The Nonclassical Norbornyl Cation

This dissertation has been microfilmed exactly as received 66-15,112

MARSHALL, James Lawrence, 1940- THE NONCLASSICAL NORBORNYL CATION.

The Ohio State University, Ph.D., 1966 Chemistry, organic

University Microfilms, Inc., Ann Arbor, Michigan THE NONCLASSICAL NORBORNYL CATION

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

James Lawrence Marshall*, B .S.

sic*###*

The Ohio State University 1966

Approved by p / J- A - Adviser Department of Chemistry ACKNOWLEDGMENTS

The author is indebted to the National Science Foundation and the Petroleum Research Foundation for partial support of this research.

ii VITA

James Lawrence Marshall, the son of Madison L. and Irene V.

Marshall,, was bora in Denton, Texas, on May 19, 1940. After gaining his elementary education in this city, his family moved to Decatur,. Alabama, in June 1952» where he attended Decatur

High School. In September 1958, he entered Davidson College,

Davidson, North Carolina, on a Proctor and Gamble Scholarship.

After two years he transferred to Indiana University, Blooming ton, Indiana, where-he received his B.S. in chemistry in June

1962. The following September he entered the Graduate School of The Ohio State University. In March 1963 > he married Julia

Clark Bechtel of Niles, Michigan.

The following appointments were held:

1962-1963 N.S.F. Cooperative Fellowship

1964-1965 " '* "

1965-1966 " " "

I 963.-I964 Petroleum Research Foundation Fellowship

In June 1966, he received his Ph.D. in organic chemistry from The Ohio State University.

iii CONTENTS

Page

ACKNOWLEDGMENTS...... i i

VITA...... i i i

TABLES ...... ^

ILLUSTRATIONS ...... x i

INTRODUCTION ...... 1

PART I . THE 2-NORBORNYL CATION...... 8

1* The problem ...... 8

Known classical norbornyl systems ...... 15 The transition s ta te ...... 19

2. Approach to the problem ...... 23

3 . R e s u l t s ...... 27

PART I I . THE 7-NQRBORNENYL C A T IO N ...... 65

1 . The p r o b l e m ...... 65

2. Approach to the problem ...... 69

3 . R e s u l t s ...... 70

EXPERIMENTAL...... 103

R e a g e n ts ...... 103

5 jL5 -Dimethoxy-l ,2 »3*4-tetrachloro- cy clo p en tad ien e C51) 104

7 >7-D im ethoxy-l>2, 3 ,4-tetrachlorobicyclo[2.2.1]hept-2-ene ($2) 104

7j7-Dimethoxybicyclo[2.2.1]heptene (jQ) ...... 105

iv CONTENTS (C ontinued)

- . Page Hydroboration of 7>7-dimethoxybicyclo - [ 2 . 2 .1]heptene (52.) ...... 105

7, 7-Diniethoxybicyclo[ 2 . 2 .1 ]heptane-exo- 2 >3-epoxide ( 67 ) ...... 106

7»7-Dimethoxybicyclo[2.2.1]heptan-exo-2-ol (54) . . . 107

7»7-Dimethoxybicyclo[2.2.1]heptan-2-one ( 66 ) .... 108

2 ,4-Dinitrophenylhydra zone of 7» 7-dimethoxy- b ic y c lo [ 2 . 2 .1 ]h e p ta n - 2-one ...... 108

7>7-Eimethoxybicyclo[2.2.1]heptan-endo-2-ol (55) . . 109

Oxidation of 7*7-dimethoxybicyclo[2.2.1]- heptan-endo- 2- o l ( 55 ) ...... 110

Reduction of 7>7-dimethoxybicyclo[2.2.1]- h e p ta n - 2-one ( 66 ) with sodium-ethanol ...... 110

Reduction of 7i7-d:methoxybicyclo[2.2.1]heptan- 2-one ( 6 6 ) with lithium aluminum hydride .... 110

Reduction of 7x7-diroe'thoxybicyclo[2.2.1]heptane- exo -2 ,.3-epoxide with lithium-ethylamine .. I l l

exo-2-Acetoxy-7,7-diraethoxybicyclo[2.2.13- heptane ( 62 ) ...... I l l

endo-2-Acetoxy-7>7-dimethoxybicyclo[2. 2 * 1 ]- heptane ( 63 ) ...... 112

7 *7-Dimethoxybicyclo[2.2,l]heptan-exo-2-ol 2-toluene sulfonate (6 4 ) ...... 112

7 >7-Dimethoxybicyclo[ 2 . 2 .l]heptan-endo- 2- o l 2 -toluenesulfonate ( 65 ) ...... 113

exo-2-Hydroxybicyclo[2.2. 1 ]heptan-7-one ( 5 6 ) .... 113

exo-2-Acetoxybicyclo \ 2 . 2 .llheotan-7-one ( 60 ) .... 114

exo-2-Acetoxybicyclo[2.2.l]heptan-7-one ( 60 ) via the hydrolysis of exo- 2- a c e toxy -7 »7- dimethoxybicyclo[ 2 . 2 . 1 ]heptane ( 62 ) ...... 114

v CONTENTS (Continued) Page

exo-2-Hydroxybicyclo[2.2.1]heptan-7-one jo-toluenesulfonate (£8 ) ...... 115

endo-2-Hydroxybicyclo [2 .2.llheptan-7-one (57) .... 115

endo-2-Acetoxybicyclo[2.2.1]heptan-7-one ( 6 l ) .... 115

endo-2-Hydroxybicyelof 2.2. 1 ]hepta n-7-one £-toluenesulfonate (£2 ) ...... I l 6

Reduction of exo-2-acetoxybicyclof2.2.11- h ep tan - 7-one ( 6 o) with lithium aluminum h y d rid e ...... 116

7-Ketobicyclo[2.2.1]heptene ( 7 5 ) ...... 117

7>7-exo-2-Trimethoxybicyclo[2.2.1]heptene ( 80) . . . 117

exo-2-Methoxybicyclor 2 . 2 . 1 Iheptan-7-one ( 78) . ... 118

exo-2-syn-7-Bicyclof2.2.1Iheptanediol ( 6 8 ) ...... 118

exo-2-anti-7-Bicyclor2.2.11heptanediol (10b) .... 119

7-t-Butoxybicyclo[2.2.1]heptadiene ...... 119

7-Acetoxybicyclo[2.2.1]heptadiene ...... 120 a n ti- 7-Rydroxybicyclor 2 . 2 .llheptene tetrahydropyranyl e th e r ...... 121 exo~2-anti-7-Bicyclor2.2.1Iheptanediol (106) . . . . 121 syn-7-Hydroxybicyclor2.2.1lheptan-2-one (107) .... 122 syn-7-Hydroxybicyclof2.2.1]heptan-2-one jo-toluenesulfonate (103) via Jones oxidation of exo-2- syn-7-bicyclor2.2.11heptanediol ( 68 ) ...... 123 syn-7-Hydroxybicyclof 2.2.llheptan-2-one £-toluenesulfonate (103) via chromic acid-water oxi dation of exo- 2-sy n - 7-b ic y c lo r 2 . 2 . 11- heptanediol ( 68 ) ...... exo-2-syn-7-Bicyclor 2 . 2 .l]heptanediol-di £-toluenesulfonate ( 108) ...... 12if

vi CONTENTS (Continued) Page

syn-7-Hyd roxybicyclo[2.2.1]h epta n-2-one ^-toluenesulfonate (103) via solvolysis o f ex o - 2- syn- 7-b ic y c lo r 2 . 2 .llheotane- diol di-£-toluenesulfonate ( 108) ...... 125

anti-7-Hydroxybicyclo[2.2.1]heptan-2-one ( 110) .... 126

anti-7-Hydroxybicyclor 2 . 2 .llheptan-2-one £-toluenesulfonate (104) ...... 126

exo-2-anti-7-Bicyclo[2.2.l]heptanediol di-£-toluenesulfonate (ill) ...... 127

an ti-7-H.yd roxybicyc lo [ 2.2 .l]heptan-2-one £-toluenesulfonate (104) via solvolysis o f exo- 2- a n t i - 7- b ic y c lo [ 2 . 2 .l]heptanediol di-£-toluenesulfonate (ill) ...... 127

s,yn-7-Acetox,ybicyclo f 2.2 .llheptan-2-one (113.) .... 129

anti-7-Acetoxybicyclo[2.2.1]heptan-2-one (114) .... 130

KINETICS...... 131

Reagents ...... 131

Kinetics procedure ...... 132

Acetolysis product analysis of exo-2-hydroxy- b ic y c lo [ 2 . 2 .l] h e p ta n - 7~one £-toluenesulfonate (38) ...... 134

Acetolysis product analysis of endo-2-hydroxy- b ic y c lo [ 2 . 2 .l] h e p ta n - 7-one £-toluenesulfonate (. 59 ) ...... 136

Acetolysis product analysis of 7» 7-diroethoxy- b ic y c lo [ 2 . 2 ,l]heptan-exo- 2-ol £-toluenesulfonate (64) ...... 136

Dioxane-water solvolysis product analysis of 7 * 7-dimethoxybicyclo[ 2 .2 »l]heptan-exo- 2-ol £-toluenesulfonate (64) ...... 137

Acetolysis product analysis of ? ,7-dimethoxy- b ic y c lo [ 2 . 2 .1 ]heptan-endo- 2- o l £_ toluenesulfonate ( 65 ) 137

vii CONTENTS (Continued.) Page

Acetolysis product analysis of anti-7-h.ydroxybicyclo[2.2.1]heptan-2-one £-toluenesulfonate (104) ...... 138

Acetolysis product analysis of syn-7-hydroxybicyclo[2.2.1]heptan-2-one £-toluenesulfonate (103) ...... 13.8

Partial acetolysis of anti-7-h.ydroxybicyclo- [ 2 .2 »l]heptan- 2-one £-toluenesulfonate (104) in acetic acid- 0-d ...... 140

APPENDIXES...... 142

A. Carbonyl stretching frequencies ...... 142

B. NMR d a ta ...... 145

C. An example of a solvolytic rate constant determination involving an infinite t i t e r ...... 149

D. An example of a solvolytic rate constant determination not involving an infinite titer ...... 151

E* The method of least squares ...... 153

F. Determination of the activation parameters . . . 156

G. Rate data ...... 160

H. Rate curves ...... 163

BIBLIOGRAPHY...... 168

viii TABLES

Table Page

1. NMR Chemical Shifts for Methoxyl Protons in 7»7-Dimethoxynorbornane D erivatives ...... 31

2. NMR Chemical Shifts for a-Protons in Norbornyl Alcohols and Acetates ...... 32

3« Methods of Obtaining exo- and endo-Hydroxy- 7»7-dime thoxynorbomane...... 35

NMR Chemical Shifts for a-Ppotons in 7-Keto- norbornyl Acetates and Tosylates ...... 36

5 . Acetolysis Rates of 2-Norbomyl and 7-Keto-2- norbornyl Tosylates ...... 38

6 . Ethanolysis Rates of 7-Keto-2-norbornyl Tosylates. . 40

7* Acetolysis Rates of 7>7-Dimethoxy-2-norbornyl and 7-Chloro-2-norbornyl Tosylates ...... 50

8 . Acetolysis Rates of 7 >7-Dimethoxy-endo-2- norbomyl Tosylate in Acetic Acid and in Acetic A cid-O-d ...... 6 l

9. Acetolysis Rates of Various 7-Norbornyl Tosylates. . 76

10. Acetolysis Rates of Various 7-Norbomyl Tosylates in Acetic A cid-0-d ...... 78

11. Infrared Stretching Frequencies in the Carbonyl Region of Various Norboranone Derivatives .... 1^2

12. NMR Chemical Shifts for Various Norbornane Derivatives ...... 1^5

13. Determination of a Solvolysis Rate Constant by Using an Infinite T ite r ...... 150

1^. Determination of a Solvolytic Rate Constant Without the Use of an Infinite Titer ...... 152

ix TABLES (C ontinued)

Table Page

15• The Method of Least Squares ...... 2.514-

16. The Determination of Activation Parameters .... 2.57

17. The Method of Least Squares ...... 2.58

18. Solvolytic Rate Constants Experimentally Determined for Various Norbornyl Tosylates . . . -^61

x ILLUSTRATIONS

Chart Page

I. Synthesis of 7»7-Dimethoxy and 7-Keto- norbomane D erivatives ...... 28

II. A Proposed Mechanism for the Solvolysis of 7,7-Dimethoxybicyclo[2.2.l]heptan-endo- 2-ol ja-Toluenesulfonate ...... 62

Graph

I. Acetolysis of anti-7-Hydroxybicyclo[2.2.1]_ heptan-2-one ja-Toluenesulfonate in Acetic Acid and in Acetic Acid-O-d ...... 79

II. A Plot'of log [H] vs. t ...... 88

III. Acetolysis Rate Curves of 2-Hydroxybicyclo- [2.2.1]heptan-7-one ^-Toluenesulfonates a t 100° ...... ■...... 16 ^

IV. Acetolysis Rate Curves of 7»7-Dimethoxybicyclo- [2.2.1]heptan-2-ol p-Toluenesulfonates at i o o °...... 165

V. Acetolysis Rate Curves of 7-Hydroxybicyclo- [2.2.1]heptan-2-one _g-Toluenesulfonates ..... 166

VI. Acetolysis Rate Curve of 7-Hydroxybicyclo- [2.2.1]heptane ja-Toluenesulfonate in Acetic Acid-0-d at 200° ...... 167

x i. INTRODUCTION

Wagner'*' in 1899 made th e re v o lu tio n a ry su g g e stio n t h a t th e difficulty experienced in molecular structure determinations in 2 the monoterpene series was due to unexpected*, but facile*, rearrange- 8 la ments occuring in certain interconversions. Meerweirr subsequently suggested that one such rearrangement, the conversion of camphene hydrochloride ( 1 ) to isobornyl chloride ( 2), occurred via an ionic process involving discrete carbonium ions. This ionic 1,2-migration process in a bicyclic compound is called a "Wagner-Meerwein re arrangement."^

The puzzling feature of this rearrangement was that no endo q. product (viz., bornyl chloride Q )) was formed when the less

^ G. Wagner,. J. Russ. Phys. Chem. Soc., JJl, 680 (1899). 2 For a review, see J. L. Simonsen and L* N. Owen, The Ter- penes, Vol. II, Cambridge University Press, 1949. 3 H. Meerwein and K. van Emster, Ber., 1815 (1920). H. Meerwein and K. van Emster, Ber., jjj), 2500 (1922). ^ Y. Poker,. "Wagner-Meerwein and Pinacolic Rearrangements in Acyclic and Cyclic Systems," Molecular Rearrangements, P. deMayo, ed., Interscience Publishers, New York, 1963* PP» 1-25*

1 2

2 hindered mode of attack would be expected to be from the endo side. 7 To explain this result* the suggestion was made that the formal carbonium ion intermediates postulated in the Meerwein conversions did not exist. Instead*, a bridged ion ( 6 ) was formed which was a resonance hybrid of the two contributing structures 2 and 8 in which the carbon atom of 6 was considered to be partially bonded to carbon atoms

Evidence supporting the endo side being the least hindered approach for attack is the fact that camphor (4) in reduction with lithium aluminum hydride experiences attack predominantly from the endo side to give isoborneol (£) [L. W. Trevoy and W. G. Brown, J . Am. Chem. Soc.» 2i» (1 9 ^9 ), D. S . Noyce and D. B. Denney, ibid., 22, 57^3 (1950)] .

H H

7 T. P. Nevell, E. dedalas* and C. L. Wilson, J. Chem. Soc. (London), 1939, 1188. 3

6 8

and C^. An approaching nucleophile would attack with consequent

Walden inversion* giving the exo product. This intermediate ( 6 ) has 8 9 been termed "synartetic»" “none las sic a 1 *” and a variety of other

-3» C l H

10 names, but "nonclassical" is the term most commonly used today to 11 denote this intermediate.

O F. Brown, E. D. Hughes, C. K. Ingold, and J. F. Smith, Nature, 168 , 65 (1 9 5 1 ). ^ J . D. Roberts and G. C. Lee, J. Am. Chem. Soc.»22.* 5009 (1951)< ^ J. A. Berson, "Carbonium Ion Rearrangements in Bridged Bicyclic Systems," Molecular Rearrangements, P. deMayo, ed.,. Interscience Publishers, New York, 1 9 6 3 , p . 119* P. D. Bartlett,, Nonclassical Ions, W. A. Benjamin, Inc., New York, 1965, p. v. Exploration into the nonclassical carbonium ion question was re- 12 sumed in 1952 when Winstein investigated the exo- and endo-norbornyl

£-bromobenzenesulfonates (£ and 10, respectively). The exclusive exo product of solvolysis of £ and 10 was consistent with the postulated nonclassical carbonium ion (11). Furthermore» racemization of product was found, presumably due to the optically inactive intermediate ( 11) .

Finally,, £ solvolyzed much faster than 10 (by a factor of 350). This fact was rationalized by claiming that the exo-aryl sulfonate (£)>

f,OBs/ _ HO Ac O B s <5 “ O Ac H 12

O B s a- OBs 10 11 ia

because of the favorable geometry of the leaving group> could take ad-

vantage of participation of the C..-C,. O-electrons during the transi- 1 o tion state ( 12); the endo analog ( 10) could only solvolyze via a

classical transition state ( 1£) to the classical ion 1A which then

S. Winstein and D. Trifin,. J. Am. Chem. Soc., £A, 11A7»- 115A (1 9 5 2 ). could collapse to the nonclassical structure 11. This type of partici pation during the transition state ( 12) was termed "anchimeric 13 assistance J'

Opponents of the nonclassical carbonium ion theory include 14 Brown* who suggested that the norbornyl cation was always classical.

This classical cation (14) could rapidly equilibrate to give racemized

H i 14

products. The fact that exo-norbornyl derivatives solvolyzed faster than the endo analogs was explained by assuming not that the exo derivatives solvolyzed unusually fast,, but that the endo derivatives solvolyzed slowly. Brown has suggested that this rate retardation in the endo derivatives could be caused by steric interaction of the leaving group with the endo-C^ hydrogen atom. The exclusive formation

of exo product could be attributed to two possible causes. Firstly,

solvent would be unlikely to attack the carbonium ion (14) from the

endo direction because of the shielding effect of the endo-C^ hydrogen

^ S. Winstein* C. R. Lindegren, H. Marshall, and L. L. Ingraham, J . Am. Chem. S o c ., £> , 1^8 (1953)* H. C. Brown, Chem. Soc. (London) Spec. Publ., 16 , 140 (1962). atom. Secondly* as l/f equilibrated between its two equivalent struc tures*. the solvent molecules would be excluded from the endo side and would thus be unlikely to react with from this direction.

It is thus seen that the controversy over the 2-norbornyl cation stems from the possibility of interpreting the available data ambigu- 15 >16 o u s, ly .

The nonclassical carbonium ion question has not been confined to 17 2-norbomyl systems. Winstein found that anti-7-no rbornenyl p- 11 toluenesulfonate (15) solvolyzed 10 times faster than 7-norbornyl p-toluenesulfonate (l 6 ). This tremendous rate difference was explained

O T s

11 iZ 18

15 ^ For a leading reference*, see S. Winstein, J. Am. Chem. Soc., 82* 381 (1965). For a leading reference, see H. C. Brown and M. H. Rei* J. Am. Chem. S o c ., 8 6 * 5008 (19&0. 17 S . W instein,, M. S h atav sk y , C. Norton* and R. B. Woodward* J. Am. Chem. Soc., 22* *1-183 (1955)* by assuming that the tosylate ljj experienced nonclassical anchimeric assistance from the Il-electrons of the double bond during the transi tion state, thereby lowering the transition state energy. The 18 resulting "bishomocyclopropenium cation" (1£) could be attacked by nucleophile only from the anti direction, resulting in Walden inver sion and retained configuration at the carbon atom.

This hypothesis has also been discredited by Brown19 who claimed that the carbonium ion resulting from solvolysis of 1£ had the clas sical structure 19 and equilibrated between the two equivalent forms.

OTs

12 12

The resulting controversy^*^0 again stems from the fact that the same results can be interpreted in two different ways.

At the present time, therefore, the question of the norbornyl nonclassical carbonium ion is unsettled. Apparently a fresh approach to the problem is critically needed to help resolve the problem.

18 W. G. Woods, R. A. Carboni, and J. D. Roberts, J. Am. Chem. Soc., £8, 5653 (1956). 19 H. C. Brown and H. M. B e ll, J . Am. Chem. S o c ., 8£, 232^ (1 9 6 3 ). 20 S. Winstein, A. H. Lewis, and K. C. Ponde, J. Am. Chem. Soc. 8 £, 232^ (1963). PART I . THE 2-NORBORNYL CATION

1 . The problem

The case for the 2-norbornyl nonclassical carbonium ion rests upon three major foundations, (a) The rates of solvolysis of exo- norbornyl derivatives are considered to be unusually fast, presumably Q due to anchimeric assistance during the transition state. (b) High exo-endo rate ratios for the solvolysis of norbornyl derivatives are considered to be due to anchimeric assistance occurring in the transi- 12 tion state of the exo derivatives,, but lacking in the endo analogs.

(c) Exclusive exo substitution in the solvolysis product is argued to be explicable only on the basis of a nonclassical carbonium ion inter mediate, as the classical cation would be expected to give a substantial 12 amount of endo product.

(a) Fast rates. It is difficult to anticipate what a "normal*" i.e ., classical, rate of solvolysis of an exo-norbornyl derivative 21 would be. Brown has suggested that the solvolytic rates of exo derivatives are actually quite normal, on the basis of comparison with the solvolytic rates of corresponding cyclopentyl derivatives. The difficulty in granting the validity of this approach is that there is no a priori means of determining how the unique strain in the norbornyl

21 H. C. Brown, F. J. Chloupek, and M. H. Rei, J. Am. Chem. S o c ., 86 , 12/1-7 (1964-).

8 system will affect the transition state. It is thus questionable whether the cyclopentyl system is an ideal model for solvolytic rate comparisons. 22 Schleyer has attempted to establish a priori the "normal*' solvolytic rates of various compounds on a semi-empirica1 basis. He has concluded that the solvolytic rates of endo-norbomyl derivatives are normal, but that the rates of the exo analogs are too fast and must benefit from some nonclassical assistance during the transition state.

The compounds included in Schleyer*s treatment cover a wide variety of types; unfortunately, a few behave inconsistently. This approach, although elegant, is therefore deemed not necessarily correct.

(b) High exo-endo rate ratios. The validity of the high exp end o rate ratio argument again seems to revolve about the question 23 whether the exo derivative is fast or the endo analog is slow. Brown argues that tertiary norbornyl derivatives would solvolyze to a clas sical cation,, since the tertiary carbonium ions would not require the special participation involved in the secondary norbornyl cation case.

Since the exo-endo rate ratio is still high for the tertiary deriva tives, he argues that this high ratio must be due to an intrinsic

feature of the norbornyl system itself. Brown suggests that the high

^ P. von R. S c h le y e r, J . Am. Chem. S o c.,. 86 , 18^4, I 856 (1964). ^ H. C. Brown and M. H. R e i, J . Am. Chem. S o c ., 86 , 5004 (1964). Oh, H. C. Brown and M. H. R e i, J . Am. Chem. S o c ., 86 , 5008 (1964). h. C. Brown, F. J. Chloupek, and M. H. Rei,. J. Am. Chem. Soc., 8 6 , 12 4 8 (1 9 6 4 ). 10 25 exo-endo rate ratios in tertiary systems may be due to two causes.

Firstly, the rate of the exo derivatives (20) would be accelerated by the bulky group in the endo-C^ position sterically interacting with the endo-C^ hydrogen atom. During the transition state (21), the

6 CH

20 21 bulky group would swing away from the endo-C^ hydrogen atom, thus

relieving interaction. The endo analog would, of course, lack this

feature. Secondly, the leaving group in the endo derivative would

sterically interact with the endo-C^ hydrogen atom, as already postu

lated for secondary systems (see p. 5 )*

Granting Brown's classical cation, it would seem that his first

suggestion concerning interaction of the methyl group with the endo-C^

hydrogen atom in 20 would be irrelevant, as the rate ratio for tertiary

norbornyl systems is not significantly higher than that for secondary

systems. The problem lies in establishing the magnitude of this effect. 26 It is known that this interaction is very serious in the ground state,

but extending this concept to describe transition states may not be valid, ----- Under equilibrium conditions, a tertiary system w ill undergo a Wagner-Meerwein rearrangement such that the methyl group is in the

1-position (see Ref. 10, pp. 131, 137)• C H 3 3 H The acceptance of a significant methyl-endo-C^ hydrogen inter action has serious consequences upon the classical cation theory.

If one assumes that this interaction is significant and that the tertiary norbornyl cation is classical, then he faces the interesting problem of explaining why the exo-endo rate ratio is not higher for the tertiary case without invoking a nonclassical secondary cation. Other wise,. there would appear to be no other effect which could be respon sible in bringing up the secondary exo-endo rate ratio fortuituously close to that for the tertiary case. Furthermore, a significant methyl-endo-C^ hydrogen interaction can be reconciled with a nonclassi cal tertiary cation. If tertiary systems were nonclassical, but the benefit from anchimeric assistance were not as great as in the second ary case,, then the exo-endo rate ratio for tertiary systems predicted on this basis alone would be greater than unity but smaller than in

the secondary case. The methyl-endo-G^ hydrogen interaction might

bring up the ratio fortuitously close to that in the secondary system.

There is no easy way to test Brown's second suggestion for a slow

solvolytic rate for the' endo-norbornyl derivative, viz., that the endo- 22 hydrogen atom would impede the endo-C, exophile. Schleyer's work

(see p* 9) has suggested that the endo-C^ hydrogen atom, instead of

hindering the leaving of the exophile, will actually help it because

the exophile is already outside the sphere of maximum interaction. 27 Schleyer has further shown that endo- 6 ,6 -dimethylnorbornyl g-toluene-

sulfonate ( 22) solvolyzes 19 times more slowly than endo-norbomyl

g-toluenesulfonate (2^). Extrapolation of the effect of the endo-C^

^ P. von Schleyer, M. Donaldson, and W. E. Watts, J. Am. Chem. S o c ., 8?, 375 (1965). 12

C b k O Ts OTs 22 22 methyl group to that of an endo-C^ hydrogen atom shows that the effect of this hydrogen atom would be quite small. This work therefore sug gests that the endo-C^ hydrogen atom cannot be held responsible for the exo-endo rate ratio of 350 for exo- and endo-norbomyl p-bromobenzenesulfonate.

Choosing perhaps the easiest rebuttal of Brown's proposals* Win- 28 stein argues that the high exo-endo rate ratios for tertiary systems merely show that they* too* solvolyze with anchimeric assistance.

(c) Exclusive exo substitution. One might think that this criter ion would be the easiest to test: A single instance of a norbornyl derivative solvolyzing to give both exo and endo products would strongly support the argument that general exo substitution is due to nonclassi cal intermediates. One could argue that this unique example solvolyzes to a classical carbonium ion while all other heretofore investigated norbornyl systems solvolyze to nonclassical intermediates. The dif- 29 ficulty is that no such system has been discovered. Therefore* it is not known whether all norbornyl systems whose solvolysis products have been investigated are nonclassical, or whether the criterion of exclusive exo substitution does not hold.

2® S. Winstein* J. Am. Chem. Soc., 8£» 381 (1985)* ^ There is one exception (see p. 17)* but for reasons to be dis cussed, the validity of the application of this example to solvolysis reactions is not certain. 13 14 Brown has argued that since exclusive exo product formation is always the case, then this phenomenon must be due to some novel fea ture of the classical system* In explaining exclusive exo substitution 28 in secondary systems, he invokes the "windshield wiper" effect of equilibrating ions which would effectively desolvate the endo side of cation (14) and render this side unlikely to react with solvent.

28 Winstein argues that such desolvation of an ion is energetically unfavorable and that this proposal is therefore thermodynamically 30,31 unsound. Furthermore, he argues, in view of certain n.m.r. studies the equilibration would have to occur at such a rapid rate that the free energy of activation for the process would be zero: an impossible state of affairs if Brown's proposal were correct that no resonance occurred in the carbonium ion intermediate. op \ Further studies by Brown-' on tertiary systems (24,23) have shown that the generated carbonium ions, presumed to be classical, gave exclusive exo attack. Since no secondary product was formed, probably no rapid equilibrating process was occurring. The conclusion

3^ P. von R. Schleyer* W. E. Watts, R. C. Fort, Jr., M. B. Comi- sarow , and G« A. O lah, J . Am. Chem. S o c ., 8 6 , $680 (1964). 31 M. Sauders, P. von R. Schleyer, and G. A. Olah, J. Am. Chem. S o c ., 8 6 , 5681 (1 9 6 4 ). H. C. Brown and H. M. B e ll, J . Am. Chem. S oc.,. 86 , 5006 (1964). 14

,OTs

e x c lu siv e exo a tta c k

was that even in classical systems that could have no "windshield

wiper" effect,, by some unique feature of the norbornyl system* only exo 14 product was formed. Brown has suggested that the endo-C^- hydrogen

atom might prohibit attack by nucleophile from the endo side. It is

conceivable that a "windshield wiper" effect may produce the 4.4 33 kilocalories energy difference in exo and endo solvent capture by a

classical carbonium ion, but it is exceedingly difficult to rationalize 28 the effect of the endo-C^ hydrogen atom to be this significant. It

is particularly difficult to justify this effect when hydride reduc

tion studies of norcamphor derivatives have shown that when the syn-Cr,

exo a tta c k

/O endo attack

33 H. L. Goering and C. B. Schwene, J. Am. Chem. Soc., 8£» 3516 (1 9 6 5 ). 15 position is substituted with a methyl group, attack is always 3^ preferentially endo.

Product analyses from the solvolysis of tertiary systems are exceedingly important to establish the fact that where the "windshield" wiper" effect cannot take place» attack of solvent is always exclusively exo. Unfortunately,, pertinent product analyses are critically lacking. 35 Although the tertiary product is initially formed, this rearranges 36 to the more thermodynamically stable secondary product unless one is very careful to ensure that kinetic conditions prevail, usually by 37-lfO keeping reaction conditions basic. The few other examples of tertiary systems solvolyzing to give the kinetically controlled tertiary product always result in exo substitution.

Known classical norbornyl systems. Apparently at least one classical norbornyl cation is known. It has been established by n.m.r. h -1 data that the l, 2-di(£-anisyl)norbornyl cation ( 26 ) is classical,

O M e O M e 26

^ See Ref. 10, pp. 1 3 ^ -1 3 5 . 3 5 j, a. Berson, Tetrahedron Letters, No. 16, 17 (i 960 ). 3^ For a summary of such examples, see Ref* 10, p. 1 3 1 . 3 7 N. J. Toivonen,. E. Siltanen, and K. Ojala, Ann. Acad. Sci., Fennicae, Ser. A, II, No. 6 k (1955)• 3® P. Beltrame, C. A. Bunton,. and D. Whittaker, Chem. & Ind. (London), i 960 557 • and the two opposing camps have agreed that this is a bona fide ex- 28 ^2 ample of a classical norbornyl cation. * Unfortunately*, solvolysis products of this intermediate have not been determined. The classical n a tu re o f 26 would lead one to conclude that if the Cg substituent on th e 2-norbomyl cation stabilizes the charge sufficiently* the cation would be classical, whether other norbornyl cations were classical or n o t.

Two other systems with D-anisyl substituents have been studied. ^■0 Bartlett has shown that l-(jD-anisyl)apoisobomyl chloride (27) solvolyzes kinetically to 2-(jD-anisyl)isocamphenilol (28), which then epimerizes to the more thermodynamically stable 2-(;g-anisyl)camphenilol

(22)» The intermediate cation, shown to be exceedingly stable, may well be classical. To adherents of Brown’s position it would seem that

.OMe OH OH

OMe Me 2Z 28

39 G. Komppa and S. Beckman, Ann.,. 509, 51 (193^)* P. D. Bartlett,. E. R. Webster, C. E. Dills, and H. G. Richey,. Jr.,. A n n .623, 217 (1959)* ^ P. von R. Schleyer* D. C. Kleinfelter, and H. G. Richey, J r . , J . Am. Chem. S o c ., 8£, h-79 (1963). ^ H. C. Brown and H. M. B ell,. J . Am. Chem. Soc.» 8 6 , 5003 (1 9 6 ^). 17 the formation of 28 would undisputably show that even classical ions suffer general exo attack. Unfortunately, this conclusion does not necessarily follow for two reasons. Firstly, the position is unsubstituted and attack should be preferentially exo regardless of one’s position on the nonclassical question. Secondly, a critical analysis was not made of the kinetically controlled product to deter mine if there was a minor yield of the endo product (2£). Undoubtedly this analysis would be difficult because of the ready conversion of

28 to 2£ .

The remaining example of a 2-£-anisylnorbornyl cation is that resulting from l-(£-anisyl)camphene hydrochloride (_ 20)» studied by 4 2 ,4 3 Brown. He has claimed that the carbonium ion (^1) arising from

30 is classical, if not from solvolytic rate arguments alone,, at least from analogy with the l,2-di(jo-anisyl)norbornyl cation (26). Brown also argues that the high exo-endo product ratio found in this case would invalidate the exclusive exo substitution criterion for nonclassical carbonium ions. However,, there was a significant amount of product derived from endo attack (J2)• What apparently did not strike

Brown as significant was that for the first time a norbornyl cation was reported to give a significant amount of kinetically controlled endo product. To an adherent of the classical cation theory, this data would be unexpected and not easily explicable. To an adherent of the nonclassical cation theory, these data are more easily ex plained: All heretofore studied norbornyl systems whose products were derived from exclusive exo attack (at least where a s.yn-Co

^ H. C. Brown and H. M. Bell,. J. Am. Chem. Soc., 86, 50C? (1964). 18

OMe OMe 2£ NaBK

2 1 (10S&) OMe 22 W ) methyl group was substituted) gave nonclassical carbonium ions, and this jo-anisyl case (,20) gave a classical cation. However, it is possible that the significant amount of endo attack may be due to some unanticipated effect not operating in an ordinary solvolysis, since borohydride reduction of a carbonium ion is not in the strict sense a solvolysis reaction.

j- j- - - — Solvolysis of 2-methylnorbomyl derivatives in aqueous diglyme gives exclusive exo attack at the tertiary position whereas boro hydride reduction of the same derivatives in the same medium gives trace quantities of product derived from endo attack at the same p o s itio n [H. C. Brown and H. M. B e ll, J . Am. Chem. S oc. , 86, 5006 (1964)]. For much more drastic differences between solvolysis and borohydride reduction in the same medium,, see "Part II. The 7- Norbomenyl Cation" of this dissertation. 19

The transition state* Part of the confusion surrounding studies o f th e 2-norbornyl cation is due to the lack of distinction made be tween the nonclassical carbonium ion intermediate and the anchimeric assistance during the transition state leading up to the carbonium 45 ion intermediate. Brown ^ has used the argument that since exo-1- 7 p h e n y l- 2-norbornyl chloride {255 solvolyzes 10 times more slowly than exo- 2-p h e n y l- 2-norbornyl chloride (^ 6 )» the same intermediate, the nonclassical ion 22., cannot be involved in both solvolyses. This argument does not necessarily hold, as the transition state is the

process reflected by solvolysis rates. According to the nonclassical

theory,, the solvolysis rate of 2 5 and 2& not be the same. In 2§.

the participation of the phenyl Il-electrons can occur before the

developing C^-p orbital on the endo side can overlap with the

O-electrons; in 2 5 Participation of the 0-electrons is necessary

before the stabilizing effect of the phenyl group can be experienced.

Apparently little , if any phenyl resonance stabilization is experienced

in the transition state of 25* ^o r 5 5 solvolyzes only 3*9 times faster

th an exo- 2-norbornyl chloride (34).

^ H. C. Brown, F. J. Chloupek,. and M. H. Rei, J. Am. Chem. 3oc.» 86, 1246 (1964). It might appear that the relative solvolytic rates of anc* 35 demonstrate phenyl stabilization in the transition state of If

the stabilizing effect of the phenyl group were real, then one would

expect a sim ilar rate enhancement of l-(£~anisyl)camphene hydrochloride

(38) over camphene hydrochloride (1). In fact, solvolyzes more

slowly than 1 by a factor of 1 . 7. The small and numerically reversed

OMe 21 theory inasmuch as the 1-aryl substituent need not enhance the rate.

It is much more difficult to explain these ratios in terms of the nonclassical cation theory. It might be argued that in the case of

36 and a secondary reaction site is involved which would desire participation more than in the case of 1 and 2§ where a tertiary reac tion site occurs; steric factors could overcompensate in the tertiary case to give the reversed ratio.

An excellent demonstration that the transition states of different derivatives leading up to the same intermediate need not be the same is that of exo- and endo-norbornyl jd-toluenes ulfonate (22 anc* 2 3 > i|»^ Ji n respectively)* and 2-(4-cyclopentenyl)ethyl £-toluenesulfonate (40). '

All three lead to the same product* exo-norbornyl acetate (41)* upon acetolysis. The respective relative rates of 22* 22* and 40 at 250 a re 280* 1 , and 1 . 7 .27,48

777 1 R. G. Lawton* J. Am. Chem. 3oc.> 82* 2399 (1961). 47 P. D. Bartlett and 3. Bank, J. Am. Chem. Soc.* 83, 2591 (1961). 22

The argument as to whether rates* and thus transition states* must reflect the nonclassical carbonium ion intermediate exactly* q. 9 regardless of the direction from which the intermediate is approached* depends critically upon how Hammond*s postulate is em ployedThe con- 28 cept in question is whether carbon bridging lags behind ionization. 51 Winstein infers that the transition state need not be the same for

norbornyl derivatives solvolyzing to the same nonclassical intermedi

ates, for even though apoisobornyl jo-bromobenzenesulfonate (42) and

exo-camphenilyl £-broraobenzenesulfonate (42) solvolyze at different

rates (by a factor of 23)» the products are identical, which could

arise from a common nonclassical intermediate (44).

-OBs BsO .

48 P. D. B artlett, S. Bank* R. J. Crawford, and G. H. Schmid* J . Am. Chem. 3 0c . , 8£ , 1288 ( 1965 ). 49 Of course, the difference in the ground state energies of the various derivatives all going to a common product must be taken into consideration* also. Even after compensating for this, at least in the c a se o f 22. and 22 the transition states differ by an internal energy of 4.4 Kcal./mole (see Ref. 33)• 23 2 . Approach to the problem

Since many attempts to establish the existence of the nonclassi cal norbornyl cation involved stabilizing the charge* e.g.* by the incorporation of methyl or phenyl groups, and since all investigations resulted in ambiguous results, it was felt that destabilizing the charge which would result if a Wagner-Meerwein rearrangement occurred might lead to more definitive results. In particular, a carbonyl group could be incorporated at the 7-position to give the cation 45.

Because of the inductive effect of the carbonyl function, neither the rearrangement to give 46 nor the nonclassical carbonium ion formation to give 4£ would be expected. Thus, the solvolysis of 44 might lead

'OTs t O

t 42

5° G. 3 . Hammond, J . Am. Chem. S oc.,. 2Z» 33^ (1955): "If two states, as for example, a transition state and an unstable intermediate, occur consecutively during a reaction process and have nearly the same energy content, their interconversion w ill involve only a small reorgani zation of the molecular structures.” ^ S. Winstein, A. Colter, E. C. Friedrich, and N. J. Holness, J . Am. Chem. S o c ., 82, 37® (1 9 6 5 ). 24 to a true classical carbonium ion (4£)» whether nonclassical cations ex isted in previous examples or not. The solvolytic properties of 44 and the products derived therefrom could be compared with the predictions from both the nonclassical and the classical carbonium ion theories.

If the clasical carbonium ion theory were correct, then by the preceding discussion the following predictions would be made. The factor presumably causing the unique solvolytic properties and products of norbornyl derivatives,, viz., the endo-C^ hydrogen atom,, s till exists in 44. Therefore, (a) the exo-endo rate ratio would still be high, and (b) the product would be exclusively exo.

If the nonclassical carbonium ion theory were correct, then the following predictions would be made. Since no anchimeric assistance

Would be expected in the exo case, then the rate of exo-44 would fall to a value approaching that of endo-44. Also, since no nonclassical carbonium ion (47) would be expected, then attack ty solvent could be from the endo direction as well as from the exo. Therefore, (a) the exo-endo rate ratio would be much lower than in the previous norbornyl cases, and (b) the products would include a significant amount of endo derivative.

Although the predictions above are relatively straightforward, the question of the absolute rate of 44 as predicted by both the classi cal and nonclassical cation theories is much more complicated. An uncertainty arises as to the overall effect of two opposing factors.

A rate-retarding factor is the inductive effect of the carbonyl func- 52 tion. A previous example, the 2-chloro-l»4-endoxocyclohexanes (48),

— J. C. Martin and P. D. B artlett, J. Am. Chem. Soc., 79 > 2533 (1 9 5 7 ). 25 illustrates the effect of an electron-withdrawing group incorporated into the norbornyl system. Solvolytic rates of 48 are slower than

those of the norbornyl chlorides (42) by a factor of ca. 1(P. On this basis alone, a retardation of the rates of the 7-ketonorbomyl tosylates

(44) over those of the norbornyl tosylates (^2 and 2^) would be ex p e c t e d .^

The o th e r opposing f a c to r is th e in c re a se d C-^-Cg-C^ angle# caused *54 by the flaring of the angle. An increase in the latter angle would facilitate ionization at the carbon atom,'*'’ and thus

53 It would seem that 48 would be an excellent example to test the nonclassical cation hypothesis; the same criteria listed above for the 7-ketonorbornyl cation (45) could be used here. Unfortunately* thic cannot be done* for 48 rearranges to 3-f‘oi*iiylcyclopentanol (50). + O OH

■> CHO 48

Thus,one cannot test the exclusive exo product formation criterion. Also, anchimeric assistance is possible in exo-48. Therefore, the question is not resolved whether the high exo-endo rate ratio ob served for 48 is due to assistance in the exo case or to the steric effect of the endo-C^ hydrogen atom in the endo case. 54 D. J. Cram and G. S. Hammond, Organic Chemistry, second ed., McGraw-Hill Book Company, Inc., New York, 1964, p. 148. 55 C. S. F o o te, J . Am. Chem. S o c ., 8 6 , I 853 (1964). 7-keto-2-norbornyl aryl sulfonates would be predicted on this basis to solvolyze faster than 2-norbornyl a:cyl sulfonates. Unfortunately* a quantitative prediction of the magnitude of this effect is virtually impossible* because the amount of angle flaring is not known.

It is known that a small difference in bond angle can make a tremendous difference in solvolysis rates .^ ’55 por instance * a bond angle dif- 36 7 ference of 6° can produce a solvolytic rate ratio of 10 for endo- 27 2-norbornyl £-toluenesulfonate (2^) and 7-norbornyl £-toluenesulfo* n a t e ^ ( 16 ).

23 OTs 16

From this discussion it becomes apparent that an a priori quanti tative prediction of the overall effect of the 7-carbonyl function on the absolute solvolysis rate of the 2-norbornyl system would be extremely difficult. The conclusion would be that no particular sig nificance need be attached to the 2-norbomyl/7-keto-2-norbomyl solvolytic rate ratios, but that the exo-endo solvolytic rate ratio

A. C. Macdonald and J. Trotter,, Acta Cryst .% 18, 243 (l?65). 27 for the 7-keto-2-norbomyl p-toluenesulfonates and the products

resulting therefrom would be the important data.

3 . R esu lts

Preparation of the desired 7-ketonorbornyl system was achieved

by utilizing the scheme of Gassman^ to prepare 7,7-dimethoxybicyclo-

[2.2.1]heptene (^2). Hexachlorocyclopentadiene (50) was converted

by methanolic potassium hydroxide to 5»5-8ime‘thoxy-l,2,3»4-tetrachloro-

cyclopentadiene^ (51 )% which underwent a Diels-Alder reaction with

ethylene to give 7»7-dimethoxy-l,2,3,4-tetrachlorobicyclo[2.2.1]hept-

2 -en e ^ (^2). Dechlorination^ of j>2 with sodium metal and t-butyl

M eO OMe

alcohol in tetrahydrofuran gave 7»7-dimethoxybicyclo[2.2.1]heptene (53)»

Hydroboration^ of jQ anc* subsequent oxidation (see Chart X) gave

a 7^e$ yield of a 78:22 mixture of exo- and endo-7,7-dimethoxybicyclo-

[2.2.1]heptan-2-ol (^4 ana respectively). The epimeric alcohols

■5? p. G. Gassman and P. G. Pape» J. Org. Chem.» 2g, l6o (1964). ^ J. S. Newcomer and E. -Tw McBee, J. Am. L'hem. Soc., 71 > 946 (1949)•■ 5^ P. E. Hoch, J. Org. Chem., 26, 2066 (1961). ^ P. Bruck, D. Thompson, and 5. Winstein,, Chem. Ind. (London), 405 ( I 960 ). ^ H. C. Brown and B. C. Subba Rao, J. Org. Chem., 2^> 1136 (1957)* 28

CHART I M € O^OMe

Me OjDMe

,OH

2 L OH

OTs pAc

60 OTs )Ac SSL 61 54 and were separated by preparative v.p.c. That and were epimeric was demonstrated by their conversion to the same ketone,

7,7-dimethoxybicyclo[2.2.1]heptan-2-one (66) , by Sarett oxidation.

Each of the two alcohols ^4 and was hydrolyzed with 5$ aqueous sulfuric acid to give the respective keto-alcohols, exo-2-hydroxybicyclo[2.2.1]heptan-7-one (56) and endo-2-hydroxybicyclo[2.2.1]- heptan-7-one (jjjZ)» The keto-alcohols $6 and were converted by the 62 procedure of Tipson to give the corresponding tosylates,. exo- and

endo-2-hydroxybicyclo[2.2.1]heptan-7-one £-toluenesulfonate (58

and respectively)» and were converted with acetyl chloride in

pyridine to the corresponding keto-acetates, exo- and endo-2-acetoxy- 64 bicyclo[2.2*l]heptan-7-one (60 and 61, respectively).

The ketal-alcohols jj4 and were converted by the above procedures

to exo- and endo-2-acetoxy-7,7-dimethoxybicyclo[2.2.l]heptane (62 and

63, respectively), and exo- and endo-2-hydroxy-?»7-dimethoxybicyclo-

[2.2.l]heptane £-toluenesulfonate (64 and 6^, respectively).

That the correct stereochemistry was assigned for the ketal-

alcohols and was confirmed in several ways. The lower boiling

point of the exo-alcohol (.54) was attributed to internal hydrogen

bonding of the hydroxyl proton with the syn-methoxyl group. Proof

62 R. 3. Tipson, J. Org. Chem., ^35 (1944). This tosylate has been prepared and its acetolysis rate deter mined by K. Mislow and W. E. Meyer. For details, see W. E. Meyer, thesis, New York University, 1964 . ^ This compound has previously been prepared by C. H. DePuy and P . R. S to ry , J . Am. Chem. S o c ., 82, 627 ( i 960 ). 30

MeO^OMe M eO -vO M e

2 1 OH

MeO OMe MeO OMe MeQ^OMe Me0^OMe

63 OAc § 1 OTs

of the intramolecular hydrogen bonding in jj4 was obtained by high dilution near-infrared studies .^5 Absorption at 1.442 p. (0.025 M in

•Me MeCX OMe

2k 2 1 O -H

carbon tetrachloride) for 2b. indicated hydrogen bonding, whereas the

epimeric alcohol (3 3 ) under id e n tic a l conditions showed only a free

hydroxyl stretching frequency overtone at 1.412 p,.

Nuclear magnetic resonance spectroscopy gave additional

65 R. Piccolini and S. Winstein, Tetrahedron Letters, No. 13, b (1939) 31

TABLE 1

NMR Chemical Shifts for Methoxyl Protons in

7 *7-Dimethoxynorbornane Derivatives

R2_ OMe OMe

H H 6 .8 2 r 6 .8 2 t H OAc 6.80 6.80 OAc H 6.80 6.81 H OH 6 .7 8 6.79 OH H 6 .6 9 6.7*1-

confirmation of the stereochemical assignments. One use of n.m.r. was to establish the positions of the methoxyl groups as shown in 57 Table 1. Whereas 7>7-dimethoxybicyclo[2.2.1]heptane> the endo- hydroxyketal (jjj5), and the exo- and endo-acetoxyketals (62 and 63* respectively) had virtually identical shifts for the methoxyl protons, the exo-hydroxyketal (5*0 exhibited a considerable downfield shift.

This shift signalized the effect of the internal hydrogen bonding be tween the syn-methoxyl and the exo-hydroxyl groups.

Further evidence for the stereochemical assignments was obtained

from the position of the a-hydrogen atoms in the n.m.r. spectra for the 66 ketal-alcohols and -acetates. Wong and Lee have shown that the

^ E. W. C. Wong and C. C. Lee, Can. J. Chem., {}£» 12*j-5 (196*0 • 32

TABLE 2

NMR Chemical Shifts for a-Protons in

Norbornyl Alcohols and Acetates

,OH OAc OH OAc 6.3^ * 5-83 5 M 5-13

MeO OMe MeO^OMe MeO^OMe MeO OMe

OAc OH

6.31 t 5-67 5-^9 ^-97

endo-a-hydrogen atoms appear at higher field than the exo-q-hydrogen atoms in norbornanols and norbornyl acetates. Comparison of the pub lished data with the values of the ketal-alcohols and -acetates are consistent with the assigned stereochemistry (see Table 2).

Chemical evidence for the exo stereochemistry of the hydroxyl group in was adduced in the following way. Epoxidation of 7*7- dimethoxynorbornene (£2) gave 7»7-dimethoxybicyclo[2.2.1]heptane- exo-2»3-epoxide (62). Lithium-ethylamine reduction of 62 gave in 33 addition to a quantity of exo-2-S£n-?-bicyclo[2.2«l]heptanediol 67 (68) which was identical to an authentic sample.

+ OH OH 68

Having firmly established the stereochemistry of the hydroxyl group in and various methods of preparing large amounts of the 68 pure epimers were explored. Epoxidation of jQ with perbenzoic acid or the more convenient m-chloroperbenzoic acid (FMC Corporation) gave

6?' in 87^ yield. Although models suggest that the exo side of the double bond in is more hindered than the endo side* the attacking reagent reacted from the exo position. The direction of this stereospecific epoxidation is probably due to hydrogen bonding of the per- 69 acid with the oxygen atom of the syn-7-methoxyl group. When the epoxide ( 67 .) was reduced with lithium aluminum hydride in refluxing tetrahydrofuran for twelve days* the exo-alcohol (5^) was obtained in

9 ^ yield. The exo-alcohol (5^) was readily converted to 67 For the synthesis of this compound, see "Part II. The 7-Nor- bornenyl Cation" of this dissertation. Zq 00 G. Braun, "Perbenzoic Acid," Organic Syntheses, Coll. Vol. I, H. Gilman* ed.» John Wiley & Sons, Inc., New York, 1932, p. ^31* 69 For examples of the role of hydrogen bonding in stereospecific epoxidation, see H. B. Henbest and R. A. L. Wilson, J. Chem. Soc., 1957, 1958; H. B. Henbest and B. Nicholls, ibid., 1957, 4608. 70 See EXPERIMENTAL of this dissertation. 34

7»7-dimethoxynorbornan-2-one (66) via the Sarett oxidation in 83# y i e l d .

Various methods of reducing the ketone (66) were investigated in an attempt to obtain the endo-alcohol (££). This alcohol was found to be most conveniently obtained by the Meerwein-Ponndorf-Verley reduction of 66. Experimental methods of obtaining the alcohols £4 and are summarized in Table 3*

MeOv OMe MeO. OMe

Proof of the molecular structure of the keto-alcohols ^6 and were complicated by the rapid decomposition of the two compounds.

Thus, attention was directed to the corresponding acetates and tosylates.

Nuclear magnetic resonance studies of these compounds indicated that the assigned stereochemistry was correct: The relative chemical shifts of the exo- and endo-a-hydrogen atoms were consistent with those of

Table 2 (see Table 4 ).”^"

Indication that the carbonyl function was at the strained 7- position was afforded by the infrared spectra of the 7-keto-norbornyl

^ The presence of a 7-carbonyl function adds not only an inductive effect,, but also a long-range field effect [L. M. Jackman* Applications of Nuclear Magnetic Resonance in Organic Chemistry* Pergamon Press* New York, 1959* pp» 122-124], but this effect should be approximately the same for both the exo- and endo-a-hydrogen atoms, both being approxi mately the same direction and distance from the Et-cloud of-the carbonyl fu n c tio n . 35

TABLE 3

Methods of Obtaining exo- and endo-Hydroxy-

7 t?-dimethoxyno rbornane

Reaction $ endo $ exo % Y ield MeO. .OMe LiAlH^ r 0 100 94

MeO. /OM e

B2H6 , 22 78 74 b 2* H202 , OH"

M e O s . X3Me

LiAlH^ s 25 75 77

b M e O ^ /O M e

r ' J 5 !______* 30 70 42 EtOH o M e O N / O M e r ^ - 1 Al(i-PrO)^ ^ 9g70 270 98

i-PrOH

O " 36

TABLE 4

NHR Chemical Shifts for a-Protons in

7-Ketonobomyl Acetates and Tosylates

I 1 O j rk H OAc H OTs

5*18 x 4.86 5 .2? 5 . 02

alcohols, acetates, and tosylates. All of these compounds had a carbonyl stretching frequency at 5*60-5.65 4, a feature diagnostic of 72 a strained carbonyl function. Furthermore, each of these spectra included a shoulder at 5*35-5*^3 4, characteristics which are considered 57 to be diagnostic of a 7-ketonorbomane derivative. The peak at 5*35-

5.43 4 was found to be of lower intensity in endo derivatives than in the corresponding exo derivatives.

Chemical evidence attesting to the correct stereochemistry in exo-2-acetoxynorbornan-7-one ( 60 ) was afforded by lithium aluminum

72 C. F. H. Allen, T. Davis, D. W. Stewart, and J. A. Van Allan, J. Org. Chem., 20, 306 (1955)* 3? hydride reduction. Reduction of 60 gave only exo-2-syn-7-norbomanediol^? (68), identical to an authentic sample. The identification of

68 established the exo nature of the acetate group in 6o.

The solvolyses of exo- and endo-2-h.ydroxynorbornan-7-one £-toluenesulfonate (j>8 and jj£) were carried out in anhydrous acetic acid (buffered with sodium acetate) at 75> 90* and 100°. The specific rate constants are listed in Table 5* Included are the specific rate constants for the 2-norborqyl £-toluenesulfonates.

It is seen from Table 5 that the acetolysis rates of 7-keto-endo- norbornyl tosylate (59) and endo-norbornyl tosylate ( 23) are virtually identical at 25°. Apparently the cancelling factors previously dis cussed, the inductive effect of the carbonyl function and the Haring of the G1-C2-C^ bond angle, are: fortuitously the same. The signifi cant datum, however, is that the exo-endo rate ratio for the 7-ketonorbornyl tosylates (^8 and jjg) is not only low, but is less than unity (viz.* 0.17). Interpreted in light of the previous discussion

(see p. 2^)» these results are more consistent with the nonclassical

carbonium ion theory.

Before this conclusion can be accepted as correct,, other rationali

zations for the unique exo-endo rate ratio for $8 and 59 should be

considered. Three other possibilities might be considered: (a) hemi-

ketal formation, (b) different mechanism* and (c) dipole-dipole

interactions.

(a) Hemiketal formation. The possibility that the rate of

solvolysis of the endo-tosylate (59) is accelerated by hemiketal

formation followed by an intramolecular Sjj2 displacement requires 38

TABLE 5

Acetolysis Rates of 2-Norbornyl and 7-Keto-

2-norbornyl Tosylates

H* K cal. / S* Compound Ref. Temp. C. Rate (s e c . mole) ( e .u .)

100.00 (1 .8 4 i . 01) x 10”^ 2 7 .1 -3 -5 90.00 (5-96 ± . 06 ) X 10"-5 OTs 75.71 (1-33 t -01) x io -5 (25) 1.-44 x 1 0-8

100.00 (4.66 + .00) x 10 24.7 - 8.1 90.00 (1 .7 9 t -03) x 10-^ 75*74 4.28 x 10-5 (25) 8.66 x 10-8

21.6 -7-2 OTS 27 (25) 2.33 x 10"5

27 ( 25 ) 8.28 x 10 -8 25.8 -4 .4

OTs 22 consideration. If this type of internal displacement were responsible 2 3 for an acceleration of the solvolysis of 22 hy a factor of 10 -10 ,

H 60

then changing solvent should product a drastic change in the absolute rate of solvolysis of 22 anc* in the exo-endo rate ratio. In factj ethanolysis of 2§ and 22 Save rates which were very close to the acetolysis rates both in absolute rate values and in the exo-endo rate ratio of 2.1 (at 100°) versus O.h- (at 100°) for acetolysis (see Table 6).

(b) Different mechanism. The possibility that J58 and 22 solvolyze by different mechanisms should be considered. For instance> 2§ could

solvolyze in a relatively straightforward manner to yield the bicyclic

ion (^2) while the solvolysis of 22 might occur via a concerted bond

cleavage process leading to the acyl cation ( 69 ) However> th e f a c t TABLE 6

Ethanolysis Rates of 7-Keto-2-norbornyl Tosylates

Compound Temp. °C. Rate (sec.“"*•)

100.00 2.56 x 10" ^ 90.00 (1.21 t .01) x 10“^

100.00 1.2 x 10"^ 90.00 (4.93 ± *03) x 10"-5

OTs 2SL that both jj8 and jjg solvolyze to give products with the 7-ketonorbornane skeleton intact in over 50 $ yield (vide infra) precludes such a possi b i l i t y .

J5§ i n OTs

(c) Dipole-dipole interactions. The question of a dipole-dipole or dipole-ion interaction is much more complex and thus much harder to resolve. It has been shown by Kwart' that not only the presence of an inductive group but also its orientation relative to the reac tion site can influence solvolysis rates. The orientation factors studied by Kwart generally changed the rates by a factor of less than 3 ten* which is far less than the factor of 10 needed to explain the unique exo-endo rate ratio of J58 and More specific comparison can be made with syn- and anti-7-chloro-exo-2-norborn.vl p-toluenesulfonate (£0 and £1, respectively)* studied by Roberts.^ It is seen —— This cation apparently exists as an intermediate in the reac tion of diazonorcamphor in dilute acid and in the solvolysis of J58 and in aqueous acetic acid and aqueous formic acid [M. Hanack and J. Dolde, Tetrahedron Letters, No. 3> 321 (1966)]. 74 H. Kwart and T. Takeshita, J. Am. ^hem. 5oc.» 86, ll6 l (196*0 . 75 W. G. Woods, R. A. Carboni, and J. D. Roberts, J. Am. ^hem. 3oc., 2§, 5653 (1956). that the relative orientation of the carbon-chlorine and carbon- tosylate bonds in 70 and 71 are very similar to the carbonyl and

carbon-tosylate bonds in and j 52 > respectively, in regard to both direction and distance. Roberts finds that the relative rate ratio of 20 and 21 is only 1.7. Therefore, dipole-dipole interactions should be of little consequence in j )8 and I t may be argued th a t 76 the carbonyl dipole, being greater than the carbon-chlorine dipole,

should produce a larger rate ratio in the case of £8 and However,

the syn-chloro compound ( 2 0 ) solvolyzes faster than the anti-chloro

analog (21)• Analogy of the relative directions of the dipoles in 20

and 2i with those in ^8 and Jj>2 suggests that a dipole-dipole inter

action, if significant in the solvolysis of j >8 and jj 2 » should enhance

the rate of the exo epimer ($ 8 ) over that of the endo epimer ( 59 ) •

The unique low exo-endo rate ratio for $8 and cannot, therefore,

be rationalized on the basis of a dipole-dipole interaction.

The conclusion is that the exo-endo rate ratio for %8 and £2 is

uncomplicated and is more consistent with the low exo-endo rate ratio

predicted by the nonclassical carbonium ion theory.

J. D. Roberts and M. C. Caserio, Basic Principles of Organic Chemistiy, W. A. Benjamin, Inc., New York, 1964, p. 165 . It remained to discover whether product analysis would support the classical or the nonclassical carbonium ion theory. When the product analysis of %8 and was carried out by vapor phase chromatog raphy of the crude acetolysis mixtures* the only detectable products were endo-2-acetoxynorbornan-7-one ( 6 l) and exo-2-acetoxynorbornan-7- one ( 60 ). The identity of these products was established by comparison of v.p.c. retention times of the products with those of authentic samples on three different columns, and by comparison of the infrared spectra of samples isolated by preparative v.p.c. with the infrared spectra of authentic samples.

The exact yields and product ratios were very difficult to assess due to the sensitivity of the endo-acetate ( 6 l) to heat, acid, or 77 base. Vapor phase chromatography of the crude acetolysis mixture resulting from $§. showed only 60 and 6 l in the ratio of 3*2. When the acetolysis mixture was neutralized with sodium bicarbonate and the sol volysis products were isolated and distilled, a 71$ yield of mixed products was obtained. This material consisted of 65 $ exo-acetate

(60 ) and 20$ endo-acetate ( 6 l) . Six minor components constituted the rem aining 13$. It was considered likely that the minor components were a result of the reaction of the endo-acetate ( 6 l) with base in 78 the neutralization of the acetic acid. To test this possibility,

77 C. H. DePuy and P. R. S to ry , J . Am* Chem. S o c ., 82, 627 (I9 6 0 ). 78 Compound 61 "hydrolyzed com pletely in a few m inutes a t room temperature with dilute base" (see Ref* 77)* a very small amount of solution was reacted under identical reaction conditions; the small amount of acetic acid allowed very rapid workup.

Vapor phase chromatography of the crude product indicated the absence of four of the six minor components observed previously (there re mained two elimination products, to be discussed below), and the exo- endo acetate ratio was 52:48, compared with the respective ratio of 69:31 for the distilled product isolated in a larger scale. This showed that four of the minor products detected in the large scale solvolysis product mixture arose during workup, and that most, if not all,, of these minor products resulted from the endo-acetate ( 6 l ) .

Analysis of the crude products from the acetolysis of the endo- tosylate (ji^) also showed a mixture of 60 and 6 l, this time in the respective ratio of 97*7*2.3. When the acetolysis products were isolated as before in the case of 5 8 , a 59 $ yield of material was obtained, which consisted of 76$ of exo-acetate ( 60 ), 3$ of endo-acetate (6l), and 21$ of other components.

It appears that the products obtained from 58 and are n°t arising from a single free carbonium ion, as this ion would require

58 and ^ yield identical proportions of 60 and 6l. Two explana tions for the reduced 6o_:6l ratio in the acetolysis of deserve

recognition. The first explanation is that hemiketal formation (see

p . 38) is playing a significant role in the acetolysis of j>2» ^or ^ e

product resulting from this process would be exclusively exo. To see

if this explanation is reasonable, one could calculate the expected

6 0 :6l product ratio on the basis of the data of Tables 5 and 6. There

is a small change in the exo-endo rate ratio as the solvent is changed from acetic acid to ethanol. If one assumes that this small change is due exclusively to a hemiketal factor which plays a role in the solvoly sis of 32 acetic acid but not in ethanol, then one easily calculates that the relative rate of 32 must decrease by a factor of 5 as the

solvent is changed from acetic acid to ethanol. The solvolysis rate of 32 acetic acid would therefore consist of a hemiketal component

(80$) and a "normal" component (20$). The hemiketal component would give 100$ exo product, and the "normal" component might be expected to give a 60:40 exo-endo product ratio , from analogy with the products obtained from the exo-tosylate (3§) • It follows that the acetolysis of 32 should give at least 8$ endo-acetate (6l). In fact, the percent age o f 6 l r e s u ltin g from 32 rouch le s s th a n t h i s ; s u re ly some o th e r

effect is responsible.

The second explanation for the reduced 6 0 :6 l product ratio observed

in the solvolysis of 32 fs that the carbonium ions formed from 38 and

32 are highly reactive since they lack the stabilization normally

derived from nonclassical cation formation. As a result both 38 and

32 may be yielding products at an early stage in the separation of the 79 ion pairs represented by £2 and 73, respectively. The ion pair 72

would be hindered on the endo side by the endo-C^ hydrogen atom and

on the exo side by the leaving tosylate anion. In contrast, 2 2 would

have both the endo-C^ hydrogen atom and the leaving tosylate anion

hindering endo substitution while the exo side would be relatively

available for solvent addition.

' The less stable a carbonium ion is, the more S^2 character the reaction involving its formation exhibits [A. Streitwieser, Jr., Solvolytic Displacement Reactions, McGraw-Hill Book Company, Inc.,. New York, 1962, pp. 60-6l]. 46

; o t s

,O T s

3§ 22

O T s h i'O T s H

The conclusion is that the solvolysis of 8 and J£, yielding both exo and endo products, is most consistent with the nonclassical carbonium ion theory. This first example of a secondary norbomyl cation being attacked by solvent from the endo direction®^ is ir reconcilable with the classical carbonium ion theory.

Additional evidence in support of the nonclassical theory was obtained by studying the elimination products of the acetolysis of 58.

When a small amount of ^8 was solvolyzed as previously described (see p. 44), vapor phase chromatography indicated the presence of only

— Displacement by strong nucleophile on exo-norbornyl derivatives can give endo product.! reaction of exo-norbornyl brosylate with lithium jg-thiocresoxide is attended with complete inversion [5. J. Cristol and G. D. B r in d e ll, J . Am. Chem. 5 o c.» _£6, 5^99 (19'5^)]« i+7 four compounds: the two keto-acetates 60 and 6l previously identified, and nortricyclanone®1 (2^) and 7-ketonorbornene^ (££) * The elimina tion products 2}± and 2 5 constituted an 8.0$ yield of the total product, and the ratio of 2}k. to 2 5 was 1*1* These results compare with a **$

.OAc + + +

O a c

^8$ ^ $ ^ .0 $ A05 . & elimination product yield from the acetolysis of exo-2-norbornyl £- Qp bromobenzenesulfonate (£) where the nortricyclanesnorbornene ratio was 98*2. Winstein argues that the preponderance of nortricyclane (76) is consistent with the nonclassical cation theory insofar as the inter mediate that would most closely resemble 25. would be the nonclassical carbonium ion (11). If this reasoning is correct, then the acetolysis of the 7-keto analog should give a reduced nortricyclane :norbomene skeletal product ratio. This is the case, as the ratio decreases fifty-fold. What is surprising, however, is that the quantity of nortricyclanone is as great as that of 7-ketonorbornene. This is

0*1 F.V. Zalar, Ph.D., 1966, The Ohio State University, generously donated an authentic sample of this compound for comparison studies. 82 5. Winstein, E. Clippinger, R. Howe, and E. Vogelfanger, J. Am. Chem. S o c ., 8£, 376 ( 1965 ). 48

,OBs

Ik 11

98 : 2

perhaps indicative of facile bridging between the C„ and C, carbon 2 o atoms even when the carbon atom is not involved. On the other

'+ H

hand, the quantity of nortricyclanone could mean that some residual

nonclassical character exists in the 7-ketonorbornyl cation (vide

i n f r a ) .

It was felt that the solvolysis of the 7>7-dimethoxy-2-norbornyl

tosylates (64 and 6£) might yield interesting results, as these com

pounds also bear electron-withdrawing groups at the 7-position. The

ketal function in 64 and 6^ was surprisingly inert to acid treatment. ^9 In acetic acid (buffered with sodium acetate) excellent pseudo first- order kinetics were observed; Table 7 includes the results.

The products from the acetolysis of 64 and 6£ were completely different (vide infra). Therefore, the meaning of the exo-endo rate ratio is unclear, as the solvolysis mechanisms need be different for the two tosylates.

It is seen from Table 7 that the rates of the exo-tosylate (64) and syn- and anti-7-chloro-exo-2-norbomyl ja-toluenesulf onates (20 and 2i» respectively) are all virtually identical. These rates are over two magnitudes slower than that of exo-2-norbornyl tosylate 27

(39)• The question may be raised whether the rates of these three tosylates are retarded because of a purely inductive effect or to lack of anchimeric assistance. Although solvolyses of tosylates with p-substituents have been studied in some detail (vide infra), ^3*84 pertinent studies of the solvolysis of tosylates with a-substituents are lacking. It is therefore difficult to predict a priori what the solvolysis rates of 64, 22 > and 2i should be if only an inductive ef fect were responsible for their retardation.

In view of the significance of the product analysis in the solvoly

sis of the keto-tosylates (58 and j$2), it seemed important to ascertain

whether any endo product resulted from the solvolysis of the ketal-

tosylate (64). Acetolysis of 64 gave, after workup and distillation,

a 66$j yield of mixed acetates consisting of 95 *5% exo-acetate (62)

and 4.5^ endo-acetate ( 63 ). There is again a substantial unprecedented

amount of endo product, which is consistent with the nonclassical cation 50

TABLE 7

Acetolysis Rates of 7»7-Dimethoxy-2-norbornyl

and 7-Chloro-2-norbornyl Tosylates

H* (K c a l./ S* Compound fvsx Temp.°G. Rate (sec.-l) mole) (e.u.)

M e O s /O M e 100.00 (1.07 i .04) X i o - 3 26.8 -0 .7 90.00 (3 .5 8 i . 13) x 10- ^ 75-85 (8.12 t . 03) x io -5 (25) 9 .5 3 X l o - 8

100.00 (6.35 i 27.5 -5 .8 M e O s / O M e 90.00 (2.05 ± 75.77 4.39 X ,-9 (25) 2.51 X O Ts

(25) 1 x 10 -7

(25) 6 x 10-8 51

Me OwOMe MeOv OMe MeO^OMe

,OTs OAc+ OAc 64 62 § 2

95*5 : 4 .5 theory. However,, the large difference in the exo-endo product ratio

in 64 and in £ 8 deserves critical analysis•

If both and 64 were solvolyzing to classical cations (4£ and

74, respectively), one would predict a lower exo-endo product ratio

OTs

MeOv nM e in the case of 64 because of the steric hindering effect of the syn- methoxyl group in 2it* In fact, the reverse is observed. Therefore, the products resulting from and 64 cannot be rationalized in terms of the classical cations 4£ and 74.

The baffling exo-endo product ratio for 64 may be explained if one assumes a gradual change from nonclassical to classical cations^; the solvolysis of 64 would then represent an intermediate case. Such a treatment would probably best be applied by utilizing a concept

Q o q z T developed by Winstein. Winstein * has defined k^ as the anchimeri- cally assisted solvolysis rate and k as the solvent assisted rate of s an aryl sulfonate. He considers the ratio k^/kg to be very high for the exo-2-norbornyl derivatives he has studied. Three experimental facts form the basis for this concept. First, an exo-2-norbornyl derivative always solvolyzes considerably faster than the corresponding endo epimer. This fact is explained by assuming that the endo deriva tive solvolyzes via only a ks contribution and that the exo derivative solvolyzes via both the kg and the k& contribution. Since k^ is much greater than k_ in typical norbornyl systems, the overall effect is a high exo-endo rate ratio. Second, the product of solvolysis of a nor bornyl derivative is always substituted exclusively exo. Analogy of QO J S. Winstein, E. Grunwald, and L. L. Ingraham, J. Am. Chem. Soc., 22» 821 (1948). & S. W instein and E. Grunwald,, J . Am. Chem. S o c ., 7 0, 828 (1948). 85 , H. C. Brown, K. J . Morgan, and F. J . C hloupek, J . Am. Chem. Soc., “ , 2137 (1965). S. Winstein, E. Allred, R. Heck, and R. Glick, Tetrahedron, 2 , 1 (1958). 53

(very large)

OTs or OTs (very small) OAc

or ,OAc

(very small) OTs

the transition states of the tosylates with the reaction of the carbonium ion intermediate with solvent suggests that the latter reaction would preferentially involve a nonclassical transition state* Thus, the resulting product would have to be exo. Third» while the solvoly sis of exo- 2-norbornyl brosylate is attended with complete racemization, the solvolysis of the endo analog occurs with partial retention 12 of optical activity. This is explained by assuming that the exo derivative, solvolyzing essentially completely via a process, ionizes virtually lOO^b to ihe optically inactive nonclassical carbonium ion. The endo derivative solvolyzes exclusively via a ks process;

S^j2 displacement is there&ne possible, explaining the optically active f r a c tio n .

Winstein's treatment is applicable to the solvolysis of the keto-tosylate (jj 8) and the ketal-tosylate (64) in the following manner.

As the stability of a nonclassical transition state is decreased (i.e.» by placing electron-withdrawing groups at the 7-position), the k^/kg ratio would be expected to decrease accordingly. Thus, the k^Jkg r a tio should be higher for 64 than for Since the ks contribution can give rise to S ^2 displacement, then ^8 could give a higher proportion of endo product. Therefore, using the k^/ks ratio concept, it is

reasonable to view the exo-endo product ratios of 64 and 58 in term s

of the nonclassical carbonium ion theory.

As it is probable that the k^/ks ratio concept is applicable to

the solvolysis of jj 8, it is possible that this concept can clarify

the exo-endo product ratio obtained in the acetolysis of 7-keto-endo- 2-

norbornyl tosylate (^£). The solvolysis of the 7-ketonorbornyl tosyl

ates may not be truly classical, i.e ., the k^ contribution may be sig

nificant in the solvolysis of the exo-tosylate (58). One may reasonably

assume that this k^ contribution cannot be preponderant because of the

large amount of endo-acetate ( 6 l) derived from jj>8. I t is th e re fo re

possible that the ks and k^ contributions are comparable in the

solvolysis of jj 8 . If one now assumes that the kg contribution would

give almost complete inversion, then he would predict that the endo-

tosylate (59) would give almost completely exo-acetate ( 6 o). Further

more, he would predict the exo-tosylate (j> 8) would give a variable

amount of both exo- and endo-acetates, the relative ratio of which

depending upon the k^kg ratio. This is essentially what is observed.

It is therefore possible that the k pks j ratio concept is the appropri

ate way to interpret the exo-endo product ratios of the keto-tosylates 55

O A c (m ajor) 60 OTs

OAcO A c (m ajor) 61

O A c

OTs (m ajor) i 2 60

58 and jj2# This interpretation necessarily involves assuming that some residual nonclassical behavior is present in the solvolysis of

5 8 . This assumption is reconcilable with other aspects of the acetoly sis of jj8, as pointed out previously in this discussion (see p. 48).

In contrast to the solvolysis of 7,7-dimethoxy-exo-2-norbornyl

£-toluenesulfonate (64)> the endo epimer (6£) upon acetolysis gave no

2-acetoxy-7»7-dimethoxynorbornane (62 or 6^). Instead, the observed 81 products were nortricyclanone (^4), exo-2-methoxybicyclo[2.2*l]- heptan-7-one (£8), and exo-2-acetoxynorbornan-7-one ( 60 ) in th e respective ratio of 29sl6:55. The identity of each of these compounds was established by comparison of their retention times with those of authentic samples on two different columns. In addition, the three 56

products were separated by preparative v.p.c* and their structures were confirmed by comparing their infrared spectra with those of authentic samples.

An authentic sample of 2§. was prepared by the following scheme. exo-2-Hydroxy-7>7-dimethox.ynorbornane (5*0 was reacted with sodium hydride in dry tetrahydrofuran to give the alkoxide salt (22) which 87 was converted with methyl iodide to 7>7-exo-2-trimethoxybicyclo-

[2.2.l]heptane (80) in 91$ yield. Hydrolysis of 80 with 5$ aqueous sulfuric acid gave £8 in 91$ yield. The n.m.r. and infrared spectra of both 80 and 2§ were consistent with the assigned structures.

87 For methyl ether formation via the Williamson synthesis, see J. Meinwald, I. C. Meinwald, and T. N. Baker, III, J. Am. Chem. S o c ., 8 6 , *4-07*+ (196*0* 57

M e O^OMe M e O ^O M e M eO OMe

'S OMe

The absence of 2-acetoxy-7,7-dimethoxynorbornane seems to exclude

the possibility of the formation of the expected carbonium ion (?4).

Instead* the presence of the 2-methoxy derivative (2§) indicates that

MeO nM e

65 OTs

methoxyl bridging during the transition state is important. The bridged 88 cation (81) could collapse to the methoxonium ion (82) which might

convert to the observed compound (2§)• Since no water is present, 89 the normal path of hydrolysis of 82 would be excluded. A mechanism

which does not require water is the addition of acetic acid to 82 to

give the ketal 8^ which could give 78 via a cyclic mechanism.

88 The author is not necessarily trying to suggest that 81 must be a distinct intermediate as opposed to a transition state between 65 and 82. Lack of this distinction, however, has no bearing upon the following discussion. 89 J. D. Roberts and M. C. Caserio, Basic Principles of Organic Chemistry, W. A. Benjamin, Inc., New York, 1964, pp. k k J -^ 6 . 58

Me O^ /OMe MeO. MeOs. Me d +

6 ^ OTs 81 OTs Me

,OMe *■ OMe

28 82

CH3 ^ ° - CH3 CH ° o \\

OCH OCH OCH.

82 59

Although 1,2-? and 1,5-methoxyl migrations^ in solvolysis reactions are known, and other examples of 1 , 2-migrations 83 .ftli-,93 are well documented, J this is apparently the first case of a

1 ,,3-methoxyl migration in a solvolysis reaction.

It was considered possible that the bridged intermediate ( 8l ) could react with solvent to form the endo-acetate ( 63 ); this acetate might be unstable under reaction conditions and would react in a

MeO Me Q '+

81 62 OAc

manner analogous to the endo-tosylate ( 65 ) to give the observed products. This possibility was eliminated by a control experiment i n which 63 was recovered under identical acetolysis conditions.

A satisfactory mechanistic explanation for the formation of

nortricyclanone ( 2 +) and exo- 2-acetoxynorbornan- 7-one ( 60 ) cannot be

made. It was first necessary to ascertain that no interconversion of

the three observed products occurred. Furthermore, it was considered

necessary to explore the possibility of the observed products arising 90 — — ~ - S. Winstein and L. L. Ingraham, J. Am. Chem. Soc.* H60 (1952). 91 7 D. S . Noyce* B. R. Thomas, and B. N. B a stia n , J . Am. Chem. Soc.» 82, 885 (I960). ^ D. S . Noyce and B. N. B astian* J . Am. Chem. Soc.» 82, 12^6 ( i 960 ). 93S. Winstein* C. Hanson, and E. Grunwald* J. Am. Chem. Soc., 70 t 812 (1 9 ^8 ). from 62* 6 ^, 84> and 80. Compound 84 m ight a r i s e from th e brid g ed intermediate (81), and 80 might arise from the reaction of the methyloxonium cation (82) with methanol displaced from another molecule.

.OMe OAc

2± 6 o

MeOvpMe MeOwOMe MeOwO^^

OAc OMe

62 62 O A c 80

Control experiments on 2!±.» 2§.> ^2> §2.> 84 and 80 under

identical reaction conditions demonstrated that none of these compounds

could convert to the observed reaction products.

A mechanism responsible for the formation of exo-2-acetoxy-

norbornan-7-one (60) might be the following. An acetic acid molecule

participates in a cyclic transition state (8^), resulting in the

expulsion of methanol and the formation of exo-2-acetoxynorbornan-7-

methyloxonium ion (86). The cation 86 could convert to 60 in a manner

analogous to that proposed for 82. It is to be noted that according

to this mechanism the oxygen-hydrogen bond in acetic acid is broken

during the transition state. It was felt that if this mechanism were 6l

CH. ,CH

C •OAc CH

correct, then substitution of acetic acid-O-d as a solvent might

result in a noticeable primary isotope effect. In fact, when acetic

acid-O-d was used as the solvent, there was no detectable decrease in

the observed rate (see Table 8 ),^

TABLE 8

Acetolysis Rates of 7,7-Dimethoxy-endo-2-norbornyl

Tosylate in Acetic Acid and Acetic Acid-O-d

Compound S o lv en t Temp. °C. Rate (sec.“^)

HOAc 100.00 6.35 x 10"-5

DOAc 100.00 6.45 x 10”5

Another possible mechanism involves the methyloxonium ion (82)

(see Chart II). If attack by solvent on 82 to form the ketal (83)

were s.yn to the methoxyl group, then the acetate function in s.yn-83

would be in a position to attack the carbon atom. Consequent

Walden inversion and expulsion of methanol (see 82) would result in

^ For more data and discussion of the acetic acid/acetic acid-O-d rate ratio, see "Part II. The Norbornenyl Cation” of this dissertation. CHART I I the cation 88. This cation could react further in two different ways.

First, expulsion of a proton (82) would give a nortricyclanone ketal

(90) which could convert to nortricyclanone ( 7^ ) v ia a mechanism analogous to that previously proposed. Second, the cation ( 88) may react with a solvent molecule with accompanying Wagner-Meerwein re arrangement (£1) to give $ 2 . Compouhd 22 could then c o n v ert to th e observed compound ( 60 ).

This proposed scheme may possess a serious drawback. One may consider it unlikely that 83 would be attacked by solvent mainly from the syn direction, which would be necessary to explain the preponderance o f 2J± an

o ^ c x C h 3

Although the exact mechanism whereby and 60 are formed is not

known with certainty, the experimental facts do lead to a significant

conclusion. Since no carbonium ion 2± is formed from 81, and since

81 instead chooses to convert to the methyloxonium ion (82), then it

follows from Hammond's postulate^ that 81 more closely resembles 82

than It als° follows that 82 is most probably much more stable 6U

M e ^ + M e O

•OMe

82 81

th a n J>±. This conclusion has direct bearing on the question on the nonclassical 7-norbornenyl cation (vide infra). PART I I . THE 7-NORBORNENYL CATION

1 . The problem

The case for the bishomocyclopropenium cation (17) rests upon two experimental facts, (a) The tremendous difference in reactivity between 1£ and l6 could be attributable only to some unique and energetically favorable phenomenon occurring in the transition state

OTs

IZ 16

17 18 95 of 1£. * * (b) Exclusive anti product could be explained only in terms of an intermediate which had bonding between carbon atom and carbon atoms C„ and C . 2 3 19 Brown argues that the intermediate resulting from 1£ would be

the=classical ion (lg). He finds that the tosylate (1£) in aqueous diglyme under sodium borohydride reduction gives substantial quan- 3 7 tities of tricyclo[4.1.0.0 * ]heptane (2^)* He argues that the

cation most likely to be responsible for the formation of would

be 2^* Under acetolysis conditions, the classical cation (12) would

react with solvent to form the tricyclic acetate (25)» which would

95 S. Winstein and M. Shatavsky, J. Am. Chem. Soc., £§.» 592 (1956). 1 5 1 2 2*

HO Ac

A r ^

-> OAc 25 26

be thermodynamically unstable and which would rearrange to the observed product,, the anti-acetate (96). 20 Winstein takes issue with Brown concerning the stability of 25*

He claims that 92 would not be any more unstable than various other compounds which he has isolated from sim ilar solvolysis conditions.

He further shows that Ijjj under borohydride attack gives not only the tricyclic hydrocarbon (2k)» but also substantial quantities of norbornene (22)* Formation of this compound (72) could arise only via borohydride attack at the C carbon atom, which would mean that there

±1 2k 21 (m ajor) (m inor) must be some positive charge centered at this carbon atom in the inter mediate. In any case Winstein infers that the borohydride reduction and the solvolysis of 1£, giving products with different carbon skeletons, merely show that conclusions drawn from the comparison be tween the two very different reactive media are invalid.

Concerning the stability of the tricyclic acetate (££), perhaps

Winstein is engaging in a pointless argument, for the nonclassical cation (12) would be a resonance hybrid of mainly the contributors

19a and 19b. Most of the positive charge would reside on the carbon

1Z i 2 i 19b minor contribution atoms and C , and kinetically controlled attack would be expected ^ 3 at either or to give the tricyclic acetate (£$)• Furthermore, it is quite possible that 25 would be unstable under the reaction con ditions, as the analogous tricyclic derivative 2§> derived by methoxide attack on 7-chloronorbornadiene (22) > ts extremely sensitive to dilute acid and reacts very rapidly in an aqueous medium to form 7-hydroxy- 96 norbornadiene (92)• !n any case, if the tricyclic acetate (9£) were unstable to reaction conditions, there would be no consequence upon either the nonclassical or the classical cation theory.

. . H. Tanida, T. Tsuji, and T. Irie, J. Am.Chem. Soc., 88, 864 (1966 ). 68

r i .OH + 'OMe H3 °

OMe 2§. 2 1

Story7' reduced 7-chloronorbornadiene (9£) with lithium aluminum deuteride in tetrahydrofuran,. and found that attack was preferentially endo to give 100. This is exactly what would have been predicted from the nonclassical cation concept. On the other hand,, classical cation r i

102 D" theory would have predicted a major portion of product resulting from exo attack to give 101, since exo attack would be less hindered. Un fortunately, lithium aluminum hydride is much more reactive than sodium borohydride; probably the free cation never exists. Instead, the reac tion mechanism is likely to be synchronous (102). Thus, whether Stoiy's work is pertinent to the free cation is unknown.

P. ft. Story, J. Am. Chem.Soc., 8^, 33^7 (19&1) 69 2. Approach toihe problem

During the course of the w riter's research, it was felt that a good model for determining the dipole-dipole interactions in the solvolysis of a^d 22 (see p. 41) would be the keto-tosylates 103 and 104, as in the latter two compounds no possibility of the com plicating nonclassical carbonium ion formation could exist.

,OTs OTs

OTs TsO

104

Accordingly, these two compounds were prepared and solvolyzed. It was observed with a great deal of surprise that 104 solvolyzed much faster than 103• It was subsequently determined that 104 solvolyzed via an enolization rate-determining step.

The solvolysis rate of the enol (105) immediately became of

interest, as the rate ratio of anti-7-norbomenyl tosylate (1£_) to

the enol (105) could be compared with that predicted by both

classical and the nonclassical carbonium ion theories. 70

T s ° T s ^

> p ro d u cts Slow f a s t o OH 104

3 . R esu lts

The starting compounds for the synthesis of the keto-tosylates 103 and 104 were exo-2-syn-7-bicyclor2.2.11heptanediol (68) and exo-2- 99 anti-7-bicyclo[2.2.1]heptanediol (106). An analogous approach was taken for both the syn- and anti-tosylates.

syn-7-Hydroxybicyclof2.2.1 ~lheptan-2 -one £-toluenesulfonate (103) was prepared by two independent routes. First, the diol (68) was oxidized to' syn-7-hydroxybicyclor2.2 .l~lheptan-2-one100 (107) by chromic acid in acetone or in water. The keto-alcohol (107) always contained a portion of unreacted diol. Because of the instability of 107, it was immediately converted to the keto-tosylate (103) by the procedure of 62 Tipson. The crude 103 contained a portion of the ditosylate (108) resulting from the diol (68) contaminant. The keto-tosylate (103) was separated from the ditosylate (108) by the Girard technique^"*" to give pure 103? m.p. 6 3 .8 -6 4 .6 ° .

— - H. Kwart and W. G. Vosburgh, J . Am. Chem. S o c ., 2 6 , 5400 (195*0. 99 J. K. Crandall, J. Org. Chem., 22* 2830 (196*0. This compound has also been prepared by K. Mislow and W. E. Meyer. For details see W. E. Meyer, thesis, New York University, 196*1. . L. F. Fieser, Experiments in Organic Chemistry, third ed., D. C. Heath and Company, Boston, 1957* pp» 88- 89. 71

OH OTs

68 107 O 103 O m .p. 6^o III V OTS OTS OTs

O T s

108 109

The second route to 103 consisted of preparing exo-2-syn-7- bicyclo[2.2.1]heptanediol di-£-toluenesulfonate (108) by the procedure 6)2 of Tipson. The solvolysis of 108 in 60:40 dioxane:water gave 2-syn-

7-norbomanediol ?-]D-toluenesulf onate (109). The crude 109 was 102 oxidized with chromic acid in acetic acid to give crude keto-tosylate

(103)« After Girard separation and recrystallization, pure 103 » m .p.

95.6-96.1°, was obtained. This material had a melting point and a KBr spectrum different from that obtained by the first route. That the two compounds were structurally identical and were allotropic was

established by n.m.r. and infrared solution spectral analysis (carbon disulfide and chloroform). Subsequent attempts to obtain the lower melting allotrope by either route failed. -j np For an example of oxidation of a hydroxy-tosylate to the keto-tosylate, see N. A. Nelson and G. A. Mortimer, J. Org. Chem., 22, lli*6 (1957). anU-7-Hydroxybicyclo[2.2.1]heptan-2-one £-toluenesulfonate103

(10*0 was prepared by two routes analogous to those for the syn

epimer (103)« First, the diol (106) was oxidized to anti-7-hydroxybicyclo[2.2.1]heptan-2-one^^ (110) by chromic acid in water. The

keto-alcohol (110), always contaminated with unreacted 106 » was

isolated by column chromatography and was converted to the keto-tosyl- 62 ate (104) by the procedure of Tipson.

The other route to 104 consisted of preparing exo-2-anti-7-

bicyclo[2.2.1]heptanediol di-jc-toluenesulfonate (111) and subsequent

solvolysis in 60:40 dioxane:water to give 2-anti-7-norbornanediol

This compound has previously been, prepared by G. H. Whitham and S. C. Lewis. For details, see d.'Cv^W is, thesis, University of Birmingham, England, 1964. • 104 This compound has previously been prepared by H. Krieger, Ann. Acad. Sci. Fennicae, Ser. A, II, No. 109 (1962). Other workers have also prepared the compound: S. C. Lewis, thesis, University of Birmingham,, England, 1964; W. E. Meyer, thesis, New York University, 1964. ?-£-toluenesulfonate (112). Oxidation of 112 with chromic acid in 102 acetic acid gave a crude sample of 104 which was purified by column chromatography with difficulty. The two compounds obtained by the two different routes were shown to be identical by infrared and n.m.r. spectral analysis, by mixed melting points, and by com parison of their solvolytic rates.

The difficulty in obtaining pure 104 by the second route was shown to be caused by 25 $ contamination of the syn epimer (103).

Quantitative assessment of this impurity was facile because of the large difference in solvolytic rates between the anti-tosylate (104) and the syn-tosylate (103), A crude sample of 104 obtained by the second route was acetolyzed to infinite tite r, which was 75$ of the theoretical amount if the sample of 104 in question were pure;. The product mixture was worked up in the usual manner and chromatographed

In addition to the acetate formed by acetolysis of 104, a quantity of unreacted syn-tosylate (103) was isolated corresponding to the theoretical amount based on the infinite titer. Probably this syn epimer arose because of some 6,2-hydride shift^0^*^^ occurring in

the aqueous solvolysis of the ditosylate (111). This hydride shift would change the orientation of the tosylate group relative to the

hydroxyl group.

105 J. D. Roberts, G. G. Lee, and W. H. Saunders, Jr.,. J. Am. Chem. S o c ., 76 , 4501 (195*0* 106 J. D. Roberts and C. C. Lee,, J* Am. Chem. Soc., 7^, 5009 (1951)* ?b

TSO TsO TsO

•OTs HO' i l l 102

That no analogous contamination of 104 was found in the syn- tosylate (103) obtained via aqueous solvolysis of the ditosylate (108) is most certainly due to the Girard separation* whose conditions were shown to destroy the anti-tosylate (104). It is thus fortuitous that the Girard separation works for the syn-tosylate (103) because this technique separates 103 from both the anti-tosylate (104) and the ditosylate (108).

5yn-7-Acetoxybicyclo[2.2.11heptan-2-one (113) and anti-7- acetoxybicyclo[2.2.1]heptan-2-one (114) were prepared fjfom the respective keto-alcohols (107 and 110) by reaction with acetic an hydride in pyridine. OH .CAC

107 m

HO AcO

no O 75 The two keto-tosylates (103 and 104) were solvolyzed in acetic acid to give the rate data listed in Table 9* Included are the rate data for 7-norbornyl ja-toluenesulfonate ( 16 ) and anti-7-norbornenyl ja-toluenesulfonate Q £). o n At 25 104- solvolyzes ca. 10 times faster than 16 when one might have expected a rate retardation because of the inductive effect of the carbonyl function. This rate enhancement observed in the solvolysis of 104 appears to be due to an unusual mechanism whereby 104 enolizes in a rate-determining step to 105. The avail able evidence indicates that the enol (105) rapidly either reconverts to starting tosylate (104) or solvolyzes to the keto-acetate (114)» where and are comparable.

The enol (105)> bearing a remarkable sim ilarity to anti-7-norbornenyl X7 18 tosylate (ljj>)» might be expected to undergo anchimeric assistance *

to give either the classical cation (115) or the nonclassical cation 76

TABLE 9

Acetolysis Rates of Various 7-Norbornyl

T o sy lates

H* (K cal. / S* Compound Temjx*°C R ate ( s e c .”^) mole ( e .u .)

OTS 200.0 (5*45 t *19) x 10“6 34.9 - 9 .8 210.0 (1 .2 0 1 . 0 0) x 10"^

90.00 (1 .1 3 t . 06 ) x 10"5 22.0 -2 0 .7

TSO 100.00 (2.78 ± .0 9 ) x 10“5

110.00 (6 .8 8 + . 01) x 10“5

(25) 1 .2 8 x 10“8

OTs l ? (200) 5 .4 9 x 10“5 35.7 - 3 .5 ✓ ✓ -15 (25) 6 .3 6 x 10

(25) 9.04 x 10“^ 23.3 5 .7

1 2 n (116). Either cation could collapse to product with overall retention at the Cr, carbon atom. Fulfilling this prediction) the anti-acetate

(114) was found to be the sole acetolysis product of 104. Upon workup and distillation of the acetolysis product of 104) a 63 $ y ie ld

°f 114 was obtained. Absolutely no other products were observed in v.p.c. analysis. However) due to the almost identical retention times of 114 and the sjrn-acetate (113) on a variety of columns (including capillary columns)) it was impossible to establish the purity of 114 beyond 97$•

Whereas in normal acetic acid 104 displayed excellent pseudo first-order kinetics throughout 72$ reaction), in acetic acid-0-d the rate of solvolysis of 104 decreased almost seven-fold before a final constant pseudo first-order rate was observed (see Table 10) Graph I).

This gradual change in the solvolytic rate reflected the fact that as

104 solvolyzed in acetic acid-0-d, there was some return of enol (105) to keto-tosylate (104); this latter process thus incorporated deuterium in the a-position. Since the rate-determining step of acid-catalyzed enolization is the removal of the o-hydrogen subsequent to proton addi- 107 108 tion to the carbonyl function) '* a primary isotope effect became 109 detectable as 104 became deuterated. 7 There was a consequent

C. G. Swain, A. J. DiMilO) and J. P. Cordnerf J. Am. Chem. Soc., 80, 5983 (1958). 108 R. P. Bell) The Proton in Chemistry, Cornell University Press, Ithaca, New York) 1959» pp» 140-15**. 109 ' For an example of a primary isotope effect in enolization, see 0. Reitz, Z. Electrochem., 4 3 , 659 (1937)» wherein perdeutero- acetone enolizes 8 times more slowly than acetone in an acid- catalyzed reaction at 25 ° . 78

TABLE 10

Acetolysis Rates of Various 7-Norbornyl

Tosylates in Acetic Acid-O-d

Initial rate Final rate Compound Temp.°C, ( s e c .”-1-) ( s e c .- '1')

TsO 110.00 (1 .3 1 t .02) x 10"4 (1.9^ t .09) X 10"5

10^

-5 200.0 (5-73 t .01) X 10 unchanged TsO

-6 OTs 200.0 5 .7 6 x 10 unchanged

103 79

GRAPH I - A c e to ly s is Rate Curves o f a n t l -.7-H.ydro x y b ic y c lo - [2.2»l]heptan-2-one je-Toluenesulfonate in Acetic Acid and in Acetic AcidiO-d.at 110°G.

CD CO

X o 00

o CM

.CM Time, Time, sec. x 10

-C£>

00 [SlOdJ 60 80 observed drop in the acetolysis rate until essentially 100$ of the remaining tosylate (104-) had become deuterated, at which time the observed rate approached linearity.

While the ratio of the acetolysis rate in acetic acid-0-d to that in normal acetic acid is 1.04- for 16 and 1*06 for 103 the ratio of initial acetolysis rate in acetic acid-0-d to that in normal acetic acid for 104- is 1.90. This much greater isotope effect in the case of 104- is consistent with other enolization rates carried out in deuterated and nondeuterated solvents wherein a general acid catalysis mechanism is responsible

If the proposed mechanism for the acetolysis of 104- were correct, then one would expect to find the tosylate isolated after increasing periods of solvolysis time in acetic acid-0-d to have increasing

amounts of deuterium incorporation; the accompanying acetate would

always be 100$ deuterated since enolization is a prerequisite for its

formation. In two runs of different solvolysis time, the resulting

acetate in both cases had identical infrared spectra which were dif

f e r e n t from u n d e u te ra te d 114-. The accompanying k e to - to s y la te is o la te d

in both cases had different infrared spectra which were both different

from that of undeuterated 104-. This is exactly what would have been

predicted. That the tosylates were not rearranged products was es

tablished by identical melting points and by the n.m.r. patterns.

In the solvolysis of aryl sulfonates where carbonium ion formation is the rate-determining step, changing from light to heavy water as a solvolysis medium has little effect on the solvolysis rate [E. R. Thornton*. Solvolysis Mechanisms, The Ronald Press Company, New York, 1965, pp. 212-2L4-]* The listed rates for acetolysis of l6 and 103 in acetic acid-0-d (see Table 10) bear testimony that this generalization carries over to acetic acid. 0. Reitz, Z. Electrochem., 4-3, 659 (1937). In addition, it was considered desirable to obtain quantitative assessment of the deuterium incorporation in the acetates and tosylates via mass spectrometry. Such analysis of the recovered tosylates showed different degrees of deuteration in the two samples. Unfor tunately , due to unanticipated technical difficulties, it was impossible to establish that the deuterium incorporation in ihe acetates was always 100$.

Since in the solvolysis of the anti-tosylate (10M the enol was the actual solvolyzing species, one might anticipate the syn-tosylate

(103) to react via an analogous enol (117). In view of work^-^ done on syn-7-norbornenyl tosylate (118), which acetolyzes with skeletal rearrangement to give 119 > the enol (117) might be expected to undergo anchimeric assistance to give 120. However, the only product observed upon acetolysis of 103 was the anti-acetate (llh ).

OAc

118 11£

/OTs i— ■» It was extremely difficult to make a quantitative assessment of the acetolysis product of 103 due to the extensive decomposition

OTs Ac O

103

caused by the high temperatures required for solvolysis of 103. After

incomplete reaction (1335 minutes at 200.0°) a v.p.c. analysis with

internal standard indicated 2% of the theoretical yield of 114.

Shorter reaction time (112 minutes at 200.0°) gave a minimum of &jj>

of the theoretical yield of 114;. The identity of the acetolysis product was established by comparing its v.p.c. retention time with

that of an authentic sample of 114 on three different columns.

It was reasoned that any 120 present in the acetolysis product

could be detected by v.p.c. since this compound should have a reten

tion shorter than that for 114. However, on both” polar and nonpolar

columns, analysis of the acetolysis product showed no compounds with

retention times shorter than that for 114.

The fact that no syn-acetate (113) is formed during the acetolysis

of 103 could be interpreted in two different ways. The first possi

bility is that the intermediate which is responsible for the formation

of the anti-acetate (114) is the same as that generated in the ■i i o R. Foltz, Battelle Institute, Columbus, Ohio has generously furnished the mass spectral data discussed in this dissertation. S. Winstein and E. T. Stafford, J. Am. Chem. Soc., 791 505 (195?)• acetolysis of the anti-tosylate (104). The required intermediate

(either 115 or 116 ) could immediately follow from the carbonium ion

122. Two different ways of obtaining 122 are shown below: The tosyl ate (103) either initially solvolyzes (to give 121) and then enolizes, or the tosylate enolizes and then solvolyzes. In either case* the tosylate exophile must be completely removed from the

121 m TsO

o r HO 103 122

HO HO m 116

reactive site before further reaction occurs* and a consequent fully

developed carbonium ion (121 or 122) must ensue at the Cy carbon

atom. One would think that such a carbonium ion developing at the 79 strained carbon atom would be extremely reactive and would not

become free to participate in the proposed mechanisms. The second

possibility to explain the anti product arising from the solvolysis

of 103 appears to be more attractive. The anti-acetate (ll*Q could

arise via an S^2-like process, giving complete inversion at the Cy

carbon atom. Postulating such an inversion process is quite m reasonable, for even the relatively strain-free cyclohexyl system 28 can give 98% inversion under acetolysis conditions.

OTs AcO

o m

It was considered to be of interest to determine the magnitude of the acetolysis rate of the anti-enol (see p. 75)» and to compare this with the acetolysis rate of anti-7-norbornenyl tosylate

(15 ). The relative ratio of these two rates could then be compared

TsO

io n 105 OH with those predicted by both classical and nonclassical theory. The remainder of this discussion section will be devoted to this task.

The approach taken to determine the ratio k /k (i.e., of ion was to study the solvolysis of 1C& in acetic acid-O-d.

As previously discussed, solvolysis of 104 in acetic acid-O-d involves the following mechanistic scheme:

TsO

TsO AcO D + HOTS

We shall first determine the value of k-^> k^» and kg/k^. For I lk convenience, let us define:

The mathematical foundation used in treating the kinetic data is discussed in A. A. Frost and R. G. Pearson, Kinetics and Mechanism, second ed.» John Wiley and Sons, Inc., New York, 1961, pp. 8-26, 160-199. 86

[H] = concentration of 104 (undeuterated tosylate)

[D] = concentration of 124 (deuterated tosylate)

[T] = concentration of 104 and 124 (total tosylate concentration)

[C] = concentration of 125 (measured by concentration of acid, HOTs)

ky = acetolysis rate of H (104) in DOAc at 110.00°

kg = acetolysis rate of D (124) in DOAc at 110.00°

[H ] = original concentration of H

From Table 10 we observe that the in itial rate of 104 in acetic acid-0-d is ^initial = (1*31 - »02) x 10**^ sec.”\ and that the final rate kfinai = (l»94 + .09) x 10 ^ sec.--*-. Since only H and no D is present at time t = 0> then ^initial = k^. Since only D and no H is present at time -* go (t = 300 minutes), evidenced by the leveling of the plot log [C] vs. t (see Graph I), then kfinal = kg. In the solvolysis of

104 in acetic acid-0-d at 110.00°, at any time t

d[G]/dt = kH[H] + kD[D] (1)

Since [T] = [H] + [D]» then substituting into (1)»

d[C]/dt = kH[H] + kD([T] - [H]) (2)

Rearranging (2) to solve for [H],

d[C]/dt - kg[T]

[H] = “kH k------kDk------<3)

Substituting the known values of k^ and kQ into (3)* we get 87

d[C]/dt - 1.94 x 10~5[T] [H] = ------(4) 11.16 x 10-5

We can solve for [H] if we know d[C]/dt and [T] at various times.

We know [T] = [Hq] - [C]. We find d[C]/dt by plotting [C] vs. t to get a graphical slope. We thus solve for [H] at different times t = t .

Now H is lost only by the reaction k^. Therefore,

-d [H ]/d t = k-jLH] (5)

-2 .3 0 3 log[H ] = k ^ t + co n st (6 )

The values for log[H] should thus form a linear plot against t, with the slope being -k^/ 2 . 303. The points do form a linear plot (see

Graph II), from which is obtained

= (1 .9 4 t *08) x 10^ sec."1 (7)

Let us assume that the [E] present in this reaction sequence

is always very low. This would follow if k^ - k2 >:>k^>k^.

Evidence for the validity of this assumption is that in the solvolysis

of 104 in acetic acid or in acetic acid-O-d there is no induction

period. Then at time t = 0, the amount of acid developing is k-^

times (kg/kg + k^)> the fraction of [E] partitioned through k^, i.e.*

k2 kH = kl ( 8) 86

GRAPH II. A plot of los[H] v s . Time

■3 Time, sec. x 10 Substituting (7) and the known value for kH into ( 8) , we g e t

2= = .677 + .038 (9) k2 + k3

kp 4, * = 2.14 _ .037 ( 10) k3

At time t -* 0 0, the amount of acid developing is k^ times

^ 2/^2 + ^ 3)* fraction of [E] partitioned through k2, i.e.,

kD = -- 2 (11) 2 + 3

Substituting (9) and the known value for k^ into (11), we get

kk = (2.88 + . 29) x 10"5 s e c ."1 (12)

As a check for this treatment, we can calculate what the [H]/[D]

ratio w ill be after t = t, and then experimentally determine what

this ratio actually is for a specific time t. Expressions relating

the rate of change of [H] and [D] with respect to time are

-d[H]/dt = k1 [H] (5)

d[D]/dt = -...,.k3 [ h 3 " — ^3 (!3) k2 + k^ L k2 + k3 90

Solving (5)»

[H] = [H0 > ' V ' (lit)

To solve (13)» we first substitute (1*0 into (13) to get

dC D j/dt = k J - k4 — ------[D] (15) x k2 + k3 ° k2 + k3

For convenience let us define

k*i ko k. = --I-?— (16) 2 + k3.

k9kii, k" = £-2----- (17) k2 + k3

Substituting (16) and (17) into (15)» we get

d[D]/dt = k*[i*o]e"’kl'k - k"[D] (18)

Rearranging (18), we get

d[D]/dt + k"[D] = k'C^le”^ (19)

We solve (19) by finding an integrating factor q

6 = e ^ Pdt (20)

For the integration of differential equations by the use of integrating factors, see A. L. Nelson, K. W. Fcflley, and M. Coral, D ifferential Equations, D. C. Heath and Company, Boston, 1952» pp* 39JWL. 91 where (19) assumes the form

d[D]/dt + P[D] = Q (21)

where P = c o n sta n t

Q = f ( t )

Solving for

jk " d t p = e

k " t (22) P = e

Multiplying (19) by (22)

(23)

The solution to (23) is

p[D] = Jp Qdt + Const. (2k)

Solving (2^),

k " t e [D] t k'[H le”^ dt + Const.

ekMt[D] = /e (k" " kl 5t k'[H ]dt + Const. ■J o

k* rTT n.Ck" - kn)t e ^ t D ] = —- LH„Je'‘ 1/ + C onst. (25) (k" - kn) “ ° i>

To find the constant in (25)> we know at t = 0, [D] = 0. Thus,

e k " t[0 ] = [Hn ]e<“ " - + C onst. (k" - kx) L ° k '[H ] C o n st. = ------(k» - kx )

Substituting (26) in (25),

ek ,,t[D] = _____ £ [H ] e (k " - kx ) t k ’ [H0 ] (k" - kj^) 0 ” (i;» . k l )

Simplifying (27),

ek..t[D]. _ k ^ L _ [e(k-. kln _

cd] ■ [e_kit - e-k,,t3

Now the percent deuterium incorporation in recovered tosylate w i l l be

TDl — 1—1------o r [H] + [D]

1 - CH] [H] + [D]

Substituting (14) and (28) into (30), we get 93

1 - (31) k* 1 + [1 - e(*l “ k,,>^ (k" - kx )

The expression (31) gives percent deuterium incorporation in the tosylate recovered at time t = t.

In the actual experiment % tosylate was recovered after 75*5 minutes of reaction time at 110*00°* i.e.* *W 53 x 10^ s e c - The values of k* and k” are* from ( 15 ) and ( 16 )»

k* = (1 .9 ^ x 104 ) -----!S2- ( 32) k2 + k3

k" = (2 .8 8 x IQ"5) (.677) ( 33)

by s u b s titu tin g th e known v a lu e s ( 7), (9)> and (12). The ratio

(k ^ /k 2 + k^) in ( 32) is easily calculated from ( 9) and ( 10) to be

ko , 2 = .325 t .20 (3>0 k2 + k3

Incorporation of (3*0 into (32) and consequent multiplication leads

k* = (1.9*4- x 10J+ ) (-325)

k* = (6 .2 8 + . 60 ) x 10“5 s e c ."1 ( 35 )

k" = (1.96 t . 31) x 10“5 sec.-1 ( 36 ) 94

Substitution of the known values for k*» k", k^» and t into the expression ( 31)>

1______

1 " 1 6*28 X 10"^ r n (19.4-1.96) X 10"5 (4.53 X 103) a ^6 :1 9 : 4 T x W * 1 " e ]

1 -

1n +, ------6 '2 8 LIr-1 - e (17.^) (4.53) X 10-2n ] -1 7 .4 4

1 - ( . 360)[1 - e ' ' ]

1 - 1 - ( . 36 o ) [ l - 2 . 20]

1 - 1 - ( . 360 )(- 1 . 20)

1 - 1------1 + .432

1 L 1.432

1 - .698

.302

30 J2$> (37) The experimental error in (37) can be computed from the expression

(31) by incorporating the error in k ', k", and k^. The maximum deviation in ( 37) occurs when the following values for k*, k"> and k^ are used:

k* = 3.67 x 10

k" = I .65 x lO "'5 (38)

= 1.86 x 10 kl

and

k* = 6 .8 8 x 10”5

k» = 2.27 x 10“ 5 (39)

= 2.02 x 10“^ kl

Substitution of (38) and then (39) into (31) gives

28. 0- 32 . 5 ^ deuterium incorporation (^K))

The actual experimental value obtained from mass spectrometric

studies was

We are now in a position to gain an idea of the relative magni

tude of k^ (see p. 85). We can determine the value of kg if we

know the value of k^» from expression (10). The value of k^ can be

determined if we know the ratio of k^/k^, which is the enolization

equilibrium constant for ~L2h in acetic acid-0-d at 110°. We shall

obtain a relative value for this equilibrium constant by relating it

to the enolization equilibrium constant for camphor. The enolization equilibrium constant for camphor in 70$ aqueous methanol at room temperature is 1.4-0 x 10”^. Enolization equi librium constants for small ring ketones appear to be a function of ring size and resulting strain difference between the ketone and the enol. For instance* the equilibrium constant of camphor lies inter- 117 mediate between those of cyclobutanone and cyclopentanone. One would thus expect the equilibrium constant of 104- in 70$ aqueous methanol to be very close to that of camphor. Furthermore* the equilibrium con stant of 104- should change little as the solvent is changed from 70$

T *1 O aqueous methanol to acetic acid ; the equilibrium constant of a ketone appears to be a function of solvent p o larity ,an(j thus qualitative comparison of the equilibrium constant in different solvents is pos sible. Thus, the enolization equilibrium constant of 104- in acetic acid should be about the same as that of camphor in 70$ aqueous methanol.

It now remains to determine what w ill happen to the equilibrium constant for the process

K at room temperature (4-1) 104- O 10J5 OH

T "I ■ A. Gero, J. Org. Chem., 12* i960 (195^)* 117 A. Gero, J. Org. Chem., 26, 3156 (1 9 6 1 ). Xl8 G. W. Wheland, Advanced Organic Chemistry, second ed., John Wiley and Sons, Inc.,. New York, 1957* P» 588. 119 See Ref. 118, pp. 607 - 609 . 97 as we change it to that for the process

T sO -|\ DOAc o a t 110 (42) * I? 12ft 6 122 OD

The change from process (Al) to process (42) involves a tempera ture factor and an isotope factor. Both of these factors should increase the relative amount of enol in process (42) over that in process (41). The temperature factor should result in more enol in

(42) because the equilibrium constant, related to the temperature by

,, . 120 the expression

k cc e - AE/ RT (43) w ill more nearly approach unity as the temperature T increases.

That the isotope factor will result in more enol in (42) follows from th e fo llo w in g tre a tm e n t. We may d e fin e a new e q u ilib riu m con

stant to be

K = K"/K* (44) which describes the process

120 K. B. Wiberg, Physical Organic Chemistiy, John Wiley and Sons, Inc., New York, 1964, pp. 273-277. 98

T>r> T<;n Ten tr ^

(45)

Now

K”/ k * = q 1-^1 Y10^ o(~h/2kT)/vlQ^ + v121 - v 122 - viq ^) Q124 Q1Q5 (46) where is the internal partition function for the species concerned and where is the vibrational frequency of the C-H, C-D, 0-H, or 121 0-D bond in question. In practice the partition function term is very close to unity, and in any case is much closer to unity than the second term. Thus, the difference in zero-point energy w ill be the main factor causing isotopic discrimination, and whether

K"/K* is greater or less than unity w ill depend upon the second term.

Incorporating the vibrational stretching frequencies 121 of the bonds in question into the second term of (46), we find that K is greater than unity, and that K" is therefore greater than K’, i.e.*. there is more enol content in process (62) than in process (41).

We thus come to the conclusion that the equilibrium constant for the process (62) should be greater than that for camphor in 70 % methanol at room temperature. That is to say, KM is greater than

1.60 x 10~3. Now

(47)

12-*- r . p. Bell, The Proton in Chemistry, Cornell University Press* Ithaca, New York, 1959» pp» 186-187* 99

From the known value of (12) it follows from (47) that k must be less than 2.06 x 10” sec.” . From the known value of kg/k^ (10) it 2 -1 is concluded that kg must be less than 4.4 x 10 sec. . p The most likely source of error for the value of kg < 4.4 x 10" sec.”^ is the value quoted by Gero^^ for the enolization equilibrium 122 constant for camphor. Bell has shown that the enolization equilib rium constants for ketones determined by the titrim etric method tend to be too large* presumably because of impurities present with the ketone. If the equilibrium constant of camphor quoted by Gero is too. large, the final result would be a greater value for kg. Therefore, the qualification must be made that k^ may be somewhat greater than

4.4 x lO-^,Spending upon the reliability of Gero's work.

The rate of acetolysis of anti-7-norbornenyl tosylate (15) in acetic acid-0-d at 110° should be, if anything, greater than 7*21 sec. ^.17,110 The conclusion is thus reached that anti-7-norbomenyl tosylate (15) solvolyzes faster than 123 in acetic acid-0-d by a factor 2 of at least 1.6 x 10 . The reliability of this value depends upon the factor discussed in the previous paragraph. It now remains to deter mine if this value is more in accord with the classical or the nonclassical carbonium ion theory.

If 1£ were to solvolyze to a classical carbonium ion (3^), then one would expect 105^2^ to solvolyze to a classical cation 115. _ _ — R. P. B e ll and P. W. Sm ith, J . Chem. Soc. (B ), 1966, 241. 123 J Strictly speaking, we have shown that 1£ must solvolyze faster than the deuterated enol (123). However, for convenience we shall re fer to the undeuterated enol (105 )t for there is no reason to believe 105 should solvolyze at a rate much different from that of 123. 100

According to classical carbonium ion theory, one might predict 115 to be formed from 105 more readily than 1£ from 1£, because in light

TsO

OH t)-H + m m

of previous work (see pp. 63-64) an oxonium ion should be formed more readily than a secondary carbonium ion. The fact that 105 solvolyzes appreciably slower than 1£ is thus inconsistent with the classical cation theory.

If 1£ were solvolyzing to a nonclassical carbonium ion (1£), then it would be difficult to predict whether 105 would solvolyze to a classical cation (115) or to a nonclassical cation (116), since 105 does not possess the symmetry of 1£. According to nonclassical cation

theory, if 105 were to solvolyze to the nonclassical structure (ll6 ),

then one would perhaps expect 105 to solvolyze more slowly than 15 j

because the electron-withdrawing hydroxyl group would render the

Il-electrons less available for bonding with the Cy carbon atom in 105

than in 1£. On the other hand, if 105 were to solvolyze to the 101

OTs

±Z

OTs

HO HO 116

classical structure (115)» it would be extremely difficult to predict

whether 1£ or 105 would solvolyze faster, as it would be difficult to

choose between the relative stability of 1£ and 115»

One might argue that if the oxonium ion (115) did exist as a

discrete-interm ediate, then he might expect the formation of the tri

cyclic ketone (126) by expulsion of a proton. No such compound was

+ 124

found in the acetolysis of 104. This may indicate either that 115

is not an intermediate, or that 126 is unstable to reaction conditions.

The latter possibility is not unreasonable, as all attempts to syn- r 124 thesize 126 have been unsuccessful.

P. R. Story, G. H. Whitham; private communication to P. G. Gassman. 102

The conclusion is that the solvolytic data is inconsistent with the classical interpretation of the solvolysis of 15 _» but can be reconciled with the nonclassical interpretation. The exact nature of the intermediate obtained from the enol (105)> however* is not known. EXPERIMENTAL

Boiling points and melting points are uncorrected. Mass spectral analyses were performed on an AEI MS-9 mass spectrophotometer. Nuclear magnetic resonance spectra were obtained on a Varian model A-60 spec trometer. Near infrared spectra were obtained on a Cary Model 1^ spectrophotometer. Vapor phase chromatographic work was performed with an Aerograph HyFi Model 600 and an Aerograph "Autoprep" Model

A-?00. Elemental analyses were obtained from, the Scandinavian Micro- analytical Laboratory, Herlev, Denmark, and the Mickroanalytisches

Laboratorium im M ax-Planck-Institut fur Kohlenforschung, Mulheim,

Germany. All per cent composition determined by v.p.c. have been corrected for the weightjarea factor by utilizing an internal standard,

except where otherwise designated. All v.p.c. analyses were done on

15% Carbowax 20M on Chromosorb W unless otherwise designated.

R eagents

Dry tetrahydrofuran was prepared by distillation from lithium

aluminum hydride in a dry atmosphere. Dry ether was obtained from

Mallinckrodt Chemical Works, and was not re-dried before use. Hexa-

chlorocyclopentadiene, obtained from Matheson Coleman and Bell, was

not purified before use. Bicyclo[2.2.1]heptene» obtained from

Enjay Chemical Company, was sublimed before use. Bicyclo[2.2.1]- o heptadiene, obtained from Matheson Coleman and Bell, was stored at 5

and was not redistilled before use. 103 lot*

5»5-Dimethoxy-l,2,3 ,4-tetrachloro- cyclopentad iene (51)58

A 3-1. , three neck flask fitted with a condenser, an addition funnel, and a mechanical stirrer was charged with 254 g. of hexachlorocyclopentadiene (J50) and 800 ml. of commercial methanol. A solution of 120 g. of potassium hydroxide in 600 ml. of methanol was added dropwise with stirring over a period of two hours. The reaction mix ture was stirred for two additional hours and then poured over 3 1. of chopped ice. After the ice had melted, the mixture was extracted with three 250-ml. portions of methylene chloride. The combined ex tracts were dried over anhydrous magnesium sulfate, concentrated to a yellow syrup under reduced pressure, and distilled to yield 213 g*

(86$) of a viscous yellow-tinted oil, b.p. 79-91° (0.6o mm.).

This reaction could be scaled up fivefold if four hours were taken for the addition of the methanolic base. No danger of an un controlled exothermic reaction existed when the temperature of the reaction mixture was maintained at 50-60°.

7,7-Dime th o x y -1 ,2 ,3 ,4 - te tr a c h lo r o - bicyclo[2.2.1~lhept-2-ene (52)39

A large pyrex gas washing bottle (Corning No. 31750) was fitted with a condenser and a drying tube. A slow stream of nitrogen and

ethylene was passed through the sinter as 213 g* of was placed in

the bottle. The reaction vessel was heated at 180-190° for six

hours as ethylene was bubbled through; the color of the liquid

changed from yellow to reddish-brown. The reaction mixture was

cooled and distilled to yield 185 g. (78$) of a yellow syrup, b.p.

75-90° (0.02 mm.). 105

7 ,7-Dimethoxybicyclof2.2.1Iheptene (53)

A 3-1., three neck flask equipped with a Hershberg stirrer, a condenser, and a constant pressure dropping funnel was fitted with a heating mantle and a nitrogen inlet capable of maintaining a slightly positive nitrogen pressure. The flask was charged with

1500 ml, of commercial tetrahydrofuran,. 130 g. of sodium metal chopped into l/4—inch cubes, and 150 g. of t-butyl alcohol. This mixture was stirred vigorously and brought to gentle reflux under a nitrogen atmosphere. When refluxing commenced, 106 g . o f $2 was added over a period of 2 hrs. After stirring under reflux for 38 h o u rs, the reaction mixture was cooled, filtered through wire screen to remove the unreacted sodium,, and refiltered through Celite. The reac tion mixture was carefully mixed with 2 1 . of chopped ice and 500 m l. of ether.. The aqueous phase was separated and the organic phase was washed with 500 -rol* portions of saturated sodium chloride solution until the washings were clear. The organic solution was dried over anhydrous magnesium sulfate, concentrated by fractional distillation, and distilled to yield g* ( 65 %) of a colorless liquidb.p.

61-71° (18 mm.), n Q25 1.4-598.

Hydrobo ratio n ^ of 7 ,7-dimethoxy- bicyclor2.2.11heptene (53)

Diborane, generated by the dropwise addition of 3.9 g« of sodium borohydride (103 ro moles) in 100 ml. of diglyme to 53*1 g. o f k7% boron trifluoride etherate (14-5 m moles) in 100 ml. of diglyme over a period of 3 hours, was passed through a solution of 10.1 g. (66 m moles) of 52. in 150 ml. of dry tetrahydrofuran at 0°. To the latter lo6 solution was added 10 g. of sodium hydroxide in 55 ml. of water and

35 ml. of 30% hydrogen peroxide. After 1 hr. of stirring, the re sulting two phases were separated and the aqueous phase was extracted with 50 ml. of ether. The organic phases were combined and diluted with 150 ml. of ether. The resulting aqueous layer was drawn off, and the ether phase was washed with 50 ml. of water, dried with an hydrous magnesium sulfate* concentrated under vacuum* and distilled to y ie ld 8.76 g . , b .p . 60 - 93° (1*3 mm.), of a clear oil which con sisted of 75$ of the exo-alcohol (54)* 20% of the endo-alcohol (55)> and 5% of an unidentified component, resulting in a 74% overall yield for 54 and 555 , combined.

7 ,7-Dimethoxybicyclof2.2.1Iheptane- exo-2,3-epoxide (^7) fiR To a solution of 0.173 mole of perbenzoic acid in 1 1. of chloroform was added 21 .56 g. of jQ with stirring. After standing at 5° for seven days the solution was washed with two 125 -ml. portions of 10% sodium hydroxide and two 150 -ml. portions of water. The organic layer was dried over anhydrous magnesium sulfate* concentrated under reduced pressure, and distilled to yield 20.92 g. ( 87%) o f 6 7 * b.p. 53-56° (0.4 mm.). Redistillation gave an analytical sample, nD25 1.4738.

Anal.125 Calcd. for C, 6 3 . 51 ; H, 8.29

Found: C, 6 3 .4 5 ; H, 8.37*

An alternate procedure for the synthesis of 67 involved the more

convenient reagent m-chloroperbenzoic acid (EMC Corporation). A

125 The w riter is indebted to P. G. Pape, M.S., The Ohio State University, 1962, for this elemental analysis. 107 solution of 151 g* of m-chloroperbenzoic acid (assay: minimum of 85$) in 1500 ml. of chloroform was added to a solution of 94.76 g. of 53 in 800 ml, of chloroform with stirring over a period of 45 minutes.

The reaction was mildly exothermic but no external cooling was neces sary. After stirring for a subsequent period of 4.5 hours, the solution was washed with two 1000-ml. portions of 10$ aqueous potassium hydroxide solution and then with two 1000-ml. portions of water. The organic phase was dried over anhydrous magnesium sulfate, concentrated under reduced pressure, and distilled to yield 91*3? g» ( 87$) o f 6 £» b .p .

6 l-6 4 ° (.1 5 mm.).

7,7-Dimethoxybic.yclor 2.2.llheptan- exo-2-ol (54)

To a slurry of 18.28 g. of lithium aluminum hydride in 600 m l. of dry tetrahydrofuran was added dropwise 22.76 g. of 6£. The reac tion mixture was stirred under reflux for 12 days. The tetrahydrofuran solution was cooled to 0° and 70 ml. of water was added dropwise over a period of several hours with vigorous stirring. The reaction mixture was then stirred at room temperature for 2 hr., filtered, and concen trated under reduced pressure. The remaining liquid was distilled

to yield 21.70 g. (94$) of pure exo-alcohol (54), b.p, 62-72° (0.6 mm.).

Redistillation gave an analytical sample, b.p. 50.5-51.0° (0.35 mm.),

nD27 1.4676.

Anal. Galcd. for G> 6 2 . 76 ; H, 9.36

Found: C, 6 2 . 7 6 ; H, 9*43* 108

7 ,7-Dimethoxybicyclc>r2.2.11- heptan-2-one (66) 126

To 170 ml. of pyridine at 0° was added with stirring 14.4 g. of

chromium trioxide. When the pyridine-chromium trioxide complex had

formed, 4.98 g. of ^4 was added. The reaction mixture was stirred

f o r 15 hr., poured into 1 1. of water, and continuously extracted with

ether for 2 days. The resulting ethereal solution was dried over an

hydrous magnesium sulfate, concentrated under reduced pressure, and

distilled to yield 4.10 g. ( 83$) of 66, b.p. 52 - 56 ° (0 .2 mm.). Re

distillation gave a pure product 1.4660, which crystallized when

cooled. Recrystallization from 35-45° petroleum ether and subsequent

sublimation yielded an analytical sample,, m.p. 33-35°«

Anal. Calcd. for : G, 6 3 , 5 1 ; H» 8.29.

Found: C, 63*37; H, 8 . 3 2 .

2,4-Dinitrophenylhydrazone of 7x7- dimethoxybicyclof 2.2.1Iheptan- 2-one ( 60 )

A sample of 66 (0.11 g.) in 3 ml* of O .25 M 2,4-dinitrophenyl- 127 hydrazine in phosphoric acid and ethanol immediately yielded a

precipitate. After standing in the reagent for 8 hr., the crystals

were collected and recrystallized from 95 % ethanol four times and

eluted through Grade IV neutral Woelm alumina with benzene to give

an analytical sample as fine yellow needles, m.p. 208. 5 -? 0 9. 0°«

■12 ^ 1—1 m S arett oxid ation : G. I . Poos, G. E. Arth,. R. E. B eylor, and L. H. S arett,, J . Am. Chem. Soc.» £5* ^ 2 (1953)* 127' L. F. Fieser, Experiments in Organic Chemistry, third ed., D. C. Heath and Company, Boston, 1957> p* 316. 109 Anal. Oalcd. for c15Hl8\0 6: C, 51.^2; H, 5 .18; N, 1 5 .9 9 .

Found: C, 51. 23; H, 5 .2 1 ; N, l6 .0 4 .

7 i 7 -.Dime th o x y b icy clo f 2.2 .1 1 - heptan-endo- 2-ol (55)^~2^

A 500 m l., three neck flask fitted with a short Vigreaux column and a dropping funnel was filled with 11.96 g . o f 66 ., 18.6 g . of crushed aluminum isopropoxide» and 600 ml. of isopropyl alcohol.

This mixture was stirred and distilled slowly until no acetone could be detected in the distillate (5 hr.). The mixture was then concen trated under vacuum to yield a residue which was dissolved in 1100 ml. o f 5f> sodium hydroxide solution. This solution was extracted with five

250-ml. portions of ether. The combined ethereal extracts were dried over anhydrous magnesium sulfate, concentrated under reduced pressure, and distilled to yield 10.95 g« (91^>) of the endo-alcohol (.55,)» b.p.

75-79° (1.0 mm.), of which the exo epimer (5+) comprised 2$. Redistil

lation of this viscous product gave an analytical sample, b.p. 77°

(0 .2 8 mm.), nD29 I . 4767.

Anal. Calcd. for Cy^Oy C, 62 . 7 6 ; H, 9.36

Found: C, 6 2 .6 3 ; H, 9 .^ 0 .

It was discovered that the rate at which the solvent was distilled

off was critically important. When the solvent was distilled at a slow

rate, considerable amounts of the exo epimer ( 5 ^) r e s u lte d . The amount

of was h e ld to a minimum when th e so lv e n t was removed ra p id ly (2-3

ml. per minute). 128 Meerwein-Ponndorf-Verley reduction: A. Vogel, Practical Organic Chemistry, Longmans, Green and Company, London, 1956* pp. 882-6. 110

Oxidation of 7>7-dimethoxybicyclof2.2.1]- heptan-endo-2-ol (55)

Addition of 0.27 g. of ^ to a mixture of 0.87 g. of chromium, trioxide in 5 ml* of pyridine gave, after dilution with 75 ml* of water, extraction with three 25 -ml. portions of ether, drying over

anhydrous magnesium sulfate, and concentration under reduced pres

sure *. an oil which proved by v.p.c. to consist exclusively of pyridine

and the ketone (66). The infrared spectrum of the ketone collected

by preparative v.p.c. was identical to that of 66.

Reduction of 7 >7-dime thoxybicyclof.2.11- heptan-2-one (66) with sodium-ethanol

Sodium-ethanol reduction of the ketone (66) was always incomplete,

even with a large excess of sodium. A typical run consisted of adding

11.0 g. of sodium in small pieces to a stirring solution of 2.22 g.

of the ketone (66) in 100 ml. of absolute ethanol over a period of

5 hr. This solution was poured over 1 1. of ice water and extracted

with three 300-ml. portions of ether. The combined extracts were

washed with 50 ml. of water, dried over anhydrous magnesium sulfate,

concentrated under reduced pressure, and distilled to yield 1 .2 7 g.

of a clear liquid, b.p. 70-73° (0.6 mm.), which consisted of 16% of

the endo-alcohol ( 55)» 3 &j° of the exo-alcohol (J4), and Lf-6% of unreacted

ketone (66).

Reduction of 7>7-dimethoxybicyclor2.2.11- heptan-2-one (66) with lithium aluminum hydride

To a stirring slurry of 0.32 g. of lithium aluminum hydride in

5 ml. of dry ether at 0° was added 0.32 g. of the ketone (66) in Ill

5 ml. of dry ether. After 2 hr. of stirring, 2 ml. of water was added dropwise. After 45 minutes of subsequent stirring» filtration and concentration under reduced pressure gave 0.25 g . (77$) of an oily residue which consisted of a 3:1 ratio of the exo-alcohol (54) to the endo-alcohol ( 55 )•

Reduction of 7,7-dimethoxybicyclof2.2.11- heptane-exo- 2 , 3-ep o x id e ( 67 ) w ith lithium-ethylamine-^9

To a solution of 2.63 g. of 62, in 47 ml. of anhydrous ethylamine a t 0° was added with stirring 1 .0 g. of lithium in small pieces over a duration of 10 minutes. After 45 minutes the deep blue reaction mixture was quenched with 100 ml. of water and extracted with four

100-ml. portions of methylene chloride. The combined extracts were washed with two 50 -ml. portions of water, dried over anhydrous magnesium sulfate, concentrated under reduced pressure, and distilled to yield

O.63 g. (24$) of the exo-alcohol (54), b.p. 53-62° (0.5 mm.). Sub limation of the pot residue gave 0.30 g . (15 $) o f exo- 2- syn- 7-b ic y c lo -

[2.2.1]heptanediol ( 68 ), m.p. 178.0-178.5°. The infrared spectrum was identical to that of an authentic sample, and the mixed melting

point was undepressed.

exo-2-Acetox.y-7, 7-dimet hoxybic.vclo- [ 2 . 2 . 1 ]heptane ( 6 2 )

To a solution of 4.30 g. of the exp-hydroxyketal (54) in 40 ml.

of pyridine was added dropwise 2 .8 g. of acetyl chloride with

swirling. After standing for 15 minutes, the resulting mixture was

poured over 400 ml. of water and extracted with 4 50 -ml. portions

j , j, Hurst and G. H. Whitham,.J. Chem. Soc., 1963 , 710. 112 of ether* The combined extracts were dried over anhydrous magnesium

sulfate* concentrated under reduced pressure, and distilled to give

4 .7 6 g . ( 89$) of 62, b.p. 83- 85 ° (0 .6 mm.). Redistillation gave an

analytical sample,, b.p. 104-105° (6 mm»)> 1.4587. 125 Anal. Calcd. for C^H^gO^: C, 61.66; H> 8.47

Found: G, 6 1 .5 1 ; H» 8. 5 0 .

endo-2-Acetox.y-7>7-dimethoxybicyclo~ r2.2.11heptane ( 63 )

Treatment of the endo-hydroxyketal (55) as above gave an 89$ y ie ld ( 1.66 g.) of the endo-acetate. ( 63 ), which crystallized when

cooled. An analytical sample was prepared by redistillation, b.p.

67° (0.3 mm.). A sample recrystallized from hexane gave white

flakes, m.p. 29. 0- 30. 1° .

Anal. Galcd. for C, 61.66; H, 8.47

Found: C, 6 1 .3 9; H, 8 .4 9 .

7 ,7-Dimethoxybicyclof 2.2.1)heptan- exo-2-ol p-toluenesulfonate (64)^2

To a solution of 1.71 g* of in 11 ml. of pyridine cooled to

0° was added with stirring 2.10 g. of £-toluenesulfonyl chloride over

a period of 20 minutes. The resulting solution was allowed to stand

at 5° for 22 hours and then was mixed with 65 m l. o f w a te r. The to s y l

ate, which immediately crystallized, was washed with three 25 -m l.

portions of water and dried in a vacuum desiccator overnight to yield

2.52 g . ( 78$) of white crystals, m.p. 83-87°. Recrystallization from

hexane gava- an analytical sample, m.p. 93«0- 9 3* 6 ° . 113 Anal. Calcd. for C^H^Q^Ss C, 58.88; H, 6*79; S, 9.82

Found; C, 58.62; H, 6.90; S, 9.90.

7 >7-Dimethoxybicyclof2.2.llheptan- endo-2-ol p-toluenesulfonate (6~

To a solution of 3*17 g. of ^ in 20 ml. of pyridine cooled to 0° was added 3*90 g. of js-toluene sulfonyl chloride over a period of 5 minutes. The resulting solution was allowed to stand at 5° for 12 hours. The solution was mixed with 300 ml. of water and extracted with three 25-ml. portions of chloroform. The combined extracts were dried over anhydrous magnesium sulfate and concentrated under reduced pressure to give a crystalline product. Recrystallization from hexane gave 3*38 g. (. 56 $) of clear cubes,m.p. 93«2-9^.0°.

Anal. Calcd. for C^ft^O^S: C, 5 8 . 88; H> 6 .7 9 ; S» 9*82

Found; C, 5 8 .9 0 ; H, 6 .80; S , 9 . 6 7 . exo-2-Hydrqxybicyclo[2.2.11- h e p ta n - 7-one ( 5 o)

A mixture of 0.68 g. of the exo-hydroxyketal (jj>4) "Snd 5 ml. of 5$ sulfuric acid was stirred vigorously for 16 hr. The resulting emul sion was extracted with three 10-ml. portions of ether. The combined extracts were dried over anhydrous magnesium sulfate and concentrated under reduced pressure to give 0*39 g* ( 78$) of a yellow-tinted oil.

Attempted distillation of a sample of this material resulted in con siderable decomposition. Sublimation and preparative vapor phase

chromatography also gave decomposition. Due to the difficulties in

purification, a satisfactory elemental analysis was not obtained. 114 exo-2 - Ace toxybicyc lo[2.2.1~l- h e p ta n - 7-one ( 6 p)

Freshly prepared from I .30 g. of ^4 was dissolved in 10 ml. of pyridine. To this solution (cooled to 0°) was added dropwise

1*0 g. of acetyl chloride with swirling. After the mixture had stood a t 0° overnight^ it was poured over 50 ml. of water and extracted with four 20-ml. portions of ether. The combined extracts were dried over anhydrous magnesium sulfate* concentrated under reduced pressure* and distilled to yield 0 .2 9 g . (23$ overall) of 6 o> b .p .

59-62° (0.20 mm.). Preparative v.p.c. (15$ Didecylphthalate on

Chromosorb P) enabled an analtyical sample to be isolated* b.p. 65 °

(0.03 mm.), nD 1.4690.

Anal. Calcd. for C, 64.27; H, 7*19

Found: C> 64.02; H> 7*32. exo-2-Acetoxybicyclor 2. 2 .llheptan- 7-one ( 60 ) via the hydrolysis of ex o - 2-A3gtoxy- 7 ,7-dimethoxybicyclo- [ 2 . 2 . 1 ]heptane (. 6 2 )

A mixture of 3*99 g. of 62 and 20 ml. of 5$ sulfuric acid was stirred vigorously for 6 hr., followed by extraction with four 10-m l. portions of ether. To the combined extracts was added 4 ml. of pyri dine. The extracts were; dried over anhydrous magnesium sulfate, concentrated under reduced pressure, and distilled to yield 0*94 g. of

6 0 , b .p . 55 ° (0.02 mm.), with a small amount of the further hydrolyzed product jji 6 . Continuous extraction of the aqueous phase with ether gave an additional 1.59 g . of 60 for a cumulative yield of 81$ . 115 exo-2-Hydroxybicyclof 2.2.llheptan- 7-one p-toluenesulfonate (58)^3

A sample of freshly prepared from 1*93 g« of 54- was converted 62 . to its tosylate according to the procedure of Tipson, as for 64 and

65 above. After standing for 21 hr. at 5°> the purple solution was

poured over 50 ml. of water and extracted with four 15 -ml. portions

of chloroform. The combined extracts were washed with two 10-m l.

portions of 20$ sulfuric acid and then with two 10-ml. portions of water, dried over anhydrous magnesium sulfate, and concentrated under

reduced pressure to a pink syrup. Recrystallization from ether-hexane

gave 1.13 g* (36$ overall) of the tosylate, m.p. 72.. 4-73.0°.

Anal. Calcd* for C^tt^O^S: C, 59*98; H, 5.75; S% 11.44

Found: G, 60.04; H> 5 . 7O; S, 11*4-0.

endo-2-Hyd roxybi c yclo f2.2.11- h e p ta n - 7-one (57)

A mixture of 1.16 g. of the endo-hydroxyketal (55) and 8 m l. of

5$ sulfuric acid was stirred vigorously for 23 hr. The resulting

solution was extracted with six 10-ml. portions of ether. The com

bined extracts were dried over anhydrous potassium carbonate and

magnesium sulfate and concentrated under reduced pressure to yield

0 .6 6 g . (? 8$) of a clear syrup which could not be obtained pure because

of rapid decomposition.

endo-2-Acetoxybicyclor 2 . 2 .1 1 - h e p ta n - 7-one ( 6 l ) ^

Freshly prepared Tro 1*1 O.63 g. of 55. was dissolved in 8 m l. o f

pyridine. To the solution was added dropwise 0.62 g. of acetyl 116 chloride with swirling. After standing at 0° for 14 hr., the pink solution was poured over 50 ml. of water and extracted with three

25 -ml. portions of ether. The combined extracts were washed with

10 ml. of water, dried over anhydrous magnesium sulfate, concentrated under reduced pressure, and distilled to yield 0.26 g. ( 31$ overall) of a clear oil, b.p. 80° (1 mm.). endo-2-Hydroxybicyclor 2 >2.1~lheptan- 7-one p-toluenesulfonate (59)

Freshly prepared fi*om 3*42 g, of ^ was dissolved in 23 ml. of pyridine. To this solution at 0° was added with stirring 4.50 g. of £-toluenesulfonyl chloride. The resulting solution was allowed to stand at 5° for 65 hr. The purple reaction mixture was poured over

750 ml. of water and extracted with three 100-ml. portions of chloro form.. The combined extracts were dried over anhydrous magnesium sulfate, concentrated under reduced pressure to a purple syrup, and chromatographed through silica gel with 10$ ether in benzene to give

2 .6 3 g. (47$ overall) of an alcohol-free clear oil. Due to the extreme difficulty encountered in rigorously purifying this compound, a satis factory elemental analysis could not be obtained. Titrimetric data indicated a purity of 88.9$ on the material used in the rate measure m ents .

Reduction of exo-2-acetoxybicyclor2.2.l1- h e p ta n - 7-one ( 60 ) with lithium aluminum hydride

A mixture of 20 mg. of 60 and 20 mg. o f lith iu m aluminum h y d rid e

in 20 ml. of dry ether was stirred at room temperature for 4 hr. The pro duct was worked up and isolated in the usual manner to yield exo-2-syn 7-norbornanediol (68), identical to an authentic sample. 117 57 7"Ketobicyclc>r2,2.11heptene (75)

A mixture of 45*86 g. of 7»7-

7 »7-exo-2-Trimethoxybicyclo- [ 2 . 2 . 11heptane ( 80)^?

A mixture of 4.02 g. of the exo-hydroxyketa1 (54) * 5 g* of sodium hydride* and 100 ml. of dry tetrahydrofuran was stirred under reflux for 22 hr. The stirring mixture was allowed to cool, and 25 g. of methyl iodide was added. After 6 hr. of stirring at room temperature,

10 ml. of water was added cautiously. The resulting mixture was washed with saturated sodium chloride solution until no more solids were carried out with the washings, dried over anhydrous magnesium sulfate* concentrated under reduced pressure, and distilled to yield 3.94 g.

(91$) of a clear oil, b.p. 73- 77° (1*5 mm.). A portion was purified via preparative v.p.c. (15$ Didecylphthalate on Chromosorb P) and subsequent distillation to give an analytical sample, b.p. 75 ° ( 1 *^ mm.)» 1*4581.

Anal. Calcd. for ctl0H18°3: C* ^ * ^ 9 ; H, 9*74.

Found: C, 64.55* H, 9*75* 118 exo-2-Methoxybicyclor2.2.11- heptan-7-one (78)

A mixture of 2.86 g. of 80 and 20 ml. of 5$ sulfuric acid was stirred vigorously at room temperature for 18 hr. and then extracted w ith two 50 -ml. portions of ether. The combined ethereal extracts were dried over anhydrous magnesium sulfate, concentrated under reduced pressure, and distilled to yield 1 .9 6 g. (91$) of a clear liquid, b.p.

78-80° (1.5 mm.). A portion was purified via preparative v.p.c. (15$

Didecylphthalate on Chromosorb P) and subsequent distillation to give

O an analytical sample* b.p. 80-81 (1.4 mm.), n^ 1.4690.

Anal. Calcd. for CgH^OgS C, 68.34; H, 8 . 6 3 .

Foundi C, 6 8 . 5 6 ; H, 8 . 6 3 . exo-2-5yn-7-Bicyclor2.2.lV heptanediol (68)98

To a stirring solution of 560 ml. of 90$ formic acid and 144 ml. o f 30$ hydrogen peroxide at 0° was added 120 g. of norbornene (77) over a period of 1 hr. After the norbornene had dissolved, the solu tion was allowed to come to room temperature and stirred for 24 hr.

A pinch of platinum black was added and stirring was continued for

1 hr. The solution was concentrated under reduced pressure to ca.

100 ml. of volume and cooled to 0°. A solution of 150 g. of sodium hydroxide in 300 ml. of water was added with stirring over a period o f 15 minutes. After ensuring that the pH was above 7, the mixture was stirred at room temperature for 12 hr. and extracted with two 200- ml. portions of ether. The combined ethereal extracts were dried over anhydrous magnesium sulfate and concentrated under reduced pressure to a dark oil. The oil was stirred thoroughly with 500 ml. of benzene. 119

The desired product (68) was precipitated from the benzene extract by the slow addition of petroleum ether with stirring to give 48 g. (35$) of nicely crystalline white solid. Vapor phase chromatography always indicated the presence of ©co-2-anti-7-norbornanediol ( 106 ) up to a percentage of 20$. This undesired epimer could not be removed by column chromatography on a conveniently large enough scale; therefore, the 68 used in subsequent reactions always contained a portion of 106 . exo~2-anti-7-Bicyclof 2.2.11- heptanediol ( 106)99

This compound was obtained completely free of the syn epimer (68) by the following sequence.

130 7-t-Butoxybicyclof 2.2.1~lheptadiene

A 1-1*, three neck flask equipped with a condenser, a constant pressure dropping funnel, and a Hershberg stirrer was fitted with a heating mantle and a nitrogen inlet capable of maintaining a slightly positive nitrogen pressure. The flask was charged with 153 g. of bicyclo[2.2.1]heptadiene, 500 ml. of benzene, and 0*350 g. of cuprous bromide. The contents were stirred and brought to gentle reflux under

a nitrogen atmosphere. Over a period of 5 hr. a solution of 122.5 g.

of t-butyl perbenzoate (Wallace & Tiernan, Inc.) in 100 ml. of benzene

was added dropwise* After stirring under reflux for an additional

hour, the solution was allowed to cool. The solution was washed with

potassium carbonate solution until effervescence ceased and the blue-

green color was extracted out. The resulting solution was dried over

130 P. R. Story, J. Org. Chem., 26, 287 (.1961 ) 120 anhydrous magnesium sulfate, concentrated by careful distillation* and distilled to yield 2 5 A g . ( 25 $) of a clear liquid, b.p. 7^-80°

(22 mm.)•

130 7-Acetoxybicyclo[2.2.llheptadiene

A solution of 25*39 g* of 7-t-butoxybicyclo[2.2.1]heptadiene,

254 ml. of glacial acetic acid, and 50.8 ml. of acetic anhydride was allowed to stand at room temperature for 0*5 hr. After being

cooled to 0°, the solution was poured rapidly with vigorous swirling

into 3 ^ g. of 70$ perchloric acid previously cooled to 0°. A red- brown color immediately ensued. After 1 minute of swirling in an ice

bath, the mixture was poured over 850 ml. of ice water; after brief

shaking, the red color disappeared. The mixture was extracted with

four 50-ml. portions of chloroform. The combined extracts were washed

generously with saturated sodium bicarbonate solution, with water, and

with saturated sodium chloride solution. The organic phase was dried

over anhydrous magnesium sulfate, concentrated by fractional d istil

lation, and distilled to yield 17*38 g* (.75$) of a clear oil, b.p.

75-83° (18 mm.).

anti-7-Hydroxybicyclo[2.2.llheptene tetrahydropyranyl ether

To a stirring slurry of ?.5 g. of lithium aluminum hydride in

150 ml. of dry ether cooled to 0° was added dropwise a solution of

17*38 g. of 7-a-cetoxybicyclo[2.2.1]heptadiene in 60 ml. of dry ether

over a period of 1 .5 h r . The m ixture was s t i r r e d a t room tem perature

for 18 hr. and then cooled to 0°. Water (30 ml.) was added dropwise.

A fte r 5 hr. of stirring, the mixture was filtered. The filtrate was 121 concentrated under reduced pressure- to yield anti-7-hydroxynorbomene as a wet solid Without further purification the solid was dis solved in 20 g. of dihydropyran with a drop of concentrated hydrochloric acid. After standing at room temperature for 15 hr., the solution was mixed with 100 ml* of ether % washed with dilute sodium hydroxide solu tion, washed with water, dried over anhydrous magnesium sulfate, con centrated under reduced pressure, and distilled to yield 1 8 .2 g . (81$) of a clear o il, b.p. 81-88° (1*5 mm.).^ exo-2-anti-7-Bicyclof 2.2.11- heptanediol ( 106 ) ^

Diboranegenerated by the dropwise addition of 255 g» of ^7$ boron trifluoride etherate to a stirring slurry of 22 g. of sodium borohydride in 400 ml. of diglyme over a period of 8 hr., was swept by a slow stream of nitrogen into a stirring solution of 81.1 g. of anti-7-hydroxybicyclo[2.2.l]heptene tetrahydropyranyl ether in 500 m l. of dry tetrahydrofuran cooled to 0°. To the latter solution was added dropw ise 135 ml* of water,. ^0 g. of sodium hydroxide in 80 ml. of water, and 205 nil. of 30$ hydrogen peroxide. The stirring mixture was allowed

to i-rarm to room temperature. After stirring for ^ hr., the organic

phase was removed and washed with saturated sodium chloride solution;

the remaining aqueous layer was extracted with ether. All organic

phases were combined, dried over anhydrous magnesium sulfate, and con

centrated under reduced pressure, to a clear syrup. The syrup was

dissolved in 500 ml. of commercial methanol with 0*5 ml. of concen

trated hydrochloric acid. The resulting solution was refluxed for

5 hr. and then concentrated under reduced pressure to a wet solid. 122

The solid was mixed with a small amount of carbon tetrachloride and heated on a steam bath until nice crystals separated. These crystals were collected and sublimed to yield 35*19 g* (66$) of the desired d i o l ( 106 ). syn-7-Hydroxybicyclor 2.2.11- heptan-2-one (107)100,131

To a stirring solution of 12.64 g. of 68 in 300 ml. of reagent grade acetone cooled to 0° was added 37 ml. of 6 M chromic acid over a period of 1*5 hr., with precautions taken to exclude external moisture. After stirring for a subsequent 0.5 hr.,, the supernatant liquid was decanted,, and the residue was washed thoroughly with acetone.

The combined acetone phases were dried with anhydrous magnesium sulfate and concentrated under reduced pressure without external heating. The

resulting residue was dissolved in 500 ml. of ether, washed with three

15 -ml. portions of saturated sodium bicarbonate solution and with

15 ml. of water, dried over anhydrous magnesium sulfate, and concen

trated under reduced pressure to a yellow oil which crystallized from

cyclohexane. Recrystallization from cyclohexane gave 2.59 g* of white

c r y s t a l s , m .p. 1*4-0-155°• Vapor phase chrom atography in d ic a te d t h a t

the product was a mixture of 68 and 103 in the respective ratio of

4 0 :6 0 .132

Attempts to increase the percent composition of 103 in the

■^1 Jones oxidation: I. Heilbron,. E. R. H. Jones, and F. Sond- heimer, J. Chem. Soc.> 1949> 604. 132 Correction has not been made for the areajweight ratio by inclusion of an internal standard. 123 product by increasing the amount of chromic acid proved to be dis astrous*, as extensive decomposition resulted in an uncharacterizable product* syn-7-Hydroxybicy clo [ 2 . 2 .1 ~lheptan-2-one p-toluenesulfonate (103) via Jones oxidation of e3Co-2-syn-7-bicyclo- f 2 . 2 «llheptanediol ( 08)

A sample of 68 was oxidized by the above procedure to yield 6 .1 g .

of crude 107» The crude product was dissolved in 40 ml. of pyridine and cooled to 0°. With stirring was added 14.1 g. of £-toluenesulfonyl chloride. The resulting solution was allowed to stand at 5° Tor 41 hr.

The resulting pink solution was mixed with 500 ml. of water and extracted with four 50-rol. portions of chloroform. The combined extracts were dried over anhydrous magnesium sulfate and concentrated under reduced pressure to a purple oil. The oil was decolorized with charcoal in

ether to yield 7*3 g» of a yellow oil* which was dissolved in 73 nil.

of 95$ ethanol. To this solution were added 4.28 g. of Girard’s "T"

reagent^'*’ (Arapahoe Chemicals, Inc.) and 1 ml. of glacial acetic acid.

The resulting solution was refluxed for 4*5 hr., cooled, and mixed

thoroughly with a mixture of 400 ml. of water and 400 ml. of ether.

The aqueous layer was drawn off and to it was added 1 ml. of concen

trated hydrochloric acid. The aqueous phase was heated on a steam

b a th f o r 3*5 hr., cooled, and extracted with two 200-ml. portions of

ether. The combined extracts were dried over anhydrous magnesium

sulfate and concentrated under reduced pressure to a yellow oil. Hot

hexane washings gave, when cooled, 1.30 g . ( 2 . 5 $ overall) of white

crystals. Recrystallization from hexane gave an analytical sample,

m.p. 63.8-64.6°. 12k

Anal. Galcd. for C1^H16°4S: G, 59‘98*’ H» S * 11 Found: C, 59.93; H, 6.05; S,. 11.6l.

Subsequent attempts to prepare 103 by this procedure resulted in the allotrope, m.p. 95*0-95*5°• syn-7-Hydroxybicyclor 2.2.llheptan-2-one p-toluenesulfonate C'l03) via chromic acid-water oxidation of exo-2-syn- 7-bicyclor 2.2.1Iheptanediol (68)

To a stirring solution of 55 g« of 68 in 250 ml. of water cooled o to 0 was added 210 ml. of 6 M chromic ^cid over a period of 0.5 hr.

After 15 minutes of additional stirring, 10 ml. of isopropyl alcohol was added. After stirring for a further 15 minutes, the mixture was extracted with four 100-ral. portions of ethyl acetate. The combined extracts were dried over anhydrous magnesium sulfate and concentrated under reduced pressure to a brown oil. Formation of the tosylate and

subsequent Girard separation as above gave after recrystallization from

hexane 1.78 g. (.1.5$ overall) of fine white needles, m.p. 93*0-9^.0°.

exo-2-syn-7-Bicyclo[2.2.1Iheptanediol di-p-toluenesulfonate ( 108)

To a solution of k.<32. g . o f 68 in 25 ml. of pyridine cooled to

0° was added with stirring 1 2 .9 g« of £-toluenesulfonyl chloride.

The resulting solution was stored at 5° for 20 hr. and then poured

over 750 ml. of water and extracted with three 125 -ml. portions of

chloroform. The combined extracts were dried over anhydrous mag

nesium sulfate and concentrated under reduced pressure to a clear

syrup. Crystallization from methanol in a dry ice-isopropyl alcohol

bath gave 10*5^ g* (77%) of white powdery crystals, m.p. 111- 115 ° . 125

Recrystallization from methanol-ether gave an analytical sample*. m.p.

1 2 1 .5 -1 2 1 .6 °.

Anal. Calcd. for C H . Q,S_ i C, 57.78; H, 5.54; S, 14.69 £1 eft O d Founds C, 57.70; H, 5 . 6 3 ;, 3, 14.53. syn-7-Hydroxybicyclof2.2.llheptan-2-one •p-toluenesulfonate (103) via solvolysis of exo-2-syn-7-bicyclor2.2.11heptane- diol di-p-toluenesulfonate (108)

A solution of 5*82 g. of 108 in 100 ml. of 60:40 dioxaneswater

(v/v) was refluxed for 37 hr. and then cooled. An excess of sodium bicarbonate was added. After effervescence had subsided*, the mixture was concentrated under reduced pressure to a sludge which was mixed thoroughly with 200 ml. of water and 200 ml. of chloroform. The

chloroform layer was drawn off* and the aqueous layer was extracted with another 200 ml. portion of chloroform. The combined organic phases were dried over anhydrous magnesium sulfate and concentrated

under reduced pressure to yield a yellow-tinted o il., 2-syn-7-

norbornanediol 7-£-toluenesulfonate (109). The crude 109 was dis

solved in a solution of 4 ml. of acetone and 7*5 ml* of glacial acetic

acid. The resulting solution was cooled to 0° as 5 .3 ml. of chromic 102 acid-acetic acid solution (prepared by dissolving 21 g. of chromium

trioxide in 17 ml. of water and 35 nil* of glacial acetic acid) was

added dropwise with stirring over a duration of 16 minutes. The cold

mixture was stirred for an additional 40 minutes, mixed with 200 m l.

of ether, washed with two 10-ml. portions of water, three 20-m l.

portions of saturated sodium bicarbonate solution,, and 10 ml. of water,

dried over anhydrous magnesium sulfate, and concentrated under reduced

pressure to a yellow oil. Treatment of the oil with Girard’s ”TM 126 reagent as above gave 0.^38 g . ( 11.756 overall) of crude crystalline

103• Recrystallization from hexane gave white crystals, m.p. 95*6-

9 6 . 1°» chemically identical to the material obtained by the first r o u te . a n t i - 7-Hydroxybicyclor 2 . 2 . 11- h e p ta n - 2-one (llO)^ *

To a solution of ^*965 S* of in 21 ml. of water kept at 20-30° was added with stirring 18 .5 m l. o f 6 M chromic acid over a period of

4-0 minutes. The mixture was cooled to 0° and stirred for 30 minutes.

Isopropyl alcohol (1 ml.) was added for an additional 10 minutes of stirring. The mixture was extracted with four 30-ml. portions of ethyl acetate. The combined extracts were dried over anhydrous magnesium sulfate and concentrated under reduced pressure to yield a clear oil.

Pressures down to 1 mm. were necessary to ensure that no residual ethyl acetate would be present in the following chromatography. The o il was chromatographed through 150 g. of silica gel with methylene chloride-ether to give 1 .0 g . (20$) of a white solid. anti-7-Hydroxybicyclor 2.2 .llheptan-7.- one p-toluenesulfonate ( 10^-)^Q^

The purity of 110 used in this preparation was found to be very important. All fractions obtained in the chromatography immediately above which were not at least semi-solid were discarded, regardless of the appearance of their infrared spectra. Inclusion of these in ferior fractions in this preparation rendered the 1(& o b tain ed by th is

route difficult, if not impossible, to purify to a satisfactory degree,

and lowered yields drastically* 127

To a solution of 1.0 g. of 110 in 15 ml. of pyridine cooled to 0° was added 1.75 g* of _p-toluenesulfonyl chloride. The resulting solu tion was stored at 5° for 40 hr. and then mixed with 150 ml. of water and extracted with three 50-ml. portions of methylene chloride. The combined extracts were washed with 30 m l. o f 20$ sulfuric acid and

10 ml. of water, dried over anhydrous magnesium sulfate, and concen trated under reduced pressure to a crystalline solid. Recrystalliza tion from hexane gave 1.46 g. (84$) of white flakes, m.p. 95*0-96.5°•

Further recrystallization from hexane gave an analytical sample, m.p. 95*0-96.5°.

Anal. Calcd. for G, 59*98;. H, 5.75; S, 11.44

Found:. C, 6 0 . 06 ; H, 5 .8 0 ; S , 1 1 . 3 3 . exo-2-anti-7-Bicyclor2.2.llheptanediol di-p-toluenesulfonate ( 111)

A 0.80 g. sample of 106 was converted to the ditosylate as described above for 108 to yield I .60 g. (59$) of 111. Recrystal lization from methanol gave: an analytical sample, m.p. 129. 5 - 131*5 °•

A nal. C alcd . f o r C, 57*78; H, 5 . 5 4 ; S* 14.69

Founds C, 57*78; K, 5 .61;; S, 14.68. a n t i - 7-Rydroxybicyclor 2 . 2 .l] h e p ta n - 2-one p-toluenesulfonate (l04) via solvolysis of exo-2-anti-7-bicyclor2.2.11heptane^~* diol di-p-toluenesulfonate (111)

A 10.6 g. sample of 111 was solvolyzed in dioxane:water and sub

sequently oxidized with chromic acid-acetic acid as for the epimer

(108) above to give 5*1 g* of crude crystalline product. This product

was recrystallized from hexane to give white crystals % m .p. 83- 90° . This material proved to consist of a 75*25 ratio of 104:103 by the following procedure. A sample was acetolyzed to give a rate constant identical to that of an authentic sample of 104. The infinity titer, however, indicated that the sample consisted of only 75$ of 104.

The impurity was isolated ty heating a solution of 596 mg. of the

83-90° sample in 20 ml. of acetic acid (buffered with 0.1 M sodium acetate) at 150 ° f o r 100 minutes (to completely acetolyze the 104 present), mixing the resulting solvolysis mixture with 400 ml. of water, neutralizing the mixture by the slow addition of 30 g. of sodium bicarbonate with stirring, and extracting the mixture with two 100-m l. portions of ether. The combined ethereal extracts were dried over an hydrous magnesium sulfate, concentrated under reduced pressure, and chromatographed through 15 g. of silica gel with benzene-ether to give two fractions. In addition to the acetate (114) (130 mg., 48$ of theoretical) arising from the solvolysis of 104, a 136 mg. sample of

103 was obtained ( 92$ of theoretical).

It was very difficult to obtain a pure sample of 104 from the

83-90° m aterial. Successive attempts at chromatography through silica gel using a benzene-ether eluant afforded white crystals, m.p. 92. 0-

95 —5° 5 mixed melting point with an authentic sample of 104, m.p.

92.5-95-7°. The n.m.r. pattern and solution infrared spectrum (carbon

disulfide) of this material was identical to that of an authentic o sample o f 104. The 92.0-95*5 material was obtained in an overall 129 sy n -7 - Ace to xybicyc lo f2 .2 .n - h e p ta n - 2-one (113)

A sample of 68 was chromatographed through Grade III neutral

Woelm alumina with ether which proved by v.p.c. to consist of a

91:9 ratio of 68:106. To a stirring solution of 1.87 g. of this material in 10 ml. of water maintained at 2 0 . 30° was added 7 m l.

o f 6 M chromic acid over a period of 15 minutes. Isopropyl alcohol

(2 ml.) was added and stirring was maintained for an additional 0 .5

hr. The resulting dark mixture was extracted with three 15-ml. por

tions of ethyl acetate. The combined extracts were dried over anhyd

rous magnesium sulfate* concentrated under reduced pressure to give

crude 107> and dissolved in 10 ml. of pyridine. This solution was o cooled to 0 and 3*5 nil* of acetic anhydride was added with stirring.

The solution was allowed to warm to room temperature for 4.5 hr. of

additional stirring. The solution was again cooled to 0° as 10 ml.

of water was added dropwise. After 15 minutes of stirring* the solu

tion was mixed with 90 ml. of water and extracted with three 25 -m l.

portions of ether. The combined extracts were washed with 15 ml* of

20io sulfuric adid and with 15 ml. of water* dried over anhydrous

magnesium sulfate* concentrated under reduced pressure* and distilled

to yield 0.673 g»> b.p. 80-100° Cl mm.). Vapor phase chromatography

(15$ Didecylphthalate on Chromosorb P) indicated a keto-acetate to 132 diacetate ratio of 6 7 :3 3 » th e e x o - 2-s y n - 7-diacetoxybicyclo[ 2 . 2 . 1]-

heptane:exo- 2- a n t i - 7-diacetoxybicyclo[ 2 . 2 .llheptane ratio being 132 8 6 :1 4 .

A portion was purified by preparative v.p.c. on the same column

and by subsequent distillation to give an analytical sample* b.p. 78° 130

(0.6 mm.), 1.4725* The analytical sample consisted of a 93:7 ratio of 113:114, as shown by v.p.c. (Capillary column Carbowax 1500,

100* x 1/ 16 ").

Anal. Calcd. for C^H-^O^: C, 64.27; H*. 7.19.

Found: C, 64.28; H, 7*33*

anti-.7-Acetoxybicyclor2.2.11- hep tan-2-one (114)

A 5*240 g. sample of pure 106 was converted to 114 as for 113

to g iv e 2.916 g. of crude distilled product, b.p. 79-85 (0.4 mm.).

Vapor phase chromatography (15$ Didecylphthalate on Chromosorb P)

indicated a ll4:exo-2-anti-7-diacetoxybicyclo[2.2.1]heptane ratio of

59:41."^ This corresponds to a 2 5 $ y ie ld o f 114. A p o rtio n was

purified by preparative v.p.c. on the same column and by subsequent

distillation to give an analytical sample, b.p. 80° (0 .7 mm.), n ^

1*473?.

Anal. Calcd. for ^4.27; H, 7.19*

Found: C, 64.48; H, 7.15* KINETICS

Reagents

Anhydrous acetic acid was prepared by refluxing a solution of acetic anhydride and sodium acetate in glacial acetic acid for 2h h r . and subsequent fractional distillation in a dry atmosphere. Standard sodium acetate in acetic acid (ca. 0 .1 M) was prepared by the careful addition of anhydrous acetic acid to a solution of anhydrous sodium carbonate in acetic anhydride* such that c£. 1$ acetic anhydride re mained after the water of neutralization was removed* followed by refluxing in a dry atmosphere for 5 hr."^ (calculated to be 1 . 325 g» of anhydrous sodium carbonate and 3*78 g. of acetic anhydride diluted to 250 ml. with anhydrous acetic acid). Anhydrous ethanol was pre pared by the procedure of Fieser.’^ ' Standard perchloric acid in

acetic acid (ca. 0 .0 2 M) used in titrating acetolysis aliquots was

prepared by the careful addition of 70$ perchloric acid to a solu

tion of anhydrous acetic acid and acetic anhydride* such that 1$ a c e tic

anhydride remained after the water was removed* followed by standing

at room temperature for 12 hr."^ (calculated to be 0.7177 g. of 70$

perchloric acid and 3 »72 g. of acetic anhydride diluted to 250 m l.

with anhydrous acetic acid). The molarity of the standard perchloric _ - — P. D. B a r t l e t t and W. P . Giddings*. J . Am. Chem. S o c ., 82, i2ho (i960). L. F. Fieser, Experiments in Organic Chemistry* third ed.> D. C. Heath and Company* Boston* 1957# p» 285* 13-1 132 acid in acetic acid was determined by titrating an aliquot of primary standard of potassium acid phthalate in anhydrous acetic acid using bromophenol blue as the indicator. Standard sodium methoxide in methanol used in titrating ethanolysis aliquots was prepared by addition of solid sodium methoxide to anhydrous methanol; the molarity of the resulting titran t was determined by dilution of an aliquot with an excess of water and subsequent titration with standard aqueous hydrochloric acid using phenolphthalein as the indicator*

Acetic acid-0-d was prepared by the following procedure. A sample of acetic anhydride was fractionally distilled in a dry atmos phere. The first cut, b.p. I 36 -I 380, was discarded. The second cut, b.p. 138-139°, was collected. A mixture of 20.00 g. of deuterium oxide

(Columbia Organic Chemical Co., Inc.), 103*98 g. of acetic anhydride, and 0.663 g* of anhydrous sodium carbonate was allowed to stand in a dry atmosphere for 58 hr. and then refluxed for 2 hr. Nuclear magnetic resonance spectroscopy indicated 2.75-2.90$ normal acetic acid. Titra tion of an aliquot with perchloric acid in acetic acid indicated

0.108 M sodium acetate content.

OO Kinetics procedure

A ca. 0.06 M solution of tosylate in acetic acid (0.1 M sodium acetate) or in ethanol was prepared in a 10 ml. volumetric flask.

Aliquots of this solution slightly in excess of 1 ml. were removed

and sealed in glass ampoules. The ampoules were placed in a constant

temperature bath held at a specified temperature, and an accurate

timer was started. The ampoules were removed at appropriately timed 133 intervals and immediately quenched in ice water (in solvolyses run at

200° or higher*. the ampoules were; quenched by merely removing them from, the bath). The ampoules were allowed to arrive at room tempera ture* whereupon exactly 1 ml. of solution was removed from each aliquot

by means of an automatic pipette and immediately titrated. Acetolysis

aliquots were titrated with standard perchloric acid in acetic acid*

using a drop or two of saturated bromophenol blue in acetic acrid as

the indicator. The end point was considered to be reached when the yellow solution turned clear. Ethanolysis aliquots were titrated with

standard sodium methoxide in methanol* using a drop or two of 0 . 1$

methanolic bromophenol blue as the indicator.^-35 The end point was

considered to be reached when the yellow solution turned blue.

In the acetolysis of 103* serious decomposition rendered the ali

quots brownish; therefore* the yellow to clear indicator change was

inapplicable. However* consistent and satisfactory pseudo first-order

results were obtained when the color change used was yellow to light

brown* titrating from the last aliquot back to the first and using the

last aliquot as the color standard upon which all other end point de

terminations were made.

All rates were determined utilizing an infinity titer* except

in the case of 6 ^» 1 6 * and 103* which yellowed somewhat before the

infinity titer was reached. All infinity titers were considered to

be reached after 10 half-lives. All infinity titers corresponded to

100$ of theoretical, except in the case of 22 anc* of impure samples

o f 10^-* previously discussed.

-*-35 s. G. Smith, A. H. Fainberg* and S. Winstein* J. Am. Chem. Soc., 82, 6l8 (1961). 13fr The ethanolysis of Jjj8 gave good pseudo first-order kinetics throughout two half-lives, but that of gave a rate which continu ally increased. Therefore, an initial rate for included, which was calculated over 0.2 of a half-life wherein the rate curve was l i n e a r . 136 Rate constants were calculated using the method of least squares'

(see Appendix E). The activation parameters were calculated from the

k, m r (kT/h) e - “ */ST

Acetolysis product analysis of exo-2- hydroxybicyclof 2.2.llheptan-7-one p-toluenesulf ona te

A solution of 2.^31 g* of £8 and 0.7^2 g. of anhydrous sodium

acetate in 50 ml. of acetic acid containing 5 drops of acetic anhydride

was heated on a steam bath for 11 .5 hr. The solution was then cooled

and mixed with 300 ml. of water. To the stirring mixture was added

in portions 70 g. of sodium bicarbonate. The resulting mixture was

extracted with six 100-ml. portions of ether. The combined ethereal

extracts were-: dried over anhydrous magnesium sulfate, concentrated under

reduced pressure, and distilled to yield 1.037 g» ( 71$) of a yellow oil,

b .p . 55-7 0° (0.2 mm.). Vapor phase chromatography indicated a com

position of 65 $ of exo-acetate ( 60 ) , 20$ of endo-acetate ( 6 l ) , and 15 $ 132 — of six other components. Preparative vapor phase chromatography

of the distilled acetolysis product mixture (30$ Carbowax 20M on

^ p. Daniels, J. H. Mathews,. J. W. Williams,, P. Bender, and R. A. A lb e rty , E xperim ental P h y sic a l C hem istry, f i f t h ed .» McGraw- H ill Book Company, Inc., New York, 195^> P« 3 3 9 . i> 5 42/60 Firebrick) allowed the isolation of pure samples of 60 and 6 l which were shown to be identical to authentic samples by infrared spectroscopy. Additional confirmation of the assigned structures of the acetolysis products was available by comparing retention times of the acetolysis products with those of authentic samples of 60 and 6 l on 15$ butanediol succinate on Firebrick No. 20.

Vapor phase chromatography of crude acetolysis mixtures in two separate runs indicated the presence of only 60 and 6 l in a respective

composition of 60 $ and 40$.

An independent experiment established that 60 did not epimerize to 6 l (to.l$) under acetolysis conditions.

A solution of 0.035 g» of ^ in 0.50 ml. of anhydrous acetic

acid buffered with 0 .1 M sodium acetate was heated at 100. 00° f o r

646 minutes, cooled, mixed with 50 ml. of water, and quickly neutral

ized with sodium bicarbonate. The resulting solution was immediately

extracted with 100 ml. of ether. The ethereal extract was dried over

anhydrous magnesium sulfate and concentrated under reduced pressure.

Vapor phase chromatography indicated the presence of only 6 0 , 6 l , 74,.

and 2jl "^h® respective ratio of 48;44:4.0:4.0. Each of these products

was shown independently not to convert to any other compounds under

identical acetolysis conditions.

137 A. A. Frost and R. G. Pearson, Kinetics and Mechanism, second ed.,. John Wiley and Sons, I n c ., New York, 1961, p. 99. 136

Acetolysis product analysis of endo- 2-hydroxybicyclor2 .2.11heptan-7- one p-toluenesulfonate (59)

A solution of 1.438 g. of alcohol-free 52 (88.9$ pure; as deter mined try titration after complete acetolysis) and 0.445 g. of anhydrous sodium acetate in 50 ml. of acetic acid containing 5 drops of acetic anhydride was heated on a steam bath for 4-.5 hr. This reaction solu tion was. worked up and distilled in a manner identical to that used f o r 58 above: to yield 425 mg. (59$> calculated on 88.9$ original tosylate) of a yellow oil. Vapor phase chromatography indicated a composition of 76$ of exo-acetate (6o)» 3$ of endo-acetate (6 l)> and

21$ of other components .^2

A sample of alcohol-free 52 was solvolyzed under identical con ditions. Vapor phase chromatography of the crude acetolysis mixture indicated only the products 6o and 6l in a respective composition of

97*7$ and 2.3$* The original sample of 52 was rechromatographed on silica gel to remove any possibly remaining traces of 5j3 or 52. Vapor phase chromatography of the crude acetolysis mixture of this sample of

59 gave the respective percentages of 60 and 6l as 97*8$ and 2.2$.

The chromatography of 52 was shown to completely remove any traces

of endo-hydroxyketone (57) or exo-tosylate (58) which might have

been responsible for the formation of the endo-acetate (6l).

Acetolysis product analysis of 7 %7- dimethaxybicyclof 2.2.llheptan-exo- 2-ol-p-toluenesulfonate (o4)

A solution of 0.476 g. of 64 and 0.140 g. of anhydrous sodium

acetate in 20 ml. of anhydrous acetic acid was heated at 100.00° for 137 11^ minutes. This reaction solution was worked up and distilled as above to yield 0.270 g . (66 %) of a clear liquid, b.p. 90-100° ( 0 .5 mm.), whose infrared spectrum was identical to that for the exo- acetoxyketal (62). This product was reduced with lithium aluminum hydride in ether to form the ketal-alcohol, as the ketal-acetates

(62 and 65 ) could not be separated by v.p.c. The crude alcohol mixture proved to consist of 9 5 -5 % of the exo-alcohol (<&) and of the

endo-alcohol (55)• An independent experiment proved that a pure sample

of the exo-acetoxyketal ( 62 ) could not account for the endo-hydroxy-

k e ta l ( 55 ).

Dioxane-water solvolysis product analysis o f 7 , 7-dimethoxybicyclor 2 .2 . 1 jheptan-exo- 2-ol p-toluenesulfonate ( 6 ^)

A sample of 6 ty in 60:40 dioxane:water (v/v) at 50° initially

formed the exo-h.ydroxyketa 1 («&) with a small amount of the endo-

hydroxyketal (55)« However, since it was observed that 5ft was thermo

dynamically unstable under the reaction conditions and epimerized to

t. * 55* no quantitative assessment of the relative amounts of 5ft and

was made.

Acetolysis product analysis of 7,7-dimethoxy b ic y c lo r 2 . 2 .llheptan-endo- 2- o l p- to'luenesulfonate ( 65 )

A solution of 0.730 g. of 65 and 0.206 g. of anhydrous sodium

acetate in 30 ml. of anhydrous acetic acid was heated at 100. 00°

for 1950 minutes. This solution was worked up and distilled as above

t o y ie ld 0.220 g. of a yellow-tinted product which proved to consist

o f 2 9-3% o f 2 2 > 1 5 * 8$ o f 2 §» and 5^*9of 6 0 , corresponding to 138 respective molar percentages of 22$, 15 $,. and 63 $. This mixture corresponds to a 70$ yield from 6 ^. Each of the products was col lected T^y preparative v.p.c. (15$ Didecylphthalate on Chromosorb P) and was identified by comparing its infrared spectrum with that of an authentic sample. Independent experiments under identical acetoly sis conditions proved that 21i» Z§, 6 0 * 62 , 6 3 , 8 4 , and 80 could not lead to any of the observed products.

Acetolysis product analysis of antj-7- hydroxybicyclor 2.2.llheptan-2-one p-toluenesulfonate (104)

A solution' of 1.227 g. of 104 and 0.400 g. of anhydrous sodium acetate in 25 ml. of anhydrous acetic acid containing 5 drops of acetic anhydride was heated at 150-160° for 90 minutes. Thissolution was worked up and distilled as above to yield 0.466 g. ( 63 $) o f an oil whose infrared spectrum was identical to that of the anti-acetate

(114). Vapor phase chromatography (Capillary column, Carbowax 1500,

100' x l / l 6 ") established that no other volatile products were present.

No syn-acetate (113) was observed. However, due to the almost identi cal retention times of 113 and 114, it was impossible to establish

the purity of 114 beyond 97$* An independent experiment showed that

113 did not epimerize to 114 under identical acetolysis conditions.

Acetolysis product analysis of syn-7- hydroxybicyclof 2 . 2 . 1 lh e p ta n - 2-one p-toluenesulfonate (103)

Because of extensive decomposition the acetolysis of 103 was

never carried beyond a few percent of reaction. A solution of 0.0113 g«

o f 103 in 1 ml. of anhydrous acetic acid (buffered with 0 .1 M sodium 139 acetate) heated at 200° f o r 5 hr, gave, when worked up as before, a residue which proved toconsist only of 114 > except for a 5 $ unidenti fied impurity with longer retention time. The basis for this identification was comparison of the retention time of the acetolysis product with that of an authentic sample of 114 on three, separate columns: (a) Capillary column, Carbowax 1500* 100* x l/l 6 ";

(b) 0.2$ Carbowax 20M on 80/120 glass beads, 10* x 1/8"; (c) 5$ S ili cone Rubber on 60/80 Chromosorb W,. 10* x l / 8".

A percent yield of 114 was determined by establishing the amount of 114 present after a measured solvolysis time and by comparing this determined value with the theoretical amount of 114 determined from the acetolysis rate. The absolute amount of 114 in an acetolysis product was determined by utilizing an internal standard, diethyl phthalatev. A solution of 0.0234 g, of 103 in 1.00 ml. of acetic acid o buffered with 0 .1 M sodium acetate was heated at 200.0 f o r 1335 minutes, corresponding to 37$ theoretical reaction. The solution was

cooled and worked up as before. Vapor phase chromatography (0.2$

Carbowax 20M on 80/120 glass beads) showed 114 was in 25 $ y ie ld o f A theoretical.

A similar experiment run for a much shorter period of time was

complicated by the possibility of a small amount of 104 being present

as a trace impurity. Since the 104 would solvolyze completely, the

114 ensuing therefrom would be significant if a very small amount of

the 103 were solvolyzed. Since the 104 would solvolyze almost immedi

ately, then if it were present in significant quantity, it could be

detected during the in itial part of an acetolysis rate determination i4 o of 103« During such a rate determination*, no such detection could be made. Experimental error dictated that the maximum percentage of 104 would be 1.2$. A solution of 0.0219 g. of 103 in 1.00 ml. of acetic acid buffered with 0.1 M sodium acetate was heated at 200.0° for 112 minutes, corresponding to 3»6$ theoretical reaction. The solution was cooled and worked up as before. Vapor phase chromatography (0.2$

Carbowax 20M on 80/120 glass beads) indicated 117$ theoretical yield of 114. The incorporation of the 1.2$ maximum impurity of 104 led to the calculated minimum yield of 114 arising exclusively from 103 to be 83$.

Partial acetolysis of anti-7-hydroxybicyclo- r2.2»llheptan-2-one p-toluenesulfonate (104) in acetic acid-O-d

Two samples of 104 were solvolyzed in acetic acid-0-d, each tinder different conditions, (a) A solution of 0.292 g. of 10*4- in

8 ml. of acetic acid-0-d buffered with 0.1 M sodium acetatevas

heated on a steam bath for several hours, cooled, and worked up as-

before, (b) A solution of 0.431 g. of 104 in JO ml. of acetic o acid-0-d buffered with 0.1 M sodium acetate was heated at 110.00

f o r 75*5 minutes, cooled, and worked up as before. In each case

the crude product was chromatographed through silica gel with

benzene-ether to give an acetate fraction and a tosylate fraction.

The acetate fraction was purified by preparative v.p.c. (15$ Didecyl-

phthalate on Chromosorb P) in the case of (a) and by distillation

(b.p. 75°» 0.25 mm.) in the case of (b). The infrared spectra of the

acetates from (a) and (b) were identical, but were different from 141 that of 114. The tosylate fractions were purified by recrystalliza tion from hexane (m.p. 97*0-98.0°). The infrared spectra of the tosylates from (a) and (b) were different from one another and from that for 104; that from (b) was intermediate between that from (a) and that of 104. The n.m.r. pattern of tosylate (b) was identical to that of 104> except that the intensity of a signal masked by the aryl methyl group was diminished. Mass spectral analyses indicated a mono- 112 deutero percentage of 54$ for tosylate (a) and 34$ for tosylate (b). APPENDIXES

A. Carbonyl stretching frequencies

TABLE 11

Infrared Stretching Frequencies in the Carbonyl

Region of Various Norbomanone Derivatives

Compound Absorption> u Phase

5.60; 5*38 (weak) Neat

5 * 6 1 ; 5*35 (weak) Neat .OAc

5 .6 1 ; (weak) Neat

O OAc

.OTS 5 .6 2 ; 5.43 (weak) KBr

5»6l; 5*^3 (weak) Neat TABLE 11 (Continued)

Compound Absorption* |J. Phase

5 . 6 5 ; 5.^-1 (weak) Neat OH O

5 . 6 2 ; 5.35 (weak) Neat

5*61; 5»36 (weak) Neat ,0 Me

Me O-v/'OMe

5*70 N eat

5-71 Neat

5*71 Neat

O OTs

5*70 KBr

TsO TABLE 11 (Continued)

Compound Absorption* p. Phase

5*66 Neat

OAc 5*70 Neat 145

B. MR d a ta

TABLE 12

MR Chemical Shifts for Various

Norbornane Derivatives

Compound Phase Chemical Shift* I M e C X X > M e Neat 6 .8 2 (6 protons; singlet) 8. 1- 8 .3 (6 protons; unresolved) 8 . 8- 8.9 (4 protons; unresolved)

Neat 3*97 (2 protons; triplet; J = 2.0 c.p.s.) 6 .8 7 (3 protons; singlet) 6 .9 5 (3 protons; singlet) 7 .3 2 (2 protons; multiplet) 8 .0- 8 .3 (2 protons; unresolved) 9* 02- 9.27 (2 protons; multiplet) Q

Neat 3.41 (2 protons; triplet; J = 2.5 c.p.s.) 7 .2 5 (2 protons; multiplet) 7* 9- 9 »0 ( 4 protons; unresolved)

M eO -^-O M e „0 Neat 6 .8 6 (3 protons; singlet) 6 .8 8 (3 protons; singlet) c a . 6.90 (2 protons; multiplet) 7 .6 8 (2 protons; multiplet) 8.2-8.9 (4 protons; unresolved)

CHC1, 6 .3 1 (1 proton; multiplet) 6.52 (1 proton; singlet) 6 .6 9 (3 protons; singlet) 6 .7 4 (3 protons; singlet) 7.9-9*0 (8 protons; unresolved) 146

TABLE 12 (Continued)

Compound Phase Chemical Shift* t

Me(XX>Me Neat 6.44 (l proton; multiplet) 6.6l (1 proton; singlet) 6.73 (3 protons; singlet) 6*78 (3 protons; singlet) 7-9-9-0 (8 protons; unresolved) MeO^OMe CHC1, 5*67 (1 proton; doublet; J = 10 c.p.s.) 6.10 (1 proton; singlet) 6 .7 8 (3 protons; singlet) 6*79 (3 protons; singlet) OH 7.8-9.1 (8 protons; unresolved) MeO-

/OMe CHClr 2.20 (2 protons; doublet; J = 8 c.p.s.) 2.64 (2 protons; doublet; J = 8 c.p.s.) 5»6l (1 proton; doublet doublet; 3*7* 4.3* 3*3 c.p.s. separation) 6.78 (3 protons; singlet) 6.81 (3 protons; singlet) 7-57 (3 protons; singlet) 7.8-9.0 (8 protons; unresolved)

M eO^/OM e CHC1, 2.22 (2 protons; doublet; J ~ 8 c . p . s . ) 2 .6 7 (2 protons; doublet; J = 8 c.p.s.) 5.08 (1 proton; multiplet) 6.80 (3 protons; singlet) 6 .8 3 (3 protons; singlet) OTs 7-57 (3 protons; singlet) 7 .6 - 9.0 (8 protons; unresolved)

\ 147

TABLE 12 (Continued)

Compound Phase Chemical Shift» T

Neat 6 .7 8 (3 protons; singlet) 6.81 (3 protons; singlet) 7.52 (2 protons; multiplet) 7.7-8.7 (6 protons; unresolved)

CHC1, 5.18 (1 proton; multiplet) ,OAc 8.03 (3 protons; singlet) 8. 0- 8 .5 (8 protons; unresolved)

CHC1, 4*86 (1 proton; multiplet) 7*83 (3 protons; singlet) 7*7-8*9 (8 protons; unresolved) OAc

CHC1, 2.23 (2 protons; doublet; J = 8 c.p.s.) .OTs 2.61 (2 protons; doublet; J = 8 c.p.s.) 3.27 (1 proton; doublet; J = 3 *6 c . p . s . ) 7.61 (3 protons; singlet) 7.9-8.9 (8 protons; unresolved)

* • CHC1, 2.17 (2 protons; doublet; J = 8 c.p.s.) . 2*65 (2 protons; doublet; J = 8 c.p.s.) 5.02 (1 proton; multiplet) OTs 7.58 (3 protons; singlet) 7»7-8.4 (8 protons; unresolved)

Neat 5*05 (1 proton; multiplet) 8.02 (3 protons; singlet) 7*4-8.6 (8 protons; unresolved) 148

TABLE 12 (Continued) Compound Phase Chemical Shift* T AcO Neat 5*14 (l proton; multiplet) 7*46 (2 protons; multiplet) 8.01 (3 protons; singlet) 7*9-8.6 (6 protons; unresolved)

OTS CHC1, 2.21 (2 protons; doublet; J = 8 c.p.s.) 2 .6 5 (2 protons; doublet; J = 8 c.p.s.) 5*19 (1 proton; multiplet) 7*57 (3 protons; singlet) 7 .4-9*1 (8 protons; unresolved)

TsO CHC1, 2.21 (2 protons; doublet; J = 8 c.p.s.) 2 .6 7 (2 protons; doublet; J = 8 c.p.s.) 5*31 (1 proton; multiplet) 7*41 (2 protons; multiplet) 7*58 (3 protons; singlet) 7.9-8.6 (6 protons; unresolved) M e O

Neat 6*81-6.87 (10 protons; unresolved) OMe 7*85-8.40 ( 6 protons; unresolved) 8*95-9*07 ( 2 protons; unresolved)

Neat 6.46 (1 proton; multiplet) ,OMe 8.79 (3 protons; singlet) 8.0-8.7 (8 protons; unresolved) 149

C. An example of a solvolytic rate determination involving an infinite titer

The determination of a solvolytic rate constant involving an in finite titer is accomplished by the completion of a table. Table 13 was the one used for an acetolysis run of endo-2-hydroxybicyclo[2.2.1]- heptan-7-one p-toluenesulfonate (59) at 100°. Exactly 1 ml. aliquots were titrated; 0.0200 M perchloric acid in acetic acid was the titrant.

Column (1) lists the times in minutes at which the ampoules were with drawn. Column (2) lists the times in seconds. Column (.3) lists the volume of titrant used for each aliquot. Column (4) lists the sodium acetate concentration in each aliquot, determined by multiplying each value in Column (3) by the molarity of the titrant (in this case,

0.0200). Column (5) lists the acid generated since the first aliquot, determined by subtracting each entry in Column (4) from the first

(in this case, 0.0956). Column (6) lists the molarity of the tosylate in each aliquot, determined by subtracting each entry in Column (5) from the last entry in Column (5) (in this case, 0.0822). Column (7) lists the logarithms of each entry in Column (6). The solvolytic rate constant is determined from the data in Columns (2) and (7) (see

Appendix E). The infinite titer, the last entry in Column (3)> is taken after at least 10 half-lives. The half-life of a reaction is T Aft determined from the relationship

tl/2 = (0.693/k) (49)

1 Aft A. A. Frost and R. G. Pearson, Kinetics and Mechanism, second ed., John Wiley and Sons, Inc., New York, 1961, p. 42. 150

This method does not require a precise value of the molarity of

the titrant or of the original solvolysis solution. However* these

values must be known accurately if one wishes to compare the original

molarity of tosylate: determined by solvolysis (Column ( 6 )» f i r s t

entry) with the actual original molarity.

TABLE 13

Determination of a Solvolytic Rate Constant

by Using an Infinite T ite r

CD ( 2) (3) (4) (5) (6 ) (7) Time* Time* Ml. NaOAc A cid .g en . M o larity * Log M m in. sec .xlO ' t i t r a n t ( m ./l.) ( m ./l.) to s y la te

0 0 4 .7 8 .0956 0 .0822 .9149

8.33 0.50 3.93 .0786 .0170 .0652 .8142

16.67 1.00 3.21 .0642 .0314 .0508 .7059

25.00 1.50 2.66 .0532 .0424 .0398 .5999

33.33 2.00 2.30 .0460 .0496 .0326 .5132

41.67 2.50 1.95 .0390 .0566 .0256 .4082

50.00 3.00 1 .6 8 .0336 .0620 .0202 .3054

346 0.67 .0134 .0822 151

D. An example of a solvolytic rate determination not involving an infinite titer

The determination of a solvolytic rate constant not involving an infinite titer is accomplished in a manner analogous to that described in Appendix C. However, the precise determination of the molarity of the titrant and of the original solvolysis solution is critical* and the measured aliquots must be exact. Table 14 was the one used for the acetolysis of 7-norbornyl jc-toluenesulfonate in acetic acid-O-d at 200°. No infinite titer was possible in this solvolysis, as the reaction solution began to yellow seriously toward the end of the reaction. The acetolysis solution was prepared by charging a 10 ml. volumetric flask with 0.2230 g. of tosylate and by the subsequent dilution with acetic acid- 0-d ( 0.109 M sodium acetate) to exactly 10 ml. of solution. The molarity of the solution was calculated to be

0.0837 (Column ( 6 ), first entry). The molarity of the titrant

(perchloric acid in acetic acid) was determined to be 0.0203* Exactly

1 ml. aliquots were titrated.

The first five columns in Table 14 correspond exactly to those in Table 13» except that no infinite titer was taken. Column ( 6 ) differs in that the first entry is obtained by the procedure des cribed above. Each entry in Column ( 6 ) is determined by subtracting

each entry in Column (5) from the first entry in Column ( 6 ) (in this

case, .0837)* Column (7) lists the logarithms of the values in

Column ( 6 ). The solvolytic rate constant is determined from the data

in Columns (2) and (7) (see Appendix E). 152

TABLE 14

Determination of a Solvolytic Kate Constant

without the Use of an Infinite Titer

(1 ) ( 2) (3) (4) (5 ) (6 ) (7) Time, Time, Ml. NaOAc Acid. gen. M o la rity , Log M m in. sec •xlO- -' t i t r a n t ( m ./l.) ( m ./l.) to s y la te

0 0 5.35 .1086 0 .0837 .9227

33-33 2.00 4 .9 2 .0999 .0087 .0750 .8751

100.00 6.00 4 .1 7 .0847 .0239 .0598 .7767

166.67 10.00 3.37 .0725 .0361 .0476 .6776

233.33 14.00 3.10 .0629 .0457 .0380 .5798

300.00 18.00 2.69 .0546 .0540 .0297 .4728

366,67 22.00 2.40 .0487 .0599 .0238 .3766

433.33 26.00 2.16 .0440 .0646 .0191 .2810 153- 136 E. The method of least squares

As an example of the method of least squares, we shall deter mine the rate constant for the acetolysis of at 100° from the values of Table 13*

The specific rate constant for a first-order reaction is experi mentally determined by the equation

-2.303 log[A] = kt + const. . ( 6 ) o r

log[A] = -(k/2.303) t + const. where [A] is the concentration of the tosylate at time t and k is the specific rate constant. A plot of log [A] vs. t should be linear with a slope of -k/2»303» The values in Columns (2) and (7) of Table 13 form such a linear plot. The best slope to fit a linear plot can be determined by the method of least squares. According to this method, the slope of m of an equation of the form

y = mx + c ( 50 ) where y and x are variables and c is a constant, may be determined

by the equation

S x .% Sxy - n

M = — ( 5 i > n

where x and y are specific values of various sets and n is the

number of such sets. 154

Comparison of equations ( 6 ) and (50) shows that

y = lo g [A]

x = t (52)

m = -k /2 .3 0 3

To determine the values to substitute into equation (51)> we set up

Table 15* From Columns (2) and (7) of Table 13» we incorporate O values of x and y» respectively} into Table 15• The values of x

and xy are consequently calculated for each set. Each column of

Table 15 is summed to find £xj 2y» and £xy. Then the values 2 — (2x) and are calculated.

TABLE 15 The Method of Least Squares

X X 2 Z 0 x 103 0 x 10° -9149 0 x 103

0.50 0.25 .81^2 0.40710

1.00 1.00 .7059 0.70590

1.50 2.25 .5999 0.89985

2.00 i+ .o o .5172 1.02640

2.50 6.25 .4-082 1.02050

3.00 9.00 .3054 0.91620

10.30 x 103 22.75 X 106 4.2617 4.97595 x 10

(Sfe) (Ex2 ) (%) ( 2 x y )

(Ex )2 = 110.25 x 106

M y = 44.74785 x 103 155

It is seen in Table 15 that there are 7 sets. Therefore> the value of n in (51) is 7* Incorporation of the known values into

( 51 ) le a d s to

l> .97595 - (44.74-785/7)] x lo 3 m .= ______■ ______[22.75 - ( 110. 25 / 7) ] x 106

-97595 - 6 .39255 ] x io3' m = ------z------[22.75 -15.75] x I0b

- 1 .4-166 x 10-3 m = ------7.00

m = - 2.024- x 10"4 ( 52 )

Since m = -(k/2*303)j we multiply (52) by -2.303 to get

k =--2.303 m = -2.303 (-2.024- x 10“^)

h. -) = 4-.661 x 10 s e c ." . 156

F. Determination of the activation param eters

The values of the activation parameters Ah* and As* a re 137 calculated by means of the equation J

kr » (kT/h) e-AH*/RT (W) where kr = specific rate constant

k = Boltzmann constant

h = Planck's constant o T = temperature K.

R = gas constant

Dividing (48) by T and taking the logarithm* we get

log (kr/T) = log (k/h) + As*/ ( 2 .303) (R) - Ah*/ (2 . 303) (RT) (53)

Substituting the known values of k, h> and R into (53 ) j we g e t

log (kr/T) = log[(1.380 x 10‘l6)/(6.624 x 10"27) ] -

AH*/[( 2 .303)(1.987)(T)] + As*/(2.303)(1.987)

log (ky/T) =10.31876 - AH*/[ (4.576) (T )] + AS*/ (4 .576) ( 54 )

It is seen from (54) that a plot of log(kr/T) vs. (l/.T) should

be linear and have a slope of -AH*/4.576. Furthermore* rearranging

(5 4 ),

As* = [log(kr/T) - 10.31876]4.576 + AH*/T (55)

The following example illustrates how the values of AH* and

As* are determined from equations (54) and 55)• From the values 15? of the specific rate constants at different temperatures for exo-2- hydroxybicyclo[2.2.1]heptan-7-one p-toluenesulfonate (j58)* Table 16 is built up. The listed values for log (kr/T) and l/T form a linear plot. The slope of this plot is determined by the method of least squares. Table 17 lists the values of x, x * y*. and xy where x = l/T and y = log(kr/T). Since we are determining a slope* the characteristic of y = log(k /T) has been changed on a sliding scale to arrive at more r convenient numbers for y.

TABLE 16

The Determination of Activation Parameters

T (°C .) 100.00 90.00 75.71

T (°K .) 373.16 363.16 348.87 kr 1.81)- x io "4 5.96 x 10“5 1.33 x 10”5 VT 4 .9 3 x 10“7 1 .6 4 x 10"7 3.81 x 10-8 lo g (k r /T) .6928 - 7 .2148 - 7 .5809 - 8 l/T 2.6798 x 10“3 2.7536 x 10“3 2.8664 x 10"3 158

TABLE 17

The Method o f L e ast Squares

X x 2 y xy 2.6798 x 10“3 7.18132804 x 10-6 1.6928 4.536365^ x lo"3

2.7536 7.58231296 1.2148 3.3^507328

2.8664 8.21624896 0.5809 I .66509176

_3 8.2998 x 10”3 22.97988996 x 10 - 6 -3 .4 8 8 5 9.5^653048 x 10

(2x) (Sx2 ) (%) (2 xy)

- 6 ( Ik )2 = 68.88668004 x 1 0 '

23x2y = 28.95385230 x 1 0 '-3

The slope m = -AH*/4.576 is thus

Iky - (5k%/n) m " 2(x2) - (Ik ) z /n

[9.5^6530^8 - (28.95385230/3)] x 10“3 m ~ [ 22.97988996 I (68.88668004/3)] x 10"6

m = - 5.9306 x 103

Since m = -AH*/4.576*

ah* = - 4 . 576 m = - 4 . 576 (- 5.9306 x 103 )

= 27>138 calories/mole

To obtain the value of AS*> we now substitute the calculated

v a lu e of Ah * into (55) at the three different temperatures. 159

At 100°>

AS* = (.6928 - 7 - 10.31876)^.5?6 + (27j100)/ 373.16

= -3.46 e.u.

At 90°>

As* = (.2148 - 7 - 1 0 . 31876)4.576 + ( 27»100)/ 363 . l 6

= -3*65 e.u.

At 75.71°>

AS* = (.5809 - 8 - 10.31876)4.576 + ( 27>100/348.87

= -3*47 e .u .

The average i s th u s

AS* = -3.5 e.u.

If the activation parameters for a particular reaction are known,, the reaction rate can be calculated at any new temperature by rearranging equation ( 53 ) 1

AH* - TAS* lo g kr = 10.31876 + logT ------( 56 ) r 4 .5 7 6 T

At 25°C. equation ( 56 ) reduces to

AH* - 298.16AS* lo g k = 12.79321 ------( 57 ) 1364.4

The units of Ah* and AS* in equations ( 56 ) and (57) are always

calories/mole and e.u., respectively. l6o

G. Rate data

Table 18 lists the specific solvolytic rate constants of various tosylates discussed in this dissertation. The error included with the rate constants is an average deviation for two runs. The tempera ture control was 75.00 and 90.00° * 0.01; 100.00° + 0.02; 110.00° *

0.03; and 200.0 and 210.0° * 0.1. The listed value of t^ j ^ i s th e largest number of half-lives from which a rate determination was cal culated for a given tosylate in a given solvolysis medium; the

solvolysis temperature at which this value of ^ / Z aPP-^es ^-s n°^

implied in the table. TABLE 18 Solvolytic Rate Constants Experimentally Determined for Various Norbornyl Tosylates

NaOAc T osylate conc. co n c. 0 Compound Solvent ( m ./l.) ( m ./l.) Temp. C. Rate (sec .”■*■) No. o f

-4 O HOAc .099 . 0690 , .0719 100.00 (1 .8 4 ± ' .01) x 10 1.6 K HOAc .099 .0604, .0594 90.00 (5.96 ± . 06 ) x 10“5 ■ \ r " HOAc .099 .0738, .0744 75.71 ( 1.33 ± . 01) x 10j? /O T s EtOH . 072^* .0724 90.00 (1.21 ± . 01) x 10 2 L— EtOH .0749 100.00 2.56 x 10

O HOAc .099 . 0822, .0670 100.00 (4 .6 6 + . 00) x 10"f 3 .0 v HOAc .099 .0 8 4 6 , .0788 90.00 (1 .7 9 ± .0 3 ) x 10-^ HOAc .099 .0531 75.7^ 4 .2 8 x10"5 f 7 EtOH .0730, .0725 90.00 (4.93 ± *03) x 10“3 .02 1 EtOH .0583 100.00 1.2 x 10"4 t )Ts

MeCL HOAc .099 . 0668 , .0797 100.00 (1 .0 7 t .04) x 10"P 2 .2 \ HOAc .099 .0 6 4 6 , .0652 90.00 (3.58 ± .13) x 10*4 r " 7 __ HOAc .099 .0748, .0824 75.85 (8.12 ± .03) x 10-5 / , 0 Is

i—* o\ H TABLE 18 (Continued) NaOAc T osylate conc. conc. Compound Solvent ( m ./l.) ( m ./l.) Temp.°C. Rate (sec.**^) No. o f t]

M e O ^ X )M e HOAc .099 .0778, .0801 100.00 (6.35 ± .01) x 1 0 ^ 1 .6 HOAc .099 .0736* .0765 90.00 (2.05 ± .13) x io -5 HOAc .099 .0781 75.77 4 .3 9 x 10-5 c DOAc .109 .0807 100.00 6.4-5 x 10-5 1 .9 §O T s icO T s HOAc .096 .0782 * .0809 200.00 (5.^5 ± .19) x 1 0 " § 0.6 HOAc .096 .0832, .0804- 210.00 (1.20 + .00) x i o " J c . DOAc .109 .0834 200.00 5.76 X 10“6 0 .3 Qo

T s O ^ HOAc .101 .0706, .0612 90.00 ( 1 .13. t . 06 ) x 10"5 1 .8 HOAc .101 .0702> .070^ 100.00 (2.78 ± .09) x 10--5 HOAc .101 . 0698, .0682 110.00 (6.88 + .01) x 10j? DOAc .109 .0805* .0795 110.00 (1.31 ; .02) x 10-4" c (initial rate) 1

DOAc .109 .0837* .0826 200.00 (5.73 t .01) X 10-5 2 .1 T s O - ^ n

On j\> H. Rate curves

The following graphs are representative rate curves for the solvolysis of various tosylates. The line connecting the points is not a least squares fit. i A H I. ol s Xt Cre o 2-Hdoybi o— lo c y ic b Hydroxy - 2 oX Curves Xate is s ly to e c A GivAPH XIX.

[ROTs] .5 .8 .2 X00°C. 2. 1] an- one £- Ieeufnts t a oIuenesuIfonates £>-T e n -o -7 n ta p e ]h .1 .2 [2 i , 10 x c e s e, Tim 3 OTs ■OTs 6 16k •og [ ROTsJ .8 .6 5 O 5 7-3 15 lO 05 e OMe MeO ie sc 10 x sec. Time, .OTs 0 MeO. 12 OTs Me

2. 1] an- ol - leeufnt a 100°C. at oluenesulfonate £-T l -o -2 n ta p e ]h .1 .2 [2 i 7>7-*Oi|Tiethoxybicyclo of Cui'Ves Rate cetolysis A IV. GRAPH O n log [R O T s] R P V Acet ysi t Cuvs 7-l oxybi o lo c y ic b y x ro d -lly 7 f o urves C ate R is s ly to e c A GRAPH V* .8 # ,8 .5 .9 2. ] an- one J Tounslostes e t oluenesulfons -T JD e n -o -2 n ta p e ]h .1 .2 2 [ Time, sec. x 10 J 10 x sec. Time, '15 8 6 3 18 Ts O Ts OTs 210 100 ~r> 0 3 log [r o t s ] - lo c y ic b y x ro d y -H 7 f o Curve ate R is s ly to e c A I. GRAPH V 9 6 3 i- d 200°C. t a -d cid-O A l ane j Tol f e in Acetc tic e c A n i te a n lfo u s e n e lu o -T je e n ta p e h ,l] 2 . 2 [ ie sc x ^ 0 1 x sec. Time, 12 DOAc 24 BIBLIOGRAPHY

1. G. Wagner». J. Russ. Phys. Chem. Soc.» ^1, 680 (1899)*

2. For a review» see J. L. Simonsen and L. N. Owen* The Terpenes, Vol. II, Cambridge University Press* 19^9*

3* H. Meerwein and K. van Emster, Ber.* 1815 (1920).

H. Meerwein and K. van Emster, Ber., 2500 (1922).

5* Y* Pocker, "Wagner-Meerwein and Pinacolic Rearrangements in Acyclic and Cyclic Systems," Molecular Rearrangements, P. deMayo, ed.,. Interscience Publishers,. New York, 1963> p p . 1 -2 5 .

6 . Evidence supporting the endo side being the least hindered approach for attack is the fact that camphor (b) in reduction with lithium aluminum hydride experiences attack predominantly from the endo side to give isobomeol (j>) [L. W. Trevoy and W. G. Brown,. J. Am. Chem. Soc., £1* l6?5 (19^9)? D* 3* Noyce and D. B. Denney, ib id ., 22 > 57^3 (1950)]•

7* T. P. Nevell, E. deSalas, and C. L. Wilson, J. Chem. Soc. (London), 1939, 1188.

8 . F. Brown, E. D. Hughes, C. K. Ingold, and J. F. Smith, Nature, 168, 65 (1 9 5 1 ).

9* J• D. Roberts and C. C. Lee, J. Am. Chem. Soc., 22* 5009 (1951)*

10. J. A. Berson, "Carbonium Ion Rearrangements in Bridged Bicyclic Systems," Molecular Rearrangements,. P. deMayo, ed., Interscience Publishers, New York, 1963, p. 119.

168 169

11 . P. D. B artlett, Nonclassical Ions, W. A. Benjamin, Inc., New York, 1965* p. v.

12. S. Winstein and D. Trifan, J. Am. Chem. Soc., 7 k, 1147, 115^ (1 9 5 2 ).

1 3 . S. W instein, C. R. Lindegren,. H. Marshall, and L. L. Ingraham, J. Am. Chem. Soc., 21* 1^® (1953)*

1 4 . H. C. Brown, Chem. Soc. (London) Spec. Publ. 16, 140 (1962).

15- For a leading reference, see 3. Winstein,. J. Am. Chem. Soc., 82, 381 (1965).

1 6 . For a leading reference, see H. C. Brown and M. H. Rei, J. Am. Chem. Soc., 86, 5008 (1964 ).

1 7 . S. W instein, M. Shatavsky, C. Norton, and R. B. Woodward, J. Am. Chem. Soc., 21* *H83 (1955).

1 8. W. G. Woods, R. A. Carboni, and J. D. Roberts, J. Am. Chem. Soc., 2§, 5653 (1956).

1 9 . H. C. Brown and H. M. Bell, J. Am. Chem. Soc., 8 5 , 2324 (1963 ).

2 0 . S. Winstein, A. H. Lewis, and K. C. Ponde, J. Am. Chem. Soc., §1* 2324 (1963).

2 1 . H. C. Brown, F. J. Chloupek, and M. H. Rei, J. Am. Chem. Soc., 86, 1247 (1964).

2 2 . P. von R. Schleyer* J. Am. Chem. Soc., 86, 1854, 1856 (1 9 6 4 ).

23- H. C. Brown and M. H. Rei, J. Am. Chem. Soc., 86, 5004 (1964).

Zk, H. C. Brown and M. H. Rei, J. Am. Chem. Soc., 86, 5008 (1964).

2 5 . H. C. Brown, F. J. Chloupek, and M. H. Rei, J. Am. Chem. Soc., 86, 1248 (1964).

26 . Under equilibrium conditions, a tertiary system w ill undergo a Wagner-Meerwein rearrangement such that the methyl group is in the 1-position (see Ref. 10, pp. 131, 137)*

CH c h 3 170

27# P. von R. Schleyer# M. M. Donaldson# and W. E. Watts# J. Am. Chem. Soc., 8 7, 375 (1965).

28. S. Winstein# J. Am. Chem. Soc.» 8 7, 381 ( 1965 )*

2 9. There is one exception (see p. 17)# but for reasons to be dis cussed# the validity of the application of this example to solvolysis reactions is not certain.

30. P. von R. Schleyer, W. E. Watts# R. C. Fort# Jr.,. M. B. Comi- sarow# and G. A. Olah# J. Am. Chem. Soc.# 8 6 # 5680 (1 9 6 4 ).

31* M. Saunders, P. von R. Schleyer, and G. A. Olah# J. Am. Chem. S o c ., 86 , 5681 (1 9 6 4 ).

32. H. C. Brown and H. M. Bell# J. Am. Chem. Soc.» 8 6 , 5006 (1 9 6 4 ).

33* H. L. Goering and C. B. Schwene# J. Am. Chem. Soc.# 8 7# 3516 (1965 )*

34. See Ref. 10# pp. 134-135*

35* J* A. Berson# Tetrahedron Letters, No. 16# 17 ( i 960 ).

3 6 . For a summary of such examples,see Ref. 10# p. 131*

37* N. J. Toivonen# E. Siltanen# and K. Ojala# Ann. Acad. Sci. Fennicae# Ser. A# I I # No. 64 (1955)*

38. P. Beltrame# C. A. Bunton# and D. Whittaker#, Chem. & I n d . (London)# 1960# 557*

39* G. Komppa and S. Beckman# Ann.# 509? 51 (1934).

40. P. D. Bartlett# E. R. Webster# C. E. Dills# and H. G. Richey# Jr.# A nn., 6 2 7 , 217 (1959).

41. P. von R. Schleyer# D. C. K leinfelter, and H. G. Richey# Jr., J. Am. Chem. Soc.# 8^# 479 (1963)*

42. H. C. Brown and H. M. Bell# J. Am. Chem. Soc.# 8 6 # 5003 (1964).

43. H. C. Brown and H. M. Bell# J. Am. Chem. Soc., 8 6 , 5007 (1964).

44. Solvolysis of 2-methylnorbornyl derivatives in aqueous diglyme gives exclusive exo attack at the tertiary position whereas borohydride reduction of the same derivatives in the same medium gives trace quantities of product derives from endo attack at the same position [H. C. Brown and H. M. Bell# J. Am. Chem. S o c ., 8 6 # 5006 (1964)]. For much more drastic differences between solvolysis and borohydride reduction in the same medium see "Part II. The 7-Norbornenyl Cation" of this dissertation. 171

45* H. C. Brown* F. J. Chloupek, and M. H. Rei, J. Am. Chem* Soc., 86, 1246 (1964).

46. R. G. Lawton, J. Am. Chem. Soc., 8^, 2399 (1961).

47. P. D. Bartlett and S. Bank, J. Am. Chem. Soc., 82» 2591 (1961).

48. P. D. Bartlett, S. Bank, R. J. Crawford, and G. H. Schmid, J. Am. Chem. Soc., 82, 1288 ( 1965 ).

49. Of course, the difference in the ground state energies of the various derivatives a ll going to a common product must be taken into consideration, also. Even after compensating for this, at least in the case of an|i Z2 the transition states differ by an internal energy of 4.4 Kcal./mole (see Ref. 33)*

5 0 . G. S . Hammond, J . Am. Chem. S o c ., 22.* 33^ (1955): "If two states, as for example, a transition state and an unstable intermediate, occur consecutively during a reaction process and have nearly the same energy content, their interconversion w ill involve only a small reorganization of the molecular structures."

51. S. Winstein,, A. Colter, E. C. Friedrich, and N. J. Holness, J. Am. Chem. Soc., 82, 378 ( 1965 ).

52. J. C. Martin and P. D. Bartlett,, J. Am. Chem. Soc., 79* 2533 (1 9 5 7 ).

53* It would seem that 48 would be an excellent example to test the nonclassical cation hypothesis; the same criteria listed above for the 7-ketonorbomyl cation (4=>) could be used here. Unfor tunately, this cannot be done, for 48 rearranges to 3-formyl- cyclopentanol (50)* Thus, one cannot test the exclusive exo

product formation criterion. Also,anchimeric assistance is possible in exo-48. Therefore, the question is not resolved whether the high exo-endo rate ratio observed for 48 is due to assistance in the exo case or to the steric effect of the endo- C^ hydrogen atom in the endo case.

54. D. J. Cram and G. S. Hammond, Organic Chemistry, second ed., McGraw-Hill Book Company, Inc., New York, 1964, p. 148.

55. C. S. Foote, J. Am. Chem. Soc., 86, 1853 (1964). 172

5 6 . A. C. Macdonald and J. Trotter, Acta Cryst*., 18, 243 ( 1965 )*

57* P* G. Gassman and P. G. Pape, J. Org. Chem., 2£, l 6 o (1 9 6 4 ).

58. J. S. Newcomer and E. T. McBee, J. Am. Chem. Soc., 946 (1949)*

59* P« E. Hoch, J. Org. Chem., 26, 2066 ( 1961 ).

60. P. Bruck, D. Thompson,, and S. Winstein,. Chem. Ind. (London), 405 (I960).

6 1. H. C. Brown and B. C. Subba Rao, J. Org. Chem., 2^, 1136 (1957)*

62. R. S. Tipson, J. Org. Chem., £, 235 (1944).

6 3 * This tosylate has been prepared and its acetolysis rate determined by K. Mislow and W. E. Meyer. For details, see W. E. Meyer, thesis, New York University, 1964.

64. This compound has previously been prepared by C. H. DePuy and P. R. Story, J. Am. Chem. Soc., 82, 627 ( i 960 ).

6 5 . R* Piccolini and S. Winstein, Tetrahedron Letters, No. 13, 4 (1959)*

66. E. W. C. Wong and C. C. Lee, Can. J. Chem., 42, 1245 (1964).

6 7 * For the synthesis of this compound, see ''Part II. The 7-Norbornenyl Cation” of this dissertation.

68. G. Braun, "Perbenzoic Acid,” Organic Syntheses, Coll. Vol. I, H. Gilman, ed., John Wiley & Sons, Inc., New York, 1932, P* 431*

6 9 . For examples of the role of hydrogen bonding in stereospecific epoxidation, see H. B. Henbest and R. A. L. Wilson, J. Chem. Soc., 1957, 1958; H. B. Henbest and B. Nicholls, ibid., 1957, 4 6 0 8 .

70. See EXPERIMENTAL of this dissertation.

71* The presence of a 7-carbonyl function adds not only an inductive effect, but also a long-range field effect [L. M. Jackman, Applications of Nuclear Magnetic Resonance in Organic Chemistry, Pergamon Press, New York, 1959* pp* 122-124], but this effect should be approximately the same for both the exo- and endo-a-hydrogen atoms, both being approximately the same direction and distance from the ri-cloud of the carbonyl f u n c tio n . 173.

72. C. F. H. Allen,, T. Davis, D. W. Stewart, and J. A. Van Allan, J. Org. Chem., 20, 306 (1955)*

73* This cation apparently exists as an intermediate in the reaction of diazonorcamphor in dilute acid and in the solvolysis of and 5 2 in aqueous acetic acid and aqueous formic acid [M. Hanack and J. Dolde, Tetrahedron Letters, No. 3» 321 (1966)].

74. H. Kwart and T. Takeshita, J. Am. Chem. Soc., 86, ll6 l (1964).

75. W. G. Woods, R. A. Carboni, and J. D. Roberts, J. Am. Chem. Soc., 5653 (1956).

76. J. D. Roberts and M. C. Caserio, Basic Principles of Organic Chemistry, W. A. Benjamin,. Inc ., New York, 1964, p. 16 5 .

77* C. H. DePuy and P. R. Story, J. Am. Chem. Soc., 82, 627 (i960).

78. Compound 6l "hydrolyzed completely in a few minutes at room temperature with dilute base" (see Ref. 77)•

79* The less stable a carbonium ion is, the more S 2 character the reaction involving its formation exhibits. W[A. Streitwieser, Jr.,. Solvolytic Displacement Reactions, McGraw-Hill Book Company, Inc., New York, 1962, pp. 60-6l].

80. Displacement by strong nucleophile on exo-norbornyl derivatives can give endo product: reaction of exo-norbornyl brosylate with lithium ]D-thiocresoxide is attended with complete inver sion [S. J. Cristol and G. D. Brindell, J. Am. Chem. Soc., 76, 5699 (195*0]-

81. F. V. Zalar,. Ph.D., 1966, The Ohio State University, generously donated an authentic sample of this compound for comparison s t u d i e s .

83* 8. W instein, E. Grunwald, and L. L. Ingraham, J. Am. Chem. Soc., 20, 821 (1948).

82. S. W instein, E. Clippinger, R. Howe., and E. Vogelfanger, J. Am. Chem. Soc., 82, 376 ( 1965 ).

84. S. Winstein and E. Grunwald, J. Am. Chem. Soc., 70, 828 (1948).

8 5 . H. C. Brown, K. J. Morgan, and F. J. Chloupek, J. Am. Chem. Soc., 8 2, 2137 ( 1965 ).

86. S. Winstein, E. Allred, R. Heck, and R. Glick, Tetrahedron, 2> 1 (1 9 5 8 ). 174

8 7. For methyl ether formation via the Williamson synthesis, see J. Meinwald, X. C. Meinwald, and T. N. Baker, III, J. Am. Chem. Soc., 86, 4074 (196*0.

88. The author is not necessarily trying to suggest that 81 must be a distinct intermediate as opposed to a transition state between 6£ and 82. Lack of this distinction»however, has no bearing upon the following discussion.

89* J. D. Roberts and M. C. Caserio, Basic Principles of Organic Chemistry* W. A. Benjamin, Inc., New York, 1964, pp. 44-3-446.

90. S. Winstein and L. L. Ingraham,. J. Am. Chem. Soc., 2!i> H 60 (1 9 5 2 ).

91* D. S. Noyce, B. R. Thomas, and B. N. Bastian, J. Am. Chem. Soc., 8 2 , 885 ( i 960 ).

92. D. S. Noyce and B. N. Bastian, J. Am. Chem. Soc., 82, 1246 (i 960 ).

93* S. W instein, C. Hanson, and E. Grunwald, J. Am. Chem. Soc., 70, 812 (1948).

94. For more data and discussion of the acetic acid/acetic acid-O-d rate ratio, see "Part II. The Norbornenyl Cation" of this dissertation.

95• s * Winstein and M. Shatavsky,. J. Am. Chem. Soc., 78, 5 92 (195& ).

9 6 . H. Tanida, T. Tsuji, and T. Irie, J. Am. Chem. Soc., 88 , 864 (1966 ).

97. P. R. Story, J. Am. Chem. Soc., 8^, 3347 (1961).

98. H, Kwart and W. G. Vosburgh,. J. Am. Chem. Soc., 5400 (1954")•

99• J* K» Crandall, J. Org. Chem., 29, 2830 (1964 ).

100. This compound has also been prepared by K. Mislow and W. E. Meyer. For d etails, see W. E. Meyer,, thesis, New York University, 1964.

101. L. F. Fieser, Experiments in Organic Chemistry, third ed., D. C. Heath and Company, Boston, 1957,. pp. 88- 8 9.

102. For an example of oxidation of a hydroxy-tosylate to the keto- tosylate, see N. A. Nelson and G. A. Mortimer, J. Org. Chem., 22, 1146 (1957). 175

103* This compound has previously been prepared by G. H. Whitham and 3. C. Lewis. For details, see S. C. Lewis, thesis, University of Birmingham, England, 1964.

104. This compound has previously been prepared by H. Krieger, Ann. Acad. Sci* Fennicae, Ser. A, II, No. 109 (1962). Other workers have also prepared the compound: S. C. Lewis, thesis, Univer sity of Birmingham, England, 1964; W. E. Meyer, thesis, New. York University, 1964.

105» J. D. Roberts, C. C. Lee, and W. H. Saunders, Jr., J. Am. Chem. Soc., 4501 (1954).

106* J. D. Roberts and C. C. Lee, J. Am. Chem. Soc., 72 > 5009 (1951)*

107® C* G. Swain, A. J. DiMilo, and J. P. Cordner, J. Am. Chem. Soc., 80, 5983 (1958).

108. R. P. Bell, The Proton in Chemistry, Cornell University Press, Ithaca, New York, 1959, pp. 140-154.

109* For an example of a primary isotope effect in enolization, see 0. Reitz, Z. Electrochem., 4^, 659 (1937)> wherein perdeutero- acetone enolizes 8 times more slowly than acetone in an acid- catalyzed reaction at 25 ° .

110. In the solvolysis of aryl sulfonates where carbonium ion forma tion is the rate-determining step, changing from light to heavy water as a solvolysis medium has little effect on the solvolysis rate [E. R. Thornton, Solvolysis Mechanisms, The Ronald Press Company, New York, 1 9 6 5 , pp. 212-214]. The listed rates for acetolysis of 16 and 103 in acetic acid-0-d (see Table 10) bear testimony that this generalization carries over to acetic acid.

111. 0. Reitz, Z. Electrochem., 43> 659 (1937)*

112. R. Foltz, Battelle Institute, Columbus,. Ohio has generously furnished the mass spectral data discussed in this dissertation.

113. S. Winstein and E. T. Stafford, J. Am. Chem. Soc., 79, 505 (1 9 5 7 ).

114. The mathematical foundation used in treating the kinetic data is discussed in A. A. Frost and R. G. Pearson, Kinetics and Mechanism, second ed., John Wiley and Sons, Inc., New York, 1961, pp. 8-26, 160-199.

115. For the integration of differential equations by the use of integrating factors, see A. L. Nelson, K. W. Folley, and M. Coral, Differential Equations, D. C. Heath and Company, Boston, 1952, pp. 39-41* 176

116. A. Gero, J. Org. Chem.,. lg., i 960 (195*0.

117. A. Gero, J. Org. Chem., 26, 3156 (1961 ).

118. G. W. Wheland, Advanced Organic Chemistry, second ed., John Wiley and Sons, Inc., New York, 1957* p» 5 8 8 .

119* See Ref. 118, pp. 607-609»

120. K. B. Wiberg, Physical Organic Chemistry, John Wiley and Sons, In c., New York, 1964, pp. 273-277.

121. R. P. Bell, The Proton in Chemistry, Cornell University Press, I t h a c a , New Y ork, 1959» PP« 184-187.

122. R. P. Bell and P. W. Smith, J. Chem. Soc. (B), 1966, 241.

123. Strictly speaking, we have shown that must solvolyze faster than the deuterated enol (123). However, for convenience we shall refer to the undeuterated enol (10 5)> for there is no reason to believe m should solvolyze at a rate much differ ent from that of 123.

124. P. R. Story; G. H. Whitham; private communication to P. G. Gassman.

125. The writer is indebted to P. G. Pape, M.S., The Ohio State University, 1962, for this elemental analysis.

126. Sarett oxidation: G. I. Poos, G. E. Arth, R. E. Beylor, and L. H. S arett, J. Am. Chem. Soc., 25* ^22 (1953)*

127. L. F. Fieser, Experiments in Organic Chemistry, third ed., D. C. Heath and Company, Boston, 1957, p. 3 1 6 .

128. Meerwein-Ponndorf-Verley reduction: A. Vogel, Practical Organic Chemistry, Longmans, Green and Company, London, 1956, pp. 882-886.

129. J- J. Hurst and G. H. Whitham, J. Chem. Soc., 1963 , 710.

130. P. R. Story, J. Org. Chem., 26, 287 (1961).

131. Jones oxidation: I. Heilbron, E. R. H. Jones, and F. Sond- heimer, J. Chem. Soc., 1949, 6 o4 .

132. Correction has not been made for the area:weight ratio by inclusion of an internal standard.

133. P* D. B artlett and W. P. Giddings, J. Am. Chem. Soc., 82, 1240 (I960). 177

1 3 4 . L. F. Fieser, Experiments in Organic Chemistry, third ed., D. C. Heath and Company, Boston, 1957* p* 285 .

1 3 5 . S. G. Smith, A. H. Fainberg, and S. W instein, J. Am. Chem. Soc., 82, 618 (1961).

136. F. Daniels, J. H. Mathews, J. W. Williams, P. Bender, and R. A. Alberty, Experimental Physical Chemistry, fifth ed., McGraw-Hill Book Company, Inc., New York, 1958, p. 339.

137- A. A. Frost and R. G. Pearson,, Kinetics and Mechanisms, second ed., John Wiley and Sons, Inc., New York, 1961 , p« 99*

1 3 8 . A. A. Frost and R. G. Pearson, Kinetics and Mechanism, second ed., John Wiley and Sons, In c., New York, 1961 , p . 4 2 .