Title: Synthetic and solvolytic studies in the bismethanonaphthalene system
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00099260/00001
 Material Information
Title: Synthetic and solvolytic studies in the bismethanonaphthalene system
Physical Description: ix, 161 leaves : ill. ; 28 cm.
Language: English
Creator: Timberlake, John Foushee, 1948-
Copyright Date: 1976
 Subjects
Subject: Methanonaphthalene   ( lcsh )
Solvation   ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Statement of Responsibility: by John Foushee Timberlake.
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 157-160.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00099260
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000178112
oclc - 03108084
notis - AAU4616

Downloads

This item has the following downloads:

syntheticsolvoly00timb ( PDF )


Full Text











SYNTHETIC AND SOLVOLYTIC STUDIES IN THE
BISMETHANONAPHTHALENE SYSTEM










By

JOHN FOUSHEE TIMBERLAKE















A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY











UNIVERSITY OF FLORIDA

1976























To Laurie


without whose love, understanding, and support the author

could not have coped with the rigors of graduate study.














ACKNOWLEDGEMENTS


The author wishes to thank Professor Merle A. Battiste

for his interest and enthusiasm in the development of this

research. It has been a pleasure to work with a research

director who knows when to let the student work out his own

problem and when assistance is needed. Appreciation is ex-

pressed to Dr. Roy W. King for his many helpful discussions.

Special thanks also go to Dr. Rocco Fiato, Dr. Donna McRitchie,

Dr. Richard Galley, Dr. Robert Posey, Dr. Lou Kapicak, Henry

Gingrich, George Kuta, and Neil Weinstein for their advice,

suggestions, and for generally making the author's stay in

Gainesville pleasant.















TABLE OF CONTENTS


ACKNOWLEDGEMENTS

LIST OF TABLES

LIST OF FIGURES

ABSTRACT

CHAPTER

I INTRODUCTION

II PENTACYCLO[6.3.1.13' 6.02'7 09'1
TRIDEC-13-YL BROSYLATE

III TETRACYCLO[6.2.1.13 6.02,7] DODEC-
4-EI211-Y, PENTACYCLO[6.3.1.03-10.-
0 .0 3' DODEC-ll-YLv AND PENTA-
CYCLO[7.2.1.04,11.05,12.06,1 ]DODEC-
2-YL BROSYLATES

IV KETO- AND KETAL-SUBSTITUTED BROSYLATES

V CIS,ANTI-4 5-EPOXY-ANTI-TETRACYCLO-
[6.2.1.13' .02,7]DODEC-l--YL BROSYLATE

VI EXPERIMENTAL

Synthesis
Kinetic Studies
Solvolysis Product Studies

BIBLIOGRAPHY

BIOGRAPHICAL SKETCH















LIST OF TABLES


Table

I Relative Rates of Solvolysis in the 7- 6
Bicyclo[2.2.1]heptyl and the 8-Tricyclo-
[3.2.1.02,4]octyl Series

II Acetate Product Distribution from the 36
Acetolysis of (9)-OBs, (15)-OBs, and
(16) -OBs

III Relative Rates of Sulfonate Esters at 41
25 C

IV Rates of Acetolysis of Several 7-Nor- 54
bornyl Brosylates















LIST OF FIGURES


Figure

1 Free Energy Diagram for the Acetolysis 42
of (15)-OBs and (16)-OBs

2 The 1Hnmr Spectrum of the Acetolysis 51
Mixture from (51)-OBs

3 The 1Hnmr Spectrum of (101) 60

4 Correlation Diagram for the Highest 63
Occupied MO's in Cyclopropane, Oxirane,
and Thiirane

5 The 1Hnmr Spectrum of the Product Mixture 65
Obtained from Solvolysis of (72)-OBs to
Greater Than 10 Half-Lives in 60% Aqueous
Acetone

6 The 1Hnmr Spectrum of (103) 67

7 The 1Hnmr Spectrum of the Product Mixture 68
Obtained from Solvolysis of (72)-OBs to
2.6 Half-Lives in 60% Aqueous Acetone

8 The 1Hnmr Spectrum of the Reduction Products 69
of a Mixture of (76) and (103)

9 The 1Hnmr Spectrum of (105) 70

10 Plot of ln Observed Versus Time for (55)- 131
OBs at 140 C

11 Plot of ln Observed Versus Time for (51)- 133
OBs at 50 C

12 Plot of In Observed Versus Time for (72)- 138
OBs at 100 C














Abstract of Dissertation Presented to the
Graduate Council of the University of Florida
in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy



SYNTHETIC AND SOLVOLYTIC STUDIES IN THE
BISMETHANONAPHTHALENE SYSTEM

By

John Foushee Timberlake

August, 1976

Chairman: Dr. Merle A. Battiste
Major Department: Chemistry

Acetolysis studies of exo,endo,exo-pentacyclo[6.3.1.13'6-

02,7.09'll]tridec-13-yl brosylate (I) have shown a rate which

is 107.7 times faster than 7-norbornyl brosylate and a mix-

ture of products containing a least eight acetates. The two

major products have been shown via extensive spectral analysis

to be exo- and endo-pentacyclo[6.3.2.03 .'10 4'1.0 5'9]tridec-

9-yl acetates rather than the expected endo-pentacyclo[6.6.-

13'10.04'1.05'9]tridec-10-yl acetate (II). Attempts to syn-

thesize II and the ketone derived from II via ring expansion

were unsuccessful.

During the synthesis of starting materials for these

ring expansion reactions, exo,exo-tetracyclo[6.2.1.13 6.02'7]-

dodec-4-en-ll-yl brosylate (III-OBs) was observed to give

both endo-pentacyclo[6.3.1.03'10.04'12.05'9]dodec-10-yl

acetate (IV)-OAc and exo-pentacyclo[7.2.1.04'11.05'12.06'10]-









dodec-2-yl acetate (V)-OAc upon acetolysis contrary to the

previously reported result that III-OBs gave only IV-OAc.

Reinvestigation of the acetolysis products showed that III-

OBs gives 91.5% of IV-OAc and 8.5% of V-OAc while IV-OBs

gives 65.4% of IV-OAc and 34.6% of V-OAc, and V-OBs gives

9.9% of IV-OAc and 90.1% of V-OAc. Furthermore, acid cata-

lyzed equilibration of a 65/35 mixture of IV-OAc and V-OAc

in acetic acid resulted in 99.5% of IV-OAc and 0.5% of V-OAc

corresponding to 3.7 kcal/mole energy difference between

these two acetates. Thus product stability, as reflected in

the transition state, is a dominant factor in determining the

relative extent of competitive carbon-carbon bond participa-

tion in the solvolysis of IV-OBs. In an attempt to synthe-

size V-OH more conveniently than previously reported, penta-

cyclo[7.2.1. 4'11 .o5'12.06'10]dodeca-2,7-diene was generated

in one step by the reaction of 7-chloronorbornadiene with

cyclopentadienyl thallium in diglyme at 150 C.

The introduction of an electron withdrawing ketal or

keto group in the 12-position of I was investigated in the

hopes of reducing both the number and extent of carbonium

ion rearrangements on solvolysis. Contrary to expectations

acetolysis of 12,12-dimethoxy-exo,endo,exo-pentacyclo[6.3.-

1.13,6.02,7.09,11]tridec-13-yl brosylate (VI) was accelerated

by a factor of 1.3 over I at 50 C, and gave at least twelve

volatile products (glpc), 58% of which could be assigned to

methoxy ketones. By contrast, exo,endo,exo-pentacyclo[6.3.-

1.13'6.02'7.09'11]tridecan-12-on-13-yl brosylate (VII)

viii









solvolyzed in acetic acid ca. 100 times slower than I at

100 C and gave endo-pentacyclo[6.4.1.03"10.04'13.05'9]tri-

decan-13-on-10-yl acetate as the major (68.7%) product. For

comparison, 12,12-dimethoxy-exo,exo-tetracyclo[6.2.1.13'.02'7]-

dodec-4-en-ll-yl brosylate (VIII) and exo,exo-tetracyclo-

[6.2.1.13'6.02'7]dodec-4-en-12-on-ll-yl brosylate (IX) were

solvolyzed in acetic acid. The rate of acetolysis of VIII

at 25 oC was at least as fast as III-OBs and gave exo-10-

methoxypentacyclo[6.3.1.03'10.04'12.05'9]dodec-12-one(X-OMe)

as the major (69%) product. Other products isolated and

identified were X-OAc (19%) and 2-carbomethoxytetracyclo-

[5.3.1.03'8.04,ll]undec-9-ene (3.3%). Brosylate IX solvolyzed

ca. 100 times slower than III-OBs and gave 85% of X-OAc and

15% of the decarbonylated brosylate.

With a view to examining the possibility of remote

participation by an oxirane ring in a solvolysis reaction,

4,5-epoxy-exo,exo-tetracyclo[6.2.1.13'6.02'7]dodec-ll-yl

brosylate (XI) was prepared and its solvolytic rate constants

determined in 60% aqueous acetone. When extrapolated to

acetic acid at 25 C these rates correspond to a 104.6

accelerative factor compared to 7-norbornyl brosylate. Par-

ticipation by the edge carbon-carbon bond of the oxirane

ring was confirmed by the isolation of endo-10-hydroxy-endo-

tetracyclo[6.2.1.0 4,11.05 9]undecan-3-carboxaldehyde.














CHAPTER I
INTRODUCTION



Nonclassical carbonium ion chemistry was thrust into

the forefront of mechanistic organic chemistry with the re-

port by Winstein and Trifan of the exo/endo rate ratio of

350 for the acetolysis of 2-norbornyl p-bromobenzene sulfon-

ates (brosylates) (1)-OBs and (2)-OBs and the complete

scrambling of Cl and C2 in the products of acetolysis of

(1)-OBs. The controversy has raged between the followers

of Winstein's theory that the symmetrical "nonclassical"









1 2 3 4

carbonium ion (3) is produced as an intermediate in the

solvolysis of (1)-OBs and the followers of Brown's theory

that unsymmetrical, equilibrating "classical" cations de-

picted by (4) are the intermediates.2'3 Though the contro-

versy may never be resolved to the satisfaction of both

camps, much work continues both in support and opposition

to the existence of nonclassical carbonium ions.

The bridged ion (3) is envisioned as arising from








participation of the CI-C6 a-bond simultaneous with the de-

velopment of positive charge at C2. The stabilization de-

rives from a bridging three-center, two-electron charge de-

localized bond rather than a simple vertical (hyperconjugative)

stabilization. Since the initial observation of the bridged

ion (3) other carbon-carbon bonding groups have been shown

to participate including the n-bonds of the carbon-carbon

double bond and the strained o-bonds of the cyclopropane.

The anti-7-norbornenyl system (5) demonstrates a

decidedly more dramatic neighboring group rate acceleration

in solvolysis reactions than does (1). An acceleration fac-

tor of 1011.2 compared to 7-norbornyl p-toluenesulfonate

(tosylate) (6)-OTs was observed for (5)-OTs whereas syn-7-

norbornenyl tosylate (7)-OTs is only 103.6 times faster than

(6)-OTs.4'5 This tremendous rate acceleration for (5) was

attributed to backside stabilization of the developing car-



X X X




5 6 7 8

bonium ion center at C-7and was expressed as the bishomo-

cyclopropenyl ion (8).6 As in the previous example some

authors preferred to write equilibrating classical ions.7

To check the generality of participation of double

bonds and to determine the effect of a different orientation

of the double bond on the solvolytic rate enhancement, Win-








stein and Hansen solvolyzed tetracyclo[6.2.1.13'6.02'7]-

dodec-4-en-ll-yl brosylate (9)-OBs in acetic acid and observ-

ed a rate which was 107 times that of tetracyclo[6.2.1.13'6.-

02,7]dodec-ll-yl brosylate (10)-OBs. A better model for the

















+
10
10-- 11 12

rate enhancement of the double bond in (9)-OBs might be

7-norbornyl brosylate (6)-OBs because (10) is accelerated by

103 over (6), perhaps via hydride participation leading to

hydrogen bridged ion (11) Collapse of ion (11) to the

carbon bridged ion (12) followed by solvent capture provides

an adequate explanation of the products of acetolysis of

(10)-OBs. Relative to 7-norbornyl brosylate (6)-OBs the

double bond in (9)-OBs then shows an accelerative effect of

10 The bishomocyclopropenium ion (13) should be initially

formed on ionization of (9)-OBs, but could subsequently re-

arrange to ion (14) prior to solvent capture. However, the

only product reported from acetolysis of (9)-OBs was exo-

pentacyclo[6.3.1.03,10.04,12.05,9]dodec-ll-yl acetate (15)-OAc.8a








Had ion (14) been involved one would expect some exo-penta-

cyclo[7.2.1.04'11.05'12.06'10]dodec-2-yl acetate (16)-OAc




XT X^z,_Sb




15 16

to be produced along with the (15)-OAc. This absence of

(16)-OAc may point to the thermodynamic stability of the bis-

homocyclopropenium ion (13) to the exclusion of rearrangement

to other ions.

More recently, the cyclopropane ring has been employed
9
as a neighboring group. The strained o-bond of the cyclo-

propane should provide greater stabilization of the develop-

ing positive charge than a "normal" carbon-carbon a-bond

because the higher energy orbitals of the cyclopropane should

allow better mixing with the empty p-orbital. The initial

example of cyclopropane participation involved the solvolytic

study of the 3-bicyclo[3.1.0]hexyl systems (17) and (18).

Formation of the symmetrical trishomocyclopropenyl ion (19)


X





17 18 19

from cis-ester (17) was implicated by product and labeling

studies but the rate enhancement was less than anticipated.10








More dramatic evidence for participation of the cyclo-

propane o-bond was demonstrated via the solvolytic study of

the 8--tricyclo[3.2.1.02'4]octyl systems (20) through (23).



X XX X X




20 21 22 23 24

These systems were chosen so as to require the bicyclol[3.1.0]-

hexyl moiety to assume a rigid chair conformation for (20)

and (21) and boat conformations for (22) and (23). Direct

comparisons of any of the systems (20) through (23) could

also be made with 7-norbornyl (6), syn- and anti-7-norborn-

enyl (5) and (7), and 7-norbornadienyl (24) derivatives.

Thus (6) was designated as the "parent model" for the series.

The relative rates of solvolysis for the 7-bicyclo[2.2.1]-

heptyl and 8-tricyclo[3.2.1.02'4]octyl derivatives are sum-

marized in Table I.

The rate of solvolysis of endo,anti-8-tricyclo[3.2.1.-

0 2' 4]octyl p-nitrobenzoate (20)-OPNB was 1015.7 times faster

than (6)-OPNB.4bll To account for this exceptionally large

rate acceleration, the authors proposed the formation of the

delocalized trishomocyclopropenium cation (25) derived from

interaction of the "edge" cyclopropane orbitals with the

developing positive charge at C-8. The rate factor of 0.4

for (22), when compared to the factor of 1015.7 for (20) pre-

cludes the importance of "face" participation by cyclopro-






TABLE I. Relative Rates of Solvolysis in the 7-Bicyclo-
[2.2.1]heptyl and the 8-Tricyclo[3.2.1.02,4]octyl Series


Compound Reference krel25
re 1


x



k


x



x
xt


1011.2


103 .6


1014.7


4b,ll








pane while edge participation is geometrically impossible









25
isme.12a
in this isomer. The meager rate acceleration of only 6

for the endo,syn-isomer (21) rules out any effective stabi-

lization of charge development at C-8 by the cyclopropane

via frontside assistance to ionization.llb The edge cyclo-

propane bond must be properly oriented to interact effective-

ly with the rear lobe of the developing vacant p-orbital.

The 103 acceleration of exo,syn-(23) over (6) has been pos-

tulated to be due to steric acceleration of ionization rather

than any stabilization of the developing positive charge by

the cyclopropane. 12b,13

With the success obtained from the 8-tricyclo[3.2.1.02'4]-

octyl derivatives, researchers looked for other systems in

which the orientational effect on cyclopropyl assistance

could be tested. One obvious choice was pentacyclo[6.3.1.-

13,6.02'7.09'11]tridec-13-yl brosylate (26)-OBs in which the

orbitals of the edge cyclopropane bond are less favorably

oriented in relation to the developing empty p-orbital than

is the edge cyclopropane bond in the endo,anti-ester (20)-

OPNB.14 The rate acceleration for (26)-OBs of 107.7 rel-

ative to (6)-OBs reflects this less favorable orientation.14

Although 107.7 is a considerable rate acceleration, the








cyclopropane in (26) is 10 times less effective as a neigh-

boring group than is the cyclopropane in (20).

Acetolysis of (26)-OBs gave a mixture of products con-

taining no brosylates and at least eight acetates, none of

which had the structure (26)-OAc.14 A possible reaction

scheme leading to six acetates is given in Scheme I in which

classical ions are used for simplicity. The rate accelera-

tion points to the initial formation of the tris-homocyclo-

propenium ion (27) upon ionization of (26)-OBs, which could

subsequently rearrange to other ions such as (28), (29), and

(30). Capture of these ions by solvent would lead to three

epimeric pairs of acetates, (31) through (36). Other prod-

ucts could arise from hydride shifts or further rearrange-

ments in the carbon skeleton.

On the basis of extensive 1Hnmr spectral analysis, the

major component has been assigned the structure (35)-OAc.15

The second most abundant component was assigned the structure

of the epimer (36)-OAc by reducing (35)-OAc with lithium

aluminum hydride, oxidation to the ketone, reduction to a

mixture of alcohols with lithium aluminum hydride, and re-

conversion to the acetates with acetic anhydride and pyridine.

This mixture of acetates had the same gas-liquid partition

chromatography (glpc) retention times as did the major and

second most abundant products.15

The extensive rearrangements observed in the solvolysis

of (26)-OBs are surprising considering the total lack of










OAc
33-OAc


AcO

N34-OAc


26-a3


26-OBs







27 28


+z-S
29


1!


QOAc

AcO -


31-OAc 32-OAc


Ac


35-OAc


+


OAC
36-OAc


SCHEME I


^








rearrangements reported in the solvolysis of the unsaturated

brosylate (9)-OBs. This dramatic difference would suggest

that bishomocyclopropenium ion (13) is much more stable than

its rearranged ion (14) while the trishomocyclopropenium ion

(27) must be less stable than or about equally as stable as

ions (28), (29), and/or (30). It is even more surprising

that (15)-OBs is reported to give only (15)-OAc upon acetol-

ysis since (15)-OBs has the possibility of direct partici-









15

pation by either the C9-C10 bond to give ion (13) or the C1-

C12 bond to give ion (14).

Recent studies have been conducted on systems similar

to (15) in which direct participation of more than one

carbon-carbon bond is possible. For example, exo-tricyclo-

[4.3.0.0 3,]non-7-yl brosylate (37)-OBs could solvolyze by


6 5 +


BsO B 3 > and/or
83

37-OBs 38 39

participation of the C3-C8 bond to give ion (38) or the

C5-C6 bond to give ion (39).16a In this case products de-

rived from ion (38) predominate by a factor of 2.2 : 1 over








products from ion (39). This is significant because ion (39)

is favored by thermodynamic considerations (product stabili-

ties) while ion (38) is favored by bond alignment.16a

Similarly, exo-tetracyclo[5.2.1.02'6.04 .8]dec-9-yl

tosylate (40)-OTs would be expected to give ion (41) by par-

ticipation of the Cl-C2 bond and ion (42) by participation

of the C4-C8 bond. However, in this case the sole isolated




TsO = Ts0JO

2

40-OTs 41 42

product from acetolysis of (40)-OTs is (40)-OAc which could

arise from capture of either ion by the solvent.

In view of the work on (37)-OBs and the extensive re-

arrangements in the solvolysis of (26)-OBs, the isolation of

only (15)-OAc from the acetolysis of (9)-OBs and (15)-OBs

seemed suspect. This is particularly true since the work was

done prior to the routine use of nmr and prior to the advent

of capillary glpc. In addition, preliminary work had indi-

cated that (16)-OBs solvolyzes in acetic acid to give ca.

80% of (16)-OAc and ca. 20% of (15)-OAc. All of the above

considerations led to the reinvestigation of the products

of acetolysis of (9)-OBs and (15)-OBs and the synthesis and

acetolysis of (16)-OBs, the results of which are discussed

in Chapter III of this dissertation.

A detailed investigation of the products of acetolysis









of (26)-OBs is essential before the nature of the ions in-

volved can be determined. One approach is that used by

Seidl, i.e., separation of the products and identification

by spectral analysis. This method is practical for the

major components but is less feasible for the minor compo-

nents, particularly considering the long synthetic route to

(26)-OBs. A more practical approach would appear to be the

authentic synthesis of the probable products of solvolysis

of (26)-OBs. In this vein one could synthesize the acetates,

alcohols, or ketones and then compare their properties with

those of the isolated components. The present study attempt-

ed to synthesize pentacyclo[6.4.1.03'10.04'13.05'9]tridecan-

12-one (43) which could be derived for (31)-OAc or (32)-OAc.









43 44 45

Several approaches to the synthesis of (43) were proposed,

all of which involved pentacyclo[6.3.1.03'10.04,12.05'9]do-

decan-ll-one (44) as a key intermediate.8a Some of the pro-

posed procedures also involved ll-methylenepentacyclo[6.3.1.-

03,10.04'12.05'9]dodecane (45) which could be prepared from

(44) .

Ring expansion of ketone (44) with diazomethane would

be expected to give (43) although complications might occur

due to addition to the "wrong" side and by multiple addi-








tions. Ring expansion of (45) directly to (43) might be

possible through the action of arenesulfonyl azides or thal-

lium(III) perchlorate. 20 Spiro{pentacyclo[6.3.1.010

04'12.05'9]dodecan-exo-11,2'-oxacyclopropane} (46) prepared

by epoxidation of (45), and the endo-isomer (47), prepared

by methylene addition to (44), should give aminoalcohols (48)

and (49) upon reaction with sodium amide in liquid ammo-

nia.21,22,23 Reaction of the individual aminoalcohols with

CH2NH2 H





46 47 48 49

nitrous acid should yield ketone (43), pentacyclo[6.4.1.03'10

04'1 .05' ]tridecan-11-one (50), or a mixture of both.23








50

The details of these synthetic attempts are discussed in

Chapter II.

Another approach to the identification of the solvolysis

products of (26)-OBs is to appropriately substitute (26)-OBs

to prevent or retard the extensive rearrangements observed.

Appropriate substitution should reduce the number of products

and simplify their separation and identification. Substitu-








tion of a ketal or ketone group in the 12-position of (26)-

OBs would appear to be a likely choice, because these elec-

tron withdrawing groups would retard rearrangements to ions

having positive charge site R to the ketal or ketone func-

tional group. The synthesis of 12,12-dimethoxypentacyclo-

[6.3.1.13'6.02,7.09,11]tridec-13-yl brosylate (51)-OBs and

pentacyclo[6.3.1.13'6.02'7.09'11]tridecan-12-on-13-yl bros-

ylate (52)-OBs could be achieved by the same route used to







MeO
MeO
51 52

synthesize (26)-OBs with minor modifications. Examination

of the rates of solvolysis of (51) and (52) might provide

some insight into the nature of the trishomocyclopropenium

ions derived from (26). The synthesis and solvolysis of

12,12-dimethoxytetracyclo[6.2.1.13,6.02'7]dodec-4-en-ll-yl

brosylate (53)-OBs, tetracyclo[6.2.1.13'6.02'7]dodec-4-en-

12-on-ll-yl brosylate (54)-OBs, 12,12-dimethoxytetracyclo-

[6.2.1.13'6.02'7]dodec-ll-yl brosylate (55)-OBs, and tetra-

cyclo[6.2.1.13'6 .02'7]dodecan-12-on-ll-yl brosylate (56)-OBs

were also undertaken for comparison to the solvolysis of the

cyclopropyl brosylates.

Work reported by Gassman and coworkers on various 7-

substituted 2-norbornyl derivatives may assist in predicting
24
the results of solvolysis of (51) and (52). Exo-bicyclo-













MeO
MeO
53




MeO
MeO


0
54




Ox
0"",V


55 56
[2.2.1]heptan-7-on-2-yl tosylate (57)-OTs did not exhibit
rate acceleration similar to other 2-norbornyl systems.1,24a
In fact, the endo-epimer (58)-OTs solvolyzed six times


0


zLx


0


0


57 58 59
faster than (57)-OTs in acetic acid at 25 'C.24a This un-
usual behavior was attributed to the normal solvent-assisted
ionization of (58)-OTs and an absence of stabilization from
the formation of bridged ion (59) in the solvolysis of (57)-
OTs, because (59) places a partial positive charge adjacent
to the carbonyl group. Analysis of the products of acetol-
ysis of (57)-OTs showed the presence of 48% of (57)-OAc,
44% of (58)-OAc, 4% of nortricyclanone (60), and 4% of








bicyclo[2.2.1]hept-2-en-7-one (61), while (58)-OTs gave 97.7%

of (57)-OAc and 2.3% of (58)-OAc.


0 0






60 61

In further study of 7-substituted 2-norbornyl derivative,

the acetolysis of the ethylene glycol ketals of (57) and

(58), (62)-OTs and (63)-OTs, was performed.24b The rate of


0 0 0 0 CH2OH




X
62 63 64

solvolysis of (62) was nine times that of (63). The products

were similar to those of (57) with ketal group intact

for both (62) and (63), except there was a large amount (57%)

of cleavage product from (63) having structure (64) after

reduction with lithium aluminum hydride. The cleavage prod-

uct complicates any correlation of the rates of (62) and (63).

Finally, exo- and endo-7,7-dimethoxybicyclo[2.2.1]hept-

2-yl tosylates (65)-OTs and (66)-OTs were solvolyzed in

acetic acid.24c Ketal (65)-OTs solvolyzes 37 times faster

than (66)-OTs in acetic acid at 25 C. The products of

solvolysis of (65) were 95.5% of (65)-OAc and 4.5% of (66)-

OAc, while (66) gave 16% of 2-methoxybicyclo[2.2.1]hept-7-one









MeO OMe MeO OMe


X OMe
X
65 66 57-OMe
(57)-OMe, 29% of nortricyclanone (60), and 55% of (57)-OAc
with no products isolated with the ketal intact.
A natural extension of cyclopropane participation is
the use of three-membered rings containing heteroatoms as
a neighboring group in solvolysis reactions. Several authors
have recently published the synthesis and solvolysis of
compounds which might be expected to show such anchimeric
assistance. Syn- and anti-9-oxabicyclo[6.1.0]non-2-yl
brosylates (67)-OBs and (68)-OBs were solvolyzed in 80%




o O


67 68
aqueous acetone. The rate of solvolysis of (67)-OBs
was 260 times that of (68)-OBs. This rate difference results
from better Cl-C8 bond alignment for participation with the
developing positive charge in (67)-OBs than in (68)-OBs.
The only reported attempt to show remote participation
of an oxirane carbon-carbon bond in a solvolysis reaction
employed cis- and trans-6-oxabicyclo[3.1.0]hex-3-yl tosylates









(69)-OTs and (70)-OTs which are exactly similar to Winstein's

3-bicyclo[3.1.0]hexyl system.10,25b Epoxide (69)-OTs solvol-

yzes 12 times faster than (70)-OTs and 11 times slower than

cyclopentyl tosylate in acetic acid at 25 OC.25b,10c The

absence of an appreciable rate enhancement for the solvolysis

of (69)-OTs as well as the absence of products involving



OTs

OTs



69-OTs 70-OTs

C -C5 bond cleavage rules out any significant participation

of the epoxide ring in the solvolytic process.25b This

system suffers from the same nonrigid stereochemistry as the

3-bicyclohexyl systems.10 Two systems which should alleviate

this problem are the epoxides (71) and (72). Both systems

freeze the epoxide ring with the carbon-carbon bond in a

geometry favorable for participation as in the cyclopropane

analogs.11,14 For the present study epoxide (72)-OBs was

chosen because of the ease of synthesis by epoxidation of






0

71 72

(9)-OBs and because of the similarity to other systems in

this study. Solvolysis in aqueous acetone of (72)-OBs would




19



give trishomocyclopropenium ion (73) which should rapidly

rearrange to ion (74) or (72)-OBs could give ion (74) direct-

ly. Capture of ion (74) by solvent (water) would give hemi-

acetal (75) which should hydrolyze to the hydroxyaldehyde






bbHO bOHC

73 74 75 76

(76). The results of the solvolysis of (72)-OBs are dis-

cussed in Chapter V.














CHAPTER II
PENTACYCLO[6.3.1.13'6.02'7.09'11]TRIDEC-13-YL BROSYLATE



Introduction

One approach to unraveling the complex mixture of prod-

ucts obtained from the acetolysis of pentacyclo[6.3.1.13'6.-

02'7.09'11]tridec-13-yl brosylate (26)-OBs is to isolate and

characterize each component separately.14 A more practical

approach to the identification of all but the two most abun-

dant components is the authentic synthesis of compounds antic-

ipated to be among the acetolysis products of (26)-OBs.

Thus the synthesis of pentacyclo[6.4.1.03'10.04'13.05'9]tri-

decan-12-one (43) was attempted via several ring expansion

reactions. Comparison of the properties of ketone (43) with

those of the ketone mixture derived from the acetolysis

products of (26)-OBs would indicate the amount of products

with structure similar to (43).



Synthesis

The synthetic route used in the preparation of (26)-OBs

followed the published procedure and is outlined in Scheme II.

Norbornadiene reacted with benzoyl peroxide in benzene using

a copper(I) bromide catalyst to give 7-norbornadienyl benzo-

ate (24)-OBz after vacuum distillation.27 Reaction of (24)-OBz









OBz OH

eMgBr >

ether


24-OBz


24-OH


LiAlH4
ether


Cl
A -OAc //

C16

150 C
77-OAc


OAC
Ac20

pyridine
5-OAc


LiAlH4
ether
-78 C


O H OH AC
C16-L Na Ac20

tBuOH pyridine
77-OH THF 9-OH 9-OAc


CH2N2
Cu (I)Br


B BsC1

pyridine
26-OBs


H LiA1H4

ether
26-OH


SCHEME II


(OCO2) 2
C(H
Cu(I)Br


OH


ks


5-OH


OAc


"1-


26-OAc








with phenyl magnesium bromide in ether gave a good yield of

7-norbornadienol (24)-OH, which afforded anti-7-norbornenol

(5)-OH on reduction with lithium aluminum hydride. 27,28 The

acetate (5)-OAc, prepared by treatment of (5)-OH with acetic

anhydride in pyridine, was reacted with hexachlorocyclo-

pentadiene in a Diels-Alder fashion to give 3,4,5,6,12,12-

hexachloro-exo,exo,anti-tetracyclo[6.2.1.13 '6.02'7]dodec-4-

en-ll-yl acetate (77)-OAc.26'29

Reduction of (77)-OAc with lithium aluminum hydride in

refluxing ether according to the reported procedure resulted

in a mixture of products (glpc) with the major component of

this mixture being the desired (77)-OH.26 The mass spectrum

of (77)-OH contained a series of isotope peaks at m/e 380

(six chlorines), 345 (five chlorines), and 309 (four chlo-

rines), with the isotope peak at m/e 311 as the base peak.

These peaks correspond to loss of a chlorine atom from the

parent ion for 345 and further loss of hydrogen chloride for

309. The second most abundant component which was also the

shortest retention time component was isolated by preparative

glpc and tentatively assigned the structure of the didechlo-

rinated alcohol (78). The shorter glpc retention time for

Cl





H Cl

78

(78) is in agreement with its lower molecular weight and








expected lower polarity. The IHnmr was essentially identical

to (77)-OH except for the presence of a two-proton singlet at

6 2.32. The mass spectrum exhibited a series of isotope

peaks at m/e 312 (four chlorines), 277 (three chlorines),

241 (two chlorines), and 202 (four chlorines) with the base

peak at m/e 204. With m/e 312 as parent ion, then m/e 277

and 241 correspond to loss of a chlorine atom and further

loss of hydrogen chloride, respectively, as in (77)-OH. The

peak at m/e 202 agrees with a retro Diels-Alder cleavage of

the parent ion to give tetrachlorocyclopentadiene radical

cation and norbornenol. The only isomer in which two chlo-

rines have been replaced by hydrogens that would be consist-

ent with a two proton singlet at 6 2.32 is (78), similar to

previously reported substitution of bridge chlorine by hydro-

gen via lithium aluminum hydride.30

These complications were avoided by reducing (77)-OAc

at -78 C and destroying the excess lithium aluminum hydride

prior to warming the sample to room temperature. In this

way (77)-OH of sufficient purity for further reaction was

produced. Dechlorination of (77)-OH with sodium and tert-

butanol in tetrahydrofuran (THF) gave (9)-OH in relatively

impure form.26 Chromatography on silica gel followed by

recrystallization from hexane gave (9)-OH which showed a

single glpc peak. Preparation of (9)-OAc was achieved by

reaction of (9)-OHu with acetic anhydride in pyridine. After

distillation (9)-OAc was converted to (26)-OAc by reaction

with diazomethane in the presence of copper(I) chloride as








catalyst. Reduction of (26)-OAc with lithium aluminum

hydride followed by treatment with p-bromobenzenesulfonyl

chloride in pyridine gave (26)-OBs in good yield.

The ketone (44) and olefin (45), as well as the

epoxides (46) and (47), were required for the attempted syn-

thesis of ketone (43) which is expected to be derived from

the products of acetolysis of (26)-OBs. The synthesis of

ketone (44) followed the published route with one modifica-

tion and is outlined in Scheme III.3a Treatment of (9)-OH

with p-bromobenzenesulfonyl chloride followed by solvolysis

in sodium acetate buffered acetic acid afforded (16)-OAc as

well as (15)-OAc contrary to the previous report. For a

more thorough consideration of this development, see

Chapter III. Treatment of this mixture with p-toluenesul-

fonic acid in acetic acid gave essentially pure (15)-OAc,

which gave (15)-OH upon reduction with lithium aluminum hy-

dride. Oxidation of (15)-OH with Jones' reagent gave (44)

cleanly.31 Attempts to prepare olefin (45) by reaction of

(44) with the Wittig reagent failed, perhaps because of steric

interactions in the transition state.32 Reaction of (44),

prepared from the crude solvolysis mixture of (9)-OBs, with

methyl lithium afforded a mixture containing two components

(glpc). The major component was isolated by preparative

glpc and confirmed to be (79) by the presence of a methyl

singlet at 6 1.19 in the 1Hnmr and an alcohol stretching

peak in the infrared spectrum at 3230 cm-1. The infrared

and glpc both showed the absence of any unreacted ketone (44).








SCHEME III AcO



H >0 L OBs 15-OAC
BsC1 B HOAc +
pyridine NaOAc
9-OH 9-OBs AcO

16-OAc




LiA1H4 HOTs

ether HOAc


15-OH


15-OAc


Jones'
Reagent


OH
0 L Me H 2
MeLi M HOTs 2

ether OH
44 79 45


Me3SOI+
NaH
DMSO 0


m-Cl)CO3 H
CH2C12


0








Treatment of (79) with toluenesulfonic acid in refluxing

benzene gave the olefin (45).33 The infrared spectrum had

a methylene stretching absorption of 1670 cm-1 and the 1Hnmr

spectrum had two one-proton doublets at 6 4.53 (J = 1.7 Hz)

and 4.35 (J = 1.7 Hz) corresponding to the two nonequivalent

vinyl protons. The mass spectrum gave m/e at 172 for the

parent ion and 91 for the base peak.

Treatment of (45) with m-chloroperbenzoic acid gave

the epoxide (46) The vinyl protons of (45) were trans-

formed into a two-proton singlet at 6 2.86 in the 1Hnmr spec-

trum of (46). Exo-addition would be expected since 2-methyl-

enenorbornane gives 86% exo-addition and olefin (45) is much

more hindered towards endo-addition.34 The attempted prep-

aration of epoxide (47) by the reaction of ketone (44) with

trimethylsulfoxonium iodide and sodium hydride in dimethyl

sulfoxide (DMSO) was unsuccessful.22



Product Study

Before the authentic synthesis of the products of sol-

volysis of (26)-OBs can be meaningful, the actual products

must be obtained for comparison. Solvolysis of (26)-OBs in

acetic acid gave a mixture of at least eight acetates (glpc),

the largest four peaks integrating for 50.8, 24.5, 9.1, and

9.0%, respectively, of the mixture. The major component was

previously assigned the structure of acetate (35)-OAc by ex-

tensive 1Hnmr spectral examination.15 Reduction of the prod-

uct mixture with lithium aluminum hydride gave a mixture of








at least ten alcohols whose four major components integrated





AcOA
AcO

35-OAC 36-OAc

for 49.4, 24.4, 7.1, and 6.7%, respectively, of the mixture.

The increased number of peaks is probably due to better glpc

separation of the alcohols and not to the production of addi-

tional products upon reduction. Oxidation of the alcohol

mixture gave a mixture of at least five ketones whose major

two components integrated for 76.7 and 13.3%, respectively,

of the mixture. When compared to the percentages for the

acetates and alcohols, the ketone percentages suggest that

the two major components are epimeric as are the third and

fourth components. Seidl assigned the epimeric structure

(36)-OAc to the second most major component after the reduc-

tion of (35)-OAc with lithium aluminum hydride, oxidation to

the ketone, and reduction to a mixture of two alcohols.15

Reconversion to the acetates gave two acetates with the same

glpc retention times as the two major components. This anal-

ysis was coupled with the same series of reactions on the

product mixture to give the same ratio of the two major ace-

tates as from (35)-OAc.15

Aside from the two major components, the further isola-

tion of the products of solvolysis of (26)-OBs is impractical

because of the similar properties and incomplete separation








SCHEME IV






-N



44 80 2



50



as well as the long synthetic route to (26)-OBs. The more

feasible approach is the authentic synthesis of expected

acetates, alcohols, or ketones derived from (26)-OBs.

The reaction of ketone (44) with diazomethane would

be expected to give intermediate (80) and then lose nitrogen

and rearrange to ketone (43) or (50) as depicted in Scheme IV.

Rearrangement by the migration of the C10-C11 bond (path a)

to give (43) would be expected to predominate because the

migration of the C1-C11 bond (path b) to give (50) would

have to go through a more strained transition state to give

(50) initially in the conformation shown in Scheme IV. The

complete absence of ring expanded products would point to

steric repulsions in the formation of intermediate (80) as

the oxide ion is forced into the vicinity of the endo-alkyl

groups. The presence of boron trifluoride should act as a

driving force since coordination of boron trifluoride to the

carbonyl oxygen would place greater positive charge on the








carbonyl carbon, but still no ring expanded products were

obtained.

The possible difficulty in the formation of intermediate

(80) could be circumvented by the action of nitrous acid

on aminoalcohols (48) and (49). This would give intermediates


CH2NH H

HIO H2NH2C




48 49

similar to those derived from endo- and exo-addition, respec-

tively, of diazomethane to ketone (44). Reaction of nitrous

acid with (48) should give ketone (50) while (49) should give

ketone (43). The reaction of epoxide (46) with sodium amide

in liquid ammonia was expected to yield (48) but gave

only starting material. The failure of this reaction is

further evidence for steric hindrance to attack at C-ll from

the endo direction in this ring system. The failure in the

attempted synthesis of epoxide (47) precludes the formation

of amino alcohol (49).

Reaction of (45) with toluenesulfonyl azide or m-nitro-

benzenesulfonyl azide gave only starting materials under the

conditions reported to give ring expansion products in other

similar systems. The reaction of (45) with thallium(III)

perchlorate was expected to give (43) and (50). Only a minor

amount of volatile material was observed (glpc) with com-

plete destruction of (45).




30



All these reactions appear to suffer to some extent

from the steric problems observed in the diazomethane addi-

tion reaction. Apparently another approach is needed to

solve the product mixture from (26)-OBs. One approach,

the substitution of electron withdrawing groups to retard

rearrangements,is discussed in Chapter IV.














CHAPTER III
TETRACYCLO[6.2.1.13'6.02'7]DODEC-4-EN-11-YL, PENTACYCLO-
[6.3.1.03',1.04,12.05'9]DODEC-ll-YL, AND PENTACYCLO[7.2.1.-
04,1.05,12.06,10]DODEC-2-YL BROSYLATES



Introduction

As previously noted, the absence of additional re-

arranged products in the reported solvolysis of tetracyclo-

[6.2.1.13'6.02'7]dodec-4-en-ll-yl brosylate (9)-OBs and

pentacyclo[6.3.1.03'10.04'12.05'9]dodec-ll-yl brosylate

(15)-OBs was surprising considering the extensive rearrange-

ments observed in the solvolysis of (26)-OBs. Contrary to

the published result, an additional product was observed

in the synthesis of (15)-OAc from (9)-OBs videe supra).

This result coupled with the finding that pentacyclo[7.2.1.-

04'11.05'12.06'10]dodec-2-yl brosylate (16)-OBs gave a

mixture of acetates with (16)-OAc predominating upon acetol-

ysis prompted a reinvestigation of the solvolysis products

of (9)-OBs and (15)-OBs and the synthesis and solvolytic

product study of (16)-OBs.17



Synthesis

Since (9)-OH and (15)-OH were available from previous

work, the only task remaining was the synthesis of (16)-OH.

The reported synthesis of (16)-OH was a rather time consuming,









SCHEME V

l T diglyme [2 + 41



150 C
24-Cl 81 82


m-Ci CO3 H
CH2Cl2
0 C




HO LiA1H4 H 2 O
Pd/C
16-OH 84 83


low yield process and required equipment not generally avail-

able in the laboratory. A more convenient synthesis of

(16)-OH was proposed and is depicted in Scheme V. Reaction

of 7-chloronorbornadiene (24)-Cl with thallium cyclopentadi-

enide in diethylene glycol dimethyl ether (diglyme) at 150 C

led to pentacyclo[7.2.1.04,11.05,12.06,10]dodeca-2,7-diene

(82) apparently via the intramolecular Diels-Alder cyclization

of (81).35 This single step process yields (82) in 8-12%

yield compared to 7.3% overall yield for the previously re-

ported multistep synthesis of (82).

Treatment of diene (82) with an equimolar amount of

m-chloroperbenzoic acid gave a mixture of unreacted diene,

the monoepoxide (83), and the diepoxidized olefin which were

easily separated via chromatography on alumina. Because

of time limitations and the generosity of Professor Paquette








in supplying a sample of (16)-OH, further work on the synthesis

of (16)-OH was discontinued. A feasible route to obtain

(16)-OH would involve catalytic hydrogenation of the un-

saturated epoxide (83) to the saturated epoxide (84) and

reduction with lithium aluminum hydride to give (16)-OH.

An interesting second product was obtained from the

reaction of (24)-Cl with thallium cyclopentadienide whose

Hnmr was strikingly similar to that of a mixture of dihydro-

as-indacenes.36 The mass spectrum was in good agreement

with this assignment, giving strong peaks at m/e 154, 153,

and 152 corresponding to the molecular ion, loss of one

hydrogen, and loss of two hydrogens, respectively, and sig-

nificant peaks at m/e 77, 76.5, and 76 probably due to

doubly charged ions. The absence of further significant

fragmentations is typical of aromatic systems.

A possible reaction pathway for the formation of the

dihydro-as-indacenes (85) and (86) is given in Scheme VI.

Attack of the thallium cyclopentadienide at the C2 position

of the bishomocyclopropenium ion (87) followed by successive

1,5-hydrogen migration, Cope rearrangement, and homo-1,5-

hydrogen rearrangement leads to the tricyclic intermediate

(88). Two different 1,5-hydrogen shifts followed by de-

hydrogenation, presumably oxidative in nature, lead to

dihydro-as-indacenes (85) and (86). Another possible route

to (86) is two consecutive 1,5-hydrogen shifts from (88)

involving only the five membered ring followed by oxidation

to give (86). The oxidizing agent in these cases could be











-Tl Cl
+


[1,5]
>y


[3,3]


[1,5]


88


[1,5]







86


an impurity in the thallium reagent resulting from air
oxidation of the thallium cyclopentadienide. This suggestion
would appear to explain the observation that dihydro-as-in-
dacenes were formed in appreciable amounts only when aged
and somewhat discolored thallium reagent was used. When a
fresh bottle of the thallium reagent was used, much less
dihydro-as-indacene was formed and what was presumed to be


SCHEME VI


[1,5]


I [








the unoxidized tetrahydro-as-indacene(s) was detected in the

product mixture.

The product studies for which the above syntheses were

attempted required the preparation of (9)-OBs, (15)-OBs, and

(16)-OBs. Treatment of the respective alcohols with p-bromo-

benzenesulfonyl chloride in pyridine gave the brosylates

with the expected downfield shift of the a-hydrogen in the

Hnmr. For comparison to possible products of solvolysis

of these brosylates, the endo-alcohol (89)-OH was prepared

by reduction of ketone (44) with lithium aluminum hydride.


OH






89-OH

The spectral properties were in agreement with the assigned

structure.



Product Studies

Acetolysis of the three interrelated brosylates (9)-OBs,

(15)-OBs, and (16)-OBs gave mixtures of the acetates (15)-OAc

and (16)-OAc which were analyzed by capillary glpc and

checked by 1Hnmr. The results of these glpc analyses are

given in Table II. Reduction of the acetate mixtures ob-

tained from acetolysis of (9)-OBs and (15)-OBs with lithium

aluminum hydride resulted in alcohol mixtures which were

also analyzed by capillary glpc. Incomplete separation of








TABLE II. Acetate Product Distribution from the Acetolysis
of (9)-OBs, (15)-OBs, and (16)-OBs


Substrate % of (15)-OAc % of (16)-OAc



Bs 91.50.2 8.50.2


9-OBs


BsO
BsO 65.40.3 34.60.3


15-OBs




BsO 9.90.3 90.10.3

16-OBs








the peaks precluded accurate analysis of the alcohols, but

the integration of the two largest alcohol peaks was in

general agreement with the acetate ratios. A third smaller

peak was observed in both alcohol mixtures. The mass spectra

of the larger peaks obtained directly by GC/MS analysis of

the alcohol mixture derived from (9)-OBs were in good agree-

ment with the conventionally obtained spectra of (15)-OH and

(16)-OH, but the glpc obtained mass spectrum of the third

component was not similar to any of the reasonable alcohol

products such as (9)-OH, (15)-OH, (16)-OH, (89)-OH or the

endo-alcohol (90)-OH. However, the third component was ob-






HO

90-OH

viously an isomeric alcohol which may either be a product

of acetolysis or a result of decomposition of the product

alcohols on the glpc column or the injection port. These

results obviously contradict the report that (9)-OBs and

(15)-OBs give a single product on acetolysis. This error

is understandable since the work was done prior to the rou-

tine use of nuclear magnetic resonance spectroscopy and

the components may have given a single peak on glpc analysis

if a less efficient column were used.

The product ratio for the solvolysis of (16)-OBs

indicates that the o-bridged ion (14) is trapped by the








solvent with little or no rearrangement to the bishomocyclo-

propenium ion (13). The extremes for the extent of rearrange-

ment of ion (14) to ion (13) range from a low of zero if

capture of ion (14) leads to 90.1% (16)-OAc and 9.9% (15)-OAc

to a high of 9.9% if capture of ion (14) leads solely to

(16)-OAc. Using these extreme product ratios for the cap-

ture of ion (14) by solvent, (9)-OBs, which must ionize ini-

tially to ion (13), undergoes 8.5 9.4% rearrangement to

ion (14) prior to solvent capture. Similar analysis indicates

that ester (15)-OBs, which has the possibility of the partic-

ipation of either the C -C12 bond to give ion (14) or the

C9-C10 bond to give ion (13), leads to a mixture of 61.6 -

65.4% of ion (13) and 34.6 38.4% of ion (14) on ionization.

To assess the relative importance of thermodynamic con-

siderations (product stability) versus bond alignment in the

formation of ions (13) and (14) from (15)-OBs, the acid-

catalyzed equilibration of (15)-OAc and (16)-OAc was performed

in acetic acid at 75 C. A starting mixture of 65% of (15)-

OAc and 35% of (16)-OAc resulted in a mixture containing

99.5% of (15)-OAc and 0.5% of (16)-OAc. This ratio corre-

sponds to a free energy difference of 3.7 Kcal/mole assuming

equilibrium was established. Thus, participation of the C9-

C10 bond in (15)-OBs would be favored by thermodynamic con-

siderations while participation of the CI-C12 bond is favored

by bond alignment since molecular models show that the CI-C12

bond is exactly anticoplanar to the leaving group while the

C9-C10 bond deviates from anticoplanarity.











B sO


15-OBS


Bs



9-OBs


A similar situation is found in the exo-twistbrendyl

brosylate (37)-OBs where participation of the C5-C6 bond




BsO 654 3


37-OBs

leads to products which are 2.24 3.13 Kcal/mole more stable

than products derived from C3-C8 bond participation, but

where participation of the C3-C8 bond is favored by bond

alignment.16a Acetolysis of (37)-OBs led to products which

favor migration of the C3-C8 bond by 2.2 : 1 over C5-C6 mi-

gration. No products with the (37)-OAc structure were ob-

served because (37)-OAc is 8.7 11.8 Kcal/mole higher in

energy than the observed products. The opposite situation

is observed in the solvolysis of (15)-OBs where the products


SCHEME VII


BsO

16-OBs







14
14


13








favored by thermodynamic factors predominate ca. 2 : 1 over

products favored by bond alignment.

Analysis of the rate for acetolysis of (16)-OBs relative

to (15)-OBs may aid in explaining this behavior. A lower

limit to the rate constant for the acetolysis of (16)-OBs

can be determined from the product study with this ester.

Acetolysis for nine hours at ca. 25 C gave complete reaction

to the limits of the 1Hnmr spectral analysis which corresponds

to at least five half-lives and a rate constant of at least
-4 -1
1.lxl0 s at 25 C. The rate of (16)-OBs appears normal

when compared to exo-2-norbornyl brosylate (1)-OBs, but un-

usually high when compared to the annelated norbornyl systems

(91), (92), and (93) (Table III). On the contrary, (15)-OBs

appears quite in line with the rates of these annelated 2-

norbornyl systems. If one assumes that the free energy dif-

ference between (15) and (16) is essentially the same at

75 OC as at 25 oC, and that the products of acetolysis from

(15)-OBs represent the relative amounts of C1-C12 and C9-C10

bond participation, then a free energy diagram for the ace-

tolysisof (15)-OBs and (16)-OBs can be constructed as in

Figure 1. Most of the 4.8 Kcal/mole difference in the free

energy of activation for the respective ionizations of (15)-

OBs and (16)-OBs to ion (14) is thus seen to result from the

higher ground state energy of (16)-OBs (3.7 Kcal/mole). By

way of confirmation of the ground state energies, molecular

models indicate that (16) is more strained than (15) and in

particular that the C1-C12 bond in (16) is more strained than







Table III. Relative Rates of Sulfonate Esters at 25 C


Compound Reference k



J / 15 8a 1


16 This Work


X

















Ox





x
,&X x


21200





2500


5.02


17.3


8.89










= 1.. Kcal/mole 0.4 Kcal/mole












14 13





-----3.7 Kcal/mole



Bs BsO-J /V

16-OBs

15-OBs



Reaction Coordinate




Figure 1. Free Energy Diagram for the Acetolysis of (15)-OBs
and (16)-OBs








the C1-C12 bond in (15). The added strain of the C -C12

bond in (16) raises the energy of the o-orbital allowing

better orbital mixing at the developing carbonium ion site.

If one assumes that the relative stabilities of (15)-OAc and

(16)-OAc reflect the relative stabilities of ions (13) and

(14), respectively, then the stability of ion (13), as re-


Bs BsO


1 12 1 2


15-OBs 16-OBs

flected in the transition state, may partially compensate

for the poorer alignment of the C 9-C10 bond in (15)-OBs.

Finally, the relative difficulty with which ions (13)

and (14) appear to interconvert may be a reflection of the

rigidity of this system. Realignment of the bonds for the

interconversion of these ions may be higher energy processes

than in the more flexible ions derived from the cyclopropyl

brosylate (26)-OBs.















CHAPTER IV
KETO- AND KETAL-SUBSTITUTED BROSYLATES



Introduction

As a result of the unsuccessful attempts to synthesize

the ketone (43), another approach to the identification of

the solvolysis products of (26)-OBs was taken. The intro-

duction of a ketal or ketone functional group in the 12-

position would be expected to reduce the number and/or

amounts of rearranged products from the solvolysis of bros-

ylates (51)-OBs and (52)-OBs because the electron with-

drawing effect of these groups would make the rearranged ions

less favorable. In addition the rates of acetolysis of the

substituted brosylates should give some insight into the

transition state of the solvolysis of (26)-OBs. Since the

products of acetolysis of the unsaturated (9)-OBs have been

thoroughly investigated, substitution of a ketal or ketone

functional group in the 12-position to give (53)-OBs and

(54)-OBs would be useful for a comparison of the acetolysis

products with those of the cyclopropane brosylates (51)-OBs

and (52)-OBs. The rates of the saturated brosylates (55)-

OBs and (56)-OBs can be used for comparison with the rates

of acetolysis of (51)-OBs and (52)-OBs.








Synthesis

The synthesis of 12,12-dimethoxy-exo,exo,anti-tetracyclo-

[6.2.1.13'6.02'7]dodec-4-en-ll-ol (53)-OH followed essentially

the same procedure used for the preparation of (9)-OH and is

outlined in Scheme VIII, along with the preparation of several

alcohols derived from (53)-OH. Heating (5)-OAc with 5,5-

dimethoxy-1,2,3,4-tetrachlorocyclopentadiene gave 12,12-di-

methoxy-3,4,5,6-tetrachloro-exo,exo,anti-tetracyclo[6.2.1.-

16,3.02,7]dodec-4-en-ll-yl acetate (94)-OAc as the only

adduct, which gave (94)-OH on reduction with lithium aluminum

hydride. Neither (94)-OAc nor (94)-OH gave a parent ion in

the mass spectrum but both showed isotope peaks for parent

minus a chlorine atom and other peaks consistent with the

structures. Dechlorination of (94)-OH with sodium and tert-

butanol in THF gave (53)-OH cleanly and in good yield, which

was in sharp contrast to the dechlorination of (77)-OH to

give (9)-OH, and must be related to the absence of chlorines

in the 12-position of (94)-OH. Hydrolysis of (53)-OH gave

the ketoalcohol (54)-OH in good yield. Ketoalcohol (54)-OH

bubbled upon melting, presumably extruding carbon monoxide

to give dienol (95)-OH. The mass spectrum showed similar









95-OH

behavior, giving no parent ion, but a spectrum consistent







SCHEME VIII


+ 7 e 150 0C
1J'4


Qc



5-QAc


-MeOAv

MeO 53-OAc


I CH2N2
Cu (I) Br


53-OH -


5% H2SO H2

N 4\ Pd/C


5 O H



54-OH


51-OAc


LiAlHe4
ether


51-OH


MeOOH

MeO
MeO 55-OH


10%
H2SO4


OH




56-OH


Meo 94-OAc



LiAlH4
ether
-78 C


52-OH








with the loss of carbon monoxide. Catalytic hydrogenation

of (53)-OH gave the saturated alcohol (55)-OH which was

hydrolyzed to the saturated ketoalcohol (56)-OH. The ele-

mental analysis for (56)-OH was consistent with partial

formation of the hydrated ketone, but there is no doubt as

to the identity of (56)-OH since a satisfactory analysis

was obtained for (56)-OBs derived from this sample of

(56)-OH.

The preparation of the cyclopropyl alcohol (51)-OH

followed the procedure previously described for (26)-OH.26

Treatment of (53)-OH with acetic anhydride in pyridine,

reaction with diazomethane in the presence of a copper(I)

chloride catalyst, and reduction with lithium aluminum hy-

dride gave the cyclopropanated alcohol (51)-OH. Hydrolysis

of (51)-OH gave the ketoalcohol (52)-OH.

The 1Hnmr spectra of (51)-OH and (52)-OH are of interest.
H H H

11 OH H10 11 H
9 9
MeO
12 0 12
MeO
51-OH 52-OH

The 1Hnmr spectrum of (51)-OH contains two one proton reso-

nances at 6 0.17 doublett of triplets, J = 4.5, 7.0 Hz) and

6 1.50 doublett of triplets, J = 4.5, 3.0 Hz). The downfield

resonance must be assigned to HI0s since the close proximity

to the oxygen would be expected to deshield H10s and since

the smaller coupling constant for the triplet would be ex-








38
pected for the HI0 which is trans to H9 and H The up-

field position of H10a is higher than expected for a normal

cyclopropane proton, but is accounted for by reciprocal shield-

ing by the electrons in the H1 0s bond which are repelled by

the oxygen. The 1Hnmr spectrum of (52)-OH contains two one

proton resonances at 6 0.23 doublett of triplets, J = 6.0,

3.0 Hz) and 6 0.72 doublett of triplets, J = 6.0, 7.0 Hz).

In this case the upfield resonance is assigned to Hl0s be-

cause of the smaller triplet coupling constant. The H10a

has a "normal" cyclopropane resonance while H10s must be in

the shielding cone of the carbonyl group.

The corresponding brosylates were prepared in the usual

manner from the respective alcohols with p-bromobenzenesul-

fonyl chloride in pyridine. The brosylates had spectra

similar to those of the corresponding alcohols.



Product Studies

Acetolysis of the unsaturated ketal (53)-OBs resulted

in a mixture of eight products (glpc), the two major products

being 69% and 19% of the mixture. These products were sepa-

rated by preparative glpc and assigned the pentacyclic struc-

tures (96)-OMe and (96)-OAc, respectively, on the basis of

spectral properties. The 1Hnmr of (96)-OAc was similar to

(15)-OAc, giving a doublet for H11 at 6 5.09 (J = 2.2 Hz)

and the acetate methyl singlet at 6 1.98. The infrared

spectrum contained the carbonyl absorptions for both the

bridge (1770 cm-1) and acetate (1740 cm-1) carbonyls. The








mass spectrum and accurate mass were also in agreement with

this structure. The 1Hnmr spectrum of (96)-OMe was similar









96-OMe

to (96)-OAc except that the resonance for H11 was shifted

upfield to 6 3.77 doublett, J = 2.2 Hz) and the acetate

methyl resonance was replaced by the methoxyl resonance at

6 3.27. In addition, the infrared spectrum showed a single
-i
carbonyl absorption resonance at 1775 cm- for the bridge

carbonyl The 13Cnmr spectrum, mass spectrum, and accurate

mass were also in agreement with this structure. The third

most abundant component (3.3%) was assigned the structure

(97) since the 1Hnmr spectrum had resonances at 6 6.40





MeO2C



97

(triplet of multiplets, J = 6.0 Hz) and 5.99 (triplet of

multiplets, J = 6.0 Hz) for the vinyl protons and a singlet

at 6 3.59 for the methyl ester. The infrared spectrum also

showed a carbonyl resonance at 1740 cm-1 for the ester

carbonyl.

By analogy, endo-2-norbornyl ketal (66)-OBs afforded







55% of (57)-OAc and 16% of (57)-OMe while (63)-OBs gave

0 T -
MfeO OMe Z4



x x

66 57 63 58

57% of the cleavage product, similar to the products iden-

tified from the solvolysis of (53)-OBs.24bc The other minor

products from (53)-OBs were not identified.

The solvolysis of the unsaturated ketone (54)-OBs gave

a mixture of products consistent with 85% of ('93)-OAc and ge-

15% of the decarbonylated brosylate (95)-OBs contrary to

the result for the solvolysis of the endo-(58)-OBs where a

significant amount of (58)-OAc was observed. Presumably,

the reason for the absence of endo products is the formation

of the bishomocyclopropenium ion from (54)-OBs and the

hindered attack from the endo direction.

Acetolysis of the cyclopropyl ketal (51)-OBs, although

anticipated to give a simpler product mixture than (26)-OBs,

gave a mixture of at least twelve products (glpc) with the

major product accounting for 21.9% of the volatile products.

The shorter retention time peaks (57.7%) can reasonably be

assigned to methoxy ketones since these would be expected

to be less polar than the corresponding acetoxy ketones and

since there are several resonances in the IHnmr spectrum

which appear to be due to methoxyl groups (Figure 2). The

57.7% of methoxy ketones from (51)-OBs is similar to the




























En

0






0


4-)


Er)
5-1
*1-I




0



4-3



4-1
0




4-)


U3

w4



E-1







69% of (96)-OMe from (53)-OBs indicating a similar degree of

methoxyl participation in the transition state. Due to the

complexity of the product mixture, the only product iden-

tified was ca. 30% of (52)-OBs formed by hydrolysis of

(51)-OBs.

As in the solvolysis of the unsaturated ketone (54)-OBs,

acetolysis of the cyclopropyl ketone (52)-OBs gave a less

complex product mixture (glpc) than the ketal. The major

component (68.7%) was isolated by chromatography on silica

gel and assigned the structure of the expected (98)-OAc on

the basis of spectral analysis. The 1Hnmr spectrum contained

the acetate singlet at 6 1.97 and H12 consisted of the ex-




AcO

O0 6 0 '

98-OAc 99 100

pected doublet of doublets (J = 6.0, 3.0 Hz) at 6 5.06

since the endo H12 should show very little coupling to H1

and should show different couplings to the H11 protons. The

infrared spectrum showed only one carbonyl absorption at

1750 cm- but the mass spectrum and accurate mass were in

agreement with structure (98)-OAc. In particular, the mass

spectrum showed a large peak for parent ion minus acetic

acid indicating an easy elimination reaction as expected

from (98)-OAc. The other products were not isolated or

identified. The main reason for the absence of extensive








rearrangements is that ions (99) and (100) would be higher in

energy than the corresponding ions from (26)-OBs due to the

proximity of the electron-deficient carbonyl group.



Kinetic Studies

The acetolysis rates of the cyclopropyl (51)-OBs and

(52)-OBs and the unsaturated (53)-OBs and (54)-OBs along

with the saturated analog (55)-OBs were determined and are

given in Table IV along with several related brosylates.

The rate of the saturated ketal (55)-OBs provides an estimate

of the rate-retarding effect of the ketal function when com-

pared to (10)-OBs. The observed ratio of only five times

faster for (10)-OBs is small compared to a factor of 33 for

the endo-2-norbornyl (66)-OTs, as expected for the increased

distance between the developing positive charge centers and

the ketal functionality. The cyclopropyl and unsaturated

ketals (51)-OBs and (53)-OBs do not show any significant rate

retardation. In fact, (51)-OBs is slightly accelerated

over (26)-OBs and (53)-OBs may be accelerated over (9)-OBs

since the rate given is only a lower limit estimate. The

rate acceleration, or at least an absence of deceleration,

and the methoxy products observed point to the participation

of ketal function in the transition state by stabilization of

the developing, delocalized positive charge by the lone

pairs on oxygen. The rate acceleration suggests that de-

localization of the positive charge must be occurring early

in the reaction sequence.





54








0 0 0 0

co












0 0
D M0 00C 0 00 0
4-1 r0 r-* -i r-i -4 1-4 r-4
-P-H 0 H H H H H-
H LO m .-- H CN rn
H4 Al


0






0 h .-- r--I Lo LO r rL r i LO


o c o H a ) IV L N l )
0 4l I4 O I I II I I 4Ia








r-I o . * * +
- ) O-I r-- o1Clrln r. I









0




N 0 0D 0 0 0 0 C

H IC D II 3
1-4






0
~p0 u) 0 U U U

I o) U


a)a)









4J

4 4-J
-

0 N 0
l CO .0
Q) En 00

S4n O o
C) (B


44

0 0
Q ) 0
4-0 4J "







4 o . C .,A 0 *
S, 0 a Q4
En 40 0

ci 0 4
o ci 44 0i













(N rz 44 0
0. 440 0 4- 0
,-H -, ,-I 4 H) 4J






I X X 3 <- -0 (.






S O O (c o m 0 0 (
m 4-1 4J









4 -
4 ( P 4 .

H od -4 Ci (C
(au t 1H i i (C 1
,3 1 H 1 4-1

















E-4 C: ti-4 ,
H) H H C-i 4 ) 0








I -XX X 0 c.
(0 C C00 N .0 Ci 0 C0
*I I C 0 0' H 0
uHn( o .0 0










4J C4 c -l >1 r--
0n 3 n 0 4









a 0- 4- (C ..

4-0 a) 4 1 4-
( d U ( ) Q ) I


















4 1-1 () 4J z z
4 0 0 0 1 ( >U
> .c. .4 (d 0- *H E 0
- oo0 1n :I 4 i 4-




QM W 4-1 1-4 I4 | -4
Xo m 0 a) V 4 Q) a) 0
'l 0 o 4i c a) Cl .Z















E(C .0 g fV 'Cl a) (4-i







Contrary to the observed rate effect with the ketals,

the cyclopropyl ketone (52)-OBs was decelerated by a factor

of 155 compared to (26)-OBs and the unsaturated ketone (54)-

OBs was decelerated by a factor of 600 compared to (9)-OBs.

These deceleration factors are not directly comparable since

the cyclopropyl rates were compared at 100 C while the

unsaturated ketone at 50 oC was compared to (9)-OBs at 25 C

assuming a similar temperature sensitivity between (9)-OBs

and (6)-OBs. Since raising the temperature has a tendency

to diminish acceleration effects, these two factors are

probably about equal. The unsaturated brosylate (9)-OBs

shows less than 10% rearrangement to ions other than the

initially formed bishomocyclopropenium ion (13) videe supra)

and, therefore, cannot derive significant stabilization

from such rearranged ions. On the other hand, extensive

rearrangements from the initially formed trishomocyclopro-

penium ion (27) are observed in the solvolysis products of

the cyclopropyl brosylate (26)-OBs. The electron withdrawing

effect of the carbonyl group would be expected to similarly

retard rearrangements from the ions formed initially during

the acetolysis of (52)-OBs and (54)-OBs. Studies of the

solvolysis products of these two brosylates help to sub-

stantiate this contention videe supra). Since similar re-

tardation effects are observed for both ketone substituted

brosylates and since (9)-OBs cannot derive significant sta-

bilization from rearrangement of the initially formed ion

(13), then (26)-OBs must be devoid of significant stabili-




57



zation in the transition state derived from further rearrange-

ments of the initially formed trishomocyclopropenium ion (27).















CHAPTER V
CIS,ANTI-4,5-EPOXY-ANTI-TETRACYCLO-
[6.2.1.37,6.02,7]DODEC-ll-YL BROSYLATE



Introduction

A natural extension of cyclopropyl participation in

solvolysis reactions is the use of three-membered rings

containing heteroatoms as a neighboring group. The most

appropriate choice for this study was the epoxy brosylate

(72)-OBs because of the structural relationship to (26)-OBs

and because it was easily synthesized from (9)-OBs. The

only other reported attempt at remote epoxide participation

resulted in no observed acceleration and no products derived

from the carbon-carbon bond participation, so if (72)-OBs

were to demonstrate either of these features it would pro-

vide the first example of remote epoxide participation in a

solvolysis reation.25b



Synthesis

The cis,anti-4,5-epoxy-anti-tetracyclo[6.2.1.13' 6.0 2,7]-

dodec-ll-yl brosylate (72)-OBs was easily prepared by reaction

of the unsaturated ester (9)-OBs with m-chloroperbenzoic

acid in chloroform. The 1Hnmr spectrum had a two-proton

singlet at 6 3.04 for the epoxide protons and other resonances








consistent with this structure. ll-Oxapentacyclo[6.4.1.03'10

04'13.05'9]tridecan-10-one (101) was required for the authen-


0
.-OBs 0-





72-OBs 101 102

tic synthesis of derivatives related to the anticipated

solvolysis products of (72)-OBs. Reaction of the pentacyclic

ketone (44) with m-chloroperbenzoic acid in a Baeyer-Villager

fashion might be expected to give the lactone (101) and the

isomeric (102) but only (101) was obtained, probably for the

steric reasons discussed previously for the reaction of di-

azomethane with (44). The IHnmr spectrum (Figure 3) of (101)

showed a doublet of doublets (J = 4.4, 2.5 Hz) at 6 4.80 for

H 10, a multiple at 6 3.18 for H1, a doublet of sextets (J =

4.4, 1.0 Hz) at 6 2.55 for H9, and a broad doublet (J = 2.5

Hz) at 6 2.42 for H3. The H10 resonance would be expected

to be a doublet of doublets in (101) since molecular models

indicate that the dihedral angle between H3 and H10 is better

for coupling than that between H9 and 1110. Decoupling of the

H10 resonance resulted in a sextet at 6 2.55 as expected for

the bridgehead proton H3 and a broad singlet at 6 2.42 as

expected for the bridge proton H9, with no change in the 6

3.18 resonance. Had (102) been produced, a more complex

resonance would have been expected for the a-portion (H1) of

the lactone. The mass spectrum of (101) had an unusually


























































C.)

44

a4

U)

C4

rz








intense peak at one unit higher than the molecular ion,

probably due to ion molecule reactions in the ion source.

The remaining spectral results for the mass spectrum, the

infrared spectrum, and the 13Cnmr spectrum were all consis-

tent with the structure of (101).



Kinetic Studies

Solvolysis of (72)-OBs in 60% aqueous acetone with

2,6-lutidine as a buffer gave rate constants of 1.68x10-4
-ilo
s-1 at 100.0 'C, 3.57xl0-5 s-1 at 85.0 C, and an extra-
-8 -
polated rate constant of 1.52xl0-8 s-1 at 25.0 C. The AH1

was 26.7 Kcal/mole and the AS was -4.8 eu. Since for simi-

lar systems the solvolysis rate for brosylates in 60% aqueous

acetone is 21.4 times faster than in acetic acid, the extra-

polated rate constant for (72)-OBs in acetic acid at 25 C
-10 -1 39
is 2.10xl0- s Thus, the rate acceleration for (72)-

OBs over 7-norbornyl brosylate is 104.6 compared to 107.7

for the cyclopropane (26)-OBs and represents 4.3 Kcal/mole

difference in activation energy at 25 oC.4a,14

The rate enhancement and solvolysis products are con-

sistent with the initial formation of the oxatrishomocyclo-

propenium ion (73) followed by collapse to the oxygen

stabilized ion (74). Capture of either of these ions by

the solvent (water) would be expected to give the hemiacetal

(75) which upon hydrolysis should afford the observed alde-

hyde (76) videe infra).














5O HO




73 74 75

One explanation for the reduced accelerative effect of

the carbon-carbon bond of the oxirane versus the cyclopro-

pane is the lower strain energy of the oxirane (13 Kcal/mole)

versus the cyclopropane (25 Kcal/mole). The lower overall

strain energy of the oxirane would be expected to lower

the energy of the occupied o-orbital of the carbon-carbon

bond thus affording poorer mixing with the developing empty

p-orbital at C.ll Additionally, the electron withdrawing

effect of the oxygen should lower the energy, and thus the

nucleophilicity, of the carbon-carbon a-bond.

The lower energy of the carbon-carbon o-bond of oxirane

compared to cyclopropane is confirmed by the photoelectron

spectra of these molecules. The correlation diagram for the

highest occupied MO's of cyclopropane, oxirane, and thiirane

is given in Figure 4.41 The most important orbitals with

respect to participation of the carbon-carbon bond in solvol-

ysis reactions are the es (one of the 2e') for cyclopropane

and the 4a, for oxirane and thiirane since these orbitals

show bonding character between the carbon atoms. Heilbronner

has shown that the 2e' orbitals of cyclopropane are split

by Jahn-Teller effects giving an average ionization potential











7





9 9.00
/ 1


10.53 1 .5/
--i 2eW 2b1
\ / 11.30
11 ,z 2b
\ 11.7__ / 11.86 2
12 4a 1l 2
/\ / 4a1
/ \ /

-13 13.2
le 2b \13 / 13.5
le 2b
-14 14.2 2 -
la2

-15






Figure 4. Correlation Diagram for the Highest Occupied MO's
in Cyclopropane, Oxirane, and Thiirane





of 10.9 eV and, further, that fusing the cyclopropane into

a norbornyl skeleton lowers the ionization potential of the

e to 9.4 eV.42 The 1.5 eV increase in the energy of the
s
e orbital is due in part to the electron donating effect
s
of the alkyl substitution and in part to the increase in

strain. Using Heilbronner's value the unsubstituted oxirane

4a1 orbital is 0.8 eV (18 Kcal/mole) lower in energy con-









tribution from ion (74) to the stabilization.

One method for distinguishing between these two possi-

bilities is the solvolysis of the thiirane analog of (72)-OBs.

If the former explanation is correct, then the thiirane will

provide less stabilization than the oxirane since sulfur is

less electronegative and, therefore, the 4a1 of the thiirane

should not be affected as greatly by alkyl substitution as

the oxirane. If the latter explanation is correct, then the

thiirane will provide greater stabilization since the 2b1

(lone pair) orbitals of the thiirane are at higher energy

and thus would interact with the developing p-orbital more

efficiently.41



Product Studies

Solvolysis of (72)-OBs for greater than 10 half-lives

gave a mixture of products whose 1Hnmr spectrum (Figure 5)

contained two aldehyde resonances. The unexpected presence

of a second aldehyde was explained by the initial formation

of endo-10-hydroxy-endo-tetracyclo[6.2.1.04'll.05 ]undecan-

3-carboxaldehyde (76) followed by isomerization to the endo,-

exo-aldehyde (103), since chromatography of the product





OHC T HO

OHC

76 103

mixture on silica gel or treatment of the product mixture
















0






0




0



4J

00


(C)

r-0

r 4J~




0


'oW


0 -Hi


C>




4JH
4)
0 0-





40


04






00
4-)


l-'.-f.-








with 15% sodium hydroxide resulted in the isolation of a

single aldehyde (Figure 6) corresponding to the minor alde-

hyde of the product mixture.

Solvolysis of (72)-OBs for 2.6 half-lives gave a mix-

ture of unreacted starting material and products whose 1Hnmr

spectrum (Figure 7) contained a single aldehyde resonance

corresponding to the major aldehyde of the above product

study.

The stereochemistry of the alcohol group in (76) and

(103) must be as shown because of the similarities in coupling

constants and chemical shifts with those of H10 in lactone

(101) and because molecular models indicate that the opposite

stereochemistry would result in coupling constants which

are small or zero. Furthermore, reduction of the aldehyde

mixture and lactone (101) with lithium aluminum hydride re-

sulted in the materials which gave the 1Hnmr spectra in

Figures 8 and 9, respectively. The diol (104) derived from

lactone (101) gave a new doublet (J = 4 Hz) at 6 3.83 re-

sulting from the new methylene group. As expected the prod-





HOHC HO0
HOHHOH 2C
HOH2C

104 105

ucts from the reduction of the aldehyde mixture gave the

same resonance for the major component and an additional




























C- ~
















0
u)




0



4-)
~0
4)







0
a)






H0)

.4-1C



W0


'4~ 0



4-)t

0>k




00)

0).

'orL'4







ta C



























LH





0



0)

04


-H







4-)
(3
0








r-o




4-

























































CD




0




04









Ul









doublet (J = 6 Hz) at 6 3.44 for the new methylene group of

the diol (105). Reduction of the endo-aldehyde (103) also

gave this resonance at 6 3.44 as confirmation of the assign-

ment. The different couplings for the methylenes in (104)

and (105) may be a result of the restricted rotation of

the relatively hindered methylene in (104). The less hindered

methylene in (105) shows a normal 6 Hz coupling (compared to

7 Hz in ether) whereas the hindered methylene in (104) may

be held in a conformation which is less favorable for

coupling. The GC-MS analyses of the reduction products of

lactone (101), the aldehyde mixture, and the endo,exo-alde-

hyde (103) were also consistent with these assignments.

The GC-MS analysis of the aldehyde mixture was in agree-

ment with the assigned structures (76) and (103) as is

shown in Scheme IX. Fragmentation of the endo,endo-alde-

hyde (76) shows facile transfer of the aldehyde hydrogen

to the alcohol group and loss of carbon monoxide to give

(76)-b at m/e 164. Alternatively (76)-a can fragment to

give (76)-c followed by loss of ketene to give (76)-d at

m/e 150. The endo, exo-aldehyde (103) can lose water to

give (103)-a at m/e 174 followed by loss of the formyl rad-

ical to give (103)-b at m/e 145. Further fragmentation of

(103)-b could lead to benzene and the cyclopentenyl cation

at m/e 67.

Reaction of the aldehyde mixture with 2,4-dinitrophenyl-

hydrazine gave a crystalline derivative whose infrared

spectrum, mass spectrum, and accurate mass analysis were









SCHEME IX


H20 H 2


OHC o0C --
76- (M') 76-a 76-b
m/e = 164



+ +.


H2-CH2CO >


O 76-c 76-d
m/e= 150


CHO
103- (M+)


+


-H20 -


I 103-b
CHO
103-a m/e = 145
m/e = 174








m/e = 67
We = 67








consistent with adducts formed from (76) and (103).

In conclusion, the accelerative effect demonstrated

by the carbon-carbon bond of the oxirane ring on solvolysis

reactions opens up a new area of interest in physical-organic

chemistry. As already mentioned, the thiirane analog of

(72)-OBs should be solvolyzed to determine which orbitals

of the heterocycle are involved in the solvolytic stabili-

zation. The aziridine analog of (72)-OBs might also be of

interest since the nitrogen is intermediate in electroneg-

ativity between carbon and oxygen. Also of interest would

be the solvolysis of (71)-OBs and its thiirane and aziridine

OBs





71-OBs


analogs to determine if the orientational effect of the

heterocycles on solvolytic reactivity is similar to the

cyclopropane orientational effect.















CHAPTER VI
EXPERIMENTAL



Synthesis

General

Infrared spectra were recorded on either a Perkin-Elmer

Model 137 B of Beckman IR-10 spectrophotometer. Solution

spectra were recorded using matched 0.1 mm sodium chloride

cells. The absorption band positions reported hereafter

are given in wave numbers (cm-1).

Nuclear magnetic resonance spectra were recorded on

either a Varian Associates Model A-60, 60 MHz spectrophotom-

eter or a Varian Model XL-100, 100 MHz instrument. Chemical

shift values for IHnmr spectra are reported in 6 units

relative to tetramethylsilane (TMS) at 6 0.00. The chemical

shift values of 13Cnmr spectra are reported in ppm from TMS.

After the chemical shift values are given in parentheses the

multiplicity of the peaks, the relative integration, the

coupling constants (J) where applicable, and the assignments

of the peaks if known. The symbols used for the multiplic-

ities are s = singlet, d = doublet, t = triplet, q = quar-

tet, p = pentet, and m = multiple.

Melting points are uncorrected and were obtained with

a Thomas Hoover melting point apparatus.









Mass spectra were recorded on an Associated Electronic

Industries (AEI) Model MS-30 mass spectrometer at 70 eV

ionizing current. Accurate mass determinations were ob-

tained using the same instrument linked with an auxiliary

PDP-8 digital computer.

Microanalyses were performed by Atlantic Microlabs, Inc.,

Atlanta, Georgia.

Analytical gas-liquid partition chromatography (glpc)

was performed with a Varian Associates Model P1440 chromato-

graph utilizing flame ionization detection. The columns

employed were

a) 100 ft. x 0.01 in. capillary coated with diethylene

glycol succinate (DEGS),

b) 100 ft. x 0.01 in. capillary coated with UCON

LB-550,

c) 100 ft. x 0.03 in. capillary coated with DEGS, and

d) 6 ft. x 1/8 in. 5% SE-30 on Chromsorb W.

Integration of analytical glpc tracings was obtained with

an Autolab 6300 digital intetrator obtained from Spectra-

Physics, Inc.

Preparative glpc was performed with a Varian Associates

Model A-90-P utilizing a thermal conductivity detector.

The columns employed were

a) 6 ft. x 1/4 in. 20% DEGS on Chromsorb W,

b) 10 ft. x 1/4 in. 20% UCON polar 50-HB-2000 on

Chromsorb W,

c) 5 ft. x 1/4 in. 5% DEGS on Chromsorb W,








d) 5 ft. x 1/4 in. 3% SE-30 on Chromsorb W

e) 5 ft. x 1/4 in. 7% OV-17 on Chromsorb W,

f) 5 ft. x 1/4 in. 5% UCON polar 50-HB-2000 on Chrom-

sorb W, and

g) 5 ft. x 1/4 in. 3% OV-225 on Chromsorb W.



Synthesis of 7-norbornadienyl benzoate (24)-OBz

In a typical reaction 4.8 g (0.033 mol.) of copper(I)

bromide and 306 g (3.32 mol.) of norbornadiene were dissolved

in 1500 ml of benzene and heated to 40 C under a nitrogen at-
27
mosphere with mechanical stirring. To this solution was added

615 g (2.54 mol.) of benzoyl peroxide in 1500 ml of benzene as

a slurry in several portions. When addition was complete, the

mixture was heated to reflux for three days, cooled to room

temperature, and separated into two portions for easier hand-

ling. Each portion was washed successively with 3 x 150 ml of

10% sodium carbonate, 2 x 150 ml of 5% ferrous sulfate, 2 x

150 ml of 10% sodium carbonate, 2 x 200 ml of 3 N hydrochlo-

ric acid, 2 x 150 ml of 10% sodium carbonate, 2 x 200 ml

of water, and 1 x 200 ml of saturated sodium chloride solu-

tion. Each set of washings was extracted with 100 ml of

benzene. The combined benzene layers were dried with an-

hydrous magnesium sulfate, filtered, and benzene removed

on a steam bath. The residue was distilled (1.5 torr,

140 oC) to give 226 g (25.0%) of 7-norbornadienyl benzoate,

(24)-OBz.27 The 1Hnmr (CCI4) contained the following reso-

nances: 6 7.8 8.1 (m, 2; ortho-aromatic protons); 7.1 -








7.5 (m, 3; meta- and para-aromatic protons); 6.71 (t, 2;

J = 2.3 Hz; H2 and H ); 6.59 (broad t, 2; J = 2.0 Hz; H5 and

H6); 4.77 (broad s, 1; H7; and 3.70 (sextet, 2; J = 2.0 Hz;

H1 and H4).



Preparation of 7-norbornadienol (24)-OH

In a typical reaction a solution of 135.5 g (0.498 mol.)

of (24)-OBz in 400 ml of anhydrous ether was added to phenyl-

magnesium bromide, prepared from 257 g (1.64 mol.) of bromo-

benzene and 40.0 g (1.67 mol.) of magnesium in 1100 ml of an-

hydrous ether under a nitrogen atmosphere, at a rate such as

to maintain reflux.27 The mixture was stirred for eight

hours after addition was complete and then poured into one 1

of saturated ammonium chloride. The ether layer was separated,

dried with anhydrous magnesium sulfate, filtered, and the

ether removed under reduced pressure. Pentane (50 ml) was

added and the solid triphenylmethanol removed by filtration.

The pentane was removed under reduced pressure and the residue

distilled (50 torr, 80 OC to 10 torr, 90 oC) to give 42.9 g

(79.7%) of 7-norbornadienol (24)-OH.27 The 1Hnmr spectrum

(CC14) contained the following resonances: 6 6.50 (t, 4; J =

2.0 Hz; H2, H3, H5, and H6); 3.76 (broad s, 1; H7); 3.38

(sextet, 2; J = 2.0 Hz; H1 and H4); and 3.2 (s, 1; OH).



Preparation of anti-7-norbornenol (5)-OH

Method A. A solution of 43.0 g (0.398 mol.) of

(24)-OH in 300 ml of anhydrous ether was added at room








temperature to 30 g (0.79 mol.) of lithium aluminum hydride

in 600 ml of anhydrous ether under a nitrogen atmosphere.28

This mixture was heated to reflux for 2 hrs., cooled in an

ice/water bath, and 100 ml of water added. The solid aluminum

salts were dissolved in 400 ml of 10% sulfuric acid and the

aqueous layer was extracted with 2 x 500 ml of ether. The

combined ether layers were dried with anhydrous potassium

carbonate, filtered, and the ether removed under reduced

pressure to yield 38.5 g (87.9%) of (5)-OH. This material

was used without further purification as the IHnmr was

essentially identical to that reported and showed greater

than 95% purity.

Method B. A solution of 25.0 g (0.166 mol.) of 7-

norbornadienyl acetate (24)-OAc (Frinton) in 75 ml of ether

was added to 10.0 g (0.264 mol.) of lithium aluminum hydride

in 400 ml of anhydrous ether under a nitrogen atmosphere.

This mixture was heated to reflux for 2 hrs., cooled in an

ice/water bath, and 10 ml of water, 10 ml of 15% sodium

hydroxide, and 30 ml of water added, respectively. This

mixture was filtered. The filtrate was dried with anhydrous

magnesium sulfate, filtered, and the ether removed under

reduced pressure to give 16.1 g (87.8%) of (5)-OH which was

greater than 95% pure by 1Hnmr.28

The materials obtained by each method were identical.

The 1Hnmr spectra (CC14) of each contained the following

resonances: 6 5.98 (t, 2; J = 2.0 Hz; H2 and H3); 3.57

(broad s, 1; H7); 2.53 (sextet, 2; J = 2.0 Hz; H1 and H4);








2.21 (s, 1; OH); 1.8 (m, 2; H5exo and H6exo); and 1.0 (m, 2;

H5endo and H6endo)



Preparation of anti-7-norbornenyl acetate (5)-OAc

Acetic anhydride (42 ml) was added to 42.3 g (0.384
29
mol.) of (5)-OH dissolved in 42 ml of pyridine. This solu-

tion was heated to 100 OC for one hour, cooled to 25 OC, and

100 ml of water added. This mixture was stirred for 10

minutes, poured into 200 ml of water, and extracted with

2 x 200 ml of ether. The combined ether layers were washed

with 2 x 250 ml of 5% hydrochloric acid and 200 ml of 10%

sodium carbonate. The ether was dried with anhydrous sodium

sulfate, filtered, and the ether removed under reduced pres-

sure to give 50.3 g (86.1%) of (5)-OAc. This was generally

used without further purification as the 1Hnmr was identical

to that reported. However, (5)-OAc could be distilled (con-

ditions unrecorded) to give a colorless liquid. The 1Hnmr

spectrum (CC14) contained the following resonances: 6 6.02

(t, 2; J = 2.0 Hz; H2 and H3); 4.33 (broad s, 1; H7); 2.76

(sextet, 2; J = 2.0 Hz; H1 and H4); 1.98 (s, 3; 02CCH3); 1.8

(m, 2; H5exo and H6exo); and 1.0 (m, 2; H5endo and H6endo).



Preparation of 3,4,5,6,12,12-hexachloro-exo,exo,anti-tetra-
cyclo[6.2.1.13'6.02'7]dodec-4-en-ll-yl acetate (77)-OAc

In a typical reaction, 20.0 g (0.131 mol.) of (5)-OAc

and 150 g (0.550 mol.) of hexachlorocyclopentadiene (Aldrich)

were placed in a glass tube, frozen in a dry ice/2-propanol








bath, pumped to 0.5 torr, sealed, and heated to 150 OC for

51 hours.26 This was cooled to room temperature, opened,

and the contents poured into 100 ml of hexane. The solid

which formed was removed by filtration and recrystallized

from hexane to give white prism, m. p. 132 133 C (lit.

132 132.5 C). The original filtrate was chromatographed

on SilicAR cc-7 with hexane to give additional solid with

a 1Hnmr spectrum identical to that of the initial solid.

The combined yield of (77)-OAc was 44.5 g (79.7%).26 The

1Hnmr spectrum (CC14) contained the following resonances:

6 4.80 (broad s, 1; H11); 2.72 (s, 2; H2 and H7); 2.50 (q, 2;

J = 2.0 Hz; H1 and H8); 1.96 (s, 3; O2CCH3); 2.0 (m, 2;

H9exo and H10exo); and 1.2 (m, 2 H9endo and Hl0endo).



Preparation of 3,4,5,6,12,12-hexachloro-exo,exo,anti-tetra-
cyclo[6.2.1.13,' 6. 02'7 ]dodec-4-en-ll-ol (77)-OH

In a typical reaction, a solution of 25.2 g (0.0593

mol.) of (77)-OAc in 300 ml of anhydrous ether was added

slowly under a nitrogen atmosphere to 5.1 g of lithium

aluminum hydride in 300 ml of anhydrous ether cooled in a

dry ice/2-propanol bath. The cold mixture was stirred for

one hour, 100 ml of water was added, and the mixture slowly

warmed to room temperature. To this was added 200 ml of

10% sulfuric acid and the mixture was stirred until all

solids had dissolved. The aqueous layer was extracted with

ethyl ether, the ether dried with anhydrous magnesium sulfate,

filtered, and the ether removed under reduced pressure to








give 21.0 g (92.5%) of (77)-OH, m. p. 131 132 C (lit.

134.5 135 C).26 The 1Hnmr spectrum (CDC13) contained

the following resonances: 6 4.08 (broad s, 1; HI1); 2.73

(s, 2; H2 and H7); 2.1 2.4 (m, 4; HI, HS, H9exo, and

Hl0exo); 1.69 (s, 1; OH); and 1.0 1.3 (m, 2; H9endo and

H10endo). The mass spectrum (70 eV) had m/e 386 (1.9%),

384 (4.3%), 382 (5.6%), 380 (2.7%), 351 (19.1%), 349 (33.0%),

347 (53.7%), 345 (33.3%), 315 (10.7%), 313 (48.9%), 311

(100%), and 309 (79.5%).

In a separate reaction a sample of the acetate was

reduced according to the literature preparation in which

the lithium aluminum hydride-ether mixture was held at room

temperature during the addition and then heated to reflux

for 30 minutes.26 The excess lithium aluminum hydride was

destroyed by the addition of water at 0 C and then treated

as before. The product obtained by this method was analyzed

by glpc using analytical column (d) (200 C) to give peaks

with the following retention times (approximate percent of

the mixture is given in parentheses): 7.5 min. (10%), 12.0

min. (5%), 13.5 min. (10%), and 17 min. (75%). Initial

reaction at -78 oC followed by heating to reflux prior to

working up produced the same mixture. The major peak corre-

sponded to (77)-OH. The product with the retention time of

7.5 minutes was separated by preparative glpc using column

(d). This component had m. p. 110.5 111.5 oC. The infra-

red spectrum (KBr) contained the following absorption bands:

3100(s), 2800(m), 1700(m), 1580(m), 1080(s), and 790(s) cm .
3100(s), 2800(m), 1700(m), 1580(m), 1080(s), and 790(s) cm








The 1Hnmr spectrum (CDC1 ) contained the following resonances:

6 3.96 (broad s, 1); 2.54 (s, 2); 2.32 (s, 2); 2.2 2.3

(m, 2); 1.8 2.2 (m, 2); 1.5 (s, 1); and 0.9 1.3 (m, 2).

The mass spectrum (70 eV) had m/e 386 (0.02%), 384 (0.03%),

382 (0.04%), 380 (0.02%), 351 (0.08%), 349 (0.19%), 347 (0.31%),

345 (0.17%), 281 (10.0%), 279 (29.8%), 277 (30.3%), 206 (53.3%),

204 (100%), and 202 (80.0%).



Preparation of exo,exo,anti-tetracyclo[6.2.1.13' 6.02]-
dodec-4-en-ll-ol (9)-OH

In a typical reaction 60 g (0.88 mol.) of tert-butanol

were added to 40 g (1.7 g at.) of sodium in 400 ml of tetra-

hydrofuran (THF) at room temperature under a nitrogen atmos-

phere with mechanical stirring. To this was added a solu-

tion of 21.0 g (0.0549 mol.) of (77)-OH in 200 ml of THF.

This mixture was heated to reflux for 36 hours, cooled to

room temperature, the excess sodium removed by filtration,

and 500 ml of water added. This was extracted with 2 x 500

ml of ether. The combined ether layers were dried with an-

hydrous magnesium sulfate, filtered, and the ether removed

under vacuum to give 11 g of crude (9)-OH. This material

was analyzed by glpc using analytical column (b) which showed

one major component with relatively large amounts of impuri-

ties. Chromatography on silica gel gave material which

showed the same impurity peaks on analytical glpc column (b)

but in smaller amounts. Recrystallization from hexane gave

pale yellow needles, m. p. 111 112 C (lit. 107 109 C,

108 109 C).26, 8a This material was found to be homo-








generous to glpc on capillary column (a).

The 1Hnmr spectrum (CC14) contained the following

resonances: 6 6.06 (t, 2; J = 2.0 Hz; H4 and H5); 4.71

(broad s, 1; H11); 2.85 (sextet, 2; J = 2.0 Hz; H1 and H8);

2.01 (t, 2; J = 2.0 Hz; H2 and H7); 1.7 1.9 (m, 2);

0.8 1.4 (m, 4); and 1.26 (s, 1; OH).



Preparation of exo,exo,anti-tetracyclo[6.2.1.13'6.02'7]-
dodec-4-en-ll-yl acetate (9)-OAc

Alcohol (9)-OH (5.30 g, 0.0301 mol.) was dissolved in

20 ml of pyridine and 20 ml of acetic anhydride and this

solution heated to 100 C for one hour.29 The reaction

solution was cooled in an ice/water bath and 50 ml of water

added. The mixture was allowed to stir for one hour and then

extracted with 200 ml of ether. The ether layer was washed

with 100 ml of water, dried over anhydrous magnesium sulfate,

filtered, and concentrated under vacuum. The product was

distilled (3.0 torr, 125 128 C; lit., 3.1 torr, 113 -

115 oC) to give 5.26 g (80.1%) of (9)-OAc. 26The 1Hnmr spectrum

(CC14) contained the following resonances: 6 6.13 (t, 2;

J = 2.0 Hz; H4 and H5); 5.45 (broad s, 1; H 11); 2.87 (broad

p, 2; H1 and H8); 0.9 2.2 (m, 10); and 1.83 (s, 3; O2CCH3).



Preparation of pentacyclo[6.3.1.13'6.02'7.09'11]tridec-13-yl
acetate (26)-OAc

Diazomethane generation was achieved by addition of

65 g of nitrosan (Aldrich, 70% in mineral oil) in small por-

tions to a flask containing 600 ml of ether, 40 ml of diethy-








lene glycol monoethyl ether, 30 ml of water, and 12 g (0.30

mol.) of sodium hydroxide which was cooled by immersion in

an ice/water bath.43 Nitrogen was passed slowly through the

ether layer containing diazomethane, through a drying tube

filled with potassium hydroxide pellets, and into a flask

immersed in an ice/water bath containing a solution of the

unsaturated acetate in ether.

Typically through a solution of 4.26 g (0.0195 mol.)

of (9)-OAc in 50 ml of anhydrous ether containing 0.5 g

(5 mmol.) of suspended copper(I) chloride catalyst was passed

diazomethane by means of the nitrogen carrier gas. After

12 hrs. of continuous treatment with diazomethane the ether

solution was filtered through a Celite pad, dried with an-

hydrous magnesium sulfate, filtered, and the ether removed

under vacuum. The (26)-OAc was used without further puri-

fication. The 1Hnmr spectrum showed the absence of vinyl

protons and was essentially identical to the reported spec-

trum as indicated by the following resonances: 6 5.79 (broad

s, 1; H13 ); 0.8 2.5 (m, 14); 1.99 (s, 3; 02CCH3); 0.4 0.7

(m, 2); and 0.12 (m, 1; H anti ).



Preparation of pentacyclo[6.3.1.13'6.02'7.09'11]tridecan-
13-ol (26)-OH

Crude acetate (26)-OAc (4.16 g, 0.0179 mol.) dissolved

in 125 ml of ether was added to 3.0 g (0.079 mol.) of lithium

aluminum hydride in 150 ml of ether under a nitrogen atmos-

phere.26 After addition was complete, the mixture was heated









to reflux for 30 minutes, cooled to room temperature, and

40 ml of saturated sodium sulfate solution was added. The

ether solution was filtered, dried with anhydrous sodium sul-

fate, filtered, and the ether removed under vacuum to give

3.04 g (89.2%) of (26)-OH.26 This crude product was recrys-

tallized from 30 75 C pet ether to give white needles, m.

p. 121.0 121.5 C (lit. 118.5 120.5 C).26 The Hnmr

contained the following resonances: 6 4.98 (broad s, 1; H13);

2.32 (broad s, 2; H1 and H8); 1.6 2.1 (m, 7); 0.7 1.2

(m, 5); 0.4 0.7 (m, 2); and 0.05 (m, 1; Hl0anti) This

spectrum was identical to that previously reported.



Preparation of pentacyclo[6.3.1.13'6.02'7.09'11]tridec-13-yl
p-bromobenzenesulfonate (26)-OBs

To 0.90 g (4.7 mmol.) of (26)-OH dissolved in 50 ml of

pyridine and cooled in an ice/water bath was added 2.5 g

(9.8 mmol.) of p-bromobenzenesulfonyl chloride and the mix-
44
ture stirred until all solid dissolved. This solution was

maintained at 10 C for 24 hrs. and thenpoured into 300 ml

of ice/water and extracted with 2 x 200 ml of ether. The

combined ether layers were dried with anhydrous potassium

carbonate, filtered, and the ether removed under vacuum. The

(26)-OBs was recrystallized from hexane to give 1.7 g (88%)

of white needles, m. p. 118.3 119.1 C (lit. 118.5 119

0C).26 The 1Hnmr spectrum (CCl4) contained the following

resonances: 6 7.58 (s, 4); 5.59 (broad s, 1; H13) ; 2.1 2.4

(m, 4); 1.6 2.0 (m, 4); 0.4 1.2 (m, 7); and 0.13 (m, 1).








Preparation of anti,endo,exo-tetracyclo[6.2.1.13'6.02'7]-
dodec-9-en-12-yl p-bromobenzenesulfonate (9)-OBs

Alcohol (9)-OH (2.43 g, 0.0138 mol.) was added to 7.1 g

(0.028 mol.) of p-bromobenzenesulfonyl chloride dissolved
44
in 50 ml of pyridine cooled in an ice/water bath. This

solution was maintained at 10 C for 24 hrs. and then poured

into 100 ml of ice/water and extracted with 2 x 200 ml of

ether. The combined ether layers were washed with 100 ml

of 3 N hydrochloric acid, dried with anhydrous potassium

carbonate, filtered, and ether removed under vacuum to give

4.3 g (79%) of (9)-OBs.8a The (9)-OBs was recrystallized

from hexane to give colorless needles, m. p. 108 109 C

(lit. 96 97 C) .a The IHnmr spectrum (CDC13) contained

the following resonances: 6 7.66 (s, 4; aromatic); 6.00

(t, 2; J = 2.0 Hz; H4 and H5); 5.42 (broad s, 1; H 1); 2.81

(m, 2); 2.0 (m, 4); 1.6 1.9 (m, 2); and 0.9 1.5 (m, 4).



Acetolysis of (9)-OBs

In a typical reaction, 4.3 g (0.11 mol.) of (9)-OBs

were dissolved in 200 ml of 0.1 M sodium acetate in acetic

acid containing 1% of acetic anhydride. This mixture was

heated to 90 C for 12 hrs. (greater than 10 x T1/2 for the

rearranged brosylate (15)-OBs), cooled to room temperature,

poured into 300 ml of ice/water, and extracted with 3 x 400 ml

of pentane.8a The pentane layers were combined, washed with

200 ml of 10% sodium carbonate, dried with anhydrous potassium

carbonate, filtered, and the pentane removed under vacuum to








give 2.6 g (110%) of a mixture of acetates which was found

by 1Hnmr to contain ca. 90% of (15)-OAc and 10% of exo-penta-

cyclo[7.2.1.04'l11.05'12.06'10ldodec-2-yl acetate (16)-OAc

videe infra). For glpc analysis see the section on product

studies.

A solution of 1.0 g (4.6 mmol.) of this acetate mixture

in 30 ml of ether was added under a nitrogen atmosphere to

0.4 g (10 mmol.) of lithium aluminum hydride in 30 ml of

ether. This mixture was heated to reflux for one hour,

cooled to room temperature, and 0.4 ml of water, 0.4 ml of

15% sodium hydroxide, and 1.2 ml of water added, respectively.

The ether solution was filtered, dried with anhydrous magne-

sium sulfate, filtered, and the ether removed to give a crude

mixture of the alcohols. This mixture was chromatographed on

SilicAR cc-7 with benzene. The (15)-OH eluted first followed

by (16)-OH. The iHnmr spectrum (CDC 3) of (16)-OH was iden-

tical to the reported spectrum and contained the following

resonances: 6 3.98 (d of d, 1; J = 7.0, 2.7 Hz; H2); 2.0 2.3

(m, 4); 1.88 (d, 1; J = 7.0 Hz); 1.1 1.7 (m, 8); and 0.9

(m, i).17



exo-Pentacyclo[6.3.1.1310.04,12.05,9]dodec-ll-yl acetate
(15)-OAc

A crude mixture of acetates (0.20 g, 0.92 mmol.) from

acetolysis of (9)-OBs was dissolved in 10 ml of glacial

acetic acid containing 1 ml of acetic anhydride and 50 ml of

p-toluenesulfonic acid. This solution was heated to 100 C

for 4 hrs., cooled to room temperature, poured into 100 ml of








ice/water, and extracted with 2 x 200 ml of pentane. The

pentane layers were combined, washed with 100 ml of 10%

sodium carbonate, dried with anhydrous potassium carbonate,

filtered, and the pentane removed under vacuum to yield

(15)-OAc.8a Analysis on capillary column (c) showed the pres-

ence of ca. 0.5% of (16)-OAc in the sample and only minor

amounts'of other impurities containing no vinyl protons.

The infrared spectrum (CC14) contained the following absorp-

tion bands: 2866(m), 1720(s), 1360(m), 1250(s), and 1040
-i
cm The Hnmr spectrum (CDC13) contained the following

resonances: 6 4.95 (broad d, 1; J = 2 Hz; H11); 1.1 2.6

(m, 14); and 1.97 (s, 3; 02CCH3). The mass spectrum (70 eV)

had m/e 218 (M 6.1%), 176 (16.0%), 174 (9.4%), 158 (54.0%),

129 (54.3%), 119 (51.1%), 91 (98.4%), and 43 (100%).

Anal.: Calcd. for C14H1802: C, 77.03; H, 8.31

Found: C, 77.03; H, 8.32.



Preparation of exo-pentacyclo[6.3.1.03'10.04'12.05'9]dodecan-
11-ol (15)-OH

A solution of 2.3 g (0.011 mol.) of (15)-OAc containing

ca. 5% of (16)-OAc in 50 ml of ether was added to 2.0 g (0.053

mol.) of lithium aluminum hydride in 100 ml of ether under a

nitrogen atmosphere. This suspension was heated to reflux

for one hour, cooled in an ice/water bath, and 10 ml of

water slowly added. After the evolution of hydrogen and heat

ceased, 50 ml of 10% sulfuric acid was added. The aqueous

layer was extracted with 2 x 100 ml of ether and the combined

ether layers were washed with 100 ml of 10% sodium hydroxide,








dried with anhydrous potassium carbonate, and filtered. Re-

moval of ether under vacuum gave 1.8 g (97%) of crude (15)-

OH.8a Recrystallization from hexane gave white needles, m. p.

82 83 C (lit. 80.0 81.5 C). 8aGlpc analysis on capillary

column (c) showed less than 2% impurities. The infrared

spectrum (CC14) contained the following absorption bands:

2960(s), 2870(m), 1120(w), and 1070(w) cm-1. The 1Hnmr

spectrum (CDC1 3) contained the following resonances: 6 4.10

(d, 1; J = 2.2 Hz; H11) and 0.9 2.4 (m, 15). The mass

spectrum (70 eV) had m/e 176 (M, 34%), 158 (23%), 119 (100%),

91 (76%), 79 (53%), 67 (42%), and 41 (39%).

Anal.: Calcd. for C12H160: C, 81.77; H, 9.15

Found: C, 81.59; H, 9.28.



Preparation of pentacyclo[6.3.1.03'10.04 12.05'9]dodecan-
11-one (44)

Jones' reagent was prepared by dissolving 26.7 g (0.267

mol.) of chromium trioxide in a solution of 23 ml of sul-

furic acid brought to 100 ml by the addition of water. A

portion of this solution was added dropwise to a solution

of 0.148 g (0.840 mmol.) of (15)-OH in 25 ml of ether until

an orange color persisted. Water (100 ml) was added and

the mixture extracted with 2 x 100 ml of ether. The com-

bined ether layers were washed with 100 ml of 10% sodium

carbonate, dried with anhydrous potassium carbonate, fil-

tered, and the ether removed under vacuum to give the ketone

(44).8a Analysis on capillary column (c) showed this sample

to be homogeneous. The infrared spectrum (CC14) contained









the following absorption bands: 2880(m), 1750(s), 1160(w),

and 1100(w) cm- The 1Hnmr spectrum (CDC1 3) contained the

following resonances: 6 2.78 (m, 1; H9) and 1.1 2.6 (m,

13). The mass spectrum (70 eV) had m/e 174 (M 59.2%),

146 (25.8%), 119 (93.9%), 108 (39.2%), 91 (100%), and 79

(82.7%).

Anal.: Calcd. for C12H140: C, 82.72; H, 8.10

Found: C, 82.65; H, 8.12.



Action of diazomethane on pentacyclo[6.3.1.03'10.04'12.05'9]-
dodecan-ll-one (44)

Method A. To a flask containing 6.0 g (0.15 mol.) of

sodium hydroxide, 15 ml of water, 20 ml of diethylene glycol

monoethyl ether, and 300 ml of ether at 0 C was added 32.5 g

of nitrosan (70% in mineral oil, Aldrich). This mixture was

stirred at 0 C for 15 minutes and then carefully distilled

into a flask cooled in an ice/water bath. This solution was

determined to be. 0.4 M in diazomethane by adding one milli-

liter of the solution to a benzoic acid solution and integrat-

ing the ester and acid peaks in the 1Hnmr. To a solution of

0.174 g (1.00 mmol.) of (44) in 10 ml of anhydrous ether was
18
added 2.5 ml (1.0 mmol.) of the diazomethane solution. This

solution was maintained at 0 C for 1 hr. and at room tempera-

ture for 3 hrs. The only substance isolated was starting

material (glpc).

Method B. An ether solution which was 0.67 M in diazo-

methane was prepared as in Method A. To a solution of 0.174 g








(1.00 mmol.) of (44) and 0.13 ml (1.0 mmol.) of boron tri-

fluoride etherate in 10 ml of ether at 0 C was added 6.0 ml

(4.0 mmol.) of the diazomethane solution. This solution was

stirred at 0 C for 15 minutes, then washed with water, dried

with potassium carbonate, filtered, and the ether removed

under vacuum. Chromatography on silica gel gave a pentane

fraction containing the ketone (44) and an ether fraction

containing two components (glpc) with no carbonyl groups

(infrared).



Preparation of 11-methyl-endo-pentacyclo[6.3.1.03'10.04'12.
05,9]dodecan-ll-ol (79)

To 1.67 g (9.59 mmol.) of ketone (44) dissolved in 25 ml

of anhydrous ether maintained at 0 C was added a 30 ml

sample of 1.0 M methyllithium (0.030 mol.) in ether under a

nitrogen atmosphere. This mixture was stirred at 0 C for

one hour and at room temperature for one hour. The mixture

was poured into 200 ml of ice/water and extracted with 2 x

100 ml of ether. The ether solution was dried with anhydrous

magnesium sulfate, filtered and the ether removed to give

1.22 g (66.9%) of a mixture of two components. Analysis

of the mixture on capillary column (a) showed a major com-

ponent (ca. 90%) at 8.6 minutes and a minor component (ca.

10%) at 9.6 minutes. The second component is probably a

result of the presence of (16)-OH prior to preparation of

the ketone. Separation by preparative glpc using column

(a) gave the major product (79) as a pale yellow solid.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs