotope effect, estimated from the effect of isotopic substitution on the 1,3- and 1,Zelimination ratio, was shown to exceed 20. The hydrogen isotope effect for the disappearance of undeuterated and deuterated salts (elimination and exchange) was approximately 7.4. These observations, as well as the influence of deuterium substitution in both the quaternary salt and the solvent on the relative rates of 1,3- and 1,2-elimination, have been shown to be in accord with an Elcb mechanism in which the rate constants for the conversion of the intermediate 3-carbanion to 1,3-elimination product and for return to reactant by abstraction of a proton from ammonia are of the same order of magnitude.

+

La reaction de I'amidure de potassium sur l'iodure de phenyl-3 propyl trimethylammonium dans l'ammoniac liquide a -55 "C donne lieu a une elimination concurrente de type 1,3 qui conduit au phenylcyclopropane et a une elimination 1,2 conduisant au phtnyl-3 proptne et aux phknyl-1 proptnes cis et trans. Les etudes sur le sel marque au deuterium au niveau du C-3 montrent que les intermediaires de l'elimination 1,3 ne sont ni un carbene ni un ylure. Un &changeisotopique sur le C-3 accompagne la reaction du sel deuterie dans I'ammoniac ordinaire; cet echange n'a pas tte detecte dans la reaction du sel non marque dans l'ammoniac-d3. Un effet isotopique sur l'azote k14/k15 = 1.022 0.001, a kt6 trouve pour 1'6limination 1,3; l'effet isotopique correspondant sur l'hydrogtne, calcule a partir de l'influence de la substitution isotopique sur le rapport de l'elimination 1,3 et 1,2 a Bte montre superieur a 20. L'effet isotopique de l'hydrogtne sur la disparition des sels non deuteries et deuterits (elimination et echange) est approximativement de 7.4. Ces observations ainsi que l'influence de la substitution par le deuterium (a la fois dans le sel quaternaire et dans le solvant) sur les vitesses relatives de l'elimination 1,3 et 1,2, sont en accord avec un mecanisme de type Elcb dans lequel les constantes de vitesse pour la conversion de l'intermtdiaire carbanion-3 en produit d'elimination 1,3 et pour le retour au reactant par arrachement d'un proton de l'ammoniac, sont du m&meordre de grandeur.

+

Canadian Journal of Chemistry, 50, 2332 (1972)

Introduction Three-membered ring formation by a 1,3elimination has been observed with a variety of reactants under widely differing reaction conditions (1). The reaction pathways followed by such reactions have been found to parallel those encountered in 1,Zelimination processes. Thus, the El pathway is characteristic of solvolyses involving reactants possessing good leaving groups (2-4), while reaction by either the E2 or Elcb mechanism is favored by the presence of strong base (4-7). A few examples of reaction 'For Part VII in this series, see ref. 39. 'Present address: Department of Chemistry, Laurentian University, Sudbury, Ontario.

by way of a carbene intermediate formed in a 1,lelimination reaction have also been reported (8). 1,3-Elimination competes most effectively with double bond formation in rigid bicyclic systems where the spatial relationship of reacting bonds is favorable for the formation of the three-membered ring (2-6). Base-promoted 1,3elimination is also found in acyclic systems, but usually, it requires the presence of electronwithdrawing groups (carbonyl(9), sulfonyl(10, l l ) , nitro (12), cyano (13), and aryl (14-18)) which facilitate the removal of the hydrogen atom. With the more powerful activating groups, there seems little doubt that the reaction proceeds by way of a carbanion intermediate, and

2333

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

WESTAWAY AND BOURNS: ISOTOPE EFFECTS. V l l l

this is one of the accepted mechanisms for the 1,3-eliminationprocess involved in the Favorskii rearrangement (7) (carbonyl activation) and the Ramberg-Backlund reaction (1 1) (sulfonyl activation). Much less certain, however, is the reaction pathway for eliminations in which activation is provided by the more weakly electron-withdrawing aryl group. Recently, Baker and Spillett (19) have observed rapid deuterium exchange at the 3-position in the reaction of a series of 2-alkyl-lmethylsulfinyl-3-phenylpropanes with dimethylsulfinyl anion in deuterated dimethylsulfoxide, thus demonstrating that the phenyl group provides sufficient activation to sustain the reversible formation of a carbanion under elimination reaction conditions. This observation, however, does not establish that the 1,3elimination is proceeding by an Elcb mechanism since, as has been pointed out by Breslow (20) with respect to 1,2-elimination reactions, the carbanion may not be on the elimination-

reaction coordinate at all but. instead. reversible carbanion formation and E2 klimination may be entirely separate and concurrent processes. An extensivelv studied 1.3-elimination involving aryl activation in an acyclic system has been the conversion of 3-arylpropyl- and 3,3diarylpropyltrialkylammonium salts to arylsubstituted cyclopropanes by reaction with sodium amide in liquid ammonia (15-17). This reaction was first reported by Bumgardner (15) who suggested (16) that it might proceed either by an Elcb mechanism, eq. 1, or by an E2 process, eq. 2, in which the transition state has considerable carbanion character. Banthorpe (21), on the other hand, has suggested the cyclopropane may be formed by the insertion into the C,-H bond of a carbene formed in a 1,lelimination process, eq. 3. A less likely pathway, but one which cannot be excluded a priori, is a 1',3-process, eq. 4, involving an $ide intermediate. The purpose of the present investigation

Elcb Mechanism 111

d

8

d

8

k,

E 2 Mechanism 121

8

d

C,H,CH,CH,CH,N(CH,)~ + NH, ~ C , H ~ ~ ~ ' H C H , C H ~ N ( C+HN, )H, ~

CbH5CH2CH2CH2N(CH3)3 + NH2+

[

"

H,N---H

+CbH5CH

/IH2 \

Y

+ (CH,),N

,

CH2

I , 1 -Elimination Mechanism 8

d

131

C6HSCH2CH2CH2N(CH3),+ NH2+C6HsCH,CH,CH. CH cbHsFL \&.+cbH5cH H

/r2

+ NH3 + (CH3),N

\ CH2

1',3-Elimination Mechanism 141

I

/C?2 CbHsCH---CH,---N(CH3)3 6-

8

+8

d

CbH5CH2CH2CH2N(CH3)3 NH2GCbH5CH2CH2CH2T(CH,),+ NH,

'

2334

CANADIAN JOURNAL OF CHEMISTRY. VOL. 50, 1972

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

TABLE I . Composition of the product formed from undeuterated and deuterated quaternary ammonium salts in light and heavy ammonia at - 55 'C Quaternary salt Undeuterated* Deuteratedt Undeuteratedz Deuterated

cis- 1Phenylpropene

Solvent

(%I

~rans1Phenylpropene

NH3 NH3 ND3 ND3

2.2

1.4 21.9

-

-

3-Phenylpropene

Phenylcyclopropane

-

2.6 22.2 -

2.3

-

96.0 53.7 >99.6 97.7

(%I

(%I

(%I

-

'Average of two experiments with agreement to within +0.5%. tAverage of two experiments with agreement to within +0.2%; reaction carried to 66% completion based on quaternary salt. $Reaction carried to 11% completion.

was to determine the mechanistic details of the 1,3-elimination reaction of 3-phenylpropyltrimethylammonium ion with amide ion in liquid ammonia. In particular, tests were made for the 1,l- and the 1',3-elimination mechanisms by examining the deuterium content of the products of the reaction of 3,3-dideutero-3-phenylpropyltrimethylammonium ion, a study was made of the extent of isotopic exchange of the deuterated salt with solvent as a means of detecting the presence of carbanionic species under the reaction conditions, and, finally, measurements were made of the hydrogen isotope effects associated with both substrate and solvent and of the nitrogen isotope effect associated with substrate in order to identify the rate-determining step of the reaction. Results and Discussion

Product Analysis The 1,3-elimination reaction of 3-phenylpropyltrimethylammonium iodide was studied at -55 "C in liquid ammonia solution using potassium amide as the base., The hydrocarbon product was quantitatively separated by preparative vapor-phase chromatography into three fractions consisting of (I) 3-phenylpropene, (2) a mixture of phenylcyclopropane and trans1-phenylpropene, and (3) cis- 1-phenylpropene. These were identified by their v.p.c. retention times and i.r. and n.m.r. spectra. The composition of fraction 2 was determined by U.V.spectroscopy and, in the case of the reaction of 3,3dideutero-3-phenylpropyltrimethylammonium ion in ordinary ammonia, the analysis was con3Bumgardner (15, 16) used the less soluble sodamide resulting in a heterogeneous system unsuitable for kinetic isotope effect studies.

firmed by v.p.c. and by n.m.r. spectroscopy. The results of the product analysis for the reactions of undeuterated and deuterated salts in ammonia and ammonia-d, are reported in Table 1. 3-Phenylpropene is the direct product of 1,2elimination, while the isomeric 1-phenylpropenes are formed by the base-catalyzed isomerization of this product during the reaction (22). The percentage of 1,2-elimination, therefore, is the sum of the percentages of each of the three olefin products. The absence of 3-phenylpropene in the product of reaction of the deuterated salt in deuterated solvent, in contrast to the approximate 1 : 1 ratio of this olefin to trans-l-phenylpropene in the products of the reactions in ammonia, is accounted for by the longer reaction time with the former. This results in a closer approach to an equilibrium composition of the three olefins in which the trans-l-phenylpropene predominates (22). No products of 1,2-elimination could be detected in the reaction of the undeuterated salt in deuterated ammonia. Based on the absence of the reddishbrown coloration characteristics of the carbanion [C,H,CH--CH--CH,le, it is estimated that the total olefin product formed in this reaction is less than 0.4%. Test for the 1 ,l -Elimination (Carbene) Mechanism This test involved a determination of the number and position of the deuterium atoms in the phenylcyclopropane produced from 3,3dideutero-3-phenylpropyltrimethylammonium iodide. Elimination by way of a carbene intermediate formed by 1,l-elimination will give a product containing two deuterium atoms per molecule, one on the methine and one on a methylene carbon, eq. 5. Any other reasonable

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

WESTAWAY AND BOURNS: ISOTOPE EFFECTS. V l l l

mechanism for the reaction will result in the loss of one of the deuterium atoms on C-3 and the formation of 1-deutero- 1-phenylcyclopropane. The trans- 1-phenylpropene - phenylcyclopropane fraction obtained by v.p.c. separation of the product of the reaction of the deuterated salt gave an n.m.r. spectrum in which the broad multiplet associated with the methine proton of phenylcyclopropane (1.55-2.00 p.p.m.) is absent and the complex multiplet associated with the methylene protons (0.45-1.10 p.p.m.) is reduced to two broad lines (23). These changes are consistent with replacing the methine hydrogen of phenylcyclopropane with deuterium. Furthermore, after subtracting the absorption associated with the phenyl protons o f trans-lphenylpropene from the total phenyl-proton absorption of the mixture, the ratio of the residual phenyl-proton absorption to that of the methylene protons of the phenylcyclopropane was found to be 5.0 :4.0; precisely that expected if deuterium enrichment did not occur in the methylene groups of the cyclopropane ring. These results were confirmed by mass spectrometry. The trans- 1-phenylpropene - phenylcyclopropane fraction was found to be 94.4% monodeuterated, 5.5% undeuterated and 0.1% dideuterated. It can be concluded, therefore, that the 1,3-elimination does not proceed by way of a carbene intermediate.

mass spectrum of the amine was identical, within the limits of measurement, to that of trimethylamine of natural isotopic abundance from which it was concluded that trimethylamine-dl did not exceed 0.5% of the product. A complication arises in the application of the test because it was observed that exchange of the N-methyl hydrogens of the quaternary salt with solvent is quite rapid in comparison with the rate of 1,3-elimination (50% exchange after only 11% reaction). Consequently, if the conversion of the ylide intermediate to phenylcyclopropane were to occur by way of the 3-carbanion,

a

@

[C6H5CDCH2CH2N(CH3)2CH2D], rather than in a one-step process, this carbanion could react with solvent to regenerate quaternary salt which in turn could lose its N-methyl deuterium by exchange before undergoing elimination. Analysis of the isotope effect results reported later in this paper, however, show that if the 3-carbanion is an intermediate in the 1,3-elimination, the ratio of the rate it abstracts a proton from solvent to the rate it loses trimethylamine to give phenylcyclopropane cannot exceed fourteen. It follows, then, that at least 7.1% of the Nmethyl deuterium-labelled 3-carbanions, if formed, must react directly to produce trimethylamine-dl . A further complication arises because the 1,3-elimination reaction of the dideuterated salt is accompanied by an almost equal amount of Test for the 1',3-Elimination ( Ylide) 1,Zelimination which, of course, produces only Mechanism undeuterated trimethylamine. It follows from The test for the 11,3-eliminationmechanism, these considerations that even if the 1,3-eliminaeq. 4, was carried out in the conventional way tion were to proceed entirely by the ylide (24) determining the deuterium 'Ontent of mechanism, only one-half of 7.1% of the trithe trimethylamine formed from 3,3-dideutero- methylamine would necessarily contain a deu3 - ~ h e n ~ 1 ~ r o ~ ~ 1 t r i m e t h ~iodide. 1 a m m oThe nium terium atom. Since the limit of detection of triof an ylide require that a methylamine-dl is at least as good as 0.5%, it deuterium atom be transferred from the 3- follows that the 1',3-elimination mechanism, if position to the N-methyl carbon and result in the involved at all in the formation of phenylcycloformation of trimethylamine-dl. propane, can only be a minor reaction pathway. The trimethylamine formed in the reaction of the dideuterated salt was separated from am- Test for Exchange at C-3 monia and examined mass spectrometrically With the carbene and ylide mechanisms using the method of Cope and Mehta (25). The eliminated as major reaction pathways for the

2336

CANADIAN JOURNAL OF CHEMISTRY. VOL. 50, 1972

2. Nitrogen isotope effects in the 1,3-elimination 1.3-elimination reaction. there remain the E2 TABLE reaction of 3-phenylpropyltrimethylamrnonium iodide and Elcb processes originally proposed by with potassium amide in liquid ammonia at - 55 Bumgardner (16). A deuterium exchange test was ~erformedin an effort to detect the car- Experiment no. Extent of reaction (k14/k" - 1) 100 banion intermediate involved in the two-step 1 53 2.13 process, eq. 1. 2 70 2.15 3,3-Dideutero-3-phenylpropyltrimethylam3 80 2.32 4 82 2.08 monium iodide containing 9.78 ) 0.01 atom % 5 84 2.21 excess deuterium was allowed to react with 6 94 2.32 potassium amide in liquid ammonia to approxiMean 2 . 2 0 f 0 . 0 9 * mately 66% completion. The unreacted qua'Error limits are given as standard deviation. ternary salt was found to contain only 9.26, 0.03 atom % excess deuterium. This corresponds where Rfis the N14/N15mass ratio of the nitroto a replacement of approximately 11% of the gen in the unreacted salt, Rois the N14/N15mass dideuterated salt by its monodeuterated analog ratio of the nitrogen in the starting material, and and establishes that 3-carbanions are present in f is the mole fraction of salt that has reacted.The the reaction mixture. The actual extent of ex- results of these experiments are shown in Table change must be considerably greater than these 2. Since 96% of the reaction leads to phenylcycloresults indicate, however, since the large isotope propane, the observed isotope effect of 2.2% is effect associated with the abstraction of the that associated with the 1,3- elimination process. hydrogen on C-3 (vide infra) will result in a more It is of value to compare this result with the rapid rate of disappearance of the exchanged nitrogen isotope effects previously observed in salt by 1,3-elimination than of the deuterium- 1,2-elimination reactions of quaternary ammolabelled reactant. nium salts. Most measurements of the latter As was pointed out in the Introduction, the have been made at 60" and it is, therefore, observation that 3-carbanions are present in the necessary to estimate the magnitude of the reaction solution, although consistent with the isotope effect for the 1,3-elimination at this Elcb mechanism for 1,3-elimination, does not much higher temperature. This was done in two rule out a concerted reaction pathway. It is ways, both of which are very approximate. First, entirely possible that reversible formation of the it was assumed that the isotope effect at 60" for carbanion and elimination are entirely separate the 1,3-elimination will be the same fraction of but concurrent processes (20). Kinetic nitrogen the theoretical maximum, calculated (29) for and hydrogen isotope effect studies were next reaction at this temperature, as the value of carried out in an effort to distinguish between 2.2% is of the theoretical maximum at -55". the E2 and Elcb reaction pathways. This suggests a value of 1.4%at 60". Second, the temperature coefficient for the isotope effect Nitrogen Isotope Effect Nitrogen isotope effect measurements were (an increase of 0.6% for each decrease of 10") made to determine whether carbon-nitrogen observed in a previous study in these laborabond rupture is involved in a rate-determining tories for the reaction of cis-2-phenylcyclostep of the elimination reaction (26, 27). Reac- pentyltrimethylammonium ion with ethoxide tions of 3-phenylpropyltrimethylammonium ion in ethanol (30), was assumed to apply to the iodide of natural isotopic abundance were car- present reaction. This method predicts a value ried to extents of completion varying from of 1.5% at 60". The isotope effect of 1.4-1.5% estimated for 53-94% based on quaternary salt and the nitrogen isotope effect was evaluated from the the 1,3-elimination reaction using amide ion in isotopic composition of the nitrogen in the initial liquid ammonia may be compared with the reactant and recovered salt using an expression results obtained for a number of 1,2-eliminations derived by Bigeleisen and Wolfsberg (28), eq. 6, using ethoxide ion in ethanol : 2-phenylethyltrimethylammonium ion, 0.9% (27); cis-2-phenylcyclopentyltrimethylammonium ion, 1.1% (30) ; cis-2-phenylcyclohexyltrimethylammonium ion, 1.2% (30); and ethyltrimethylammonium ion,

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

O C

WESTAWAY AND BOURNS: ISOTOPE EFFECTS. V l l I

TABLE 3. Ratio of 1,3- and 1,2-eliminationfor reactions of undeuterated and deuterated quaternary salts in ammonia and ammonia-d, at - 55 "C Quaternary salt

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

Undeuterated Deuterated* Undeuteratedt Deuterated

Base and solvent KNH, KNH, KND, KND,

% 1,3-Elimination % 1,2-Elimination

in NH3 in NH, in ND, in ND,

96.0: 4 . 0 53.7:46.3 >99.6: 0 . 4 97.7: 2 . 3

1,3-Elimination 1,2-Elimination 24 :1 1.2:1 >250 :1 43 :1

'Reaction carried 66% to completion based on salt. tReaction carried 1 1 % to completion based on salt.

1.7% (27). It can be seen that the isotope effect found in the present study is at least as large as the effects found in 1,2-elimination reactions proceeding by the E2 mechanism with reactants in which there is phenyl activation of the abstracted hydrogen. From this it can be concluded that the 1,3-elimination reaction must be proceeding by a mechanism in which carbonnitrogen bond rupture is involved in the ratedetermining step, that is, either by an E2 process, eq. 2, or an Elcb reaction, eq. 1, in which the rate of return of the carbanion to reactant is at least as large as its rate of decomposition to elimination product (26, 3 1). Hydrogen Isotope Eflects If hydrogen exchange with solvent at C-3 does not occur, the isotope effect for 1,3-elimination is given directly by the ratio of 1,3- and 1,2elimination products from the undeuterated salt divided by this ratio for the products from the 3,3-dideutero-salt, since the rate of 1,2-elimination, except for a small secondary isotope effect, will not be affected by isotopic substitution. The cyclopropane: olefin ratio for the unlabelled reactant was found to be 24 and that for the dideuterated salt 1.2 (see Tables 1 and 3), giving an isotope effect for 1,3-elimination of 20. The elimination reaction, however, is accompanied by exchange resulting in the conversion of a considerable amount of the dideuterated salt to monodeuterated salt which, because of the hydrogen isotope effect, will yield a much higher cyclopropane: olefin ratio than that for the original reactant. This means that the observed value of 20 actually represents a lower limit and that the actual isotope effect for phenylcyclopropane formation must be considerably higher than this. Even for a reaction at -55", this isotope effect is large and suggests the possibility of tunnelling (32).

Of even greater significance with respect to the elucidation of the mechanism of 1,3-elimination is the hydrogen isotope effect associated with the rate of disappearance of the undeuterated and dideuterated quaternary salts. For the undeuterated salt, reaction is by 1,2- and 1,3elimination only, since isotopic exchange with solvent simply regenerates the original reactant; for the dideuterated salt, reaction involves not only elimination but also exchange giving monodeuterated salt.4 This isotope effect was determined by the isotopic competitive method (28). A mixture of undeuterated and dideuterated salts containing 4.75 0.01 atom % excess deuterium was allowed to react until 57.7% of the salt had been converted to elimination products. The recovered quaternary salt was found to contain 8.48 0.02 atom % excess deuterium. The hydrogen isotope effect for the disappearance of dideuterated and undeuterated quaternary salt was then evaluated using eq. 6. In applying this equation, there is a problem in arriving at an accurate value for R,,the ratio of undeuterated and dideuterated salt in the recovered reactant, because of the presence of an unknown amount of monodeuterated salt formed by isotopic exchange with solvent. An upper limit to the amount of exchange is the 11% found in the deuterium exchange experiment (vide ante) in which 66% of the dideuterated salt had been converted to hydrocarbon compared to the less than 25% conversion of this salt in the present experiment. Assuming this extent of exchange, the composition of the recovered salt is as follows : undeuterated 10.4%, monodeuterated 9.7%, and dideuterated 79.9%.

+

+

4Further exchange produces undeuterated salt, but this will be small because of the intramolecular isotope effect associated with the abstraction of hydrogen from the monodeuterated compound.

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

2338

CANADIAN JOURNAL OF CHEMISTRY. VOL. 50, 1972

The total extent of reaction of the undeuterated and dideuterated salts to produce hydrocarbon and monodeuterated salt becomes 61.8%. These data give an isotope effect of 7.3. On the other hand, if it is assumed that the recovered reactant consists only of undeuterated and dideuterated salt (no monodeuterated salt formed by exchange), its composition would be 15.2% undeuterated and 84.8% dideuterated, the extent of reaction 57.7% (vide ante) and the isotope effect 7.5. This close correspondence in the calculated isotope effects using the two extreme limits for the extent of exchange may be fortuitous; the significant fact is that there is a large isotope effect associated with the disappearance of the isotopic reactants. An isotope effect of this magnitude, of course, is consistent with the 1,3-elimination reaction proceeding by the E2 mechanism, eq. 2. It also can be accommodated by an Elcb process, eq. 1, provided that the rate of return of the carbanion intermediate to quaternary salt is not very large relative to its rate of decomposition to phenylcyclopropane. The reason for this is that carbanion return in the case of the undeuterated salt reaction generates starting compound whereas return in the dideuterated salt reaction ~roduces a reaction product, namely monodeuterated salt. It follows that the greater the extent of carbanion return the faster the rate of disappearance of dideuterated salt relative to that of the undeuterated salt and the smaller the observed isotope effect. At large rates of return, the isotope effect, so defined, would actually favor the deuterated reactant. In summary, both the nitrogen and hydrogen isotope effect results are in accord with a concerted mechanism for the 1,3-elimination reaction. They may also be accommodated by a twostep process involving a carbanion intermediate provided that certain restrictions are placed on the relative rates with which this intermediate abstracts a proton from solvent to regenerate quaternary salt and loses trimethylamine to form the cyclopropane product. The nitrogen isotope effect of 2.2% would require that this ratio be greater than unity, whereas the hydrogen isotope effect of approximately 7.4 associated with the disappearance of undeuterated and dideuterated salts would require that this ratio not be large. (This will be discussed in more quantitative terms later in this section.)

Studies in Ammonia-d3 The reactions of both undeuterated and dideuterated quaternary salts were carried out in ammonia-d3 in an effort to distinguish between the two reaction pathways accommodated by the isotope effect results. Although the deuterium exchange found in the reaction of deuterated salt in ordinary ammonia demonstrated the presence of 3-carbanions in solution, it was impossible to even estimate the extent of exchange since the resulting monodeuterated salt undergoes 1,3-elimination much faster than the starting compound. It was reasoned that a better estimate of the amount of exchange could be obtained if the exchange test was performed by reacting undeuterated salt in ammonia-d3 where the monodeuterated salt would react more slowly than the starting compound and would, therefore, accumulate during reaction. A reaction of the undeuterated salt in ammonia-d3 was carried to 1 1% completion and the recovered salt was examined by both i.r. and n.m.r. spectroscopy. It was found that extensive exchange had occurred at the N-methyl carbons (50%) and at C-1 (3573, but deuterium was not detected at either C-2 or -3. The absence of exchange at the 3-position in this reaction compared to the extensive exchange observed in ordinary ammonia cannot be understood in terms of a concerted elimination process which is accompanied by an irrelevant formation of 3-carbanions. This is because the relative rate of carbanion formation and E2 elimination should be approximately the same in the two cases, and if the 3-carbanions are not intermediates in the elimination reaction, but only react with solvent to produce exchanged salt, the relative rate of exchange and elimination cannot be significantly different whether a hydrogen is being abstracted from ammonia or a deuterium from ammonia-d, . On the other hand, the results of the two exchange reactions can be readily understood in terms of an Elcb process, eq. 1, provided there is a large isotope effect associated with the abstraction of a proton from the solvent by the carbanion. If, for the reaction in ordinary ammonia, the rate with which the carbanion abstracts hydrogen from solvent is not much greater than the rate of its decomposition to phenylcyclopropane, then for the reaction in

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

WESTAWAY AND BOURN S: ISOTOPE EFFECTS. V l l l

deuterated ammonia, a much slower rate of carbanion return will result in an elimination rate which is much greater than the rate of exchange. Since exchange was extensive in ammonia and not within the limits of detection in ammonia-d,, the isotope effect for proton abstraction from solvent must be extremely large. This is not surprising, however, since, on the basis of zero point energy consideration (29), the isotope effect associated with the rupture of a N-H bond will undoubtedly be larger than that for a C-H bond (33,34) and it has already been demonstrated that the isotope effect for the abstraction of the hydrogen from C-3 is greater than 20. The conclusions based on the results of the exchange experiments find confirmation in the relative amounts of 1,3- and 1,2-elimination in the reactions of deuterated and undeuterated salt in ammonia and ammonia-d,, see Tables 1 and 3. It is seen that the reaction of the undeuterated salt in deuterated solvent gives a very much higher 1,3- to 1,2-elimination ratio than the reaction of this salt in ordinary ammonia. This is difficult to explain if the 1,3- as well as the 1,2-elimination reactions are concerted, since the secondary and solvent isotope effects associated with a change in the isotopic nature of the base and solvent should be approximately the same for the two elimination reactions. The same would apply for the reaction of the deuterated salt in the two base-solvent systems; yet, again, a much higher 1,3- to 1,2- ratio is found in ammonia-d,. On the other hand, the results are readily understood in terms of an Elcb mechanism, since the large primary isotope effect associated with the return of the carbanion to reactant would result in a much faster overall rate of 1,3-elimination for the reaction in ammonia-d,. Similarly, the increase in the 1,3- to 1,2elimination ratio from 24: 1 to 43: 1 in going from the reaction of the undeuterated salt in ammonia to that of the deuterated salt in ammonia-d, can only be understood in terms of the Elcb mechanism. If the 1,3-elimination is concerted, the primary hydrogen isotope effect associated with the quaternary salt would strongly decrease the rate of 1,3-elimination, while secondary and solvent isotope effects resulting from a change from ammonia to ammonia-d, would be approximately the same for

2339

the two processes. The result would be a large decrease in the 1,3- to 1,2-elimination ratio, contrary to observation. On the other hand, these results are fully in accord with an Elcb reaction pathway for - phenylcyclopropane formation. The decrease in the rate of 1,3-elimination resulting from the substitution of deuterium for hydrogen on C-3 of the quaternary salt will be compensated by an increase in the rate of this elimination as a consequence of a much smaller tendency for carbanion return in the deuterated solvent. In fact, the primary isotope effect associated with solvent is seen to outweigh the effect associated with the substrate, resulting in an overall increase in the 1,3- to 1,2-elimination rate ratio. Other comparisons of the elimination ratios shown in Table 3 are equally in accord with the El cb mechanism for the 1,3-elimination. Relative Rates of Carbanion Return and Elimination Upper and lower limits can be established for the ratio of the rates of carbanion return and conversion to phenylcyclopropane from the hydrogen isotope effects observed for the disappearance and 1,3-elimination of the undeuterated and dideuterated salts in ordinary ammonia. It may readily be shown that the rate constant, kH, for the disappearance of the undeuterated salt is given by eq. 7, where klP2is the rate

constant for 1,2-elimination, ky is the rate constant for the formation of the 3-carbanion from undeuterated salt, k-I is the rate constant for carbanion return in ordinary ammonia and includes the concentration of the solvent, and k, is the rate constant for conversion of the carbanion to phenylcyclopropane (see eq. 1). Also, the rate constant, kD, for the disappearance of the dideuterated salt by elimination and exchange in ordinary ammonia is given by eq. 8, where ky is the rate constant for carbanion formation from the deuterated salt. (In this treatment, secondary isotope effects are neglected as is the formation of undeuterated salt by a double exchange of the dideuterated reactant.) Combining eqs. 7 and 8 and substituting the

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

CANADIAN JOURNAL OF CHEMISTRY. VOL. 50, 1972

FIG. 1. Apparatus for the reaction of quaternary ammonium salts with potassium amide in liquid ammonia.

observed values of 7.4 for kH/kD,the isotope effect for the disappearance of undeuterated and deuterated salts, and 24 for kyk,/k,,,(k-, k,), the ratio of 1,3- and 1,2-elimination from the undeuterated compound (see Table 3), gives eq. 9.

+

It has been shown (vide ante) that the isotope effect for 1,3-elimination from undeuterated and deuterated quaternary salts is greater than 20. This represents a lower limit for kylky, the isotope effect for carbanion formation, and, on substitution into eq. 9, it gives a lower limit of 1.0 for the ratio of k-, lk,. An upper limit for this ratio can be established by substituting into eq. 9 the maximum probable value for kylk?. The largest hydrogen-deuterium isotope effect for a proton transfer reaction has been reported by Lewis and Funderburk (35) who observed a value of 24.8 for the reaction of 2,4,6-trimethylpyridine with 2-nitropropane and 2-nitropropane-2-d at 25". A calculation based on the activation parameters determined by these authors gives a value of approximately 150 for this effect at -55". Substitution of this value for ky/ky in eq. 9 gives a k-,/k, ratio of 14. Neither of these limits represents probable ratios for the rates of carbanion return and elimination. The lower limit of 1.0 has been based on an experimentally observed value for

the isotope effect for 1,3-elimination which has not been corrected for the extensive exchange which accompanies the reaction of the deuterated salt; the upper limit of 14 is based on an isotope effect which is undoubtedly much too high for a reaction in which there is relatively little steric hindrance to proton abstraction (35). Although it is not possible to specify narrower limits with any degree of certainty, a range of 3-10 would seem reasonable.

Experimental Preparation oJ3.3-Dideutero-3-phenyltrimethylammonium Iodide This compound was prepared in four steps from 1,ldideuterobenzyl chloride (8.6 g, 0.066 mol) which was made by the method of Buncel and Bourns (26). The chloride was converted to the Grignard reagent which was treated with ethylene oxide to give 3,3-dideutero-3-phenylpropanol(36). The alcohol was converted to the tosylate and thence to N,N-dimethyl-3,3-dideutero-3-phenylpropyamine by reaction with dimethylamine (37). Reaction of the amine with methyl iodide gave the quaternary salt (8.0g) which was recrystallized from ethanol-ether to a constant m.p. of 178.0-178.5' (uncorrected); lit. m.p. 178.5-180" (37). The . ~ undeuterproduct contained 1.96 atoms D / m o l e c ~ l e The ated salt was prepared by the same sequence of reactions. General Experimental Procedure The apparatus used for the reaction of the quaternary ammonium salt with potassium amide in liquid ammonia is shown in Fig. I . Reaction flask A contained a magnetic stirrer and was fitted with a Dry Ice-acetone condenser, D,, and a vial, El, for adding potassium metal. Flask B was 'Deuterium analyses were performed by Mr. J. Nemeth, 303 Washington Street, Urbana, 111.

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

WESTAWAY A N D BOURNS: ISOTOPE EFFECTS. V l l I

fitted with an air-tight stirrer, a Dry Ice-acetone condenser, D,, and a vial, E,, for addition of ammonium chloride. Condenser, Dl, was connected to nitrogen and ammonia purification trains and D, to a barium oxide drying tube, F, whose exit tube passed into a bubbler, H, containing oil. Ferric nitrate nonahydrate (10 mg) was placed in flask A and the quaternary salt (usually 2 g) was added to flask B. Following a thorough flaming of the apparatus and flushing with nitrogen, 200 ml of ammonia was condensed into flask B and 100 ml into flask A. Potassium metal (usually 0.57 g) was added slowly with stirring to A. Thirty minutes after the blue color of the potassium had disappeared, D l was replaced by an adapter joined to the nitrogen train, both flasks were cooled to - 55 l o in a Dry Ice-diacetone alcohol bath, flask A was inverted, and the amide solution was pushed by nitrogen through a glass wool filter, G, into B. The reaction was stopped by adding an excess of ammonium chloride. Ether (200 ml) was added to flask B, which was equipped with a water condenser connected to a gaswash bottle containing ether, and the ammonia was evaporated slowly. Finally, 100 ml of water was added to B to dissolve the residual salts, the ether layer was separated and the aqueous layer extracted with additional ether. The combined ether solutions contained the hydrocarbon products and the aqueous solution the unreacted quaternary salt.

+

Product Identification and Analysis The ether solution was concentrated by distillation using a spinning band column and the residue analyzed by gas chromatography using a 318 in. x 20 ft column of 30% SE-30 on 45/60 mesh Chromosorb P at 150". Three components were separated but only component 2 was recovered in sufficient quantity for identification. Both its n.m.r. and i.r. spectra were identical to those reported for phenylcyclopropane (23,24). The minor products were present in much larger amounts in the reaction of the deuterated salt. Component 1 was identified as 3-phenylpropene from its retention time on the gas chromatograph and its i.r. spectrum. Component 2 had the same retention time as phenylcyclopropane and trans-1phenylpropene which is known to be formed by isomerization of 3-phenylpropene (component 1) under the reaction conditions (22). Also, its n.m.r. spectrum corresponded to that expected for a mixture of 1-phenylcyclopropane-1-dand trans-1-phenylpropene-1-d.Identification of component 2 as a mixture of these two hydrocarbons was confirmed by the observation that three different methods of quantitative analysis, each assuming this composition, gave identical results: integration of the n.m.r. spectrum, gas chromatographic analysis using temperature programming on a 114 in. x 40 ft column of 5% Ucon Polar: and U.V.spectroscopy at 250 mp, where the molar extinction coefficient for trans-1-phenylpropene was found to be 16 000 and for phenylcyclopropane 153. Finally, component 3 had a retention time which was identical to that of cis-1-phenylpropene, which is a minor product of the base-catalyzed isomerization of 3-phenylpropene.

2341

quantitative recovery of unreacted quaternary ammonium iodide. The water layer from the reaction mixture was taken nearly to dryness to remove ether and ammonia and the concentrated solution was diluted with water and passed through an anion exchange column (Dowex 1-X8, 100/200 mesh) in the chloride form. The eluant was taken to dryness and the residual chloride salts were thoroughly dried. The salts were shaken for several hours with excess anhydrous acetone and the insoluble potassium and ammonium chlorides were separated by filtration. The quaternary chloride, recovered from the filtrate by removal of acetone, was dissolved in water and the solution was treated with silver oxide and filtered. The filtrate was brought to a p H of 5.5 with hydriodic acid, filtered and taken to dryness. Control experiments in which the quaternary iodide was recovered by this procedure from a solution of liquid ammonia containing more than twice the amount of potassium and ammonium chloride and iodide salts that would be present in reaction mixtures showed that the error in the extent of reaction would not exceed 3%. Test for the Carbene Mechanism Deuterium analyses were done on the phenylcyclopropane-trans-1-phenylpropene product (component 2) from the reaction of the deuterated salt using n.m.r. and mass spectroscopy. For the latter, the ionization voltage was 8.5 eV and the relative intensities of the mass peaks were compared with those for a synthetic sample of undeuterated hydrocarbon of the same composition (71% phenylcyclopropane and 29% trans- 1-phenylpropene). The percentage of undeuterated, monodeuterated, and dideuterated hydrocarbon in the product was calculated by the method described by Biemann (38). Test for the Ylide Mechanism T o a solution of 3,3-dideutero-3-phenylpropyltrimethylammonium iodide (0.44 g, 0.0014 mol) in 45 ml of ammonia was added a solution of potassium amide prepared by adding potassium (0.30 g, 0.0077 mol) to 65 ml of ammonia. The reaction was quenched with ammonium chloride after 30 min and 100 ml of ether was added. The ammonia and trimethylamine were distilled into 11 of stirred ice-cold hydrochloric acid. The hydrochloride salts, recovered by removal of excess acid under reduced pressure, were dried and the trimethylammonium chloride was extracted from the large excess of ammonium chloride by stirring successively with 600 ml of ethanol and two 550 ml portions of chloroform. The salt, recovered by evaporation of the combined solutions, was further purified by a second chloroform extraction. It was then dissolved in ethanol, the solution made basic to phenolphthalein with sodium ethoxide and the resulting trimethylamine purified by gas chromatography using a 10 ft x 318 in. column of 20% amine 220 on 60180 mesh, base-washed Chromosorb W7 at 55'. Its mass spectrum was compared with that obtained from a purified sample of trimethylamine of natural isotopic abundance.

Extent of Reaction The extent of reaction, required for both hydrogen and nitrogen isotope effect measurements, was determined by

Test for Isotopic Exchange at C-3 of 3,3-Dideutero-3phenylpropyltrimethylammonium Iodide in Liquid Ammonia Two deuterium exchange experiments each using 1.1 g (0.0035 mol) of deuterated salt were carried out as described

6This analysis was done by Varian Aerograph, Walnut Creek, California.

7The column was obtained from Varian Aerograph, Walnut Creek, California.

2342

CANADIAN JOURNAL OF CHEMISTRY. VOL. 50, 1972

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

under General Experimental Procedure. The reactions were quenched after 9 min, the unreacted salt was recovered, recrystallized from ethanol-diethyl ether, and analyzed for d e u t e r i ~ mThe . ~ extent of reaction in each case was 66 f 2%. A control experiment showed that no deuterium is lost in the recovery.and purification procedures. Hydrogen Isotope Effect Determinafion by the Competitive Mefhod A mixture of undeuterated and dideuterated quaternary salts (0.395 g, 0.0013 mol and 0.358 g, 0.0012 mol, respectively) in 300 ml of liquid ammonia was treated with base prepared by adding potassium (0.309 g, 0.0075 mol) to 150 ml of liquid ammonia. The reaction was quenched after 92 s by addition of ammonium chloride and the unreacted salt was recovered, purified, and analyzed5 for deuterium. Nitrogen Isotope Effect Determination The nitrogen isotope effect in the elimination reaction was determined by comparing the N14/NlS ratios for the quaternary salt recovered from reactions carried to known extents of completion with this ratio in the initial reactant (28). Samples of salt were decomposed in a Kjeldahl digestion to ammonium sulfate which was then oxidized to molecular nitrogen for mass spectrometric analysis (27). A control experiment showed that there is no isotopic fractionation in the recovery and purification procedures. Studies in Ammonia-d, (a) Reaction of Undeuterated Quaternary Salt 3-Phenylpropyltrimethylammonium (1.11 g, 0.0036 mol) in 100 ml of ammonia-d, (Merck, Sharpe and Dohme) was treated with base formed by the addition of potassium (0.114 g, 0.0029 mol) to 75 ml of ammonia-d,. The reaction mixture was colorless rather than the deep reddish-brown of all other reactions. The reaction was quenched after 45 min with ammonium chloride and the unreacted quaternary salt recovered and purified. The i.r. spectrum of the recovered salt in a potassium bromide disc showed none of the absorptions associated with the C,-D bonds while the intensities of the C,-H absorptions were the same as those for the starting compound. On the other hand, the absorptions associated with the N-methyl C-H and C,-H bonds were much less intense and several new absorptions associated with Nmethyl C-D and C,-D bonds were observed. Analysis of the 100 MHz n.m.r. spectrum of the recovered salt in deuterium oxide confirmed that there had been no exchange of C-2 and -3 hydrogens but that 50% of the N-methyl hydrogens and 35% of the C-1 hydrogens had been replaced by deuterium. Gas chromatographic analysis of the hydrocarbon product failed to show the presence of any of the isomeric phenylpropenes. An upper limit of 0.4% phenylpropenes was established by showing that the amount of 3-phenylpropene which would have been formed by this percentage of j-elimination produced a distinct coloration whereas the reaction solution was colorless. (b) Reaction of Deuterated Quaternary Salt 3,3-Dideutero-3-phenylpropyltrimethylammonium iodide (0.97 g, 0.0032 mol) in 100 ml of ammonia-d, was treated with base prepared from potassium (0.35 g, 0.009 mol) in 75ml of ammonia-d,. The reaction was quenched with ammonium chloride after 90 min. Neither 3-phenylpropene

nor cis-I-phenylpropene was detected by gas chromatography while only 2.3% of trans-1-phenylpropene, the phenylpropene which predominates in basic solution under equilibrium conditions (22), was shown to be present by U.V. spectroscopy. A control experiment showed that the latter olefin does not undergo significant polymerization under the conditions of the experiment. The authors are greatly indebted to the National Research Council of Canada for the financial support for this investigation and the Department of University Affairs of the Province of Ontario for scholarships (to K.C.W.). 1. A. NICKONand N. H. WERSTIUK.J. Am. Chem. SOC. 89,3914 (1967). 2. S. WINSTEIN, E. CLIPPINGER, R. HOWE,and E. VOGELFANGER.J. Am. Chem. Soc. 87, 376 (1965). 3. B. M. BENJAMIN, B. W. PONDER,and C. J. COLLINS. J. Am. Chem. Soc. 88, 1558 (1966). 4. A. NICKONand N. H. WERSTIUK.J. Am. Chem. Soc. 89, 3915 (1967); A. NICKONand N. H. WERSTIUK.J. Am. Chem. Soc. 89,3917 (1967). and J. K. STILLE. J. Org. Chem. 31, 5. F. M. SONNENBERG 3441 (1966); J. K. STILLEand F. M. SONNENBERG. Tetrahedron Lett. 4587 (1966); H. KWART, T. TAKESHITA, and J. L. NYCE. J. Am. Chem. Soc. 86,2606 (1964). C. SWITHENBANK, and A. LEWIS. J. Am. 6. J. MEINWALD, and J. K. Chem. Soc. 85, 1880 (1963); J. MEINWALD CRANDALL.J. Am. Chem. Soc. 88, 1292 (1966); S. J. CRISTOL and B. B. JARVIS. J. Am. Chem. Soc. 88,3095 and M. S. (1966); S. J. CRISTOL,J. K. HARRINGTON, SINGER. J . Am. Chem. Soc. 88, 1529 (1966); S. J. CRISTOLand B. B. JARVIS. J. Am. Chem. Soc. 89, 401 (1967). 7. H. 0. HOUSEand F. A. RICHEY,JR. J. Org. Chem. 32, 2151 (1967); H. R. NACEand B. A. OLSEN. J. Org. R. R. FRAME, Chem. 32,3438 (1967); F. G. BORDWELL, R. G. SCAMEHORN, J. G. STRONG,and SEYMOUR MEYERSON.J. Am. Chem. Soc. 89,6704 (1967); F. G . BORDWELL and M. W. CARLSON.J. Am. Chem. Soc. 92,3370 (1970). and J. G. BERGER. J. Am. Chem. Soc. 83, 8. L. FRIEDMAN and W. VON E. DOERING.Tetra492 (1961 ); W. KIRMSE hedron, 11,266 (1960). 9. R. A. RAPHAEL.In Chemistry of carbon compounds. Vol. IIA. Edited by E. H. Rodd. Elsevier Publishing Co., New York, N.Y., 1953. p. 24; J. WEINSTOCK. J. Org. Chem. 21, 540 (1956). 10. W. E. TRUCEand L. B. LINDY. J. Org. Chem. 26,1463 (1961). 11. N. P. NEUR~ETER and F. G . BORDWELL.J. Am. Chem. SOC.85, 1209 (1963). 12. P. G. BAY. Chem. Abstr. 60,421 (1964). and L. C. 13. J. B. CLOKE,E. STEHR,T. R. STEADMAN, WESCOTT. J. Am. Chem. Soc. 67, 1587 (1945); C. GILBEAT,M. C. ROUX-SCHMITT, and J. SEYDEN-PENNE. Bull. Soc. Chim. Fr. 2405 (1970). 14. H. M. WALBORSKY and C. G. PI^. J. Am. Chem. Soc. 84,4831 (1962). 15. C. L. BUMGARDNER. Chem. Ind., 1555 (1958); C. L. J. Am. Chem. Soc. 83,4420 (1961). BUMGARDNER. 16. C. L. BUMGARDNER. J. Am. Chem. Soc. 83, 4423 (1961).

~ I

1

1

i

1I

I

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 190.107.183.238 on 08/26/15 For personal use only.

1

~

I

I

1 I

I

WESTAWAY A N D BOURNS: ISOTOPE EFFECTS. VIII

2343

17. C. L. BUMGARDNER. J. Org. Chem. 29, 767 (1964); 27. G. AYREY, A. N. BOURNS, and V. A. VYAS. Can. J. C. L. BUMGARDNER. Chem. Commun. 374 (1965); Chem. 41, 1759 (1963). and M. WOLFSBERG.Advances in chemiC. L. BUMGARDNER and H. IWERKS.J. Am. Chem. 28. J. BIGELEISEN cal physics. Vol. 1. Interscience, New York, N.Y., 1958. SOC.88, 5518 (1966). 18. R. BAKERand M. J. SPILLETT.J. Chem. Soc. B, 581 p. 15. (1969). 29. L. MELANDER.Isotope effects on reaction rates. 19. R. BAKERand M. J. SPILLETT.J. Chem. Soc. B, 880 Ronald Press, New York, N.Y., 1960. p. 20. 30. A. C. FROSST. Ph.D. Thesis. McMaster University, (1969). 20. R. BRESLOW.Tetrahedron Lett. 399 (1964). Hamilton, Ontario, 1968. pp. 116-1 18. 21. D. V. BANTHORPE.Elimination reactions. Elsevler 31. A. N. BOURNS and P. J. SMITH. Proc. Chem. Soc. 366 (1964); P. J. SMITHand A. N. BOURNS.Can. J. Chem. Publishing Co., New York, N.Y., 1963. p. 105. 22. T. W. CAMPBELL and W. G. YOUNG. J. Am. Chem 48, 125 (1970). Soc. 69, 688 (1947); E. A. RAB~NOV~CH, I. V. ASTAF'EV, 32. R. P. BELL. The proton in chemistry. Cornell Univerand A. I. SHATENSHTEIN. Zh. Obshch. Khim. U.S.S.R. sity Press, Ithaca, New York, 1959. p. 21 1. J. Chem. Phys. 37,2138 32, 748 (1962). Chem. Abstr. 58,6673e (1963). 33. S. BRATO?and M. ALLAVENA. 23. Varian Associates. Varian high resolution N.M.R. (1962). spectra catalogue, Vol. 2. Instrument Division, Palo 34. L. J. BELLAMY.The infra-red spectra of complex molecules. J. Wiley and Sons, New York, N.Y., 1958. Alto, California. 1963. Spectrum 528. J. Am. Chem. 24. F. WEYGAND, H. DANIEL, and H. SIMON. Chem. Ber. 35. E. S. LEWISand L. H. FUNDERBURK. SOC.89,2322 (1967). 91, 1691 (1958); G. AYREY,E. BUNCEL,and A. N. BOURNS.Proc. Chem. Soc. 458 (1961); W. H. SAUN- 36. R. C. HUSTON and A. H. AGETT. J. Org. Chem. 6, 123 DERS,JR. and D. PAVLOVIC.Chem. and Ind., 180 (1941). J. Am. Chem. 37. A. C. COPEand C. L. BUMGARDNER. (1962). 25. A. C. COPEand A. S. MEHTA. J. Am. Chem. Soc. 85, SOC.79,960 (1957). 1949 (1963); A. S. MEHTA. Ph.D. Thesis. Massa- 38. K. BIEMANN.Mass spectrometry, organic chemical applications. McGraw-Hill, New York, N.Y., 1962. chusetts Institute of Technology, Cambridge, Mass., 1963. Chapt. 5. and A. C. FROSST. Can. J. Chem. 48, 26. E. BUNCEL and A. N. BOURNS.Can. J. Chem. 38,2457 39. A. N. BOURNS (1960); P. J. SMITHand A. N. BOURNS.Can. J. Chem. 133 (1970). 44,2553 (1966).

Suggest Documents