iv

a . yr??? .3: __.§..ax.,cn.n.e.q..n.

  

..
nol..
..,..
.
.
£.

 
 
 

PART I
NONLINEAR. FREE ENERGY BEHAVIOR OF
2- ARYL - 2 - NORBORNYL CATIONS

PART II ,
195 AND 1H. NMR STUDY OF AN
EQUILIBRIUM AMONG
.ARYLBICYCLOOCTYL moms

I Thesis for the Degree of ‘Ph , D.
MICHIGAN STATE UNIVERSITY
ANTHONY DAVID WOLF
1972

 

M 4
LI [3 RA " 3'
. I‘vlichigan State .
University

p
3 f

0‘“

I

This is to certify that the

thesis entitled

PART I. NONLINEAR FREE ENERGY BEHAVIOR OF
2-ARYL-2-NORBORNYL CATIONS

19F AND 1H KMR STUDY OF AN EQUILIBRIUM '
AMONG ARYLBICYCLOOCTYL CATIONS
presented by

PART II:

Anthony David Wolf

has been accepted towards fulfillment
of the requirements for

Ph.D. Chemistry

degfiehl

 

Date September 21, 1972 "2'

 

ABSTRACT

PART I

NONLINEAR FREE ENERGY BEHAVIOR OF
ZflARYL-Z-NORBORNYL CATIONS

PART II

19F AND 1H NMR STUDY OF AN EQUILIBRIUM

AMONG ARYLBICYCLOOCTYL CATIONS

By
Anthony David Wolf
It is shown that in the graph of the H(l) versus H(3)

chemical shifts for a series of Z-aryl-Z-norbornyl cations

lg, the linear relationship expected for a series of completely
6x

M

2 Anthony David Wolf

classical cations is not observed. Instead, nonlinear
behavior ensues with cations less stable than Z-phenyl-Z-
norbornyl cation. It is also shown that while the non-
linearity is consistent with a o delocalized or carbon-carbon
hyperconjugated description of those ions, it is not
consistent with a description based on rapidly equilibrating
classical ions.

In Part II the following cations were characterized at

O)
O O 0

CF F3 F3

R 2% he

-90°C by a 1H nmr study:

2-p~trifluoromethylphenyl-2-bicyclo[Z.2.2]octy1 cation ga,
Z-p-trifluoromethylphenyl-2-bicyclo[3.2.1]octy1 cation 2a and
6-p-trifluoromethylphenyl-6-bicyclo[3.2.l]octyl cation lla.
Warming any of these cations to -60°C produced the same

19F nmr the order of stability was

equilibrium mixture. By
found to be Ila > éé >> ga. The equilibrium constant between
ga and Ila at -80°C is 3.2. Other thermodynamic parameters

are presented and a mechanism for the equilibrium is discussed.

PART I

NONLINEAR FREE ENERGY BEHAVIOR OF
Z-ARYL-Z-NORBORNYL CATIONS

PART II

19F AND 1H NMR STUDY or AN EQUILIBRIUM

AMONG ARYLBICYCLOOCTYL CATIONS

By

Anthony David Wolf

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Chemistry

1972

To Lizzie For Her

Love And Understanding

ii

ACKNOWLEDGMENT

I would like to express my sincere appreciation to
Professor Donald G. Farnum for his expert guidance, for
the freedom which he always afforded to me in the laboratory

and for the opportunities which he has made available to me.

To Professors David Lemal, Thomas Spencer, Charles Braun
and Gordon Gribble and their graduate students ("68"-"70")
at Dartmouth College, Chemistry Department, who have
generously and enthusiastically contributed to my education

while in the military service, I am deeply indebted.

My fellow graduate students have contributed much to
this thesis. Their varied interests have served to broaden

my own.

Actual laboratory contributions to this study were made
by the following former undergraduates: Miss Irene Pupko,
Mr. Steven Fettinger and Mr. Robert Bruggenmeier. Their

enthusiasm was encouraging.

The successful completion of the nmr studies would not
have been possible without the assistance of Mr. Wayne

Burkhardt and Mr. Eric Roach.

iii

ACKNOWLEDGMENT (Continued)

Over the past two and one half years I have had the

following support for which I am thankful.

March 1970 - June 1970 Teaching Assistantship

June 1970 - September 1971 Research Assistantship (NSF)
September 1971 - December 1971 Teaching Assistantship
January 1972 - June 1972 Research Assistantship (NSF)
June 1972 - September 1972 Departmental Fellowship

iv

TABLE OF CONTENTS

PART I

NONLINEAR FREE ENERGY BEHAVIOR OF
Z-ARYL-Z-NORBORNYL CATIONS

INTRODUCTION.
RESULTS . . . . . . . . . .

General Discussion of Experimental Data in
Tables 1 and 2 . . . . . . . . . . . . . . .

Hydride Shifts in Z-Aryl-Z-norbornyl Cations
DISCUSSION. . . . . . . . . . . . . . . .
EXPERIMENTAL. .

Carbonium Ions . . . . . . . . . . . . .

Carbonium Ion Precursors . .

General Procedure for Alcohol Synthesis Using
the HMIR O O O O O O I O O O O I O O O O O O

Norbornyl System . . . . . . . . . . . . . .
Z-p-iodophenyl-Z-endo-norbornanol . .
2-p-methoxyphenyl-2-norbornene.

Bicyclooctyl System. . . . . . . . .

Z-p-fluorophenylbicyclo[2.2.2]octan-2-ol.
Z-p-methoxyphenylbicyclo[Z.2.2]octan-2-ol .

REFERENCES. . . . . . . . . . . . . . . . . .

Page

15

15
27
34
SO
SO
51

51
SS
55
SS
55
55
SS
56

TABLE OF CONTENTS (Continued)

PART II

19F AND 1H NMR STUDY or AN EQUILIBRIUM

AMONG ARYLBICYCLOOCTYL CATIONS
INTRODUCTION. . . . . . . . . . . . . . .
RESULTS AND DISCUSSION. . . . . . . . . . . . .
Proton NMR Studies . . . . . . . .
Fluorine NMR Studies . . . . . . . . . . .
EXPERIMENTAL. . . . . . . . . . . . . . . . . . . .
General Procedures . . . . . . . . . . . . .
Carbonium Ions Preparations. . . . . . . .
General Synthetic Procedure . . . . . . . . .
Ketones . . . . . . . . . . . . . .
Bicyclo[3.2.1]octan-6-one . . . . . . .

Addition of dichlorocarbene to Z—norbor-
nenone ethylene ketal lg. . . . . . . . .

Formation of IQ by catalytic hydrogenation

of lég and lék. . . . . . . . . . . . . .

Hydrolysis of ketal lg. . .
Synthesis of Alcohols

Labeling Experiments . . . . . . . . . . . .
3,3-dideuteriobicyclo[2.2.2]octan-2—one .

3,3'-dideuteriobicyclo[3.2.l]octan-2-one.

vi

Page

61
63
63
76
83
83
83
84
84
84

84

86
86
87
88
88
88

TABLE OF CONTENTS (Continued)

Page

7,7'-dideuteriobicyclo[3.2.1]octan-6-one. . 91
Recovery Experiments. . . . . . . . . . . . 91
Elimination Experiments . . . . . . . . . . 91

REFERENCES. . . . . . . . . . . . . . . . . . . . . . 95

vii

LIST OF TABLES

TABLE Page
PART I

NONLINEAR FREE ENERGY BEHAVIOR OF
Z-ARYL-Z-NORBORNYL CATIONS

l H(l) and H(3) Chemical Shifts in

Z-Aryl-Z-norbornyl Cations . . . . . . . . . 16
2 H(l) and H(3) Chemical Shifts in

2-Ary1-2-bicyclo[2.2.2]octyl Cations . . . 17
3 Proton Chemical Shifts (T values) in

Related Z-Norbornyl Cations. . . . . . . . 20
4 The Effect of a 100° Temperature

Difference (-60° to +40°) on the Chemical

Shifts of H(1) and H(3) in 1%. . . . . . . 22
5 Data on Hydride Shifts in I; at 100 MHz. . 32
6 00 Values Calculated From Various

Reactions. . . . . . . . . . . . . . . . . 38
7 Chemical Shift Deviations in p-Halogen

Cations. . . . . . . . . . . . . . . . . . 40
8 Preparation of 2~Aryl-2~endo-norbornanols. 53
9 Preparation of 2- -Aryl- 2- -bicyclo[2. 2. .2]-

octanols . . . . . . . . . . . . S4

viii

LIST OF TABLES

TABLE Page

PART II
19F AND 1H NMR STUDY or AN EQUILIBRIUM
AMONG ARYLBICYCLOOCTYL CATIONS

19

1 Thermodynamic Data from P NMR

Experiments at 94.1 MHz. . . . . . . . . . . 77
2 Proton Chemical Shifts in Arylbicyclo-

octyl Cations. . . . . . . . . . . . . . 85
3 Preparation of 2-Aryl-2-Bicyclo[2.2.2]-

octanols . . . . . . . . . . . . . . . . . . 89
4 Nmr and ir Data of Isomeric Arylbicyclo-

octanols . . . . . . . . . . . . . . . . . . 90
S Quenching Products From Several Isomeric

Arylbicyclooctyl Cations . . . . . . . . . . 92

6 Nmr Spectra of Representative Olefins
Prepared by Dehydration. . . . . . . . . . . 94

ix

LIST OF FIGURES

FIGURE Page

PART I

NONLINEAR FREE ENERGY BEHAVIOR OF
2-ARYL-2-NORBORNYL4

1 Graph of H(l) vs. H(3) for classical
norbornyl cations with no charge
leakage to C(l). . . . . . . . . . . . . . . 12

2 Graph of H(1) vs. H(3) for norbornyl

cations with charge leakage to C(l). . . . . 13
3 Graph of H(l) vs. H(3) chemical shifts

in Z-aryl-Z-norbornyl cations lg . . . . . . 35
4 Graph of H(l) vs. H(3) chemical shifts

in 2-ary1-2-bicyclo[2.2.2]octyl cations $4 . 36

PART II

19F AND 1H NMR STUDY or AN EQUILIBRIUM

AMONG ARYLBICYCLOOCTYL CATIONS
1 100 MHz spectrum of £3 at -90°C. . . . . . . 64
2 100 MHz spectrum of 23 at -100°C . . . . . . 68
3 100 MHz spectrum of 113 at -90°C . . . . . . 74

PART I

NONLINEAR FREE ENERGY BEHAVIOR OF
Z-ARYL-Z-NORBORNYL CATIONS

INTRODUCTION

Thorough reviews on the subject of norbornyl cations
are already available in the literature."5 We will simply
restate the controversial aspects of the problem which have
been formulated elsewhere‘»2 and we will focus on those
experiments which are relevant to our results.

The nonclassical ion problem has been formulated so as
to require a distinction between two alternative descriptions
of the norbornyl cation.1 That is the norbornyl cation is a
rapidly equilibrating pair of classical ions la and lb“ or

it is a o delocalized cation7 g.

 

 

Te De

Three main foundations have been summarized as support
for the latter description of the ion."2
(1) unusually fast rates of solvolysis for exo norbornyl

derivatives.

(2) Predominantly exo substitution.
(3) Large exo/endo rate ratios.

The first point no longer seems to be tenable when
suitable model systems are chosen for comparison.“8 Neither
does the preponderance of exo substitution in solvolysis
demand a nonclassical description of norbornyl cations.l
Tortimmfl.effects seem to control the stereochemistry observed
in many reactions of norbornyl systems.9 The remaining point
concerns the exo/endo rate ratios observed for norbornyl
derivatives. Brown argues that tertiary norbornyl derivatives
give rise to exo/endo rate ratios on the order observed for
the parent norbornyl system. For example it has now been
established that Z-phenyl-Z—norbornyl cation is a classical

cation,l yet the solvolysis of the exo/endo p-nitrobenzoates

in this system yield rate ratios which are not significantly
different from the ratio observed in the parent norbornyl
system.ll Furthermore it has been shown that these differences
arise predominantly in the transition state for solvolysis
rather than in the ground state.12"3 Thus either the
norbornyl cation is classical or there must be a unique inter-
pretation for the origin of the exo/endo rate ratio in this

cation if the nonclassical description would obtain. Schleyer

has offered a resolution.” He suggests that tortional and
nonbonded interactions in the tertiary endo derivatives might
largely be responsible for the observed exo/endo rate ratios.”
Thus the original basis of support for nonclassical character
in norbornyl cations must be rejected, since it has been
shown that these criteria are not specific in indicating
nonclassical character.

Deuterium isotope effects in the transition state for
solvolysis of norbornyl derivatives and studies of stable
norbornyl cations are not adequately explained by classical
norbornyl cations. In the following we cite examples of
the latter which are pertinent to our results.

Studies of kinetic isotope effects in the solvolysis
of the following systems was reported simultaneously by

two different groups.”'15 No isotope effect was observed

H OBS D OBS H D
Re éle Se tie

for solvolyses of the endo brosylates ég and ék.‘“,’” It
was reported that this result was consistent with the y

isotope effect expected for solvolysis via a classical

6&0

cation.”:15 In the axe brosylates 4% and 4b however, a 10

positive isotope effect was observed for deuterium in either
the exo or the endo position at C-6."’tls It was concluded
that these isotope effects were too large to be accounted for
by classical transition states.”'15

It seems significant that exo-6-D 4g and end0-6-D AR 2-
exo-brosylates both have large y isotope effects. It is
difficult to account for these on the basis of solvent effects
and steric interactions in the solvolysis of the exo
derivatives.‘”

The absence of a kinetic isotope effect in the solvolysis

of the endo brosylates 33 and 3b is interesting* since other

 

* Kinetic isotope effect studies by Karabatsos and Smith‘6
on 2,6,6-trimethyl-endo~norbornyl-p—nitrobenzoates é

CH
3
CH3

PNB

é

have indicated that a substantial isotope effect is operating
(based on enthalpy only) in both the exo-6-trideuteriomethyl

CH3

CD CH3 CH
3 3
CH3 PNB CD3 OPNB
2e QR
system Ea (kH/kD .78) and the endo-6-trideuterio methyl
system QR (kH/kD = .74). This isotope effect is only revealed

however when the kinetics are obtained as a function of

3 ”CH

temperature. The enthalpy contribution to the isotope effect
is offset by a compensating entropy of activation. kH/kD
values based on free energy are greater than 1.00 over the
temperature range studied with one exception!

 

experiments have indicated that steric deceleration may
play at least some role in the solvolysis of endo norbornyl
derivatives.“17 An interesting tricyclic derivative of

brexane 1a offers the opportunity to evaluate the relative

. e

H . n
X

 

1% 1R

importance of anchimeric assistance or steric deceleration
in the transition state for solvolysis of era and endo

” This system is

norbornyl derivatives respectively.“l
unique in that the leaving group X in 1b suffers similar
tortional restraints and non-bonded interactions as does
endo-norbornyl, while at the same time it is set up
geometrically for anchimeric assistance as is exo-norbornyl.
This compound was found to solvolyze at 66% of the rate of
exo-norbornyl (or 33% of the rate of ionization of exo-
norbornyl) or 225 times faster than endo-norbornyl.s A
crude correlation of these results with the rate ratio

observed with the parent norbornyl derivative leads to an

estimate of 66% anchimeric assistance in exo norbornyl and

33% steric deceleration in endo norbornyl in the transition
state for solvolyses of these compounds.*

If it is assumed that a carbonium ion in a medium which
does not solvate positive charges very well is structurally
identical to a carbonium ion formed in solvolytic media,
then the results obtained from the study of carbonium ions
in super acid has direct bearing on the nonclassical ion
problem.

The direct observation of the rapidly equilibrating
and presumably classical 1,2-dianisylnorbornyl cations 8a

and gbl” supports the proposal of rapidly equilibrating

 

 

* Swartz has indicated that other factors may be
contributing in part (or in total) to the observed rate of
solvolysis of the Z-brexyl derivative 13. For example, the
slower rate of solvolysis of 2-brexyl relative to exo-
norbornyl might in part be due to a poorer geometry for
anchimeric assistance in the former case. It is also
pointed out "that relative to endo-norbornyl the Z-brexyl
compounds have less tortional relief of 1,2-ec1ipsing strain,
more hinderaibackside solvation of the incipient ion and
possible extra endo hindrance to ionization."s

 

structures for norbornyl cations. Indeed the interconversion
between 8a and 8b is rapid on an nmr time scale even at -70°C.
It has been argued that the behavior of this cation is not
unexpected. The carbonium ion center is greatly stabilized
by the p-anisyl group and should not require 0

delocalization.2°

This latter proposal has been rejected on
the basis that absence of nonclassical character in this
system would represent a loss of > 6 kcal/mole nonclassical
resonance energy.1 At the same time it was proposed that
steric repulsion between the anisyl groups might inhibit
nonclassical resonance, but that the steric problem should
be absent in the 1,2-dimethylnorbornyl derivatives.1

The 1,2-dimethy1norbornyl cation is indeed a most

interesting cation. It has been characterized as a partially

C-6H o delocalized equilibrating cation 93 and 9b on the

   

13

basis of a comparison of C and 1H nmr Spectra with model

compounds.21

The equilibrium in a related system 10% and
10b has been observed directly and does not depend on the

use of model compounds.22 The equilibrium between 10a and

 

 

CH3 I CH3

e > I 3 H

H3 3

H

CH3 CH3 CH3 3

Re $918

lgk is fast on an nmr time scale even at -130°C but it is
detected by the temperature dependence of the methyl
resonances in these two cations of slightly different

structure.22

This observation lends support to the proposed
equilibrium between 9% and 2b.

With the above examples of rapidly equilibrating
cations in mind one might wonder if there is indeed a
rapid equilibrium between la and lb. The energy barrier
to interconversion of 8a and 8b must be greater than it is
for the interconversion of 9a and 9b as no broadening of
the nmr resonances is observed even at -140°C in the latter
case while broadening is observed at -70°C in the former
case. Thus one might argue that the expected barrier to
interconversion of la and lb would be even lower than it
is for lga and lgb either on an electronic basis, a steric
basis or both.

The cnmr and pnmr studies23 of the parent norbornyl

cation however indicate the presence of a species which

cannot be accounted for by two rapidly equilibrating classical

10

structures. Further support for this conclusion comes from
the photoelectron spectrumz“ of the norbornyl cation. On
the basis of the above studies it was concluded that the
norbornyl cation is best described as a symmetrical o
delocalized nonclassical ion l.2”

If one assumes for the moment that nonclassical
character has been demonstrated for norbornyl cations then
the following experiments reported by Brown and Takeuchi
present a dilemma.2”'2‘a They have correlated the rates of
solvolysis of a number of Z-aryl-Z-exo-norbornyl derivatives
ll with 06 values.2”:‘“”a They found that the correlation was

linear with none of the curvature expected if 0 delocalization

Us

were an important factor.26a One would expect that o
delocalization would alter the linear dependence of log k
on 09. It was pointed out that the studies were carried
out over 60% of the reactivity range from p-anisyl-Z-
norbornyl to norbornyl without any indication of o

delocalization.26a These observations coupled with the

observed exo/endo rate ratios led the authors to conclude

11

that participation if present in norbornyl cation cannot

be large.26a It seems intuitively unreasonable to us that
the norbornyl cation should be the only derivative of
norbornane exhibiting charge leakage to C(l). Our primary
goal was to try to resolve this inconsistency. Accordingly

we have prepared a number of Z-aryl-Z—norbornyl cations ll

Ar

1%

in which the aryl group is varied so as to make increasing
electron demands at the carbonium ion center which in turn
makes increasing electron demands on the norbornyl skeleton
especially at C(2) and C(3). The effect of increasing
positive charge on the chemical shift of H(1) and H(3) is
a shift to lower field for these protons.

On the basis of earlier work1° our hypothesis was that
a graph of the chemical shift of H(l) versus the chemical
shift of H(3)* should be linear with increasing positive
charge (Figure 1) provided there is no change in the mechanism
for transmission of the charge to each of the sites under

consideration. This can be stated in another way; the

 

* We have plotted the average shift of H(3) era and
H(3) endo.

 

12

A

"3
increasing 3
positive
charge Th(1) "1 Ar
at
C(Z)

 

 

H(3)
increasing positive charge at C(Z)

——->

Figure 1. Graph of H(l) vs. H(3) for classical norbornyl
cations with no charge leakage to C(l).

chemical shift difference between H(l) and H(3) will be
constant for classical arylnorbornyl cations.

A deviation in the linear relationship between H(l)

and H(3) will be observed if charge leakage to C(l) occurs
by 0 delocalization ll or by a mechanism which alters the
mode of charge transmission to C(l) relative to C(3) or
vice versa (Figure 2). This can be stated differently;
the chemical shift difference between H(l) and H(3) should

increase relative to this same difference in a classical

13

 

 

 

I N
:2."
b, A
530 «1% Ar IRE/I AT
3'3???
O'HCUU
seem
.Ee F
ii Ar
%
H(3)

increasing positive charge

.5)

 

Figure 2. Graph of H(l) vs. H(3) for norbornyl cations
with charge leakage to C(l).

cation if charge leakage to C(l) becomes an additional factor
in the origin of these particular chemical shifts. Charge
leakage to C(l) will have the effect of increasing the
chemical shift of 3(1) relative to its value in a classical
cation while decreasing the chemical shift of H(3) relative
to its value in a classical cation. Thus the chemical shift
difference between H(l) and H(3) should increase.

In the interest of having a model system for comparison
to ll we have also examined the H(l) and H(3) chemical
shifts in a corresponding series of 2-aryl-2-bicyclo[2.2.2]-

octyl cations ll. This system is known to have less tendency

14

to exhibit those properties usually associated with o

 

delocalization.‘° Thus a plot of H(l) vs. H(3) with
increasing positive charge at C-2 should be linear or any
nonlinearity observed should "lag behind"‘° any non-

linearity observed for the arylnorbornyl systems ll.

RESULTS

General Discussion of Experimenufl Data in Tables 1 and 2
Standard procedures were used to prepare the cations
for this study. The details are presented in the experimental
section. The chemical shifts of the protons of interest
i.e. H(l) and H(3) in cations ll and ll are listed in tables
1 and 2 respectively. The data were obtained at 100 MHz.
The chemical shift of proton H(l) in cation lla is
reported as a range of values I 5.68—5.71 because this
resonance appeared as a shoulder (actually the lower field
portion of a broad doublet) of the methyl resonance of
the p-methoxy group. That this assignment was correct was
verified by decoupling experiments. It was reported
previously that in Z-phenyl-Z-norbornyl cation lll, H(l)

is coupled to H(6) exo with an unusually large coupling

H6

J = 6 HzI:\_”
1 C6HS

«Iii.

constant.‘° When lla was irradiated in the region in which

H(6) era was expected to absorb the shoulder due to H(l)

15

16

H(l) and H(3)a Chemical Shifts

b

 

 

Table 1. in
2-Aryl-2-norborny1 CationsC
Cations ll Aryl Group H(l) H(3) Av(Hz)c
a6 p-CH3OC6H4 5.68-5.71 6.72 brSa
R p-HOC6H4 5.64 5.59 br sd
g 3,4-(CH3)2-C6H3 5.34 6.48 9
g p-CH3C6H4 5.32 6.46 10
g p-FC6H4 5.27 6.42 11
p-ClC H 5.22 6.38 14
6 4
g p-BrC6H4 5.22 6.47 19
hf p-ICGH4 5.25 6.63 28
l C6H5 5.17 6.34 14
l m-BrC6H4 5.09 6.32 16
l m-C1C6H4 5.08 6.32 17
l p-CF3C6H4 4.93 6.25 22
m 3,5-(CF3)2C6H3 4.75 6.20 26
6'
n 3,5-(C1)2-4-N(CD3)2H 4.63 6.34 -
p 3,5(Cl)2-4-CN

 

 

 

 

 

 

aAverage chemical shift of H(3) era and H(3) endo.

b1 values relative to internal standard of tetramethyl-
ammonium tetrafluoroborate (T 6.87)26

cDetermined at -60°C unless otherwise indicated (H(3)ex0-H(3)endb)-

d

”gig (50°), 2n (-100°).

f

Literature values.

(100 MHz).

Center of H(3) exo, H(3) endo multiplet.

 

17

Table 2. H(l) and H(3) Chemical Shifts8 in
2-Aryl-2:bicyclo[2.2.2]octyl Cationsb

 

 

 

 

 

Cations ll Aryl Group H(l) H(3)
g p-cusoc6n4 6.03 6.42
b 3,4-(CH3)2C6H3 5.75 6.17
g p-CH3C6H4 5.74 6.15
g p-FC6H4 5.68 6.09
g p-c1c6H4 5.66 6.09
g p-Brc6H4 5.65 6.19
g p-Ic6H4 5.66 6.36
kc C6H5 5.57 6.02
i m-Brc6H4 5.53 5.96
i m-c1c6H4 5.52 5.95
R” p-CF3C6H4 5.33 5.83
l 3,5-(CF3)2-C6H3 5.15-5.29 5.78

 

 

a

T values relative to internal standargbof’tetramethyl-

ammonium tetrafluoroborate (r 6.87).2

bDetermined at -60°C unless otherwise indicated.

CLiterature values.

dlgg (-80°C).

 

18

collapsed under the p-CH30 resonance thereby confirming the
assignment. The range T 5.68—5.71 is the limit over which
the center of the H(l) resonance should be found.

The chemical shift of H(l) in cation lll is also reported
as a range of values albeit for a completely different reason.

Careful ionization of olefin ll in FSO3H/SOZC1F to which was
CF3
CF

lé 3

added SbFS (to insure complete ionization) at -120°C gave

a mixture of cations as noted by the nmr spectrum at -100°C.
Since the cation of interest was a minor component of the
mixture identification of H(l) was not possible. However

the range of T values in which H(l) should absorb is reported.
The broad singlet (2H) for H(3) in lll (T 5.78) could be
identified.

The observation that a mixture of cations results on
ionization of ll has led us to make a thorough study of the
equilibration that occurs among certain aryl bicyclooctyl
cation isomers. We have reported these results elsewhere.27

All chemical shifts are reported at -60°C for cations

ll and ll with the exceptions of cations lll, llh,
(literature values‘°) lla, lln and llk.

19

The chemical shifts of cation lla mentioned above are
reported at -30°C. This was due to the fact that at this
temperature H(l) is better resolved from the p-CH30
resonance. At -60°C resolution of H(l) is quite poor,
because of viscosity broadening by FSOzH.

The chemical shifts of cation llk are reported at ~90°C,
since at ~60°C, this cation is rapidly converted to a mixture
of cations in which lll is a minor component.27 The
activation energy for rearrangement of lll is less than
it is for cation llk. Thus we were able to determine its
chemical shifts prior to rearrangement.

Cation lln undergoes a rapid Wagner-Meerwein, 6,2-
hydride shift,2° 2” Wagner—Meerwein rearrangement.

At -60°C the H(l) and H(3) resonances are broadened due to
averaging with other protons in the system. This interesting
rearrangement will be discussed in detail in a later section.

It has been reported that under some conditions of

ionization 2,2-dichloronorbornane l6 yields a mixture

1b.. “‘> “

C1 C1

$8 41 «IA

of both Z-chloronorbornyl cation ll and protonated 4-chloro-

Z-norbornyl cation ll.3° We have never been able to observe

20

cation llp. Ionization of olefin lg in FS03H/SbF5/SOZC1F

 

even at temperatures lower than -100°C produced cation l0.
This assignment was based on a comparison of the nmr
spectrum of l0 with that reported for ll and the Z-norbornyl
cation l (Table 3).3°

Table 3. Proton Chemical Shifts (T values) in
Related Z-Norbornyl Cations

 

 

 

Cation Hfiflgig:—ggzlngg 'HISEIJHI5)‘ H(4) Ar
lga 4.70 7.58
g3 4.65 7.80 6.85
ggb 4.62 7.52 2.14

 

 

 

 

 

 

 

3Chemical shift relative to capillary tetramethylsilane.

bChemical shift relative to internal tetramethylammonium

tetrafluoroborate.

21

There is a broad singlet (2H) at 2.14 which is assigned
to the two aromatic protons.* There is another broad
singlet (4H) at lower field which is assigned to H(l),

H(Z), H(6) era and H(6) endo. These protons are equivalent
as a result of a rapid 6,1,2-hydride shift.3° As a result
of this same rearrangement H3, H5 and H7 are also
equivalent. Thus this spectrum is consistent with the
assigned structure and analogous to those of l8 and l which
have already been characterized.”°

Since we found it necessary to report some of the data
at temperatures other than -60°C and since the chemical
shift difference between H(l) and H(3) for ll and ll is
important to the interpretation of our data we determined
the change in these chemical shifts over a 100°C temperature

range (-60° to +40°). These data are reported in Table 4.

 

* We do not know for certain if the -CN group in lg
is protonated. The chemical shift of the aromatic protons
in 20 is .51 ppm to lower field compared to the aromatic
protons in olefin l9 and .21 ppm to lower field compared
to the corresponding 2-aryl-2-endo-norbornanol. Olah and
coworkers have reported that alkyl nitriles are completely
protonated in FSOSH/SbFSISOZ.31 This report coupled with
the shift of the aromatic protons to lower field compared
to starting materials suggests that the CN group is
protonated.

 

22

 

 

Table 4. The Effect of a 100°C Temperature Difference
(~60° to +40°) on the Chemical Shifts of H(l)
and H(3) in ll.
ation (9&3!) (_ggg) (Egél) ($23”: AH(1)b AH(3)c
llg - 6.75 - 6.72 - 3
lld 5.33 6.46 5.37 6.51 4 S
llg 5.27 6.42 5.32 6.47 S S
lll 5.22 6.41 5.28 6.46 6 S
llg 5.22 6.47 5.29 6.54 7 7
llh 5.25 5.63 5.31 6.70 6 7
lll 5.16 6.35 5.19 6.39 3 4

 

 

 

 

 

 

 

 

aAverage chemical shift of H(3) era and H(3) endo.
bH(l) (+40°) - H(l) (-60°) (value in Hz).
”H(3) (+4o°) - H(3) (-60°) (value in Hz).

 

23

There is a distinct shift to higher field for protons
H(l) and H(3) in every cation in Table 3 when the temperature
is raised 100°C. The explanation for this observation is not
clear. What is clear however is that this effect is
operating to the same magnitude and in the same direction
(i.e. higher field with increasing temperature) for both
H(l) and H(3). Thus the chemical shift difference between
protons H(l) and H(3) is constant over a wide temperature
range and our interpretation of the data in Table 1 should
not be significantly affected by any temperature variation.
It is interesting nevertheless that the magnitude of this
effect is variable from one cation to another. We will
return to this point when we discuss the p-halogen cations.

A noteable observation concerns the multiplicity of
the resonance due to the H(3) era and H(3) endo proton in
cation ll. As mentioned earlier we reported the average
chemical shift here since an AB quartet having a variable
chemical shift difference is observed. For example in cation
lle and llb the H(3) era and H(3) endo protons give a broad
singlet in the nmr spectrum. However in all the other
cations of the aryl norbornyl series the H(3) endo proton
is the lower field component of an AB quartet having a

chemical shift difference dependent upon the magnitude of

24

the positive charge at C(2)* (except for cations lll, llg
and llh which will be discussed). As more potent electron
withdrawing aryl groups are placed at C(Z), the magnitude
of the chemical shift difference Av between H(3) axe and

H(3) endo increases as shown in Table 1.**

 

* We would like to correct several chemical shifts
reported earlier for Z-phenyl-Z-norbornyl cation.‘° H(3)
era and H(3) endo should be interchanged. This is based
on the observation that in the nmr spectrum of cation ll
produced from partially deuterated olefin ll the higher

 

 

4%

field portion of the AB quartet due to H(3) era and H(3)
endo has collapsed and is of greater intensity than the
lower field component of the quartet. Assuming exo
protonation by FSOSH H(3) endo must be at lower field. In
extending this assignment to the other cations in series
ll (except lll, llg and llh which we are not assigning) it
is assumed that no factors are operating that would inter-
change H(3) era and H(3) endo.

Proton H(6) endo should also be interchanged with an
H(S) proton in cation lll.1° This reassignment is based on
decoupling and DISST experiments (to be described),in cation
lll. We assume that these assignments will hold for 2—
phenyl-Z-norbornyl cation.

** We will show later that this effect is independent
of charge leakage to C(l).

 

25

A possible explanation for the selective deshielding
of H(3) endo (or shielding of H(3) exo) might be due to
selective hyperconjugation of H(3) endo over H(3) era with
the p orbital at the carbonium ion center. Models indicate

however that introduction of an sp2

center into a perfectly
eclipsed norbornane ll leads to a cation ll in which the

p-orbital is symmetrically diSposed with respect to these

H(3) exo

CZ-C bond

H
H

62 £4

H(3) endo

two protons. It is difficult in this case to see why H(3)
endo should be affected differently than H(3) era by a
variable positive charge of C(Z).

Twisting the cation causes the six-membered ring to
approach a twist boat conformation. In cyclohexane the
twist boat is an energy minimum with respect to the boat

2 In fact X-Ray diffraction studies and

conformation.3
calculated conformations for a number of 2-substituted
norbornanes indicate that these molecules are twisted. Thus
there is at least a possibility that the cations ll are also
twisted."3

Counterclockwise rotation around the Cz-C3 bond in

ll has the effect of aligning the p-orbital at C-2 with

26

H(3) endo while its relation to H(3) exo approaches an

orthogonal one ll. Thus in the counterclockwise twisted

 

H(3) endo

cation there is a greater potential for interaction of the
p-orbital at C(Z) with H(3) endo.

Whether or not relief of eclipsing interactions would
actually warrant twisting is speculative at best, especially
since there does not seem to be any reason to prefer the
countercflockwise twisted ion over the clockwise twisted ion.*

It seems likely that some other factor is responsible
for the observed variation in AU for H(3) era and H(3) endo
in cations ll. One possible explanation might be due to
the effect of solvation. As charge builds up at C(Z)
there should be a tighter association of the cation center
with the solvent. This could have the net effect of shielding

H(3) are relative to H(3) endo.3”

 

* In fact, since exo H migrates faster than endo
H in the 3,2-shift, there is reason to prefer the other
conformation.

 

27

Hydride Shifts in 2-Aryl-2-norbornyl Cations

 

One of the most interesting phenomena encountered in
this study concerns the Wagner-Meerwein, 6,2-hydride shift,
Wagner-Meerwein rearrangement which we have observed for a
number of cations in the norbornyl series ll. This

rearrangement is formally depicted in Scheme l,and it is

Scheme 1

   

1% Iéii

degenerate. A considerable simplification of the nmr
spectrum is observed because of the symmetrization which
is imposed when the hydride shift in iéii is rapid on an
nmr time scale as shown below for cation lll. The
complicated multiplets at higher field for protons on
C(5) and C(7) collapse to an AB quartet centered at 7.95,
while the aromatic protons and H(4) remain invariant.

The H(3) era and H(3) endo AB quartet collapses to a singlet.

28

7.9(m) 7.95(ABq)
H

    
  
 

 
 
  
  

6.73(br s) 6.73(br s)

 

(L36
7 9(m). H
8 4(m)H 6.14 “.3 ‘1 3‘33). }6.25(br s)
6 7(m)H fi" 6.7 I
8 4(m) H H

w. 3

H(6) endo and H(l) average to a singlet with the same
chemical shift as H(6) exo. Thus it was not possible to
tell whether H(6) endo, H(l) and H(6) exo were all
exchanging or if only H(6) endo and H(l) were exchanging
while H(6) era was remaining invariant. We were able to
distinguish between those two possibilities by using The
Double Irradiation Spin Saturation Transfer (DISST) Technique
that was applied recently by Sorensen, Huang and
Ranganayakulu to methyl substituted norbornyl cations.35
The success of the DISST technique depends upon the
half-life for chemical exchange being shorter than the
half-life for proton relaxation. At -10°C the coupling of

H(6) are to H(l) in cation lll is nearly wiped out as a

29

 

result of the onset of chemical exchange. If H(l) is
irradiated at this temperature (exchange is slow), the

two proton multiplet at highest field decreases by 25%

(or 50% of the integrated intensity of H(6) endo).* The
integral in the region in which H(6) era and H(4) absorb

is increased36 by 25% which corresponds to an increase of
approximately 50% for H(6) exo! Presumably this represents
an increase in the intensity of H(6) are rather than H(4).
Thus H(6) era is not exchanging with H(l) or H(6) endo.
H(6) exo also increases in intensity by 45-50% on
irradiation of H(6) endo in the absence of chemical exchange.
This is a large positive Nuclear Overhauser Effect. 50%
is the maximum enhancement for protons that is predicted

from theory.36

 

* This observation allows us to assign the chemical

shift of H(6) endo - see footnote, p. 24.

 

30

A recent review363 cites many examples of chemical
systems in which the Nuclear Overhauser Effect (NOE) has
been observed. However there were no cases of molecules
cited having geminal methylene groups exhibiting an NOE
enhancement. This may simply be a result of a practical
problem. A substantial chemical shift difference between
the observed and irradiated protons (25 Hz)36b is necessary
for a successful determination of an NOE enhancement. In
most structures having geminal methylene protons the chemical
shift difference between these protons is generally small.

In Z-substituted norbornyl cations however the chemical shift
difference between the geminal protons at C(6) can be
substantial. For example in cation lll the difference
between H(6) era and H(6) endo is 170 Hz. The closest
analogues reported in the literature are structures having
geminal methylidene protons. In these cases the protons

are also very close together (in space as are geminal
methylene protons) and the NOE enhancements are large. For
example in methyl methacrylate ll irradiation of H(a) gives

rise to a 48% increase in the integrated intensity of H(b).37

 

31

Table 5 summarizes the approximate coalescence temperatures for
6,2 llm in several cations. Cation lag presumably would have a 6,2 H»
with a temperature dependence which parallels norbornyl
cation.23

We also list in Table 5 the temperature for which
we observe the onset of a second reversible process. We
believe that this broadening is a result of a 3,2-hydride
shift completely analogous to the 3,2 Hm in norbornyl
cation.’” The cations decompose however before this process
results in complete coalescence.

Cation llm,already undergoing rapid 6,2 H'9 at +69°,
undergoes further changes due to a 3,2 HW. The AB quartet

due to protons on C5 and C7 collapses to a very broad signal

 

MUCH

(58 Hz — width at half height). At lower field the signals
due to protons on C1, C3, C4 and C6 collapse to a broad

singlet also. This process is reversible. The expected

32

Table 5. Data on Hydride Shifts in ll at 100 MHz

 

 

Cation 6,2 H~a 3,2 Hm
141 *4°° ‘

448 +80° '

Iii +20 +60°b

Tie -40° +70°c

lit -40°-0° +30 to +40°b'
422 <-80° -

 

 

 

 

 

aApproximately lowest temperature for which complete
coalescence is observed as a result of 6,2 Hm.

Temperature at which broadening is observed as a
result of 3,2 Hm.

CSee text.

33

singlet which would result from total exchange of all
protons in the system is not observed since the sample

decomposes.

DISCUSSION

Figures 3 and 4 contain in graphic form the information
presented in Tables 1 and 2 respectively. Before discussing
the meaning of the data we will first note several features
wich are apparent from the graphs. First the ordering of
the points in general follows the 0° values for these
substituents.3” A strong electron releasing group places
relatively little positive charge at C(Z); hence the nmr
chemical shifts of protons H(l) and H(3) occur at relatively
high field while a strong electron withdrawing group places
relatively more positive charge at C(Z); hence the nmr
chemical shifts of H(l) and H(3) occur at relatively lower
field. There are several exceptions to this trend which
are noteworthy. For example in the arylnorbornyl series
of cations (Figure 3) a p-methoxy substituent stabilizes
a positive charge better than a p-hydroxy group. This is
opposite to what would be expected on the basis of 00

values.””

One possible explanation involves a rapid
protonation equilibrium between cations llk and llbll.

Since it would be expected that llbll would have H(l)

34

.35

 

 

 

 

 

9 I
e‘ \+ 3,5-(C1)2-4-N€CD3)2H
4.7 - ,’
*' 3,5-/(CF3)2
lo. ’
9 I. +'p-CF3
/
5.1 _ ’+ m-Cl
81(1)
5.3 _
5.5 b
5.7 -
J l l I, 1
6.6 6.4 6.2 6.0 5.8

H§(T)

Figure 3. Graph of H(l) vs. H(3) chemical shifts
in 2-aryl-2-norbornyl cations ll.

36

 

 

 

 

 

5.2 - + 3,5-(CF)
I 3 2
/
/
'/ + p-CF
/ 3
5.4 - /
/
+ m-Cl
+ m—Br
5.6 I-
H1(t)
5.8 I-
6.0 .-
1 44 l
5.8 5.6

 

Figure 4. Graph of H(l) vs. H(3) chemical shifts
in 2-ary1-2-bicyclo[2.2.2]octyl cations

i4-

37

FSO H ,/H
3 a a;
<+——-—- il;I <§ii>I ”an

ték Iékii

 

and H(3) chemical shifts at considerably lower field than
llb, the averaging due to the rapid equilibrium would
cause the observed values of H(l) and H(3) for llb to be
at lower field than expected.* This argument will hold
as long as the extent of protonation of cation llb is
greater than it is for lll. It is not at all clear however
why this should be the case.

A second possible explanation for the above observations
lies in the origin of the (T0 values. Brown and Okamotoa”

have reported 06 values for p-HO and p-CH30 substituents

 

* The chemical shift difference between H(l) and
H(3) in the averaged cation should be the same as that
difference in cation llb (as long as charge leakage to
C(l) is not extensive in the other component of the
equilibrium i.e. cation llbll). The net effect would
be a displacement of the point for p-HOC6H4 - derivative
to lower field along the straight line position of the
curve shown in Figure 3.

 

38

which they have calculated from the data of various reactions.

These results are shown in Table 6.

Table 6. 0’6 Values Calculated From Various Reactions”

 

 

Reaction Aa BB CC
0 ~12.l4 -4.32 -3.44
p-HO - .969 - .933 - .833
p-CH30 - .826 - .736 - .721

 

 

 

 

 

 

aUncatalyzed bromination of aromatic derivatives in
acetic acid at 25°C.”°

bProtonolysis of substituted triphenylmethylsilanes
by perchloric acid in aqueous methanol at 51.2"."1

C o o o o o o
Ionlzatlon of triphenylcarblnols 1n aqueous sulfuric
acid.“"’3

In the case of reaction A it has been postulated that
the observed substituent effects simply reflect more
effective hyperconjugation by H versus CH3.“° On the
other hand to explain the results in reaction B Earborn
has claimed that equally important to any hyperconjugative

effect would be a hydrogen bonding interaction between

39

solvent and hydroxyl group in determining the relative
00 values for p-HO and p-CH30.“ An explanation of our

result based on solvent effects would indicate that in llb

F

H“"(|)SOS

H
06

we

weakening of the protonated oxygen-hydrogen bond by the
weak base FSOSH would be much less important than it would
be for solvents used in the determination of 06 values in
Table 6. Thus there might be a reversal of these two 00
values in FSOSH.

A curved line* in the case of Figure 3 and a straight
line in the case of Figure 4 satisfactorily accommodate
the bulk of data in Tables 1 and 2 respectively. There

is some unusual effect which is present in the case of the

 

* We have not included the point for lln in the curve.
Since the chemical shift of H(3) is not known with complete
certainty. Labeling experiments are necessary to verify this
assignment. The present assignment is based on the fact that
at ~100°C the 6,2 Hm in lln is slow on an nmr time scale, but
H(3) exo, H(3) endo and H(6) exo are undistinguishable. At
higher temperatures a broad singlet is formed at T 6.34 before
H(l) (and presumably H(6)endo) coalesce to a broad singlet
in the same region of the spectrum. This singlet is probably
due to the averaged H(3) era and H(3) endo protons.

 

40

p-halogen substituted cations however which causes those
points to fall "off the line" in both the aryl norbornyl

and arylbicyclooctyl series. We can estimate from 09 values
that the p-halogen derivatives should fall between phenyl
co”

and H(3) chemical shifts at about I 5.12 and r 5.32

- 0)” and m-bromo (o9 = .405) .””and would have H(l)
respectively. The approximate chemical shift differences
between observed and expected values are shown in Table 7*

along with the chemical shift differences between H(3) exo

Table 7. Chemical Shift Deviations in p-Halogen Cations

 

 

+3 F C (I

E—Halogen o AH1 (Hz) AHé¥(Hz) Av
-C1 (lll) .114 9H2 5 Hz 14 Hz
-Br (llg) .150 9H2 15 Hz 19 Hz
-I (llh) .135 lZHz 31 Hz 28 Hz

 

 

 

 

 

 

aValues from Ref. 39.

b1 H(l) observed - T H(l) expected.

CT H(3) observed - T H(3) expected.

dChemical shift difference between H(3) era and H(3) endo.

 

* The chemical shift difference between observed and
expected values for H(l) and H(3) in the p-halogen bicyclo-
octyl analogues is of about the same magnitude as it is for
the norbornyl series. For simplicity only the norbornyl
derivatives are discussed.

 

41

and H(3) endo. While H(l) is relatively constant, H(3)
is variable and in the case of the p-iodo derivatives the
magnitude is rather dramatic. H(3) is shifted to higher
field approximately 31 Hz. We have also included the
chemical shift difference between H(3) era and H(3) endo
Qv)to point out further that there is some unusual effect
operating in the p-halogen derivatives. We pointed out
earlier that Av increased with increasing positive charge
at C(Z). A comparison of the p-halogen derivatives with
the other data given in Table 1 indicates that for p-Br
and p-I Av is larger than expected. This does not mean
that there is more positive charge at C-2 in these ions,
since the chemical shifts for H(l) and H(3) are at higher
field than expected.

It also should be pointed out that the general features
of the p-halogen cation nmr spectra are typical of other
2-ary1-2-norbornyl cations. We noted earlier (p. 22) that
the temperature effects on chemical shifts in cations lll,
llg and llh were larger than for the other derivatives
studied.

While we have not discovered the nature of the effect
causing the deviations observed in the p-halogen cations,
the above observations point to a temperature dependent
equilibrium which favors the p-halogen cation (or its nmr
equivalent). The other species [X] in the equilibrium in

the case of llb for example should have the effect of

42

W12: [X]
We

decreasing charge at C(Z) so that H(l) and H(3) appear at
higher field than expected whi1e at the same time Av should

be larger than expected on the basis of 09 values and the
trend set for Av by the other cations in Table 1. Furthermore
the equilibrium contribution of [X] should be much less
important for the meta-halogen derivatives.* A possible
structure for [X] (thoUgh not a very satisfying one) is the

"trapped dimer" l9. We have investigated this possibility

050 F

£2

by a dilution study. The following concentrations of
2-p-iodophenyl-Z-endo-norbornanol were dissolved in 1 ml

of FSOSH; 150 mg, 30 mg, and 3 mg. The nmr spectrum of

 

* Meta-halogen derivatives in both series 12 and 14
cannot be largely different from their true values.
(See Figures 3 and 4).

 

43

each of these solutions was obtained at ~20°C and a
comparison of H(l) and H(3) chemical shifts was made.
There were no detectable chemical shift differences between
the 150 mg and 30 mg solutions. The time averaged nmr
spectrum of the 3 mg/ml solution showed H(l) and H(3)
shifted to lower field by about 2 Hz. This result is
certainly not conclusive. A more dramatic effect on H(3)
as compared to H(l) would have been expected. More dilute
solutions were not studied as the limits of the
instrumentation available have been reached.

Deno has calculatedtga‘values for a number of
substituents from four independent reaction series and he

has noted that 06

values calculated for p-halogena do not
show good agreement."3 Thus unusual behavior for p-halogen
derivatives is not unique for the arylnorbornyl and aryl-
bicyclooctyl cations.

It was pointed out in the introduction that charge

leakage to C(l) would be detectable in a series of Z-aryl-

2-norbornyl cations ll by nonlinear behavior in a graph of

b

Ar

M

44

H(l) vs. H(3) chemical shifts. Figure 3 illustrates just
this relationship. The fractional distribution of charge
at C(Z) remains constant from 2-p-methoxyphenyl through
2-phenylnorbornyl cation.* Substituents more electron
withdrawing than phenyl however lead to a "break" in the
linear relationship between H(l) and H(3). This effect
is rather dramatic in the case of cations lll and llm.
For H(l) chemical shifts of 4.93 and 4.75, the H(3) chemical
shifts should be within 8 Hz in each case of the straight
line. They are off the line by 8 Hz and 17 Hz
reSpectively!

By contrast the behavior of H(l) and H(3) chemical
shifts in 2-aryl-2-bicyclooctyl cations ll with increasing

positive charge is linear from p-methoxyphenyl through

Ar

 

* We have made a least squares fit of this portion of
the data (except cations llg, llg and llh). For a given
value of H(l), H(3) was fitted by a least squares method.
The standard deviation for 6 points was 8 Hz. The slope
was .74

 

45

meta-chlor0phenyl. With cation lll it seems likely that

we are observing the onset of charge leakage to C(l).

This effect is not nearly so dramatic in this case as it

is in the norbornyl system. H(l) at T 5.33 should have
H(3) within 1 Hz of the line.* H(3) is 4 Hz from the line.
Due to the uncertainty in the H(l) chemical shift of lll,
the data do not permit any further conclusion concerning
the electronic structure of these cations.

Could the "break in the line" in Figure 3 be associated
with any other factor besides charge leakage to C(l)? In
particular could the break be associated with the curious
behavior of H(3) era and H(3) endo with increasing positive
Charge. The onset of a detectable chemical shift difference
between these two protons occurs before cation llg (Av =
9 Hz), well before any nonlinearity is observed so that
it does not seem reasonable that the "break" should be
attributable to this effect.

There is some possibility that the break may be due to
an anisotropy effect of the phenyl ring whereby the phenyl
rotates slightly as charge increases at C(ZL perhaps as a
result of some better orientation of solvent. If this is
the case this effect in the 2-aryl-2-bicyclooctyl systems
fortuitously lags behind the effect in the 2-ary1-2-norborny1

cation.

 

* A least squares fit of the points for cations lla,

Aék: AAfi. Léé. $43 and llh has a slope of .87. The

deviation of these points is 1 Hz.

 

46

Having established that the nonlinearity in Figure 3
can be associated with charge leakage to C(l), what is the
mechanism by which this occurs? The following two
possibilities have already been presented:

(1) Rapidly equilibrating classical cations ll and Ifii

Ar r

11% RBI

(2) A o delocalized ion ll

6.
ii Ar

A third and currently acceptable possibility is also

considered:

(3) A carbon-carbon hyperconjugated cation l0.““’“”

 

47

Each of the three cases can superficially accommodate
our results. We could expect that in each case the mode
of charge transmission relative to a classical ion would be
attended by the interaction shown. By what means do we
distinguish these possibilities? Assuming that the aryl
group will not significantly alter the rate of a 3,2 hydride

shift in llm-I relative to the known rate for norbornyl cation

 

 

1a; ' 42w.

we can estimate the percentage of llmll that could be

present at +70°C. The rate constant for a 3,2 Hm at +70°C

5 -1,

is 6.5 x 10 sec For cation llm we observe broadening

at +70°C (see Table 5). Assuming a rate of 71 sec-1M the ratio

 

* Extrapolated fmn data in Ref. 23.

** The AB quartet (Av = 32 Hz) at high field for protons
on H(S) and H(7) has almost collapsed to a broad singlet.
We can estimate k = 71 sec’1 based on the equation for
coalescence T - J2 (Item-1."6

 

48

of ICECI/iém would be 1.1 x 10‘4 (i.e. 7.1 x 10/6.5 x 105).
Based on a chemical shift of 10 ppm"7 for a proton on a
carbon with a full positive charge a concentration of
1.1 x 10'4 for limit would
lead to a shift for H(l) in llm by .11 Hz or .001 ppm.
Assuming the deviation from the line is due to equal
contributions from a downfield shift of H(l) and an upfield
shift of H(3), then H(l) is .1 ppm* off the line - 100 x
more than accounted for by equilibrating classical ions.

We pointed out earlier that a Hammet sigma-rho plot
for solvolysis of 2-aryl-2-exo-norbornyl p-nitrobenzoates

5 Yet we observe nonclassical character in the

is linear.2
cation before it appears in the solvolysis. It seems
intuitively reasonable that the electronic demands made by
a developing cationic center should be less than the
demands made by a fully developed cation.

We have observed a remarkable effect on the classical

shift of H(6) exo protons in cations ll. For example in

Ar

I4

 

* This value is obtained by dropping a line (which makes
a 45° angle with a vertical line from llm) from the point for
llm to the extended line for a classical ion. The point of
intersection has coordinates for llm if it were classical.
This point is .1 ppm from the vertical or horizontal lines

from llm.

 

49

llg H(6) exo appears at T 7.47, while in llm it appears at
T 6.5. The chemical shift difference between them is about
100 Hz. This is the same value as the chemical shift
difference between H(l) in these two systems! While it
is difficult to distinguish H(6) endo it seems unlikely that
there is a chemical shift difference of more than 40 Hz in
these two systems. Clearly some factor which depends upon
the magnitude of the charge at C(Z) is responsible for
the effect on H(6) exo. Perhaps C(6)H o delocalization21
or a hyperconjugative interaction is operating here.

In view of the greater effect of the positive charge
on H(6) exo relative to H(6) endo, it is interesting that
the isotope effects in la and lb are the same.

In summary our data have shown that 2-ary1—2-norbornyl
cations with stabilizing substituents (09 <0) cannot be adequately
represented by classical electronic structures. Destabilizing

substituents (00

>0) lead to cations having increased
contributions from o delocalized or carbon-carbon hyper-

conjugating interactions.

EXPERIMENTAL

Melting points (uncorrected) were measured on a
Thomas-Hoover capillary melting point apparatus. Infrared
spectra were taken on a Perkin—Elmer 137 instrument. NMR
spectra were measured at 60 MHz on a Varian T-60 or A 56/60 D
or at 100 MHz on a Varian HA-100 instrument. (Tetramethyl-
ammonium tetrafluoroborate (T - 6.87)26a was used as an
internal standard for carbonium ion spectra.) Mass spectra

were measured on a Hitachi Perkin-Elmer RMU-6 instrument.

Carbonium Ions

 

The acid to be used was chosen so as to ensure complete
ionization of the carbonium ion precursor; essentially
identical spectra were obtained in acids differing
significantly in acidity.

The carbonium ions were formed in the following way.
The carbonium ion precursor was dissolved in a suitable
solvent and then added slowly dropwise to the rapidly
stirred acid (under N2 atm.) For example, CFCl3 was a
convenient solvent for many of the carbonium ion presursors

when FSO3H was used. If very low temperatures were desired

50

51

and/or any SbFs was added to the FSOSH, then SOZClF (25-75%)
and/or SOZF2 (25-75%) was used as the solvent for the
precursor. About a 5 wt % final solution concentration

of the carbonium ion was found to be ideal in most cases.

The sample was then transferred to the nmr tube.

Carbonium Ion Precursors

 

The alcohols used for this study were formed by the
reaction of the desired organometallic reagent with the
appropriate ketone - 2-norbornanone or bicyclo[2.2.2]-

“°:“” The organometallic reagent was formed

octan-Z-one.
by the Grignard Reaction5° or the halogen-metal inter-

conversion reaction51 (HMIR).

General Procedure for Alcohol Synthesis Using the HMIR

 

Ether was distilled into a dried 50 ml. 3-neck flask
containing a weighed amount of the desired aryl halide.
The flask was equipped with a low temperature thermometer,
a magnetic stirring bar, N2 inlet and 15 ml constant
pressure addition funnel. About a 10-30% by volume solution
of the aryl halide in ether was used. The solution was
cooled to about -40° using a Dry Ice-acetone bath. To
this solution was added dropwise a 10% molar excess (over
the halide) of commercially available butyllithium in
hexane. The reaction was kept between -30° and -40° during
the addition and then allowed to warm to -10° for about five

minutes. After the addition of the butyllithium was

52

complete, the temperature of the reaction mixture was
again lowered to -40°C. A solution of the ketone (one
equivalent) in anhydrous ether was added at a rate such
that the temperature of the reaction mixture did not rise
above -30°. Following the addition ofthe ketone, the
Dry Ice-acetone bath was permanently removed and the
temperature of the reaction mixture was allowed to warm
slowly to room temperature. The reaction mixture was then
hydrolyzed with saturated aqueous NH4C1 solution and
extracted with ether. Drying the ether solution with MgSO4
followed by evaporation of the ether led to a crude oil or
solid which was then purified.

The desired organometallic reagent was formed from
the aryl bromide except in the case of p-iodophenyllithium.
In this case p-diiodobenzene was used. All the aryl halides
were commercially available with the exception of 4-bromo-
2,6-dichlorobenzonitrile and 4—bromo-2,6-dichloro-N,N-
dimethylaminobenzene. The latter compound was prepared in
95%)deld by methylation of 2,6-dichloro-4-bromo-aniline52
with dimethyl sulfate. 4-Bromo-2,6-dichlorobenzonitrile
was prepared according to the procedures of Gassman and
Fentiman.”2

Tables 8 and 9 are a summary of the pertinent synthetic
and Spectral data on compounds prepared by the methods

mentioned above.

53

.50». canopaz ou~.caocaqa as» a» “Love. x: mmuooo- 38—05 «no. we: carnage; as» nosvu ppo u< .oooua cog» no: acouox och

.u.oo- :o_un ss.=u.——xu=n cu ou.:ocn .xuo «nu we’ve. an octagon” no: «caucus Is.:».. as»: Avoewauno no: no: uvuapaco Paucueo—o acouuaem.uom
.uonm—ocm you “on .mua mom «noun. cog» no: «cocaccoatoz-N .ueom- on uocoxop no: asauocoauou on» vo>pommpv oco~coaovo..v-m on» .0 .p. to».
.t.o~+ o» no“... coca no; mcsueeoasou ugh .uoo- no ocu~cooouo..u-m a» woven no: ocoxo: c. 53.2»..pxuam. "cowuu-ut utocoptw on» »a uvcoaocn
«a: acomooc u—FpouoEocooco och: Aco_uo—_.umvv sued utosmo A ogauacogao» soot any ocococcoacoc-N we covu.vv~ oz» o» to.cn «can; ~ toe
«cauacoaeou soon «a voter». can» «a: u— .uoo we vocuaoca «a: “co no; -=.=a.— oghe «Aocauaconnuu aspect use an. «conneconaoe-w mo co—apuu-
on» on toeca .mc; x~ to. eons—cos son» no: u_ .uoo an eocaaugn no: undone; uswcuvp echo ”on .eou come Agree—cu Avooau.u=. omwxsoguo

.mu_c= copuuaac co.mcu>coucouc. popes-cooo—ag ago no cocoaoaa an: “cameos o._p~uoeocenco we»; "nouoovvc. om.xcoguc mmopca pogoopo anew.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.8... 2.8... 2.. e... 2.2 2.2 3.. 3.. a... a... 2.32:... .93.-..8. ...~ ehwwmww
. . . . - . - . - . N 2 2 i. 3 . ~38 233
mmN Pom oaw pom am e on .7 pm n~ n. o no mm oz .0 z u onm—-m— a on p-.- A—uv-m.n
.Nm .2 . - 8.2 2.2 o... 2..I 3...... 3...... 0......2. 8.2.8.2 A... ”zoo-«33-...
.3 8...... - - 2.2 2.2 a... .2... 2.3 3...... own—z... 8.3-2.. 22 .23....
838 3.....2 . - 8.2 3.2 a... 8... 2.3 2.»... 2.212. 8.31.... N... {nus-e
2.1.2 -~..- - - 2.2 - 2.. - 2.2 2.2:... .98-... a b.8281.

- - - - - - - . . - of... .2-.. 22 3......
2... e2 - - 2.. z... 3.. 2.2 . 3.2 8.2 9.2sz .3-8 N... .282
-~..- 22.2 - - 2.2 a... N: 2.. . 8.3 8.2 2.2:... .98-...» a... ego—ox
332 .262 - - 3.2. 3.2 a... 8.... 12...... 2.2 2.2.2. 92..-: N2 .3895...—

.: :2; - - 2:0. :6. 3.. 3.. _ 8... S... 22:28 .8... N... Laos-e—
N2 2 - - - - 8.. a.» 2.8 2.8 02...; 2.-.. w... anagrafi
a... .3 - - - - 8.. N... No... 2.8 02...... ”Mn-mm.— ..mm M:..-%§-..n
.42 .2 - - - - a... 2.. 2:2 3.2 $2sz .32-... 2 8......uoz-J
- - - - - - - - - - of... 222.-.... 8 3.2.8.5-
....Nfiiwmaw 5e .8; . . .58.. awn... new... a“;

 

 

 

 

 

 

 

 

 

 

n.~m~ococconcoz-cMrm- -_>g<-N mo cowuacaqoca 4w.oPmMH

54

.aa.. cano..= ou...nogaa. on» a» «no... xo .oga..o.a vacate. Love: as... uc.—_on as» o» nevus-g x: vam~= «a ooeouosa ago :.

vouasv»;ov nu: .osou—a as»: no. .»u. «on. uucovu-nvp—qunauu pogo.uo... at. c. noose-s an: 00:99.; ass: we o—o.» as» .ocnxo: a... .co.u-——-u..uuou

8.5 .3... a... 2538 .2. .28.. .5 5...: 8... 2.... .5 .28.. .8 £32.. . 8.82. 8.3... 8.5 .5 8 852...... 53...:
25 $5 .38... 5:. 2...... 8.. .38.. .3... 8 2.... 22. .8555... 5.. to...
mun... cog. no: o:o--=.uuon~.~.~uo.uxu..

a

.8525... .52... 85...»... .22.... 33...... .2.
8...- 3 .223. 3.. £388.... .5 823.... «82.38:.-. .5 2. .52 8.2. 5 8.... 55
u.) u.=u..~.nmu on. .u.o- u. acouconouo..u-m o. um... «a: genus: c. aa—cawppauang ._osou—. as. uc.».o¢.o»o x. veg-gussoau .co.u-pp.a..o .o—auopn:

.Aogauaguasou 200. u.. u:o--c.uuon~.~.~Ho—u»u—n ho co.u.uv- u‘u no so... ...g m so» usauauos-ua noon an was...» cos. «a: u—
...m... 2...... as». .co.uu.os ugocm... on» »a wagon... no: acouaog u._—-uuaocaaso ugh: “ac. .20. can. :o.u._—.»..v ac.gav vomzu :u.3 vouaguagoou.
.c....o. .vo».u.uc. o..:uoguo .uo.c= co.uu.os :o...~>cousuuc. .auuau..oo..g as» a: nos-nus. «so: upocoo—< .uuuau_uc. u..:.o:uo nus—c. pocou.<.

.03...

.

v

.u.o a. vegan... n.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.2 .2 - - 2... 3... 8.. 2.. 3... 8... .28.... ....-.. .... ..p.m..~..3.-.w
2.. .3 - - - 3.2 3.. 3.. 8.2 8... ....:2. b...._... .2... .8... ..........M
.888 .888 - 2.2:... .5. .2... 2 3%....“—
8..8. 8.8.. - - 2... 8... 8.. 2.. 2... 2... 2.2.... .l 3...... 3 - .z....-L
- - - - - - - - - - ...:... B. .52.... .8. -.:.._
.8 2.. - - 2.. 2.. 2.. 8.. 2.2 8.2 22.... .3-.. L... {.24
.888 .888 - - - - - - - - 2.2.... .5 2.3.... 3-.. ..........$
8.8.. - - - - - - - - 2.2:... .l 22.... - L......&

.2 .2 - - - - - - - - 22:... .8. 7.2.8. o. ...8.2......‘-.‘
.2 .2 - - - - 2... 2.. 3... 3... 2:... .E 8...... 2 .88....93
.8 .8 - - - - - - - - o-.... .5 3.8.-.. .2 ..:..-..5.-.8
.8 .8 - - - - - - - - ....... .23-... .8. 3.8....85-4
was... 8.1.8 8E... 8...... .3... 8...... ...::. .8

 

 

...238...23.8.2923.-. .8 Esta... .. 2...

55

The nmr and it spectra of the alcohols and olefins in
the norbornyl and bicyclooctyl series of compounds were
consistent with the assigned structures. The following are

typical nmr spectra for an alcohol and olefin in each series

of compounds.

Norbornyl §xstem
Z-p-iodophenyl-Z-endo-norbornanol
Nmr (CCl4) T 2.51 (AA'BB', Av = 26 Hz, J
7.70 (m, 4H), 8.43 (m, 7H).

AB = 9 Hz, 4H),

2-p-methogyphenyl-2-norbornene

Nmr (CC14) T 3.04 (AA'BB', Av = 31 Hz, J = 9 Hz, 4H),

AB
3.96 (d, J = 3.5 Hz, 1H), 6.28 (s, 3H), 6.76 (br s, 1H),

7.06 (br s, 1H), 8.12-9.0 (m, 6H).

Bicyclooctyl System
Z-p-fluorOPhenylbigyclo[2.2.2]octan-2-ol
Nmr (CC14) T 2.58 (m, 2H), 3.09 (m, 2H), 7.5-8.8
(m, 13H)

2-p-methoxylphenylbicyclo[2.2.2]oct-2-ene*
2.98 (AA'BB', Av = 31 Hz, J

AB 9 Hz, 4H), 3.62 (dd,

J = 7 Hz, 2 Hz, 1H), 6.22 (s, 3H), 6.98 (br s, 1H), 7.38
(br d, J = 7 Hz, 1H), 8.52 (AB q, Av = 17 Hz, J = 9 Hz, 8H).

 

* Decoupling experiments showed that the 2 Hz coupling
at 3.62 is between the olefinic proton and the broad singlet
at 6.98. Comparison of a 100 MHz spectrum with a 60 MHz
spectrum of this compound indicates that the absorption
centered at 8.52 is indeed an AB quartet (or two over-

lapping AB quartets with similar chemical shifts).

 

PART I

REFERENCES

10.

11.

12.

REFERENCES

H. C. Brown, Chem. Brit., 199 (1966).

ibid., Chem. Eng. News., 45, 87 (Feb. 13, 1967).

G. D. Sargent, Quart. Rev. Chem. Soc., 20, 299 (1966).

P. D. Bartlett, "Nonclassical Ions," W. A. Benjamin,

New York, N. Y., 1965.

T. D. Swartz, "Solvolytic Studies of Brendyl, Brexyl

and Norbornyl Systems," Ph.D. Thesis, Johns Hopkins
University, Baltimore, Maryland, 1966.

H. C. Brown, "The Transition State," Special Publication,
No. 16, (The Chemical Society, London, 1962), pp. 140-158,
174-178.

S. Winstein and D. S. Trifan, J. Amer. Chem. Soc., 71,
2935 (1949).

G. B. Cream, Rev. Pure and Appl. Chem., 16, 25 (1966).

P. van R. Schleyer, J. Amer. Chem. Soc., 82, 701 (1967).
D. G. Farnum and G. Mehta, J. Amer. Chem. 800., 91,

3256 (1969).

H. C. Brown, F. Chloupek, and Min-Hon Rei, J. Amer. Chem.
Soc., 86, 1248 (1964).

H. L. Goering and C. B. Schewene, J. Amer. Chem. Soc.,

gz, 3516 (1965).

S6

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

S7

Min-Hon Rei and H. C. Brown, J. Amer. Chem. Soc., 22,

5335 (1966).
B. L. Murr, A. Nikon, T. D. Swartz, and N. H. Werstrik,

J. Amer. Chem. Soc., 22, 1730 (1967).

J. M. Jerkunica, S. Borcic, and D. E. Sunko, J. Amer.
Chem. 500., 22, 1732 (1967).

V. F. Smith, Ph.D. Thesis, Michigan State University,
1972.

P. von R. Schleyer, M. M. Donaldson, and W. E. Watts,

J. Amer. Chem. 300., 22, 375 (1967).

A. Nikon, H. Kwasnik, T. Swartz, R. 0. Williams, and

J. B. D. Giorgio, J. Amer. Chem. Soc., 22, 1613, 1615
(1965).

P. von R. Schleyer, D. C. Kleinfelter and H. G. Richey,
Jr., J. Amer. Chem. 500., 22, 479-81 (1963).

S. Winstein, Reaction Mechanism Conference, Brookhaven,
N. Y., Sept. 5, 1962.

G. A. Olah, J. R. DeMember, C. Y. Lui, and R. D. Porter,
J. Amer. Chem. Soc., 22, 1442 (1971).

T. S. Sorensen and K. Ranganayakulu, Tetrahedron Letters,
2447 (1972).

G. A. Olah, A. M. White, J. R. DeMember, A. Commeyras,
and C. Y. Lui, J. Amer. Chem. Soc., 22, 4627 (1970).

G. A. Olah, G. D. Mateescu and J. Louise Riemenschneider,
J. Amer. Chem. 300., 24, 2529 (1972).

H. C. Brown and K. Takeuchi, J. Amer. Chem. Soc., 22,

2691 (1970).

26.

27.

28.

29.
30.

31.

32.

33.

34.
35.

36.

37.

58

(a) K. Takeuchi and H. C. Brown, J. Amer. Chem. Soc.,
22, 2694 (1970).

(b) D. G. Farnum, J. Amer. Chem. 300., 22, 2970 (1967).

D. G. Farnum and A. D. Wolf, submitted for publication.

J. A. Berson and P. W. Grubb, J. Amer. Chem. Soc., 22,

4016 (1965).

B. M. Benjamin and C. J. Collins, ibid., 22, 1556 (1966).

G. A. Olah, P. R. Clifford, and C. L. Jewell, J. Amer.

Chem. 300., 22, 5531 (1970).

G. A. Olah and T. E. Klousky, J. Amer. Chem. 300., 22,

4666 (1968).

B. L. Eliel, "Stereochemistry of Carbon Compounds,"

McGraw—Hill Inc., New York, N. Y., p. 206.

C. Altona and M. Sundaralingam, J. Amer. Chem. Soc.,

22, 1995 (1970).

G. J. Karabatsos, private communication.

B. Huang, K. Ranganayakulu and T. S. Sorensen, J. Amer.

Chem. 500., 94, 1779, 1781 (1972).

(a) J. H. Nogle and R. E. Schirmer, "The Nuclear

Overhauser Effect," Academic Press, New York, N. Y.,

(1971).

(b) R. A. Bell and J. K. Saunders, Can. J. Chem., 22,

1114 (1970).
T. Fukumi, Y. Arata, and S. Fujiwara, J. Mol. Spectros,

g1, 443 (1968).

38.

39.

40.

41.

42.

43.
44.

45.

46.

47.
48.

49.

50.

51.

52.

59

M. Saunders, P. von R. Schleyer, and G. A. Olah,

J. Amer. Chem. 500., 22, 5679 (1964).

H. C. Brown and Y. Okamoto, J. Amer. Chem. 500., 22,
4979 (1958).

P. B. D. de la Mare, J. Chem. 500., 4450 (1954).

C. Earborn, ibid., 4858 (1956).

N. C. Deno and A. Schriesheim, J. Amer. Chem. 500.,

22, 3051 (1955).

N. C. Deno and W. L. Evans, ibid., 22, 5804 (1957).

T. G. Traylor, W. Hanstein, H. J. Berwin, N. A. Clinton,
and R. S. Brown, J. Amer. Chem. 500., 22, 5715 (1971).
R. Hoffmann, L. Radom, J. A. Pople, P. von R. Schleyer,
W. J. Hebre, and L. salem, J. Amer. Chem. 500., 22,

6221 (1972).

J. D. Roberts, "Nuclear Magnetic Resonance," McGraw-Hill,
Inc., New York, N. Y., 1959, p. 64.

D. G. Farnum, J. Amer. Chem. 500., 22, 2970 (1967).

P. K. Freeman, D. M. Balls, and D. J. Brown, J. Org.
Chem., 22, 2211 (1968).

J. Hine, J. A. Brown, L. H. Zalkow, W. E. Gardner, and
M. Hine, J. Amer. Chem. 500., 22, 594 (1955).

D. C. Kleinfelter and P. von R. Schleyer, J. Org. Chem.,
22, 3740 (1961).

R. G. Jones and H. Gilman, "Organic Reactions," Wiley,
New York, N. Y., 1951, p. 339.

P. G. Gassman and A. F. Fentiman, Jr., Tet. Let., 1021

(1970).

PART II

19F AND 1H NMR STUDY OF AN EQUILIBRIUM

AMONG ARYLBICYCLOOCTYL CATIONS

60

INTRODUCTION

Recently we have found it desirable to examine certain
chemical shifts in a series of Z-aryl-Z-norbornyl cations 2

and Z-aryl-Z-bicyclo[2.2.2]octyl cations 2.1 During the course

AG,

22, =p-CF3
2b, X=p-F

 

of this study we observed that upon ionization of their
respective precursors at -78° cations of type 2 in which

2 was a strongly electron withdrawing substituent e.g. 22,

X = p-CFS, gave proton nmr spectra substantially different
from cations in which X was electron donating e.g. 22, X =
p-F.l Since chemical shifts in the former case indicated
that the aryl group(s) was still at a carbonium ion center,
it seemed likely that 22 might have rearranged to one or more
new arylcarbonium ions under the ionizing conditions. Since
these circumstances would permit the determination of

equilibrium constants among those cations present, we explored

61

62

the hypothesis further. Thus, isomeric arylbicyclooctyl
alcohols were prepared and examined under similar conditions
of ionization. The results of these studies are described

below.

RESULTS AND DISCUSSION

Proton Nmr Studies

 

Careful ionization of p-trifluoromethy1bicyclo-

[2.2.2]octan-2-ol 22 in FSOSH/SOZClF at temperatures

   

 

 

ionization\ Quenching \ ‘m H
/ 7

A X A X 4, X

22 22, X = CF3 22, X = CFS

82 AR. X = F 22, x = F
22, X = H

lower than -100°C gave rise to the 2-p-trif1uoro-
methylphenyl-Z—bicyc1o[2.2.2]octy1 cation 22. The nmr
spectrum of 22 is shown in Figure l. The AA'BB' system
centered at T 1.45 (Av = 67 Hz, J = 9 Hz, 4H) is assigned to
the p-trifluoromethylphenyl group. That this group is at a
carbonium ion center2 is evident when one compares the large
chemical shift difference (Av = 67 Hz) between the ortho and
meta protons in cation 22 with the smaller difference in the

starting material 22 (Av = 0). Furthermore the chemical shift

63

oocofl- pm ww mo eneuuoam ”:2 OOH .N ot=MAm

 

64

 

 

 

 

 

mvé

 

 

 

8. 8n Sn 8‘

65

of the aryl group in 22 is greater than 1 ppm to lower field
than in the alcohol 22 (r 2.50). The broad singlet (1H) at
r 5.33 is assigned to the bridgehead proton H(l) a to the
carbonium ion center, while the broad singlet (2H) at T 5.83
is assigned to the methylene group a to the carbonium ion
center. Deuterium labeling at C(3) in 22 gives rise to a
spectrum in which the absorption at 5.83 is essentially wiped
out. There are nine other hydrogens at higher field which
were not assigned. This spectrum is quite similar (as
expected) to the previously reported and more stable cation
263

That cations 2 do indeed have the 2-bicyclo[2.2.2]octyl
skeleton was confirmed by recovery experiments. Quenching
of cation 22 in pentane/aq. NaZCO3 at 0°C gave bicyclo[2.2.2]-
octene (22) in 55% yield and bicyclo[3.2.l]oct-6-ene (222)
in 45% yield.* The origin of 222 will be discussed later.
Quenching of 22** gave essentially pure 22, which can be
produced by simply overheating alcohol 22 on distillation.
The nmr spectra of olefins 2 are quite characteristic of their
structure. For example, in 22 there is a multiplet (4H)

centered at r 2.85 assigned to the p-fluorophenyl group. A

 

* See Experimental Section.
** The fluorophenyl derivative 22 gives the same basic
spectrum as 22. It is more stable to rearrangement than 22

but it also will rearrange when warmed to 0°C.

 

66

doublet of doublets T 3.55 (J = 7 Hz, 2 Hz, 1H) is assigned
to the single olefinic proton. A broad singlet r 6.97 (IR)
is assigned to H(l). Irradiation of this broad singlet leads
to collapse of the 3.55 absorption to a doublet (J = 7 Hz).
A broad doublet (J = 7 Hz) at r 7.35 (1H) is assigned to H(4).
Irradiation at T 3.55 leads to collapse of the broad doublet
to a broad single verifying the coupling between H(3) and
H(4). There is an "AB quartet" (8H) at-higher field (I 8.50)
which is assigned to the remaining four pairs of methylene

protons on C C

5,
AB quartets on symmetry grounds one for C5, C8 and (one for

6’ C7 and C8. While one would predict two

C6’ C7L in actuality the chemical shifts of these quartets
are so similar that no further resolution of the observed
singlet "AB quartet" was obtained even at 100 MHz.

Warming 22 to -60°C in the nmr probe produces material
with an nmr spectrum identical with that of the material
obtained by ionization of 22 at -78°C. The chemical shift
of the aryl group in this material indicated that one or more
new aryl carbonium ions were present. It is well known from

solvolytic studies of bicyclooctyl derivatives 2 that they

67

generally yield approximately equal molar mixtures of
compounds 2 in which the initial skeleton is retained and 2

in which the initial skeleton has rearranged to the [3.2.1]
system.“ With the consideration that 22 might have rearranged
to 32 we attempted to prepare that cation. While ionization

of 22* at -78°C did indeed give material having a spectrum

 

x
§ 2 $8

éfi, X = CF; 23. X = CF3 222, x = F, Y = H

fik» X = F RR: x = F 222, x = F, Y = 9

identical to that of rearranged 22, ionization of 22 at
<-100°C produced material having a different spectrum! That

spectrum, assigned to structure 22:“ is shown in Figure 2.***

 

* Probably a mixture of exo and endo isomers.
** There are other materials present besides 22. This
will be discussed in the section on F19 nmr.
*** We are not assigning a conformation to cations 2 as
our data do not permit distinction between the chair or boat

conformation. It has been reported recently that the 2-methy1—

2-bicyclo[3.2.1]octyl cation exists in the boat conformation.5

 

68

 

 

_ r--.—

 

 

 

 

 

 

mw.m

u.

by)

 

 

 

we;

Unom- um ww mo aauuuomm Nmz OCH .H whamfim
l - -.M--+.7.-.-W...-W--4..m--.q.lmjfi+m+q-.-I_-¢-...--.L----z..I”-
. _ A _ 4.4“... A .

69

There is a p-CF3C6H4 group (T 1.45) at a carbonium ion
center. The broad singlet at r 5.10 (1H) is assigned to the
bridgehead proton H(l).* The methylene protons o to the
carbonium ion center are not magnetically equivalent as they
were in structure 22 and they give rise to a highly coupled
AB pattern with one component at T 5.86 (1H) and the other
at T 6.36 (1H).

Since 22 is rather unstable towards rearrangement,
labeling and quenching studies were carried out on the more
stable cation 22. This cation gives the same basic spectrum
as cation 22.

Labeling at C(3) in 22 gives rise to a spectrum in
which the highly coupled AB pattern is nearly wiped out.

In an attempt to make nmr assignments for each of the
a-methylene protons and to obtain information on the
conformation of 2 deuterated olefin 222 was prepared. Treat-
ment of 222 with FSOSH/SOZClF at -105°C gave deuterated 22

having a spectrum in which the a-methylene signals were about

 

* The broad singlet of low intensity at T 5.33 is due
to 22. It has a counterpart of double intensity at r 5.83
under the low field portion of the "AB quartet" of 22. It
has not been possible to produce this cation completely free
from 22. Even in the case of the more stable cation 22,
there is some 22 which is always present on ionization of

22 even at temperatures as low as -110°C to -120°C.

 

70

equally reduced in intensity and in which there was a
sharpening of that portion of the AB system at lowest field.
Apparently 222 is protonated equally well from both top and
bottom. Thus we were not able to assign the methylene protons
on the basis of these labeling experiments.

In order to convince ourselves completely that cations
22do have the 2-substituted [3.2.1] skeleton we quenched 22
in pentane/NaZCO3 at ~78°C and obtained the two olefins 222
and 22 in 55% and 17% yield reSpectively.* There was also
32% of unidentified material which was composed of at least
two compounds having Vpc retention times longer than the
olefins. Olefin 222 could also be produced by treatment of
22 with p-toluenesulfonyl chloride-pyridine at 110°C in a
sealed tube for 10 hrs.**

We have reported elsewhere1 that Z-aryl-Z-norbornyl
cations 2 in which X is a strongly electron withdrawing group
undergo the well known "Wagner-Meerwein, 6,2-hydride shift,"
rearrangement very rapidly on an nmr time
scale. In cations 2 this rearrangement is degenerate. It
is interesting to examine the outcome of this same rearrange—
ment in cations 2. In this case,as is shown in Scheme 1, a
new cation 22 having the 6-substituted bicyclo[3.2.l]octyl

skeleton is produced.

 

* See Experimental Section.
** Based on a procedure for dehydration of isomeric

methylbicyclooctanols.6

 

71

Scheme 1

 

Ar

% %&

In order to examine the possibility that 22 (and 22)
rearranged to cation 222 we prepared alcohol 222 as shown

below.

/ 1) P-XC6H2Li \ h h
2) ' o .
3 X
. o

X

 

éé %% %%
222, X = CF; .222, X = CF:5
22k: X = F kkk’ X = F
Although there are literature procedures available for
the synthesis of 227'1° we have independently prepared it in
good yield (48% based on Z-norbornenone ethylene ketal 2)
according to Scheme 2. By integration of the two olefinic

protons (having J = 7.6, 7.2 Hz) or the two endo C4 protons

72

Scheme 2

I
CCIZ‘COZEtMaOCH:5 > 0 C1
0 pentane/O°C < ‘ 1
L") «14

 

H m

+

. ‘b / EtSN, 95% EtOH (:3 b
<— \ 10% Pd/C, H2 | 11
0 {WI H306 H We
X
* 1%

 

 

(having J = 2.7, 2.8 Hz) it was possible to determine that
CCl2 addition* to 22 produced 222 and 222 in about a 1:1
ratio. The stereochemical assignment at C(4) was based on

a comparison of the observed coupling constants to those for
the adduct between norbornene and CClz12 (J H1, H2 = 7.0 Hz,
J H3, H4
and 222 the chemical shifts were not assigned. The mixture

= 2.8 Hz). Because of the similarity between 222

was not separated since each of the components yield the
desired product 22 on reduction. Hydrolysis of 22 yielded
ketone 22, which, on treatment with p-CF3C6H4Li followed by

 

* Same conditions that were used for CCl2 addition to

norbornene.11

 

73

hydrolysis, gave alcohol 222.*

Ionization of 22a in FSOSH/SOZCIF <-100°C produced
cation 22%. The nmr spectrum of 22% at -90°C is reproduced
in Figure 3. There is an AA'BB' system (Av = 65 Hz, J = 9 Hz,
4H) centered at T 1.45 attributable to the p-CF3C6H4 group at
the carbonium ion center. The broad singlet at r 5.13 (1H)
is assigned to the bridgehead proton H(l) a to the carbonium
ion center. The methylene protons on C(3) give rise to the
AB quartet centered at T 5.83 (J = 24 Hz) in which the higher
field component is further coupled. Deuterium labeling at
C(3) in 222 essentially wipes out the AB pattern. Preparation
of olefin 22b (partially deuterated) and ionization of this

compound below -100°C. gave a spectrum in which the lowest

 

222, X = CF3, Y = H
22k, X = CF3, Y = D
228, X = F, Y = H

 

* Although we have no evidence on the stereochemistry of
22 it seems likely that it is the endo alcohol. The reactions
of p-XC6H4Li (X = F, CF3), with ketone 22 yield predominantly
one product with a sharp melting point. Z-Norbornanone, a
ketone of similar structure is attacked preferentially from
the exo side of the molecule by nucleOphiles.

 

74

 

'—

 

.Uoom- um www mo Eswuoomm um: OOH .m ouowfim

T‘

1 1 I W q 4

i

,.
,.
l

4 1 .

. 32%

1 1 1 - d J 4‘ 1

 

_. . . .
sigma: .2, rise... a? .
_ a

s2 .

 

 

 

 

mw.m Hm.m

 

Uocml I B

 

 

 

_

u11<~+a
m _

 

oom

00v

 

.—-

 

 

 

"’ fi' "T..-

75

field component of the AB quartet was greatly reduced in
intensity while the highest field component collapsed to a
broad singlet. Assuming predominantly exo protonation by
FSOSH in 22b,* the endb proton is then the lowest

field proton while the exo proton is at highest field.

While we have not quenched cation 22% directly, it
was mentioned earlier that olefin 22% is produced on
quenching cation 22. We believe that cation 22a is actually
the source of olefin 22%. 22% was identified by comparison
with identical material produced by dehydration of 22a with
p-toluenesulfonyl chloride in pyridine.6 Furthermore
quenching cation 22b yields olefin 22g (80% of recovered
material) as the major product.

When cation 22% was warmed to -60°C in the nmr instrument
the spectrum obtained indicated that 222 had produced the
same material as cations 22 and 22 on warming to -60°C. It
was now evident by a comparison of the nmr spectra of the
mixture and 222 that cation 22% was a major component in the

mixture.

 

* Similar behavior is observed for Z-p-fluorophcnyl-

2-norbornene of similar structure.1

 

76

Fluorine Nmr Studies.

 

From the proton nmr studies of the mixture

produced by either cations 2a, 32 or 22% it was
not possible to tell exactly how much of each of the cations
were present or if any other materials were present.

19F resonance (a sharp singlet) of the p-

Monitoring the
CF3C6H4 group in the cations proved to be a convenient
method for following the formation of the mixture from each
of the cations and for obtaining quantitative information
on both the total number of cations (and other materials
having a CF3C6H4 group) and the relative amounts of each
present.* Using this method we were able to determine that
the mixture produced from 22 and 22a is virtually identical
and consists almost exclusively of these two cations. The
equilibrium constant measured at -80°C by integration of the
19F resonances is 3.2 favoring cation 22g.

Ionization of 2% at <-100°C produces cation 2% along
with several other materials (up to 40%) which have not been

19

identified and which have F resonances at lower field than

 

* While the p-FC6H4 derivatives of the cations were also

examined, this group proved to be a less convenient probe
(at 56 MHz - the conditions of exploratory runs), since the
19F resonance was broad. This was due to coupling with the

aromatic protons.

 

77

2g, 32 or 222. We believe these latter compounds probably
arise from irreversible reactions of the starting material
during the ionization process since the amounts of these
materials produced were variable in different experiments
while their total percentage remained relatively constant
on warming 33.*

When 22 is warmed to -80°C it disappears and produces
the equilibrium ratio of 22 and 222. A summary of the

19

quantitative data obtained from the F studies in the

equilibrium between these cations is shown in Table l.

19

Table 1. Thermodynamic Data from F NMR Experiments

at 94.1 MHz

 

Cation 32 + 9H2a + 22 + 7H2 + 222
% (-80°C) <2 30 70

K = 3.2 AG = -450 cal

K+SO = 1.6 AH = -700 cal
13. o -

hmy - 200 AS - -1,5 eu

 

 

 

19

aChemical shift differences between F resonances

bInversion temperature

 

* The equilibrium mixture produced from 22 or 222 was
essentially free from of these materials.

 

78

Note that the figure of <2% of 22 in the equilibrium
mixture is certainly an upper limit as 2% would have been
detectable. The equilibrium constant between 22 and 222 was
measured between -90°C and —40°C and the thermodynamic
constants were extracted from this data. Note that there
is an entropy factor whose origin is not completely obvious
which favors cation 22. To the extent that these extra-
polations are correct we have calculated that above +200°C
(the inversion temperature) cation 22 would be more stable

than 222!

From both the 19

F and 1H nmr studies we have obtained

data which support the mechanism shown in Scheme 3.

Scheme 3

 

 

79

First we re-emphasize that we are dealing with a true
equilibrium. Thus, cations 22, 22 and 222 can each be
observed prior to any significant rearrangement, and, on
warming 22, 22 or 222, the same mixture* of cations in the
same ratio is produced.

A second important fact related to the mechanism was
obtained by observing the approach to equilibrium starting
from cation 22 (19F method). We were able to detect that
an initial buildup of cation 22 occurred while 222 became
predominant as the equilbrium was approached. This indicated
that cation 22 was formed at a faster rate than 222, but 222
was more stable. The secondary cation 22 seems to be a
reasonable intermediate in the transformation of 22 into
22 and 222. One would expect that the rate of partitioning
of 22 between 22 dnd 22 (which rearranges to 222) would
favor 22 as observed. The partitioning of 22 to 22 involves
a Wagner-Meerwein rearrangement which yields an arylcarbonium
ion ie. 22. The partitioning of 22 to 22, a secondary
carbonium ion, involves a 7,2 hydride shift (7,2 Hm). Some
of the stability of the arylcarbonium ion 22 ought to show up
in the transition state to its formation. Even if cation 22

is more stable than 22, one would expect a larger activation

 

* Though 22 is not detectable in this mixture it very

likely is an accessible intermediate.

 

80

energy for this 7,2 Hm than for the W,M transformation* of
2% + @3-

A third observation is that in the transformation of
22 into 222, the methylene group a to the carbonium ion
center remains bonded to the carbon containing the aryl group.
Equilibration of 22 labeled with deuterium at C(3) produces
222 with the C(7) methylene a to the carbonium ion center
also labeled with deuterium.

An alternative mechanism which accounts for the observed
results and avoids a direct 7,2 hydride shift is of the type

22 I 22 I 222 as shown in scheme 6. The required 7,2 hydride

Scheme 6
/ “5m! : I ’41.!!!»
e
3 Ar
Ar

2% R £46 £2 We

 

shift is accomplished indirectly via the cyclopropane inter-
mediate 22. Evidence against intermediates of type 22 was

obtained by Parker and coworkers13 in their studies of the

 

* It has been suggested elsewhere that the absolute
rate of bond shifting in bicyclic systems is greater than

the rate of hydride shifting.“

 

81

buffered acetolysis of the unsubstituted bicyclo[3.2.11-6-

toluene-p-sulfonates 22 (Scheme 7). They observed products

resulting from trapping of 22, 22 and 22. They considered

that leakage of 22 into 22 and 22 might be occurring indirectly
via the cyclopropane 22 rather than by direct 7,2 hydride
shift. However, a comparison of the products obtained when

22 was submitted to the acetolysis conditions with those
obtained from 22 indicated that the mechanism involving 22

could be operative to no more than 20%.13

Assuming that the
aryl group does not significantly enhance proton loss from
222, then the arylcyclopropane 22 should not be an important
intermediate in the conversion of 22 to 222. Furthermore
one would expect that proton loss to form a cyc10propane

would compete even less effectively with 7,2 hydride shift

in F803“ than in buffered acetic acid. Hydride shifts of

82

this type are unusual, but nevertheless are well documented

1'0 15

in the structurally related norbornyl systems.‘» :

EXPERIMENTAL

General Procedures

Melting points (uncorrected) were measured on a Thomas-
Hoover capillary melting point apparatus. Infra-red spectra
were measured on a Perkin-Elmer Model 137 instrument. Mass
spectra were measured on a Hitachi Perkin-Elmer RMU-6
instrument. Proton and fluorine nmr spectra were taken at
60 MHz on a Varian T-6O or A 56/60 D or at 100 MHz on an
HA 100 instrument. Proton chemical shifts are reported in
T values versus tetramethylammonium tetrafluoroborate

(T 6.87)2 as an internal standard.

Carbonium Ion Preparation

The fluorophenylbicyclooctyl cations were generally
formed by ionization of the respective starting materials
at -78°C in FSOSH. The p-trifluoromethylphenylcarbonium ions
were ionized at temperatures lower than -100°C. In the latter
case the alcohol or olefin to be ionized was dissolved in
sulfurylchlorofluoride (1 ml) and cooled to -78°C. This
solution was added slowly dropwise to a rapidly stirred
mixture of F803H Clml) in SOZCIF (1-2 ml) (NZ atm.) at

<-100°C. About a 5 wt % final solution concentration of the

83

84

 

 

 

 

 

 

 

 

.Nm ¢~ . m AIH .m any fi=~ .ev on.~
.cm<v ~H.© No.m .flmN .sv NH.H no u x .mww
Amm Ame .Nm m u a
.Nm «N u n fiza .m “av N: me u ><
.um<v mm.m Hm.m ..mm.<<v e¢.H m u x .mww _H.~.m_.ot
rflmfl .ev no.0 nmfi .m “my fizN .ev hm.N
fimfi .ev so.o mm.m .fimN .sv no.H m u x .mw
has .N: m u n
“ma .ev em.e fizfi .m any N: me u >< m
Hfiza .sv em.m OH.m ..mm.<<v mq.H no u x .ww ”H.~.ma-~
AmN .m Hag Ara .m any fiz~ .ev mm.~
mo.o we.m .fimm .ev mH.H m n x .mw
Ame .N: m u a
has .m “av Azfi .m any .N: no a >< m
mm.m mm.m ..mm.<<v me.H me u x .ww .N.~.NH-N
mGOHOHm SOHOHQ v O
o:o~>:uo2-a vmonomwwhm m ux-m coaumu Eoumxm

 

 

 

mcoflumu quoooaoxofinfizh<

:fi muwfigm Hmufisosu :ououg

.N oHan

85

carbonium ion was prepared. The sample was then transferred
to nmr tubes which were maintained below -lOO°C by a liquid
Nz-pentane mixture.

The pertinent chemical shifts of carbonium ions prepared

in this study are listed in Table 2.

General Synthetic Procedures

 

Ketones
Bicyclo[3.2.l]octan-2-one was commercially available.
Bicyclo[2.2.2]octan-2-one‘62l7 and bicyclo[3.3.0]octan-2—one18

were prepared according to literature procedures.

Bicyclo[3.2.l]octan-6-one
Although there were literature procedures available
for the synthesis of the title compound,"10 it was prepared

independently and in good yield by the following method.

Addition of dichlorocarbene to Z—norbornenone ethylene

 

ketal 22

Dichlorcarbene was generated from ethyltrichloroacetate
according to Jefford'sprocedurell for the addition of :CCl2
to norbornene. The crude product was poured into H20 and
extracted 3 times with pentane. The pentane extracts were
dried with MgSO4 and pentane was stripped off on a rotary
evaporator. The unreacted starting materials were distilled
from the product at 90°/.3 mm leaving a reddish-black oil

which was carried on to the next step. 64 g of ketal yielded

86

100 g of the crude oil. The nmr spectra of the oil indicated
that the major products were (222) and (222) one of which

was slightly predominant. Nmr (mixture CC14): T 3.80, 3.95

(two doublets, J = 7.5 Hz, 1H), 5.50, 5.80 (two doublets,

J = 3 Hz, 1H), 6.15 (m, 4H), 7.50 (m, 2H), 8.10 (m, 2.8 Hz).

Formation of 22 by catalytic hydrogenation of 222 and 222

 

The 100 g of crude material from the preceeding step was
split into two 50 g samples which were each hydrogenated in
the following way: 50 g of a mixture of crude 222 and 222
was dissolved in 100 ml of EtSN and 150 ml of 95% EtOH. To
this mixture was added 4 g of 10% Pd/C. The mixture was
hydrogenated on a Paar apparatus (The pressure drop was 53
psi). The crude product was poured into H20 and extracted
three times with pentane. The crude products from each run
were combined and the crude ketal was distilled from polymeric
material and then redistilled. There was some hydrolysis of
the ketal in this process, but this was not critical since
the entire mixture was carried on to the next step for

hydrolysis. Nmr (purified sample of ketal, CC14) T 3.84,
(m, 4H), 2.4-1.0 (m, 12H).

Hydrolysis of ketal 22

 

The entire product from the preceeding step was treated
with 300 m1 of pentane, 200 m1 of H20 and 2 m1 of cone. ”2804.
The mixture was stirred for 6 hrs at room temperature. The

crude product was poured into saturated aqueous NaHCOS. The

87

solution was extracted three times with pentane and the pentane
solution was then dried over MgSO4 and the pentane solvent
evaporated. The crude product was sublimed at 82° at water
aspirator pressure. The yield was 25 g based on 64 g of
norbornenone ethylene ketal (48% overall yield). Nmr (CC14)
7.5 (br s, 1H), 7.65-9.0 (m, 9H). ir vneat 2990, 1750

max

(Lin,13 v 1743 cm‘l). mp 149-153° (Lit.f 155-157°).

max

Synthesis of Alcohols

 

The alcohols used in this study were all synthesized by
reaction of the desired organometallic reagent with the
appropriate ketone. The organometallic reagents were all
formed by the halogen-metal interconversion reaction. In a
typical experiment ether was distilled into a dried 50 ml
3-neck flask containing a weighed amount of the aryl bromide.
The flask was equipped with a low temperature thermometer a

magnetic stirring bar, N inlet and 15 m1 constant pressure

2
addition funnel. About a 10-30% solution of the aryl halide
in ether was used. The solution was cooled to about -40°C
using a Dry-Ice acetone bath. To this solution was added
dropwise a 10% molar excess (over the halide) of commercially
available butyllithium in hexane. The reaction was kept
between -30° and -40°C during the addition. The mixture was
then warmed to -10°C for about five minutes. After the
addition of the butyllithium wss completed, the temperature
was again lowered to about -40°C. A solution of the ketone

bne equivalent) was added at a rate such that the temperature

of the reaction mixture did not rise above -30°C. Following

88

addition of the ketone the Dry-Ice acetone bath was
permanently removed and the temperature of the reaction
mixture was allowed to warm slowly to room temperature.
The reaction mixture was then hydrolyzed with saturated
NH4Cl solution and extracted with ether. Drying the ether
solution with MgSO4 followed by evaporation of the ether
led to a crude oil which was then purified.

The following Tableszue a summary of the pertinent
synthetic and spectral data on compounds prepared by the

method mentioned above.

Labelingngperiments

 

3,3-dideuteriobicyclo12.2.2]octan-2-one

 

2.0 g of bicyclo[Z.2.2]octan-2-one,‘°:l7 6 m1 of

dioxane, 3 m1 of 99.77% D O and a catalytic amount of NaOH

2
were refluxed for 45 hrs. The reaction product was diluted
with H20, then extracted with pentane. The pentane extract

was evaporated and the residue was sublimed (150° at 3 mm).

3,3'-dideuteriobicycloj§.2.1]octan-2-one

 

1.5 g of commercially available ketone was treated with
10 ml of 99.77% D20 and 1 m1 of trifluoroacetic anhydride.
The sample was heated at 60°C for 12 hrs, then poured into

aqueous NaHCO Extraction with pentane followed by

3.
sublimation of the crude product obtained by evaporation of

the pentane yielded 1.3 g of deuterated material.

E35)

.Ahvu a. con—touaooe
.3232. our. use on. yo 23.3.- a 3532..»
.Anvu as 3:30:80
6032—33“. 5230329
63323 «ex...

.3893 53-32338 3.833.: «5 5 3088
an: 339:. 0.5.. yo 33>. 2:. 6:38.. It» «SS-3:33.89. 8.2» 53..- 33 3538 an: 3.8—- 05 52: 8.: 5.8—o E; 3:82.
.6 28.3.. a 333» 3393 8.5 .5 we 83:32:. 52638. .82» $5 55 53.3.5 5.! ap—oauo an: 389:. 0.5.8 .3 32» 2.!

6:33.. 8..» 32:33.5 9.2. 8:3 :5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

32 SN
. u_.ne .hu e o n
. ne..m -~ -~ .m.-.e~ «.mm a = u no-8
asu on~ on._~ __.F~ _~.o Pn.o o~.oo ou.no .m.mh-m~ Na .zmuncu-m
o- o- - - - - - - ..~-n~ as uezoua.a H_.~.nH
Au_.mVo-
.Aaeapvphu . o n
.Aum.m~v-~ -~ m.a_v.-~. ». .oo ~.oo a = u cu-m
o- oa~ op..~ ~_.P~ on.» Pn.o .m.oo no.0u c..~-.~ a.- .zoumuu-u
Aae.~ o-
. ne.~. _- c o
.Aa~.oov-~ -~ u:.. n. u. .oa o a uu-m
o- ow - - - - - - be... m. as .8 uexmuah ”3.3-“:
Aa~.m o-
..um.¢m F- c o n
.Awm.cmv-~ ~N~ o=_- n. ». .ou ~.n. o z 8 $9-8
oa~ o- - - - - - - use. a. an .oo ~.~m oezounuu-m
cum o- ~0.m ~¢.m as.“ .o.~ o—.oh a~.m~ .mm-mm .rxmm.» «zoom-m n~.~.~g-
“nanotmuuflmmnmmwuuvohe :8 [.3 :3 a 2:5... 3.mo_B an .m .3 .unf Bo; 33.6 :92 gnaw—

 

.m_o=.qun~.~.~uopuxu.u-~-P»u<-~ co coau.t~aota .n upsok

90

uco>aom ya: on» we com: mm: Nm-eHH_EX§m

m m

uqo>Hom us: can we wow: max Huauo
Haze Hohsze
muuooam Ham pom vumvcmum use Hmnuoucfi mm wow: mmz mzHo

 

 

 

 

 

 

 

 

 

 

vopmofivcw omflzuonuo mmoacs uno>aom he: can mm: eHoop
vopmuwucfi omwzuosuo mmoaca oHQEMm kw umocm
Anna .sv m. -N.a Aswan» .fisvmmm .fizvoeofl e o m
.fime .mv ¢.~ .fimucmmN .Azuomem e m u mu-m
nmmfl .sv o.m-m.a flevmflm .nstNw
.fimN .sv Nm.N .Aegmew .nevmmm .AavoNOH a o
.fimN .ev om.~ .nmvomwm .fiscommm dam u>-d HH.N.mH-2
ammfl .av c.m-n.a nmvoem .AmVoNOH v o m
.3: .3 mm; .383 .388 «a m 0 "BA
mzmm .sv o.m-¢.a .
AmN .ev «c.m fimvmfim nmvmmm .fimvomofl e o
.3: .E 34. .hmpomfi #88: Fund SHE
fizmfl .av o.m-¢.a mmvmmw .fisvomOH e o m
.fime .mv om.~ .fimvomaN .nmvomqm : o mu-m
mama .av m.m-m.h
”AmN ”av mo.m mmvmflw .fimvfimw e 0
“mm av mm.~ .flscoem .fimvommm .Amvommm (m m om-m _~.~.NH-~
o.nfipu nee mnH-Eov kw macho axu< Eoumxm
maocmuoooaoxownaxu< owhoEomH mo mama Aw paw uez .v oHan

91

7,7'-dideuteriobigyclo[3.2.1joctan-6-one

 

This material was deuterated using a procedure similar
to that of Farnum and Mehta3 for the preparation of 3,3'-
dideuterionorbornan-Z-one. 3.0 g of bicyclo[3.2.l]octan-
6-one, 20 m1 of 99.77% D20 and 2 m1 of trifluoroacetate
anhydride were heated for 60 hrs at 100°C. The product was
neutralized with N82C03’ and then taken up in pentane solution.
The pentane solution was washed with two 10 m1 portions of
D20. It was then dried with MgSO4 and the pentane was
evaporated on a rotary evaporator. The crude product was
sublimed (80°/water aspirator pressure) yielding 2.1 g of

labeled material.

Recovery Experiments

 

In a typical experiment the nmr sample itself was
quenched. This was done by quickly decanting the sample
into a quenching solution which was maintained at the desired
temperature. The quenching solution was agitated vigorously
with a vibro-mixer. The quenched material was then extracted
with a suitable solvent. The following table summarizes the
extractable products which were identified in a series of

experiments.

Elimination Experiments

 

In a typical experiment 250 mg of 2-p-f1uorophenyl-
bicyclo[3.2.l]octan-2-ol was treated with 250 mg of absolute
pyridine and 500 mg of p-toluenesulfonylchloride.6 This

mixture was heated in a sealed tube for 10 hrs at 110°. The

92

 

.couonq ofinfimoao can mo cowumumoucw no woman

w

.Aom-mm *oNV mnfimoao on» case osfiu coaucouon Homeoa m mcfi>mn mxmon

03¢ one: muons

owumfiuouomnmgu onu mo eofiumumoucw no woman ma mane

.mvcsanou o3“ ammoa no mo vomwuano ma Hmfluoume vowmfiucouwca echo
.Hmwuopme mo xno>ooou woo

a

.Eoumxm many a“ cououm oacflmoao

.muoawoua oanmuoouow nonuo ozm

 

mausvona voflmaucovflna

awn cam cavemu o:o-o-uuo
-Ha.~.mHoaoxownaxnonnonoSHm-m-o

.wa-\mouNmz\:ooz

emoum-m

HH.N.mH-o

 

o. muoawonn vowwwucovflca
awn paw Awnau o:o-~-uoo
-—~.N.NHoHoonnaxconaouoSHm-Q-~

mwflmv u:o-~-uoo
-"a.~.mHoauxownaxconmouosam-m-m

owa-\nouNmz\a:au:ua

e o

: um-m

HH.~.m_-~

 

flames o:o-o-uuoafl.~.ng
-oHoxowpflxconmaxnuoaouosamwuu-m-o

flammu o:o-~-uuoHN.~.~H
-oHozuwnaxconmaxguosouozamwuu-m-N

o:o-~-uuo
-_~.N.NHOHUsuManeonawhozfim-m-~

.o\nooNoz .em\o=au:ma

owa-\mou~mz\o=mueoa

v o m

z u no-8

v o

: om-m

H~.~.~H-~

 

muoawoum

 

 

mGOMufivcou magnocoso

 

nacho qu<

 

Eoumxm

 

mcofiumu azuuooaozowpazu< ofluosomH

Hmuo>om Seam muos©0Hm mcfinocmso

.m oHQMH

 

93

crude reaction mixture was poured into NaHCO3 solution.
Extraction of the bicarbonate solution with pentane, drying
with MgSO4, and removal of the solvent with a rotary
evaporator yielded a crude oil. This material was pumped at
.3 mm pressure in order to remove the last traces of
pyridine. Nmr analysis of the product indicated a mixture
of 80% of the desired Z-p-fluorophenylbicyclo[3.2.1]oct-2-ene
and 2% of Z-p-fluorophenylbicyclo[2.2.2]oct-2-one. The
desired olefin can be obtained essentially pure by
preparative gas chromatography on 1,2,3-tris(cyanoethoxy)—
propane at 135°.

Nmr (CFC13) r 2.65 (m, 2H), 3.05 (m, 2H), 4.35 (m, 1H),
6.9-9.0 (m, 10H).

The following table gives the nmr spectra for a

representative olefin from each bicyclooctyl system studied.

 

94

.vnmwcwum Hmchopnfi we won: me: make

 

 

 

 

 

 

 

 

 

w

Ame .ev a.m-m.m
.fimN .av H.m-m.a .Amfl .m any on.“ .fimfl .m you
c.“ .fimH .N: o n a .eu me.m .fime .mu me.~ .eauo mww HH.N.m_-
fizOH .av
o.m-o.a .fimfl .su mm.e .fimq .au mw.~ .mfiumu mww HH.~.nH-N
“mm .e m<v
om.” .AmH .Nz a n m .e Hag mm.“ .fizfl .m new ha.o
.fimfi .Nm N .Nm 5 u a .eeg mm.m .mze .ev mm.~ .vauu mw HN.~.~_-N
moduem Hmuflsogu :flmmfio supmsm
cowumawxnon >9 woummohm mcfimofio o>Humusomoumom mo mhuoomm H52 .0 oases

PART II

REFERENCES

10.

11.

12.

13.

REFERENCES

D. G. Farnum and A. D. Wolf, submitted for publication.
D. G. Farnum, J. Amer. Chem. Soc., 99, 2970 (1967).

 

D. G. Farnum and G. Mehta, J. Amer. Chem. Soc., 99,

 

3256 (1969).

 

H. Kwart and J. L. Irvine, J. Amer. Chem. Soc., 99,

5541 (1969).

G. A. Olah, G. Liang, J. R. Wiseman and J. A. Chang,
J. Amer. Chem. Soc., 99, 4927 (1972).

. Kraus and R. DeWald, Ann. Chem., 999, 21 (1965).

 

. N. Ipatieff, J. E. Germain, W. W. Thompson and
. Pines, J. Org. Chem., 99, 252 (1952).

 

. B. Wiberg and B. A. Hess, Jr., ibid., 99, 2250 (1966).
. Szarkowska-Szpaczek, Roczniki Chemii, 99, 235 (1962).

 

UZWE<Z

. Komppa, T. Hirn, W. Rohrmann and S. Beckmann, Ann,,
999, 242 (1936).

C. W. Jefford, J. Gunsher, D. T. Hill, P. Brun, J. LeGras
and B. Waegell, Org. Syn., 99, 60 (1971).

C. W. Jefford, S. Mahajan, J. Waslyn and B. Waegell,

J. Amer. Chem. Soc., 99, 2184 (1965).

 

R. A. Appelton, J. C. Fairlee, R. McCrindle and W. Parker,
J. Chem. Soc., 9, 1716 (1968).

 

95

14.

15.

16.

17.

18.

96

G. A. Olah, A. M. White, J. R. DeMember, A. Commeyras
and C. Y. Lui, J. Amer. Chem. Soc., 99, 4627 (1970).

 

M. Saunders, P. von R. Schleyer and G. A. Olah, ibid.,
99, 5680 (1964).

J. Hine, J. A. Brown, L. H. Zalkow, W. E. Gardner and
M. Hine, J. Amer. Chem. Soc., 99, 594 (1955).

 

P. K. Freeman, D. M. Balls and D. J. Brown, J. Org.
Chem., 99, 2211 (1968).
A. C. Cope, H. H. Lee and H. E. Petree, J. Amer. Chem.

 

Soc., 99, 2849 (1958).

 

 

J1

IIIIIIII ST ATE UNIV SITY LIBRARIE

IIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

 

2093 4212