.IA 1.3.).-
o- a I vulhzfi.
| I O a ‘ w . l
.5 v I. . y . 102. [Not .
O 4 O I ll”. LA .Ah I A. . I In” vl‘
U ' J\ I I IV \ 7 . fill. )I.‘ Ill " A 0‘ Ilth II
I .9 ‘ yet-ll“! “out... 5. WWII-(PH ‘ {flu-hung il- .A . Q
. .II n \ . m. . v. o v "Hollovt- Lug“ . {‘11:} l ’{Xfl.l. . A: r n. u .c
l 10 1 I ‘ . v..." v. . Ir’.‘ . vi 1, H .
a I i I I I . III-‘OAJI‘IIIO "Tt‘J-Lrlg‘n In 41‘ I a JIIJI-II o oblHkHI'IIO .\_|’
. . o . VLNHHMHV «I: ’u». . . :‘ltllu
\ -

- . 9111!”! Jun»: If“: t
'17}! E ‘ K -- -
1:11.}... riggrcf!f! .... 5.3...

-. .. -

1!}

 

 

   
 
 

HENRY
Mews-'31: Etate
University

WV ‘7 ‘ v v

This is to certify that the

dissertation entitled

"An Approach to the Synthesis of Dodecahedrane"

presented by

Theresa Ann Monego

has been accepted towards fulfillment
of the requirements for

Ph . D. degree in Chm.

W

Major professor

Date September 10L 1982

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

MSU

LIBRARIES
m.

~

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from

 

‘ your record. FINES will

be charged if book is
returned after the date
stamped below.

 

 

 

 

AN APPROACH TO THE SYNTHESIS OF DODECAHEDRANE

BY

Theresa Ann Monego

A DISSERTATION
Submitted to
Michigan State University
in partial filfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1982

ABSTRACT

AN APPROACH TO THE SYNTHESIS OF DODECAHEDRANE

BY
Theresa Ann Monego

A novel synthetic route to dodecahedrane l involving a
12 n electron reorganization in hexaene lg is proposed.
A key intermediate in the synthesis of hexaene lg, gig,
eggng,6-2i§(hydroxymethyl)bicyclo[3.3.0]octa-3,7-diene 4Q,
has been realized. The synthesis of diol éé proceeds in
.3 steps in 34% overall yield from the known gi§,ggdgf2,6-
dihydroxybicyclo[3.3.0]octa-3,7-diene éé' This sequence
employs a highly stereoselective chelation-controlled SNZ'
ring opening of tricyclic ether Zé with "methanol dianion".
Related studies in this area demonstrate that the [1,2]
Wittig rearrangement rather than anionic [2,3] sigmatropic
rearrangement occurs in functionalized bicyclo[3.3.0]octen-
ols. Finally, model studies on the saturated analog of
diol éé have served to elucidate further transformations

required for the synthesis of $3.

To my Mother
and to the Memory of My Father
and to My Husband, Peter,

for Their Love and Understanding

ii

ACKNOWLEDGMENTS

I would like to thank Professor Donald G. Farnum for
his support and guidance during the pursuit of my degree.
His critical insight and chemical knowledge have greatly
enhanced my understanding of organic chemistry. I grate-
fully acknowledge the advice of Professor Eugene LeGoff,
who served as second reader, and I also thank the remainder
of my committee for their interest and suggestions concern-
ing this work. My stay here at Michigan State University
has been rendered immeasureably more pleasant by the com-
panionship of my colleagues, especially those in both past
and present Farnum and LeGoff groups. I cannot express my
'gratitude nearly enough to my husband, Peter, who has made

it all worthwhile.

iii

TABLE OF CONTENTS

Chapter

LIST OF TABLES. . . . . . . . . . . . .
LIST OF FIGURES . . . . . . . . . . . .
INTRODUCTION. . . . . . . . . . . . . .

RESULTS AND DISCUSSION. . . . . . . . .

1. C10

tion Approach . . . . . . . . .

2. Sulfide Contraction/Elimination
Approach. . . . . . . . . . . .

2.1. Model Studies . . . . .

2.2. The Fate of One Bis-Wittig Re-

arrangement . . . . . . .

2.3. Another Wittig Rearrangement.

Lactone (or Diester) Dimeriza-

2.4. The Synthesis of Cis, endo-2,6-

 

bis(hydroxymethyl)bicyc15f3.3.0]

octa-3,7-diene 3Q . . . . . . .

EXPERIMENTAL O C O O C O O O O O O O O 0

General . . . . . . . . . . . . . .

Bi(2-oxa-3-oxotricycloE7.2.1.0
undeca-6,9-dien-4-y1) . Oxidative
Dimerization of Lactone 2g Lithium
Enolate . . . . . . . . . . . . . .

2,6-Bis-methylenebicycloE3.3.0]-
octane g2 . . . . . . . . . . . . .

gi§,endo-2,6-bis(hydroxymethyl)-
bicyEloE3.3.oTBEtane g . Hydro-
boration of Diene g2 w1th Disiamyl-
borane. . . . . . . . . . . . . . .

iv

5'11]-

Page

. viii

ix

16

16

20

30

43

58

80
88
88

89

91

93

Chapter

Hydroboration of Diene 6% with
BB 3 - THF COmp 1 ex 0 o o o o o o o o o o o o

Cis,endo-2,6-bis(hydroxymethyl)-
bicycloi3.3.0]octane Dimethane-
sulfonate 6%. . . . . . . . . . . . . . .

Cis,endo-2,6-bis(iodomethyl)bi-
cycloi3.3.0]octane, éé. . . . . . . . . .

Cis,endo-Z,6-bi§(bromomethyl)bi-
cyclol3.3.0]octane 69. Reaction

of Diol with Carbon Tetrabromide/
Tripheny phosphine. . . . . . . . . . . .

Cis,endo—2,6-bis(amidinothio)methyl-
bicyclol3.3.0loctane Dibromide
The Addition of Thiourea to Dibromide

g3. . . . . . . . . . . . . . . . . . . .

Cis,endo-2,6-bis(mercaptomethyl)bi-
cyclo[3.3.0]octane 88' Hydrolysis of
Bis-thiouronium Bromlde Q2. . . . . . . .

 

The Reaction of Dithiol 66 with Di-
bromide 62. . . . . . . . . . . . . . . .

Cerous Chloride Mediated Sodium Boro-
hydride Reduction of Bicyclo[3.3.0]octa-
3,7-diene-2,6-dione zg. Cis,endo-2,6-
dihydroxybicyclo[3.3. Jocta-3,7-diene éZ'

Temperature Effects in the Sodium
Borohydride/Cerric Chloride Reduction
of Dienedione 1g. . . . . . . . . . . . .

Iodomethyltri-n-butyltin. . . . . . . . .

Cis,endo-2,6-bis[(tri-n-butylstannyl)-
methoxyIbicyclol3.3.0]octa-3,7-diene SQ .

2-0xa-tricyclo[5.2.l.o4'lojnona-5,8-
diene i6. Reaction of 38 with Two
Equiva ents of neButylllthium . . . . . .

Low Temperature Quenching of Dianion
i Derived from gig-stannane and n-
Butyllithium. Cis,endo-2,6- 1methoxy-
bicyclo[3.3.0]octa-3,7-diene 2g . . . .

Page

94

96

96

97

99

100

101

102

103

104

104

105

107

Chapter Page

2-Oxa-tricyclc[5.2.l.04’lojnona-

5,8-diene, Z . Reaction of 8

with One Equlvalent of n-ButyI-

lithium . . . . . . . . . . . . . . . . . . . . 108

Cis,endo-Z- hydroxy-6- [(tri- n-butyl-
stannyl)methoxy]bicyc10[3. 3. 0]octa-
3, 7- diene 87. . . . . . . . . . . . . . . . 109

 

n-Butyllithium Exchange with Cis,

endo-2- -hydroxy-6- (tri- n-butylstanny1)-
methoxybicyclo[3. 3. 0]octa—3, 7- diene Z.

Cis,endo-Z- hydroxy- 6- -hydroxymethy1bi-

cyclol3. 3. 0]octa-3, 7- diene g, and 2-
Endo-hydroxymethyl- 6- -buty1b1cyclo[3. 3. 0]-

octa-3, 7- diene 82. . . . . . . . . . . . . . 110

n-Butyllithium Cleavage of Tricyclic
Ether é. Endc-Z- (hydroxymethyl)- -6- -n-
butylblcycloIB. 3. 0]octa-3, 7- diene 82. . . . . 112

Sodium Borodeuteride/Cerous Chloride

Reduction of Bicyclo[3. 3. 0]octa-3, 7- diene-

2, 6- dione . Cis, endo- 2, 6- -dihydroxy-

bicycloEB. .O]octa-3, 7- diene-2 d2 12. . . 113

Alkylation of[2, 6- dJDiol 17 with One Equi-
valent of Iodomethy tri-n-butyltin. . . . . . . 114

2-Oxa-tricyclo[5.2.1.04’10]deca-5,8-
diene-1,6-Q2. . . . . . . . . . . . . . . . . . 115
Cis, endo-Z- -hydroxy-6-(hydroxy-
methyl)bicyclo[L 3. 0]octa-3, 7- diene-

g§ and endo-2- (hydroxymethyl)-
6- :n-baty bicycIo I3. 3. O]octa-3, 7- diene-

812181 1.............15

Cis, endo- 2, 6- -bis(hydroxymethy1)-

bicycloIL 3. OISEta-B, 7-diene 6.

Nucleophilic Ring Opening of ri-

cyclic Ether 16 with Methanol Di-

anion . . . . . . . . . . . . . . . . . . . . . 116

Cis, endo- 2, 6- -bis(acetoxymethy1)-
bicyclol3. 3. 0|octa-3, 7- diene 112. . . . . . 119

SUMMARY . . . . . . . . . . . . . . . . . . . . . . 121

vi

Chapter Page

APPENDIX. . . . . . . . . . . . . . . . . . . . . . 124

REFERENCES. . . . . . . . . . . . . . . . . . . . . 153

vii

LIST OF TABLES

Table Page
1 Predicted Energies for Various
Dodecahedrane Inclusion Compounds . . . . 4
2 1H NMR Chemical Shifts and Multi-
plicites in Diol fig . . . . . . . . . . . 61
3 Coupling Constants in p—Nitro-
benzoate 2Q (Hz). . . . . . . . . . . . . 62
4 Temperature Dependence of [1,2]

25. [2,3] Rearrangements of 3% and
gé. . . . . . . . . . . . . . . . . . . . 70

13

5 Comparison of C Chemical Shifts

of Bridgehead Carbons . . . . . . . . . . 86

viii

Figure

LIST OF FIGURES

The tetrahedron, hexahedron, octa-

hedron, dodecahedron and icosahedron.

Platonic solids or derivatives

which have been synthesized: tetra-

E-butyltetrahedrane, tetralithio-
tetrahedrane, cubane and dodeca-
hedrane . . . . . . . . . . . . .
Proposed routes to hexaene 1% . .
Summary of sulfide to olefin re-
actions . . . . . . . . . . . . .
Proposed route to disulfide fig. .
1H NMR spectrum (250 MHz) of tri-
cyclic ether Zé (top) and its
1,6-_d_2 analog, 1% (bottom). . . .

1H NMR (250 MHz) of diol gg (top)

and its 2,6—d2 analog, 2% (bottom). . .

Examples of anionic [2,3] sigma-
tropic rearrangements . . . . . .
1H NMR spectrum (250 MHz) of

butylated product g2 (top) and

ix

Page

. 1

18

49

65

69

analog, 191 (bottom) . . . . . 82

Figure Page

10 Summary of the sulfide contrac-
tion route to hexaene 1%. . . . . . . . . 122

Al Proton NMR spectrum (CDCl 60 MHz)

5,11]_

3!
of i(2-oxa-3-oxotricycloE7.2.l.0
undeca-6,9-dien-4-yl 2x . . . . . . . . . 126

A2 Proton NMR spectrum (CDC13, 250
MHz) of 2,6-bis-methylenebicyclo-

[3.3.0]octane 6% . . . . . . . . . . . . . 128

A3 Proton NMR spectrum (CDCl 250 MHz)

3!
of cis,endo-2,6—bis(hydroxymethyl)-
bicyclo[3.3.0]octane 63 . . . . . . . . . 129

A4 Proton NMR spectrum (CDCl 250

3,
MHz) of gi§,e§gg-2,6-bi§(bromo-
methyl)bicyclo[3.3.0]octane 62. . . . . . 132

A5 Proton NMR spectrum (CDC13, 250 MHz)
of gi§,§2dgf2,6-bi§(mercaptomethyl)-

bicyclo[3.3.0]octane 66 . . . . . . . . . 134

A6 Proton NMR spectrum (CDCl 250 MHz)

3!
of the major product (93. 30%) in the
coupling reaction of dibromide 62 and
dithiol 66. . . . . . . . . . . . . . . . 136

A7 Proton NMR spectrum (CDCl 60 MHz)

3!
of cis,endo-2,6-bis[(tri-n-butyl-
stannyl)methoxy]bicyclo[3.3.0]octa—3,7-

diene ég. . . . . . . . . . . . . . . . . 138

Figure

A8

A9

A10

A11

A12

A13

A14

Page

Proton NMR spectrum (CDC13, 60 MHz)

of 2-oxa-tricyclo[5.2.1.04’10

Jnona-
5,8-diene Zé. . . . . . . . . . . . . . . 140

Proton NMR spectrum (CDC13, 250 MHz)

4'lojnona-

of 2-oxa-tricyclo[5.2.l.0
5,8-diene] Zé o o o o o o o o o o o o o o 142

Proton NMR spectrum (CDCl 60 MHz)

3,
of Z-ggggfhydroxymethyl-6-butylbi-
cyclo[3.3.0]octa-3,7-diene 82 . . . . . . 144
Proton NMR spectrum (CDC13, 60 MHz)

of gi§,gndg-2-hydroxy-6-(tri-n—butyl-
stanny1)methoxybicyclo[3.3.0]octa-

3,7-diene 8;. . . . . . . . . . . . . . . 146
Proton NMR spectrum (CDC13, 250 MHz)

of gi§,gngg-2-hydroxy-6-hydroxy-
methylbicyclo[3.3.0]octa—3,7-

diene 88. . . . . . . . . . . . . . . . . 148
Proton NMR spectrum (CDC13, 250 MHz)

of gi§,gndg-2,6-bi§(hydroxymethyl)-
bicyclo[3.3.0]octa-3,7—diene £6 . . . . . 150

Proton NMR spectrum (CDCl 250 MHz)

3'
of cis,endo-2,6-bis(acetoxymethyl)-

bicyclo[3.3.0]octa-3,7-diene 11%. . . . . 152

xi

INTRODUCTION

The Platonic solids comprise a set of five regular
polyhedra: the tetrahedron, hexahedron, octahedron, dodeca-
hedron and icosahedron shown in Figure 1. Of these, the
structural framework of three can be translated to a molecu-
lar framework of carbon-carbon bonds without demanding a
radical alteration in the directionality of sp3 hybridized

bonding, or introducing pentavalent carbons.

 

 

 

 

 

Figure 1. The tetrahedron, hexahedron, octahedron, dodeca-
hedron and icosahedron.

The synethsis of organic equivalents of the tetrahedron
(tetrahedrane), hexahedron (cubane), and dodecahedron (do-

decahedrane) has been realized; the first as the tetra-E7

l 2

butyl (Maier, 1978) and tetralithio (Schleyer 1978)

derivatives, the latter two as the parent hydrocarbonsB’4

(Eaton, 1966 and Paquette, 1982, respectively) (Figure 2).

 

U
-—r—
_,___ LI ---— LI

 

 

Figure 2. Platonic solids or derivatives which have been
synthesized: tetra-Efbutyltetrahedrane, tetra-
lithiotetrahedrane, cubane and dodecahedrane.

The very recent achievement of dodecahedrane perhaps

reflects the degree of difficulty encountered in its

5

synthesis. As a member of the Ih point group, it has 120

symmetry operations; each of twenty methine units is identi-
cal to the remainder. The high level of symmetry in this

4'6 and

molecule results in some interesting properties
predictions: it shows no sign of melting up to tempera-
tures of 450°C (for comparison, eicosane-C20H42- has a melt-

1 13

ing point of 36-38°C); the H and C NMR spectra each

consists of one singlet; and calculations indicate very
little strain energy.5'7

One of the most fascinating features of this hydro-
carbon ball is the significant cavity "inside" the carbon
framework. Assuming each carbon atom is a point, the in-
ternal cavity along one of the 3-fold axes measures 4.32 A.
As literal examples of cage compounds, various encapsulated
species have been considered theoretically.8 Molecular
orbital calculations predict the energies of these "en-
'capsulanes" range from a net destabilization of 48 kcal/mol
for a trapped hydrogen molecule, to a net stabilization of
519 kcal/mol for a trapped berylium atom (Table 1). These
calculations indicate considerable charge transfer between
the trapped species and dodecahedrane sigma bonds in the
case of Li+, Be and Na+.

Such trapped species would have interesting physical
properties and potentially valuable uses. The inclusion
compound would simulate an element or iron in the vapor
phase. It would represent an organic soluble metal cation.
The benefits of a caged radioactive cation in medical re-

search range from chemotherapy to organ imaging.

 

Table 1. Predicted Energies for Various Dodecahedrane
Inclusion Compounds.a

 

 

 

X AE kcal/molQ
e‘ +149
H+ -130
H° - 30
H" - 52
He + 22
Li+ -162, -1833
Be -519
Na+ +420.3
H2 + 48

 

 

aTaken from Reference 8.
b«X at dodecahedrane midpoint 0 relative to X at infinite
separation.

CFrom CNDO and INDO closed shell calculations, respectively.

It is the goal of the work described herein to explore
a novel synthetic route to dodecahedrane and metal en-
capculated dodecahedranes.

If dodecahedrane 1 resisted synthesis for so long,
it was not from lack of effort.5 The main avenues of syn-
thetic design are readily classified as either convergent
or linear sequences.

The synthesis of triquinacene 29 in 1964 led to efforts

to effect its dimerization to dodecahedrane. Attempts along

/\ @
$4 ”"- $537

4

these lines have failed however; for example, photolysis
of 2 effects only polymerization,10 while heating 2 at
400°C for two hours under 40,000 atmospheres of pressure
resulted in recovery of starting material.11
Connecting two triquinacenes greatly enhances the pos-
sibility of isomerization to dodecahedrane by removing

translational and rotational degrees of freedom from the

monomer. However, the cyclization of bivalvanes like 3

 

3H

have not proved successful.5’12

Alternatively, the synthesis of another triquinacene
dimer, 6, was attempted in the hope that homolytic scis-
'sion of one of the cyclobutane bonds would result in a
13

diradical which would then cyclize to dodecahedrane.

Unfortunately, the target molecule 6 was never realized.

 

A doubly linked triquinacene dimer 6 has recently been
14

synthesized. Efforts are now underway to transform 6 to

 

symmetrically substituted dodecahedranes 1.
Linear approaches to dodecahedrane have been pursued
by the Eaton and Paquette groups.

Eaton began with bicyclo[3.3.0]octanedione g and

through successive cyclopentane annulations gained entry

to the "peristylane" series, 2.15 Attempts to cap

<01? @7

Q 2

r@

COOMe

COO Me

 

LOOMe
CKKMHe

Us

well-functionalized peristylanes are in progress.

The only successful synthesis of dodecahedrane to date
uses the domino Diels-Alder reaction between 9,10-dihydro-
fulvalene 1Q and dimethyl acetylene dicarboxylate to gen-
'erate pentacyclic adduct 11, which is then elegantly trans-

formed into C16-hexaquinacene 12.16 Variations on the

A,

   

Cl6-hexaquinacene approach eventually led to the synthesis

4

of secododecahedrane 13. Dehydrogenation led to the

 

long-sought-after dodecahedrane.

The penultimate molecule in our approach to dodeca-
hedrane is hexaene 14. There are two distinct conforma-
tions 14 can adopt. One is a compact conformation with

'the two bicyclooctadienes directly facing each other (14a),

 

10

whereas the extended conformation (14%) has one bicyclo-
octadiene twisted 90° with respect to the other. Both
conformations have D2 symmetry, however 14a is geometrically
similar to dodecahedrane, disregarding distortions due to
differences in C-C single bond and double bond lengths and
angles. Hexaene 14a is in fact a hexakissecododecahedra-
hexaene. By a series of electron reorganizations, 14a can

isomerize to dodecahedrane 1. While this type of bond

~L

 

reorganization might have some of the quality of Equation 10
in Reference 5, there are some intriguing points well worth
considering.

The dodecahedrane precursors in this synthesis would
all have an element of symmetry, which may result in a
certain economy of steps in attaining the product. The
hexaene 14 itself could coordinate a small metal ion,
thus trapping it inside the dodecahedrane cavity upon cycli-

zation.

11

The final cyclization is a 12n electron reorganization,

thus a concerted [n23 + 1T2S + n23 + 1T2s + Tr2S + 1T28] hexa-
merization requires photolysis, in accord with orbital sym-
metry rules.17 On the other hand, triplet sensitization
may promote cyclization gig a diyl intermediate.

The hexamerization of six nonconjugated olefins could
conceivably be accomplished in the presence of a transition
metal catalyst. Studies on the copper(I) photo-assisted di-

merization of norbornene lg suggest that a 1:2 c0pper-olefin

complex is the direct precursor of the dimer, 16.18 (Scheme_l)

 

f \

ems

Scheme 1

12

It has been demonstrated that unstrained olefins such as
cyclopentene and cyclohexene undergo c0pper(I) promoted di-

19

merizations. Recently, several natural products have

been synthesized in which the key step was copper(I) catalyzed

photobicyclization.20

Thus the 1,6-diene 17 afforded 18
’b’b ’b
upon photolysis, which was then transformed to a-pana-

sinsene 19 and B-panasinsene 20.

II ” .-
$2 22

A copper-olefin complex in hexaene 14 could stabilize
conformation a which upon irradiation could cyclize to
dodecahedrane. Molecular models indicate the two trans
double bonds in 14% are close enough to dimerize to a cyclo-
butane 21. Interestingly, an olefin methathesis reaction of
these two double bonds would only serve to alter the con-

formation of hexaene 14! This is due to the equivalence

13

 

of the four cyclobutane carbons; the metathesis, however, is
not degenerate.

Finally, cationic or free radical induced cyclization
of 14% could lead to dodecahedrane l, or a seggfderivative,

22.
mm

 

14

The most serious drawback to the proposed cycliza-

tion of 14 to dodecahedrane concerns the dependence of

 

cyclization on only one of several conformational isomers.
Molecular models indicate the olefin protons of the trans
double bond in 14a are proximate to the ring olefin; simple
twisting of the trans double bond in this conformation does
not relieve this interaction. In addition, the starred
carbon-carbon bond in 14a exists in a sort of s—gis
arrangement with respect to the trans olefin, as opposed

to g-Egans in 14b. The conformation 14% is simply a

combination of 14a and 14b. Conformation 14a, however,

15

is ideally suited for a metal-olefin "internal" complex,
and may prove to be the more reactive conformation. It
is also possible that rotation about carbon-carbon bonds

7:

in the activated complex 14 would serve to establish the

conformation required for dodecahedrane.

RESULTS AND DISCUSSION

1. £10 Lactone (or Diester) Dimerization Approach

Dissection of hexaene 14 shows it to be two bicyclo-
octadienes linked by a trans olefin. The immediate goal

of this project was then synthesis of these two halves,

 

 

and dimerization to hexaene 14. Each half must be of the
same chirality to result in 14; two halves of opposite
chirality would lead to the isomeric hexaene 23, a megg
form which cannot be converted to dodecahedrane. To achieve
maximum efficiency then, a resolution should be performed
somewhere in the synthesis prior to the dimerization
step.

The initial approach to hexaene 14 centered on the

c0pper(II) oxidative dimerization of diester 24 di-

21

lithium enolate, or of lactone 26 lithium enolate

16

17

[‘15 1‘45
659 " m

R

(Figure 3). In the first instance, the basic carbon skele-
ton of 66 is generated in one step. In the second case,
each of the carbon bridges in 16 would be formed sequen-
tially.

The results in this area were discouraging, however.
Both diester 26 and lactone 26 were difficult to obtain
(66: 9 steps from commercially available material, 6.3%
overall yield; 26: 9 steps, 2.5% overall yield);22 more-
over, 26 required a lengthy chromatography to effect
separation from numerous by-products accompanying its
formation.23

Addition of diester 26 to two equivalents of LDA, fol-
lowed by addition of cupric triflate gave a complex mixture
of products. Analysis by GLC/MS showed the two highest
molecular weight components (besides recovered diester 26)
had parent ions at m/e 276, strongly suggesting an intra-

molecular reaction of the bis-enolate resulting in diesters

18

 

cog: a
2 sq IDA A on
TH?» '78"? Lia
no; no): : Cu(II)
I
a: w
: 31130.: H’.) g. '5.
\- 2) 1>t>(om:),,\~£
A: V - a:

 

0 uo’. H“
1) IDA A a ______ -9
o 2) Cu(II) . 2

O

 

£8 31

1) IDA
.4, <--—---—- 1‘
2) Cu(II) no; __

3) Ho': 11" V
a) Pb(0Ac)u CH3C(or:t)3

 

Figure 3. Proposed routes to hexaene 16.

19

61. This type of product is not unreasonable, given the

:==§flfi

s on 1
63(11)
A
/ 0,5:
IJO

é}

 

V

proximity of the enolate anions and the entropy factor
operating in favor of intramolecular reaction.

Addition of lactone 26 to a solution of LDA, followed
by addition of cupric triflate yielded the dimer, 61, in

addition to recovered starting material. 1

H NMR, IR, and
MS spectra all supported the assignment of structure 21 to
this compound. The isolated yield of crystalline 21 is
only 23- 15%, however. In reactions using cupric bromide
or cupric chloride, the d-bromolactone 6% or a-chloro-
lactone 66 was observed to the extent of 36% (isolated
yield) and 14% (1H NMR yield), respectively. The use of

cupric triflate24 for the enolate dimerization effectively

CG

1%" o x

ggx c1 0

suppressed this side reaction.

Br

20

2. Sulfide Contraction/Elimination Approach

 

Hexaene 66 can also be approached from bis-sulfide 66.

A number of rearrangements may serve to transform the sul-

5

fide linkage to an olefin (see Figure 4).2 Fortuitously

: ----- -> —
w \e
84. $33,

in this molecule, the rearrangement-elimination sequence

is forced to proceed to a trans olefin, as molecular

models indicate that a trans double bond is the only olefin
geometry 16 can adopt. A gig olefin at this juncture would
be impossibly sterically crowded.

The approach to 16 described herein has been used suc-
cessfully in the synthesis of various cyclophanes. The
cyclodimerization is achieved by reaction of dithiolate
66 with dibromide 66 under conditions of high dilutions.
Ring contraction of 61 gig a Stevens rearrangement, fol-
lowed by Hofmann elimination of dimethylsulfide yields
the cyclophane 69.26

This sequence has several advantages over a hypothetical

21

1) BuLi

2) CH3 I

(Ref. 327) .l
1) r(on 30)zcn+ BF“-

2) t- BuOK

 

(Ref. 27)
(Ref. 25)
5"“ SPh
5°"! 5 Ph
1) [OJ
1) (CH30)20H+ BFu- 2) 11
2) NaH (Ref. 28)
(Ref. 25)
C§§uzf>
14m .3' mm
f' 2) BuLis arZ' r' -soz r"

(Rot. 29)

Figure 4. Summary of sulfide to olefin reactions.

22

(If)

2) g-nuox

 

sea,

1) lo} 0+ n?“-
2) run

SCH,
R 38

NM

 

one step dimerization (2.3;, a Wittig reaction between ap-
propriate components). The two subunits are brought to-
gether gradually with three atoms separating the bulk of
each molecule from the other, thus minimizing steric ef-
fects in the crucial dimerization step. The starting
materials are usually obtained from the same source, since
the dibromide is easily converted to the dithiol. Finally,
sulfur extrusion may be accomplished in a number of ways
(Figure 4).

Recently, while this work was in progress, dithia-

30

hexahydro[3.3]paracyclophane, 66, and the two diastereomers

23

of dithia[3.3](2,6)triquinacenophane, 61 and 66,14 were

synthesized by this route.

In addition, there has recently appeared a publication
which details an improved synthesis of sulfide macrocycles,
based on the reactions of cesium dithiolates with dibro-

mides.31

The improved yields over literature values in
some reactions were truely spectacular, as in the synthesis
of 1,7—dithiacyclododecane, 66, prepared in 80% yield as

compared to 0.8% previously reported.32

CSCO3
BrCH2(CH2)nCH2Br + HSCH2(CH2)mCHZSH ————)

WSW
(cm). (cm)...

L.) n-

4%

The obvious route to bis-sulfide 66 is by reaction of

:3
ll
m
E
II
m

10

ll
m
E

II

dithiol 66 with dibromide 66. The dithiol can be derived

24

from dibromide 66, while it was anticipated that diol 66
would serve as a convenient precursor to the dibromide.
As all of these compounds were unknown, attention was
focused on the synthesis of diol 66. Most notably, 66 is

a homologated version of diol 61, preserving the cis, endo

H0) H5»
BK} \\
‘\
OH 8' SH

3, 4,2 :4,

configuration at C-2 and C-6. The utilization of diol 61
to introduce gig-endo alkyl substituents via a [3,3] oxy—
Cope rearrangement (Claisen rearrangement) had proved suc-
cessful in other work in this laboratory (Scheme 2).23’33
Consideration of various methods to degrade products from
these Claisen rearrangements led to the unsavory prospect
of multistep degradative sequences34 in which some of the
intermediates (E;E;r d-silyloxyaldehyde 66) or products
(229;! dialdehyde 66) might be unstable.

The most promising route to 66 was homologation of
diol 61 via a bi§[2,3] sigmatropic rearrangement (Equa-
tion 1). In this way, a side chain with the correct

number of carbons (one) would be introduced, thus

25

 

 

 

cog:
.0" /
mama), A
H+ ’
[3.31
HO
6?. 5:0,c/
8%
cuz—cuzoczas
115mm);
1 OH
\L p ‘f ‘ '
*1 «w.» A  '
I [3.3] ’
o HOJ
BtzA101°PPh3
[3.31
/CH0

8'

”\

OH

Scheme 2

26

Me,SiO
g CHO EHO
Mask) i die
\CHo
51

alleviating the need for degradation.

The Wittig [2.3]
sigmatropic rearrangement of allyl ethers appeared ideal

Y=sx
[231
be

CO

x “0!“

‘—Y
for the purpose at hand.

While the classical reaction in-
volves treating an allyl benzyl ether with g-butyllithium

35

"W

(Equation 2), Still has reported a [2,3] rearrangement

:n and

Ph
2) 320

| (2)
‘OH

27

in a-alkoxyorganolithium reagents derived from allylic

alcohols. Generation of the carbanion exploits the ready

exchange of tin for lithium in organostannane 61.36

 

\ 1):!!! A \

2) masncazz I’

 

 

0" 66 OVSn Bu,
Bum, -78 C
TH?
(Ni ‘V
1) -78 C
/ ‘1___ 2) 11+ \ + BuuSn
3% con, Li

[2,3] Rearrangement of the a-alkoxy anion leads to homo-
allylic alcohol 66 in good yield. This methodology was
used in a highly stereoselective synthesis of the C13

CecrOpia juvenile hormone, 66.37

Although this rearrangement had been demonstrated to
work well in some cyclic systems, octalols 66 and 61 were

reported to give a mixture of [1,2] and [2,3]rearranged

28

\ 1) m __\
at: 2) m35ncazx ’

‘OH (OH

 

 

\
occ
Bum/MgA
-20 cfi
Bufihl lhhsn 79’
HO) EN)
\ \ \
E 6
55
wm
. . 36
products in low yield.
93‘
= o-OH

 

B-OH

29

The application of this rearrangement in cyclopentenol
systems appears unexplored, however. Indeed, this aspect
of anionic [2,3] sigmatropic rearrangements is not cited
in the literature at all. Thus an Opportunity presented
itself for evaluation of endocyclic anionic [2,3] sigma-
tropic rearrangements in a cyclopentene ring system
(Scheme 3).

The remainder of this thesis pursues this idea, as well
as investigating reactions in model compounds which would

transform diol 66 to sulfide dimer 16. All of the following‘

0". e0\\

_ _ _ _ _ 9 3035,, SnBu,

Ho’ of
.972 I
I
I
/°" $

7; FOR,“
-32 33:2]- ., _
g 2)H+
H02 ”CH5

Scheme 3

30

reactions were performed on racemic material.

2.1. Model Studies

 

To demonstrate the feasibility of the synthetic approach
to hexaene 1 and to work out reaction conditions for some
of the required conversions, the synthesis of diene 66 was
undertaken. The immediate goal was sulfide dimer 66 which
would serve as the precursor to diene 66. The starting

material in this synthesis is bicyclooctanedione 61, which

is available in hundred gram quantities in three steps from

38 The route devised to convert dione

dimethylglutarate.
61 to sulfide dimer 66 is straightforward, and is outlined
in Figure 5.

Methylenation of the dione was anticipated to be the
most troublesome step, given the well-known propensity of
cyclopentanones towards enolization in the presence of

base, leading to self-condensation reactions. The methoxy-

methylenation of cyclOpentanone gave 14.5% of the desired

31

IMjF-GQZ
-------->
0 cu,

3% .62
I
: 1’ 32%
. 2) H202. hon
I
W
pm Samoa
1) I801
2)N¢I
<— ------- (I)
Icu, HOCH,
85. 8%
I
I
I
-33 I
I
I
l
w
pmsu
:
—————— S
+ g; 9 <<::§E§;7
HSCH, 66
'V'v

Figure 5. Proposed route to disulfide 66.

32

vinyl ether, along with a 33% conversion to 2-cyclopentyli-

dene cyclopentanone when the ylide was prepared in the

 

o I ocu, o
O W“ A O + C55
I
14.5% 33%
39

traditional way with gfbutyllithium in THF.
Several improvements or modifications of the original
procedure have been devised to deal specifically with

easily enolizable and/or hindered ketones.40

It appears
that the choice of a base and a solvent for generation of
the ylide component greatly affects the success of the re-
action. For example, sodium i-amylate deprotonation of
methyltriphenylphosphonium bromide in xylene gives the
phosphorane, which reacts with cyclopentanone to yield

81% methylenecyclopentane.38a

The presence of teamyl-
alcohol in the reaction is thought to render the enoliza-
tion process reversible. Several intermediates in the
synthesis of triquinane natural products such as mode-
hephene 6141 and isocomene 6642 were successfully prepared
in this way.

A solution of sodium ifamylate in toluene was prepared,

and used to generate methylene triphenylphosphorane from

methyl triphenylphosphonium bromide. Addition of the

33

m. 44-) .
.

diketone in toluene and refluxing overnight, followed by

 

  

the usual workup led to diene 61, readily identified by
the four proton AB quartet at 64.84 in the 250 MHz 1H
NMR, and a sharp absorption at 1650 cm-1 in the IR. This
particular preparation gave a low isolated yield of the
diene (g3. 20%) and was contaminated by ifamyl alcohol.

It was found that potassium i—butoxide in benzene gave

map-CH2
lhfl.ru£uu ’
. 81-8
0 9f

 

34

improved results, although deprotonation of the phosphon-
ium salt proceeded at a slower rate owing to the insolu-
bility of the base in this solvent. The isolated yield of
diene (bp 170°C) improved considerably when it was dis-
covered that it codistilled with the aromatic solvents
under reduced pressure, but not at atmospheric pressure.

A substantial excess of ylide (93. 1.7 equivalents per
carbonyl) gave the best results. With these modifications
pure diene 61 could be prepared on a ten gram scale in
81-89% yield.

It was anticipated that hydroboration/oxidation would

 

transform diene 61 to the cis, endo diol 66.43 Treatment

of 61 with two moles BH -THF complex followed by alkaline

3
hydrogen peroxide gave diol 66 in quantitative yield.
Analysis by GLC showed the crude product was a mixture

of stereoisomers. The major isomer had the longer reten-

tion time on a QF-l column, and predominated to the extent

.CHzO CH OH

1) mtg-m
*2] H202. N10)! 7

 

008

HOCH,

I ”I;
N

HOC
8%

of 6:1 over the minor isomer when the reaction was run at
-22°C. Lowering the reaction temperature to -45°C increased
this ratio to 9.5:1, or 90% one isomer. Thus, the major

isomer was assigned the cis, endo stereochemistry while

 

35

the minor product was presumed to be the trans isomer.
Acylation of the 6:1 diol isomer mixture followed by
chromotography gave a product which displayed a small

acetate methyl at 62.06 in the 60 MHz 1

H NMR, and a larger
one at 62.03 in approximately an 11:1 ratio, or a 6:1

ratio of cis:trans. Considering that the ultimate coupling

 

reaction required gig, endo material and that both the
dibromide and dithiol were to be derived from diol éé'

even 10% of the trans isomer would limit the cyclized
sulfide dimer to a maximum of 80% yield. Therefore it
would be advantageous to increase the exgzendg selectivity
in the hydroboration step. The 9-BBN°THF complex has been
used as a highly hindered hydroborating reagent,44 but
removal of the oxidation product of 9-BBN, 1,5-cyclooctane-
diol, would require a chromatography. Finally, the use of

disiamylborane4S (Sia BH) resulted in virtually exclusive

2
exo attack on diene gg (>98% selectivity by GLC). The
amyl alcohol oxidation by-product (bp 111°C) could be

conveniently removed under high vacuum at room tempera-

ture. When this was done, the diol was pure enough to

9H,OH

1) sugar: m
:;
2) 11202. man

1001 ~
3% HOCH, '93,

"'I,

 

"a,

36

use directly in the next step, but it could also be
kugelrohr distilled to provide an analytical sample.
Treatment of this diol with mesyl chloride in pyridine
gave a 95% yield of the dimesylate g3, which slowly
solidified on standing. Exposure of this material to
sodium iodide in refluxing acetone provided diiodide Qé in

63-70% yield after chromatography.

 

9H,0H ‘ ngMs
. I301
4;
m pyridine I
95$ .
HOCH2 "5°C": 6
63
M»
N11
19"" acetone
= 63-70!

"I.

a
.s=~

Alternatively, a one step conversion of diol g; to the

dibromide fig could be accomplished with triphenylphoSphine/

46

carbon tetrabromide. When a large scale (22' 10 9)

reaction was conducted in ether, the amount of precipitated

37

triphenylphosphine oxide was sufficient to interfere with
the mechanical stirring. Removal of triphenylphosphine
oxide gave a yellow oil which contained large quantities
of bromoform and carbon tetrabromide. The former could

be removed under high vacuum at ambient temperature (bp
146-150°C). The latter contaminant was clearly visible in

the 13

C NMR of the crude product at -29.52 ppm, initially
appearing as a fold-over at 12.52 ppm. Removal could only
be effected by chromatography, at which point pure di-

bromide 57 was obtained in 55% yield. Changing the solvent.

SCH 20H SCH 28'

”If/031'“ A
man or c1130): 7

now, 55‘7““ area,

 

0%

to acetonitrile resulted in a slightly improved yield, 74%
on a small scale, 61% on a larger (2 g) scale.

In any event, either the diiodide or dibromide could be
transformed to dithiol fig any number of ways. Direct
treatment with sodium sulfide or sodium hydrogen sulfide
was not attempted, in anticipation of unwanted intramolecu-
lar displacements yielding sulfide ég. Alternatively,

treatment of either the dibromide or diiodide with sodium

38

 

thiolacetate followed by lithium aluminum hydride reduction,
or with thiourea followed by hydrolysis,47 would effec-
tively circumvent sulfide formation. The latter classical
method was used due to the availability of thiourea in

the lab. Heating a 0.30 M solution of dibromide $9 in 95%
ethanol with thiourea in a closed tube at 100°C led to the
crystallization of bis-thiouronium bromide 11 from the
reaction mixture. Chilling the solution and filtration

of the solid gave a 95% yield of the salt. Alkaline

 

Bre 3H2
9H,Br $2st—
3 .« m.
“2" N“2 2
non :\’
f 5‘ NH,\ 5
. 9 :
BrCH, c—s CH,
69 / 7
NH, Bra «A1.

hydrolysis was achieved by refluxing the salt in aqueous

sodium hydroxide, followed by acidification of the

39

hydrolysate. In this respect, it is desirable not to
include unreacted thiourea in the hydrolysis, due to the
distinctive smell of hydrogen sulfide liberated during
acidification. The crude dithiol was isolated as a clear,
viscous oil after flash chromatography, possessed of the

characteristic thiol smell. Protection from light and

"'00: n
I:
"in
I:

21% 1) Nam-I, reflux%
2)H+

 

f
HSCH,

oxygen was essential to inhibit the formation of disulfides;
even storage under nitrogen in the refrigerator was not
sufficient to completely preclude this process.

The preceeding steps are summarized in Scheme 3.

Initial studies on the reaction of dibromide £9 with
dithiol QQ followed the general procedures used in the
synthesis of dithiacyclophanes, i;g;, slow addition of
the dihalide and dithiol to a dilute solution of base
over a period of several days in order to achieve high

dilution conditions.14'26’30

The high yield synthesis of
some sulfur containing macrocycles via cesium thiolates

appeared a promising route to the desired sulfide dimer Q0.

40

 

 

 

 

H2
mamnanr
_t_-BuOK 7
Dawn“
0 CH
8 2
61 9‘ 5 93
'VD
1) 813235
2) 11202. than
1001
\L
CH,Br 5: H10"
4 map/car“
033cm
451-7“
HOCH
arcu, ‘ 8%
.5.
‘1) nzncrmz
2) N301!) 11+
89!
w
9H,SH
g. .
HSCHI

Scheme 3

41

Thus dibromide £2 and dithiol gé in DMF were added syn-
chronously to a well-stirred suspension of cesium carbonate
in DMF at 95°C. The addition took four days, after which
time analysis of an aliquot indicated complete consumption
of QQ and g2. Filtration of the excess base and removal

of DMF led to the isolation of a yellow-white precipitate.
Column chromatography led to the isolation of three distinct
components in 1%, 3% and 30% yields, respectively, as well
as small amounts of polymeric material. The smallest
fraction eluted first from the column, and is tentatively
identified on the basis of its mass spectrum as the intra-
molecularly oxidized product, disulfide 2% (Scheme 4).

The second component to elute from the column, isolated in
3% yield, was not readily identified by its 1H NMR spectrum.
The mass spectrum indicated this product was a trimer which
incorporated four sulfur atoms. This product might be the
diastereomeric cyclic trimers with a disulfide, instead

of a sulfide, linkage at one juncture. The last and
largest fraction also had no distinguishing features in

its 1H NMR, and additionally, was not sufficiently volatile
to be analyzed by mass spectroscopy. The molecular weight
of this component will have to be assessed by other means,

which may include freezing point depression measurements,

or osmometry.

42

9,13, EH25}!
5 §
CD + CO
BIC“; a "SCH! 8%

032003

DEF

95‘0

Jr

 

. 1»

Scheme 4

43

2.2. The Fate of One Bis-Wittig Rearrangement

 

In extending the reactions performed on the model com-
pounds to the unsaturated system, the synthesis of diol 4Q
had to be addressed. As outlined previously, a bistittig
rearrangement in a derivative of diol 41 was envisioned to

proceed to diol 45. Diol 47 is available in multigram

OH
on {

3.
HO/

4'3 :6

quantities from the diisobutylaluminum hydride reduction of

22,23

dienedione 13, and is obtained in 88% yield as a 4:1

mixture of cis, endo:trans stereoisomers. The cis isomer

 

can be fractionally recrystallized from acetone and is ob-
tained pure in 57% overall yield. Recent endeavors in this
lab have improved on the synthesis of diol 47 by using a
cerous chloride mediated sodium borohydride reduction of
dione 73.48 This combination of reagents results ex-
clusively in carbonyl reduction, and diol 47 is easily
prepared in quantitative yield. The reaction is fast,
clean, and relatively easy to work up; moreover, the use

of an alkyl aluminum reagent is avoided. In a brief study,

44

the ratio of gig, endo : trans isomers was studied as a
function of reaction temperature. Thus, at 0°C, this
ratio was 6:1; at -10°C, 7.6:1; and at -15°C, 8:1 (GLC
analysis, two consecutive runs). (Interestingly, this
temperature dependence parallels that observed in the

BH3°THF hydroboration of saturated dione 61.)

 

o P”
Km
00 1 6: a co
CBC ' O
C: 032 (88‘)
° 3 Ho
-15°C

31?. 4‘72

9H

+ co

H0)

Treatment of diol 41 with two equivalents of potassium
hydride in THF, followed by iodomethyltri-n-butyltin49 gave
the bigftnbgfbutylstannylmethyl ether 58 in 95% yield, after
workup and chromatography. Although the alkylation is not
complete until after 36 h at room temperature, it has been
noted that l8-crown-6 will enhance the rate of alkylation

of hindered secondary alcohols.36

45

Addition of bis-stannane 58 to two equivalents of n-
butyllithium in THF at -78°C produced dianion i, which could
be protonated at low temperature to give the dimethyl
ether 24. Upon warming the solution of dianion i to -30°C
for 36 h, the reaction changed from clear and colorless
to a deep reddish-orange. Warming of this solution to room
temperature, followed by workup and chromatography to
remove tetra-nebutyltin, gave a clear colorless oil homo-
geneous by TLC, which exhibited a complex 60 MHz 1H NMR
spectrum. Both the absence of exchangeable protons in the
NMR, and the absence of an -OH stretch in the IR indicated
the product was not an alcohol, although a peak at 64.95
in its 1H NMR was a characteristic chemical shift for the
allylic proton a to oxygen in diol 41 and bigfstannane 58.
The appearance of a parent ion at m/e 134 confirmed the
suspicion that the product was an ether of structure 74 or
76 (see Scheme 5). Moreover, only two out of ten protons

1H NMR, making

exhibited overlapping signals in the 250 MHz
single frequency homonuclear decoupling feasible. The
distinguishing features in the NMR include one proton at
64.92 (dd, J=8 Hz, J=2 Hz), protons at 63.93 and 63.68
comprising part of an AMX system, and one proton at 63.49
(q, J=8 Hz) (see Figure A9). The signal at 64.92 was
easily identified as H2, and the AMX splitting almost

surely arose from the diastereotOpic protons on C-9 in

15 or C-lO in 16. The quartet at 63.49 could be ascribed

46

(M1 £p~—\

1) KH m SnBu;
:% l3 s
2) BUBSnCHZI "3"

L:

 

 

 

 

 

" 96% 0
$1 2%
n-BuLi
THF, -78'c
W
o ‘ gOCH,“
O 1* {D
C ‘
o 70% .
ZR LICH,O .
1
or
__ H+, -78°C
0/’ v
23 SW3".
CH4;
2%

Scheme 5

47

to only one proton out of both structures, that of C-10

in 15. Irradiation at the frequency corresponding to 63.49
collapsed the doublet of doublets at 64.92 to a doublet

(J=2 Hz), but had no effect on any of the olefinic pro-
tons. Conversely, irradiation of the doublet of doublets at
64.92 transformed the quartet into a triplet (J=8 Hz),

and removed a small (2 Hz) coupling constant from the pro-
ton at 65.68. Therefore, the proton which appears as a
quartet at 63.49 is not allylic, and structure 15 was
assigned to the product.

Later experiments proved that 75 was the isomer ob-
tained. 2,6-g2 Biol 11 was synthesized from sodium boro-
deuteride/cerous chloride reduction of dienedione 73.
Alkylation of 17 with tri-n—butyliodomethyltin resulted in
18, which when treated with nebutyllithium produced a
single product 79 whose 1H NMR clearly indicated the ab-
sence of protons at 64.92 and 65.75 (see Scheme 6 and
Figure 6).

As mentioned before, dianion i could be trapped at low
temperature. gig-stannane 58 was added to nebutyllithium
in THF at -78°C. After 15 minutes aqueous acid was added,
and the solution warmed to room temperature. Workup and
analysis showed only the presence of dimethyl ether 71,
>98% by NMR. Thus dianion i was being formed at -78°C,
but some other reaction intervened before rearrangement

could occur.

48

.Asouuonv ww .moncm

mmuo.H mun cam Agony w» umcum oflaosowuu mo lam: ommc asuuommm mzz ma

.m wusmwm

m musqflm

‘0
C
In
D

 

 

 

 

 

 

 

 

 

49

 

 

 

3)
g
V

 

 

 

 

 

 

 

 

50

 

 

 

' Dyna
nun“ .3 -+
00013: 61:20 7
0 011301) HO“; D
«7,3 «7,2
1)KH
2) Bu’SnCHzI
I)
{‘II'IIIF> z;r Edd
D ‘ W
1O

 

Scheme 6

The failure of dianion i to undergo rearrangement at the
low temperatures characteristic of [2,3] sigmatropic pro-

cesses (vide infra) indicates that at -30°C, the dianion

 

reacts exclusively by a pathway lower in energy than even

the normally facile [2,3] rearrangement. During the

51

course of the reaction, it was noted that the solvent
gradually assumed a deep orange-red color, suggesting THF
was being slowly degraded. It therefore seemed probable
that dianion i was deprotonating THF to give anion ii,

which quickly cyclized to tricyclic ether 15.

 

 

gocmu (pen,
THF
~ %
5 i
' H ‘ 2 t
L“: J) (3 CFQLI
_i_ .
E
«moon3

As a test of this mechanism, bigestannane 58 in ether
was treated with one equivalent of nebutyllithium at -78°C,
and allowed to warm to 0°C over the course of several hours.
An aliquot analyzed by TLC showed no less than six com-
ponents! Of these, five were identified by comparison

with authentic isolated samples (see Scheme 7). The sixth

52

O

8

5'18“. 3) 1g Bum. -78.C-0'C% Buusn 0 “i8
2) H+
I3u£5n

L:

 

a
3g
ocu Qj
£ §j
cap - cm;
74 75 '39
«A: «A:
. BuBSnCHZOH
6%

 

1 1 m. as
3% i ) 9q 7;

2) 0°C. 6 h:
22°C, 8 h

3) m 0

Zé

+ BuQSn + BHSSnCHZOH

3%

Scheme 7

53

component, which possessed the third highest R in the

f
group, was most probably stannane 88. Allowing the reaction
to stir at 0°C for six hours and at room temperature over-
night, produced a greatly simplified TLC analysis, showing
only three components: tetra-nfbutyltin, hydroxymethyltin
8k, and tricyclic ether 18 (Scheme 7). The various inter-
mediates had evidently funneled off to tricyclic ether 88

by scrambling of tri-gfbutyltin in bis-stannane 88 and
dianion i. After chromatography to remove the tin com-
pounds, Zé was isolated in 72% yield, 69% from diol 88.

This method represents a convenient preparation of 18, and
supports the prOposed mechanism for its formation.

In another experiment designed to suppress protonation
of dianion i by the solvent, big-stannane 88 was added to
2.3 equivalents of nebutyllithium in ether at -78°C, al-
lowed to warm to 0°C, and finally stirred overnight at
room temperature. Workup of the reaction led to two pro-
ducts (besides tetra-g—butyltin), tricyclic ether 88 and
another compound which had a much longer retention time
on GLC and TLC; both products were isolated by column
chromatography. The new compound, obtained in approximately
30% yield, showed an -OH stretch in the IR, a doublet at

63.7 in the 60 MHz 1

H NMR, and what appeared to be a con—
densed AB quartet for the olefinic protons at 65.6, as well
as some higher field multiplets (60.6-l.6) attributed to

contaminating tetra-n-butyltin. It appeared that simply

54

changing the solvent had produced the desired rearrange-
ment of i to diol 88!

This new compound exhibited some worrisome characteris-
tics, however. These included the "wrong" migratory ap—
ptitude on TLC for a diol in a given solvent system (too
large) and the observation that the new compound could be
easily analyzed by GLC under conditions where the saturated
analogue, diol 88, failed completely to elute. A GLC/MS
proved the intuitive conclusion of these observations, £;E;I
that the new compound was not diol 88. A parent ion at m/e.
192 did not represent diol 88 (MW=166) but rather, tri-

cyclic ether 7 (MW=134) plus butane (MW=58), or something
’1:

H

of structure 82.
’b’b

°\...

38

Again, quenching experiments performed at -78°C on bigf
stannane 88 added to nebutyllithium in ether gave almost
exclusively dimethyl ether 11 and some 18 (96:4, 18:18,
from GLC), indicating almost complete formation of di-
anion 1. Interestingly, the amount of 81 isolated in the
previous reaction was almost identical to the amount of ex-
cess nebutyllithium employed. Since the -CH20H group was

positively present in the molecule, it appeared that

55

dianion 1 closed to tricyclic ether 18 as before, but also
that excess nfbutyllithium cleaved 18 in some way.

To examine this possibility, tricyclic ether 18 in
ether was subjected to nebutyllithium at 0°C. Following the
reaction by TLC and GLC showed the formation of a compound

with the identical Rf and retention time as 81. It appeared

 

I:
C)
\.«I“

that some of the nebutyllithium was destroyed by the sol-
vent, as more than a stoichiometric amount was required to
complete the conversion (approximately three equivalents).
Workup of the reaction gave an oily compound identical to
81 isolated earlier.

The stereo- and regiochemical features of 81 were not

readily apparent from inspection of its 1

H NMR (Figure A10).
Overlapping multiplets for the bridgehead protons suggests
the 3,7-diene and not the 3,6-diene is formed, as the

latter isomer would be expected to have quite different

bridgehead proton chemical shifts.

While organolithium.reagents generally do not attack

56

unconjugated olefins,50 there is precedent for cleavage of
allylic ethers. Broaddus demonstrated in 1965 that allylic
ethers 88 refluxed in hexane with nebutyllithium undergo a

displacement reaction.51 Complexation of the alkyllithium

l)BuLi/hexane reflux

 

ROCH CH=CH > ROH + BuCH CH=CH
2 2 2 2
2)H20
8% 60%

is considered the activating factor in these additions. The
proposed mechanism involves a six membered cyclic transi-
tion state resulting in SNZ' displacement, but as only

allyl ethers were used in this study, a direct displace-

ment cannot be ruled out. It was noted that lower reaction
temperatures (BO-40°C) resulted in significant isomeriza-
tion to the alkyl gis-propenyl ether, which was stable to the
reaction conditions.

Other examples of alkyllithium displacements include the
addition of isopropyllithium to allylic ether 88 and
homoallylic ether 88.52 Allylic rearrangement was not
demonstrated in 81 however. Moreover, it is well known
that alkyllithium reagents add to the strained double bond

53

in norbornene. In 88, chelation of the ether oxygen with

lithium is essential for the addition, and directs the

54

regiochemistry of attack. The relative configuration of

57

i-PrLi

4¢x\//ocn,
«36

A i-Pl’Li
OCHJ A 54%

85
M»

1,
g.

80%

(L

the addend in 88 was assumed, however, and not rigorously

 

 

 

proven.
R” A, ”’° % .
ocu3 Li'!‘ OCH. OCH:
«‘39
RLi
I 7‘ NR

That nebutyllithium adds easily to tricyclic ether 18
perhaps reflects the ideal geometry for a six membered
transition state when the lithium chelates with the ether

oxygen. The same type of process may also be responsible

58

® 1‘;

\‘ b/R
‘Li—CH,C.H, - 31—33”

for the facile intramolecular displacement of alkoxide in

ii.

2.3. Another Wittig Rearrangement

 

An alternate route to diol 88 would involve two sequen-
tial rearrangements. It seemed clear that suppression of
cyclization of i to 18 could be achieved by converting one
of the side chains to a poor leaving group. This neces-
sarily destroys the symmetry of the molecule, and moreover
renders what was a one-step reaction a multi-step sequence.
But it would be interesting to see if a [2,3] rearrangement
actually could take place in this system.

The best protective group with a poor leaving ability
was conveniently already part of diol 81; elimination of
L120 appeared unlikely (however, see below). Hydroxystan-
nane 81 was prepared in 26% yield from the reaction of diol
81 with potassium hydride and one equivalent of iodomethyl-
tri-nebutyltin (Scheme 8). Separation from bigestannane
88 (31%) and unreacted diol 81 (23%) was achieved by

column chromatography. Addition of hydroxystannane 81 to

59

OH

3
1)KH

‘4—3- Egaliu'
2) 13“3‘°"“’“21 ' + 3&3 + «473

HO ,6, 31¢
'63

20am
8t2°- ~78°c

 

.OCH,Li
f

Lid

1) O'C- 22°C
2) 11+

 

9".on
3

°’ CC)
HO

HO cHon
88
'Vb

38

Scheme 8

6O

nebutyllithium in ether at -78°C, followed by gradual warm-
ing to 0°C, and finally to room temperature, produced a
heavy white precipitate. Acidification and work up of the
reaction led to a multi-component crude product. Analysis
by TLC showed two spots with the "correct" Rf for a diol,
perhaps indicating regioisomers resulting from [1,2] and
[2,3] rearrangement. The presence of other products was
mysterious, especially as GLC analysis of the crude reac-
tion mixture showed these were not trace impurities, but
major by-products (4 components, 2 of which were not com-
pletely resolved, in approximately 2:1:l:3.5 relative
ratio). The various products were isolated by column
chromatography. The major product, formed to the extent
of 40%, was a diol of either structure Qg or g2. TLC, GLC

1H NMR (250 MHz) indicated the major product was a

1

and
single isomer. Several features of the H NMR argued in
favor of structure gg (see Table 2 and Figure A12); most
notably, the protons at 63.42 and 63.56 assigned to the
bridgehead carbons both appeared as triplets with a large
(93. 8 Hz) coupling constant. Structure $2 on the other
hand would be expected to exhibit a quartet for one of the
bridgehead protons, as in tricyclic ether 2%. The trans
isomer of fig would be expected to have a smaller J5,6

(approximately 4 Hz, see below); therefore H would appear

5
as a doublet of doublets (J=8 Hz, J=4 Hz), and this is not

observed.

61

q
gaps-I

,~,

.
H3 .8

1"»

 

 

 

Table 2. 1H NMR Chemical Shifts and Multiplicites in
Diol 88.
W
Chemical Coupling Constants
Proton Shift (6) Multiplicity (Hz)
1 3.56 t of q Jt=8, Jq=2
2 4.81 d J1'2-8
3 5.94 ddd J3, =6, J3'5=l,
J2, =0.5
4 5.74 d of t J3'4=6, Jt=l
5 3.42 t of m Jt=8
6 3.00 br m -------
7 5.64 d of t J7,8=6' Jt=2
8 5.80 d of t J7,8=6' Jt=2
9A 3.68 dd JAB=11’ JA,6=5
9B 3.62 dd JAB=11, JB,6=7

 

 

62

The coupling constants in the penitrobenzoate 29 were

determined with the use of a europium shift reagent (Table

3).55 In this case, J1 2 is 4 Hz, although interestingly,
I

OPNB

39

Table 3. Coupling Constants in p-Nitrobenzoate 20 (Hz).

 

 

Jl,8 2.0 J3,4 5.6
J7,8 5.4 J2'3 2.3
J6a,7 2.3 J1,2 4.0
J68,7 2.3 J1,5 7.2
J4'5 2.0

 

 

J6a,7 = J68,7 = J2,3 = 2.3 Hz. The one diagnostic 1H
NMR spectral feature in the determination of configuration
at C-2 is the coupling constant corresponding to J1,2 in
29.

Finally, it should be noted that one of the upfield
triplets observed in the 1H NMR of the rearranged product is
not due to the -C§CH20H proton. This proton is coupled

unequally to the methylene protons, as observed in the

63

AMX system of the latter.

Initial assignment of chemical shifts to individual
protons was accomplished with decoupling experiments.

These experiments showed 1) the protons at 64.81 and 63.56
are coupled with J=8 Hz, 2) protons at 63.00 and 63.42

are also coupled with a large (23' 8 Hz) coupling constant,
3) irradiation of the signal at either 64.81 or 3.42 affects
olefinic protons at 65.94 and 65.74, and 4) irradiation of
the signal at 63.00 collapses the AMX system at 63.68 and
3.62 as well as affecting olefinic protons at 65.64 and
65.80. The observation that irradiation of signals at 63.42,.
3.00 and 4.81 all affect two olefinic protons suggests the
double bonds are at C-3 and C-7 (the signal at 63.56 could
not be irradiated, because it was obscured by adjacent pro-
tons). In contrast, irradiation of one bridgehead proton
in alternate structure 89 would affect all the olefinic
protons, whereas irradiation of the other would have no
effect.

As a final proof of structure, 2,6-d2 diol 77 was mono-
alkylated with iodomethyltri-n—butyltin yielding 91, which
was then treated with n—butyllithium as before to effect
rearrangement (Scheme 9). Workup and chromatography led
to the same mixture of products observed earlier; the
component with the identical Rf as 88 on TLC was isolated,

l

and analyzed. The H NMR of this compound clearly showed

the absence of signals at 64.81 and 63.00 (Figure 7), thus

64

.AEouaonv ww .moncm

NW!

m.m mun can Agony ww H036 «0 Ana: ommc mzz m

H

.5 musmflm

 

 

 

 

 

 

65

 

 

 

 

66

Dye c H, Li

2¢m_mnd
31:20
-78°C

V?

\
o \“\‘

[JO

 

Scheme 9

validating the earlier structural assignment. The major
product in this reaction therefore results from a [1,2]
rearrangement of carbanion iii. Moreover, the proton at
63.42 in 88 (t of m, Jt=8 Hz), was now in 92 a broad
doublet, J=8 Hz, thereby supporting the assigned configura-
tion.

The by-products in the rearrangement of 81 to diol 88
were isolated, and identified on the basis of spectral

properties to be compounds 82, 93, and 47, in 12%, 7%,

67

and 14% isolated yields, respectively. Thus the other

so -00..

product with a "diol Rf" on TLC was not an isomer of diol
88, but rather the starting material one step earlier.
The origin of these products will be discussed later; it
is significant to note, however, that 88 represents the

only rearranged diol.

 

Anion iii therefore rearranges exclusively by the [1,2]

pathway. The [2,3] rearrangement is depicted as a concerted

56,57

process (path A in the scheme below) whereas the [1,2]

b (Ref. 56)

Q.

0

68

rearrangement is formulated as a nonconcerted radical

56,58 In

dissociation/recombination process (path B).
systems where either a [1,2] or a [2,3] rearrangement may
occur, studies have shown that: l) the [2,3] rearrangement
proceeds rapidly and smoothly at low temperatures (-78°C

to -20°C) and 2) that the [1,2] rearrangement is a competing,

56 A few examples of anionic [2,3]

higher energy pathway.
sigmatrOpic rearrangements are outlined in Figure 8. In
all of these examples, rearrangement takes place at -20°C
or below.

In a study of the benzyl ethers of substituted allylic
alcohols, Baldwin found that the reaction temperature af-
fected the relative amounts of [2,3] 35. [1,2] rearranged

products (Table 4).56

(OH

ems/\OML (“Haw

c." ADM

95
Mb

In both cases, the yield of [2,3] rearranged product in-
creased with a decrease in temperature. In the rearrangement

of 28 at 0°C, the [1,2] process occurred to the extent of

LDA/THF c Lo -LiCN
Qg-WC ——_')_§ o<"‘N—-)\_2J__) __)\_73j

90% (Ref. 59)

 

 

C’H" \ 2 eq LDA/THF j \ '78 C 7"
0 fl euo MMcoH
j

com co,e Lao
(Ref. 60)

OH! C.H,

BuLi A m aka c'"’
D THF -80°c7 D ———)
_ CH, >___)\cn, / “cu,
CH, CH,

(Ref. 57)

 

 

+ 20% [1,2] (Ref. 61)

Figure 8. Examples of anionic [2,3] sigmatrOpic rearrange-
ments.

70

Table 4. Temperature Dependence of [1,2] vs. [2,3] Rear-
rangements of 94 and 95.3

 

 

Reactant Temperature (°C) Product Distribution (GLC %)

 

96 97
mm mm
94 25 95 5
mm
94 -10 98 2
mm
95 25 23 77
mm
95 -20 17 83
mm

 

 

aTaken from Reference 56.

14%, while at -80°C, this value fell to 0 indicating com-

plete dominance of the [2,3] rearrangement.57

 

3111.1
D O 4, MO“ " p“.
M
Ph I)
gag [1.2] product [2.3] product

That a [2,3] rearrangement of carbanion iii was not
observed at all then was surprising, given that in most
cases, [2,3] rearrangement predominated over [1,2] rear-
rangement. A concerted pathway in iii_might be precluded

by the geometrical requirements of a [2,3] transition

71

state. In this bicyclo[3.3.0]octadienyl system, the
"bridged" transition state geometry required resembles an

anti-Bredt norbornene. The activation energy needed to

iii

aggain this conformation could very well exceed that for a
dissociation/recombination mechanism. The approximate dis-
tance between the bond forming terminii in iii is 2.8 A,

as measured on molecular models, quite a bit larger than

the normal carbon-carbon single bond length of 1.54 A.

To date, there are no literature reports of an anionic [2,3]
rearrangement occurring in a cyclopentene ring, although
sulfenate 99 rearranges to the sulfoxide,62 albeit slug-

gishly when compared with acyclic systems. The different

H u

 

 

72

bond lengths in this example, as well as the use of the lone
pair of sulfur or oxygen, might account for the difference
in reactivity. Indeed, the geometrical factor may not be
all that important in [2,3] rearrangements of sulfur ylides,

as there are several reports where the double bond is re-

 

placed with an acetylene.63
cc": éc/cofls c/CJ'I,

It is also interesting to note that the selenium dioxide
oxidation of allylic alcohols appears to proceed by a radi-
cal dissociation/recombination mechanism in five- and six-
membered rings, but through a concerted [2,3] process in

64 Endocyclic

acylic and large ring systems (Scheme 10).
olefin oxidations in small rings often give poor yields of
allylic.alcohols, with concurrent formation of numerous
allylic ether and peroxide by-products. These can be
accounted for if the allyl radical intermediate traps the
various nucleophiles in solution.

In addition to the geometrical restraint in iii, the

system is so arranged that a [2,3] transition state results

in two negative charges being brought into close proximity

73

0*
\\ \ ”one“ [2 3] \
Sc + ———9 HOSe ' 9
o” > g /\n
H HOSeO
ROH
\
HOSeOR +
HCD
O’
0’ PK) 4*
H .
"O’Sefl ___> m ———) products
A, a: fi, A!
Scheme 10

to each other. Whether the incipient charge repulsion is
sufficient to drive the reaction exclusively through a

[1,2] pathway is unknown. Elimination of this factor may
of course be accomplished by attempting the rearrangement

in a suitably substituted cyc10pentenol.

D o

\| BuLi \ 9
Q SnBu3 ’ '

The other products in the rearrangement of carbanion

 

iii to 88 are 88, 88 and 81. The hydroxymethyl ether 88 is

74

\wm
II
Om
I:
C)

HO)

simply the protonation product of anion iii. The formation
of compounds 88 and 88 were totally unexpected, however.
The presence of diol 81 among the products indicated either
a cleavage reaction, A, or a disproportionation, B, was

occurring. In view of the ease of destannylation effected

A {OW

cm:
1
SnBu, m % co 4. csxusnm3
Li-

 

LK) ER 33
‘OLi
i
B
SL‘W
Sn Bu, ”0 8N7

 

(a
.+

E
0..
,7

$2 SnBu,
8035"

75

by nfbutyllithium, it is diffucult to rationalize the dis-
placement depicted in A. In both A and B, the presence of
81 is invoked to account for the formation of diol 88.
At the elevated temperatures ‘23- 0°C) required to effect
rearrangement, the initially formed carbanion iii must

participate in the equilibrium shown below.

QCH,Li
1

09 +

U0

iii

.0
f 7‘] +- BuLi

 

There is ample precedent for carbanion-tetraalkyltin equilib-
ria. For example, the reaction of vinyl tributyltin with
phenyllithium led to 73% tributylphenyltin and 12% tri-
butylvinyltin, while the reaction of vinyllithium with tri-
butylphenyltin gave 5% tributylvinyltin and 75% tributyl-

phenyltin (Equation 3).65

76

EtZO
Bu SnCH = CH2 + PhLi ‘___——_-£ Bu

3 2 SnPh + CH = CHLi

3 2

(3)

The equilibrium of stannyl vinyllithium 888 with dilithio-

ethylene 888 has also been proposed to account for anomalous

Bu,5n H Li H
)c =c< Bu Li (:3 >c=c< + Bu,sii
«1,99 49%

66 In

reactions of the dilithioethylene thus generated.
this last example, it is not known whether the production of
dilithioethylene is hindered by thermodynamic factors, or
by kinetic considerations.

In the event of an equilibrium which produced 88, dis-
proportionation to 88 and 81 could follow along the lines
of equation B. The fate of the other product, bis-stannane
88, would have to account for the butylated product 88 also
isolated. The disproportionation of two molecules of
88 to 88 and 88 would liberate two molecules of g-butyllithium.
It is already known that big-stannane 88 reacts with one
equivalent of E—butyllithium to give tricyclic ether 88;

thus it is conceivable that the second‘equivalent was

reacting with the tricyclic ether initially formed to give

77

butylated product 88, as previously demonstrated (Scheme

11).

QCH,5nBu,

Qj .
ms. m sm’ 3“” 9 m

L!

 

ODIN"

\Em

v83 75

'VM

8%

 

\n‘“

Scheme 11

Another mechanism would serve to account for the forma-
tion of butylated product 88, in which a [2,3] rearrangement
occurs followed (or preceded) by an SN2' displacement of
lithium oxide with gfbutyllithium (Scheme 12). It is
known that allyl alcohol adds alkyllithiums to yield 2-

substituted l-propanols 888, however the reaction of

78

 

.OCHzLi
BuL1 ,
4.120 LIO
ii; [zgn
_ Lid CHzoLi
89
'Vb
[zan
BuLi
no}
82
'Vb
C‘Hg c4...”
102 Li 103
m
0 OH

104
Wmm

SCheme 12

79

cyclopentenol and (under forcing conditions) cyclohexenol
with n—butyllithium gives exclusively olefin formation

67 Allylic rearrangement

from displacement of lithium oxide.
accompanies displacement, as demonstrated in the reaction

of 2-cyc10pentenol-l-d lgé (Scheme 12). The difference in
regioselectivity for these additions is not well understood,
although it may result in part from the greater stability

of a primary gs. a secondary carbanion in the case of allyl
alcohol.

This scenario is less likely, however, given that no re-
arrangement products corresponding to diol 82 were isolated,
and that 3 would have to rearrange exclusively gig a [2,3]
process, as Opposed to the [1,2] observed in iii.

This type of displacement reaction introduces the pos-

sibility that tricyclic ether Zé is formed by elimination

of Lizo, followed by addition of n-butyllithium.

 

Clearly, there is a surfeit of possible mechanisms to
rationalize the formation of butylated product §%' Only
the disprOportionation scheme, however, accounts for the

formation of both 8% and diol 41.

80

In the rearrangement of 2,6-d2 hydroxystannane 9%,

products lgé, lgz, lgg, and £1 were isolated. The position

Dd“) D ACHJDH
«'BUU

\\

    
 

 

U,
a
W
.5
Jr

   

of the deuteriums in lgz (Figure 9) unfortunately does not
distinguish between the proposed mechanisms for its forma-
tion, as all involve allylic rearrangement. The isolation
of $81 does, however, confirm the assigned 2,6- regiochem-

istry of the substituents.

2.4. The Synthesis of Cis, endo-2,6-bis(hydroxymethyl)-

 

bicyclo[3.3.0]octa-3,7-diene ég

 

In any case it seemed worthwhile to explore the reaction
of other nucleophiles with tricyclic ether 1Q, in an effort
to utilize this ring Opening reaction to generate diol QQ.
Since Seebach had used methanol dianion in some carbonyl

alkylations with success, this appeared to be a good

81

ml .AEouuonV www .ooamcm

pum.¢ mufl paw Amouv ww poopoum concamvsn mo Aux: ommv Eonuommm mzz ma .m madman

 

ma."
. m wusmflm . v . .
w 1 q q Id

dr-
~r~

 

 

 

82

 

\n

 

83

nucleophile to try.
CH,O Li

H ‘LacanQ; o L,

 

Dilithiomethanol was generated by the addition of
freshly prepared hydroxymethyltributyltin 8%49 to two
equivalents n-butyllithium in ether at -78°C. Addition
of tricyclic ether lg to this cold, ethereal solution of
LiOCHZLi, followed by warming to 0°C resulted in the ap-
pearance of a thick, white precipitate. Neutralization, work-
up and chromatography gave a single product, which on the
basis of its spectral properties was identified as the

desired diol 4%

‘OH
09 *LiCHzou; .0
7

/‘

“N" \

 

H0
«39

In addition to 3%, a small amount of butylation product
8% was again observed; retitration of the n—butyllithium
indicated the previously determined molarity was correct.

Thus it appeared another tin exchange equilibrium was

84

occurring (Equation 4). If this were the case, the equilib-
rium could be shifted to the left by increasing the concen-

tration of llg in the reaction. A 2:1 ratio of llg:lgg

0°C
LiOCHzLi + Bu4Sn F===3 LiOCHZSnBu3 + BuLi (4)
109 110
mmm mmm

completely suppresses the butylation reaction and diol 3%
can be isolated in 47% yield by column chromatography. An
excess of dilithiomethanol lgg relative to tricyclic ether
Z8 was used, and one of the products isolated after a long

reaction time appears to be stannane ill. This contaminant

can only be separated from the product diol by careful
column chromatography. Pure diol éé slowly crystallizes
on standing.

The relative configuration of the added hydroxymethyl

1H NMR. One substi-

substituent was determined from the
tuent is necessarily endo from the tricyclic ether start-

ing material, and an endo orientation of the incoming

85

nucleophile would result in a compound with C symmetry

2v
whereas exo attack results in a compound with C1 symmetry.

13

The C NMR spectrum of diol 39 showed only five signals.

In addition the diacetate llg was synthesized from 3%,

l

and its 250 MHz H NMR showed one sharp singlet for the

13C NMR spectrum of diacetate

acetate methyl groups. A
llg showed only seven signals. When the crude reaction
mixture is worked up with acetic anhydride, a small amount
(<5%) of another acetate was observed in the 1H NMR.

Although this material might be the isomeric trans-diacetate

%%%' the 13C NMR of this mixture is not consistent with
OM: OAc
/
i
Z
ACO/ ACO )
egg ll3
Mv»

this hypothesis, as the few small miscellaneous signals
were not close in value to the more intense signals. Fin-

ally, H in diacetate llg was observed to be a triplet,

l
with J=8 Hz.

l3C chemical shifts for the bridge-

A comparison of
head carbons of various bicyclo[3.3.0]octadienes studied
in this project is found in Table 5. It is evident that

the 3,7-dienes have similar bridgehead carbon chemical

cupu

Co

/

H0
.39

:1? 5‘3
1° 0 '2)

A00

 

39 age

Table 5. Comparison of 13C Chemical Shifts of Bridgehead

Carbons.a

 

 

Chemical Shift

 

Compound Carbon (ppm)
47 C-1 51.59
24 C-1 51.03
88 C-1 51.04
88 C-5 53.52
46 C-1 51.22
112 C-1 50.52
26 C-8 57.18
26 c-11 40.92 (40.52)b
75 C-7 57.01
75 C-10 50.75 (50.63)

 

 

al3C spectra obtained at 62.8 MHz in CDC13.

bIndefinite assignment.

87

shifts around 51 ppm; on the other hand, the 3,6-dienes
exhibit a large difference in chemical shift between the
two bridgehead carbons. The doubly allylic C-S appears at
57 ppm, similar to that found in triquinacene (57 ppm).
Both structures 1,5 and fig have one olefinic carbon at
138 ppm which is significantly different than the 131-134
ppm range found in the 3,7-dienes.

The evidence therefore strongly supports the assignment
Of regiochemistry in diol £6.

The selectivity in the ring Opening Of tricyclic ether
75 with dilithiomethanol is surprising in view of the sig-
nificant sterric bias in 15 towards exo attack. gngg selec-
tivity may be due to chelation Of lithium with the ether
oxygen similar to that proposed for the nebutyllithium

cleavage of 82.

A: ___, _

“~Li—cu,0u °"‘ up”

EXPERIMENTAL

General. Melting points were measured in Open capil-
laries with a Thomas-Hoover apparatus and are uncorrected.
Proton nuclear magnetic resonance (NMR) spectra were ob-
tained on Varian T-60 or Bruker WM-250 spectrometers at
60 MHz and 250 MHz, respectively. Proton homonuclear de-
coupling experiments were performed on a Bruker WM-250
spectrometer. Carbon-13 NMR spectra (proton decoupled) were'
determined on a Bruker WM-250 spectrometer at 62.8 MHz.
Chemical shifts are reported in parts per million downfield
from tetramethylsilane internal standard. Infrared spectra
(IR) were measured on a Perkin-Elmer 137 spectrometer and a
Perkin-Elmer 237 spectrometer, and were calibrated with the
polystyrene 1601 cm—1 peak. Mass spectra (MS) were Ob-
tained on a Finnigan 4000 instrument with an ionizing voltage
Of 70 eV, or when noted, with ionized methane (CI). Gas
liquid chromatography (GLC) was performed on an F & M model
700 chromatograph equipped with a thermal conductivity
detector. Component ratios were determined by a comparison
of peak areas (determined by triangulation) and are uncor—
rected for detector response. Columns used for analysis
were 6' x 8" aluminum columns packed with the following

stationary phases on Chromosorb G (acid and base washed and

88

89

silanized, 60/80 mesh): column A, 5% Carbowax 20M; column
B, 4% QF-l; column C, 3% SE-30. Thin layer chromatograms
(TLC) were run on Machery-Nagel Polygram SIL G/UV254 pre-
coated 0.25 mm silica gel plates, and were developed with

either iodine vapor, or an anisaldehyde spray reagent.68

Flash column chromatography69

was accomplished with What-
man LPS-2 silica gel (37-53 pm). Normal column chromatog-
raphy was performed on EM Reagents Silica Gel 60, 70-230
mesh (EM cat. #7734) rated at activity 2-3.

Unless otherwise noted, reagents and solvents were
reagent grade materials and were used as received. Ether,
THF and benzene were dried by distillation from sodium or
potassium benzophenone ketyl. Diisopropylamine and DMF
were dried by distillation from calcium hydride. g-Butyllith-
ium was Obtained from Aldrich as a 1.6 M solution in hexane,
and was titrated according to the method of Watson and

Eastham.70

5,11

Bi(2-oxa-3-oxotricyclo[7.2.1.0 ]undeca-6,9-dien-4-yl)

 

21. Oxidative Dimerization Of Lactoneggé Lithium Enolate.

 

 

Cupric triflate was prepared according to the procedure
of Jenkins and Kochi.24 Diisopropylamine (0.15 mL, 1.07
mmol) in dry THF (5 mL) was cooled to -78°C and B—butyl-
lithium (0.83 mL of a 1.23 M solution, 1.02 mmol) added
slowly gig syringe. Lactone 2623 (166 mg, 1.02 mmol) in

THF (5 mL) was slowly added to the solution of base and

90

the reaction stirred at -78°C for 30 min. Then cupric
triflate (369 mg, 1.02 mmol) in dry DMF (5 mL) was added,
and the reaction stirred for 8 h at -78°C. The reaction
was quenched with 2 N HCl (3 mL), more water was added (5 mL),
and the whole extracted with ethyl acetate (3 x 50 mL).
The organic layer was washed with water (1 x 10 mL) and
saturated NaCl solution (1 x 10 mL). Drying over MgSO4
and removal Of solvent left a brown oil. Crystallization
Of lactone dimer 27 was effected by dissolving the Oil in
iSOprOpanOl/chloroform and cooling tO 0°C overnight.
Dimer 27 was obtained as small colorless cubes (41 mg,
25%), with the following properties: mp 223-224 C; IR

1 1

(nujol) 1727 cm’ ; H NMR (CDCl 60 MHz) 66.18 (2H, dd,

3,
J=6 Hz, J=2 Hz), 5.76 (4H, m), 5.36 (4H, m), 3.73 (2H, m),
3.30 (4H, m), 2.93 (2H, m); MS m/e 323 (M+1, 7%), 322

(M+, £3. 2%), 276 (21%), 232 (23%), 161 (93%), 117 (100%),
79 (27%).

In runs where cupric bromide or cupric chloride were
used, the a-halolactone 32 or 33 was also obtained. Fol-
lowing the same procedure as before but using cupric
bromide, lactone 26 (51 mg, 0.31 mmol) was transformed
to a mixture of products which could be separated by
column chromatography on silica gel packed with ether.
Recovered lactone 26 (8 mg, 16%) and a-bromolactone 32
(27 mg, 36%) were isolated. The latter product was identi-
l

fied on the basis Of its spectral prOperties: H NMR

(CDCl 60 MHz) 66.23 (1H, dd, J=6 Hz, J=2 Hz), 5.86 (2H,

3!

91

condensed AB q, J=6 Hz), 5.58 (1H, d, J=8 Hz), 5.46 (dd,
1H, J=6 Hz, J=2 Hz), 4.33 (1H, d, J=4 Hz), 04.00-2.93
(3H, m); MS m/e 243 (M+l), 242 (M+), 241 (M+l), 240 (M+),
161, 117, 104.

Similarly, the use of cupric chloride converted lactone
26 (125 mg, 0.77 mmol) to a mixture Of products containing
a-chlorolactone 33 to the extent Of 14% (NMR yield, from
integration of the doublet at 04.40 (J=4 Hz)). The dimer
27 was Obtained by crystallization Of the crude product

from chloroform/THF (17 mg, 14%).

2,6-Bis-methylenebicyclo[3.3.0]Octane 6%. Methyl tri-

 

phenylphosphonium bromide (Aldrich, 97 g, 0.27 mol) was
dried overnight in a vacuum oven at 80°C, and then trans—
ferred to a dry 1 L three necked flask equipped with a
mechanical stirrer, reflux condenser, addition funnel, and
provision for inert atmosphere. The salt was suspended in
dry benzene (500 mL) and potassium E-butoxide (Aldrich,
27.3 g, 0.24 mol) added all at once. A yellow color formed
as the slurry became very thick and difficult to stir.
Continued stirring at room temperature for 1 h resulted

in a bright yellow solution which was quite fluid, and
easily stirred. Dione 61 prepared by the method of Haga-

38 (9.35 g, 0.067 mol) in dry benzene

dorn and Farnum,
(100 mL) was added dropwise, and the well-stirred reaction

refluxed under nitrogen for 24 h. After cooling to room

92

temperature, the precipitated solid was gravity filtered
through a large funnel and washed with an additional por-
tion of benzene (100 mL). The benzene was removed at am-
bient pressure through a 30 cm Vigreaux,and the Oily resi-
due transferred tO a l L Erlenmeyer flask while hot. Pen-
tane (300 mL) was added to the cooled residue, whereupon
triphenylphosphine oxide separated as a light tan solid.
The lumps were crushed and the flask cooled to 5°C for
several hours. The precipitate was filtered through celite
packed in a fritted funnel, and washed thoroughly with

cold pentane. The solution was concentrated by distilla-
tion Of pentane at ambient pressure to a volume Of 93.

150 mL, transferred to a separatory funnel, and washed with
water (2 x 30 mL) (to remove unreacted ylide) and saturated
NaCl solution (1 x 30 mL). The organic phase was dried
over Na2504, and concentrated as before. Repeated crystal-
lization Of triphenylphosphine oxide from pentane served

to remove most of it from the product, while the last
traces were removed by flash chromatography (75 g Whatman
silica gel packed with pentane). Diene 62 (7.83 g, 95%)
was isolated as a clear, colorless, Oil with a strong
Olefin smell. Analysis by GLC showed one component with

a retention time of 5.2 min (column C, 90°C, 11 mL/min).
Diene 62 had the following properties: bp 55°C (mlO mm
aspirator pressure); IR (neat) 3180 (w), 2950, 1675 (m,

doublet), 1440, 875 (s) cm'l; 1H NMR (CDC13, 250 MHz)

93

64.83 (4H, AB q, Av=45 Hz, J=l7 Hz), 2.94 (2H, m), 2.30

13

(4H, m), 1.92 (2H, m), 1.63 (2H, m): C NMR (CDCl 62.8

3’
MHz) 157.79, 104.80, 49.10, 33.87, 32.58 ppm; ms m/e 135
(M+l,7%), 134 (M+, 59%), 120 (7%), 106 (31%), 91 (100%),

80 (17%).

Cis, endo-2,6-bis(hydroxymethy1)bicyclo[3.3.0]octane

 

63. Hydroboration Of Diene 62 with Disiamylborane. 2-

 

 

Methyl-Z-butene (Aldrich) was distilled from lithium alum-
inum hydride before use (bp 37.5-38°C). A 500 mL three
necked flask fitted with an addition funnel, serum cap,
stir bar and provision for inert atmosphere was flame
dried under vacuum. A nitrogen atmosphere was established,
and BH3-THF solution (Ventron, 105 mL of a l M solution,
0.105 mol) was introduced by a syringe. This solution was
cooled in a NaCl-ice bath, and 2-methyl—2-butene (24 mL,
0.22 mol) in dry THF (26 mL) added dropwise over a period
Of 30 min. Stirring at ice bath temperature was continued
for 2.5 h, whereupon diene 62 (6.19 g, 46 mmol) in THF

(50 mL) was slowly added. The reaction was stirred at 0°C
for l h, then allowed tO warm to room temperature overnight.
The reaction was cooled to 0°C, and carefully quenched with

1:1 THF:H O (30 mL). This was followed by the addition of

2
aqueous NaOH (35 mL Of a 3 M solution) and the slow addi-
tion of H202 (34 mL Of a 30% solution), keeping the tem-

perature below 50°C with an ice bath. After the addition

94

was complete, the reaction was warmed to 55°C and stirred
for l h. Upon cooling, the layers were separated, and the
aqueous phase extracted with chloroform (1 x 100 mL).

This was combined with the THF portion, and dried over
MgSO4. The solution was filtered, and the solvent removed
to give an Oil which was subjected to high vacuum for 1 day
to remove the amyl alcohol by-product. DiOl 63 was Ob-
tained as a somewhat cloudy Oil (8.19 g, 104%). Kugel-
rohr distillation (110°C, 0.05 mm) gave an analytical sam-
ple of the clear, colorless Oil, which showed two peaks, in
a relative ratio Of 1:123 (99.1% one isomer) when analyzed
by GLC (column B, 155°C, 60 mL/min). DiOl 63 had the fol-
lowing properties: IR (neat) 3325 (br), 2950, 2875, 1660
(w), 1475, 1065 (s) cm‘l; 1H NMR (CDC13/D20, 250 MHz)

03.68 (2H, dd, part of an AMX system, J =11 Hz, J

AM AX:
7 Hz), 3.61 (2H, dd, part Of an AMX system, JAM=11 Hz,
JMX=7 Hz), 2.60 (2H, sextet, J=6 Hz), 2.19 (2H, d of q,

=20 Hz, Jq=7 Hz), 1.72 (2H, septet, J=6 Hz), 1.53 (4H,
13

Jd

m), 1.18 (2H, d of t, Jd=18 Hz, Jt=8 Hz); C NMR (CDCl

3!
62.8 MHz) 64.51, 45.99, 45.75, 29.17, 24.34 ppm; MS m/e
171 (n+1, 1%), 151 (2%), 134 (13%), 121 (95%), 106 (20%),

93 (100%), 79 (84%).

Hydroboration of Diene 62 with BH3-THF Complex. The

 

 

following procedure is representative of the general method

used in the hydroboration Of diene 6% with BH3-THF. A

95

three necked round bottom flask equipped with a magnetic
stir bar, addition funnel, serum cap and provision for
inert atmosphere was flame dried under vacuum. A nitrogen

atmosphere was established, and BH -THF (7.60 mL Of a l M

3
solution, 7.60 mmol) was introduced gig syringe. The solu-
tion was cooled to either -23°C or -45°C, and diene 62

(462 mg, 3.45 mmol) in dry THF (15 mL) was slowly added
dropwise to the stirred solution. The reaction was crystal—
clear initially, but eventually the borane adduct separat-
ed as a gelatinous slurry. Stirring was continued for 3 h
at the indicated temperature, and then the reaction warmed
to 0°C and carefully quenched with 1:1 ethanol:water (10
mL). Stirring was continued at room temperature for 2 h

to insure complete hydrolysis and then NaOH (3.5 mL Of a

3 N solution) and H (0.75 mL Of a 30% solution) were

202
added. The reaction was heated to 50-55°C for 1 h and then
cooled to room temperature. The mixture was extracted with
ethyl acetate, the organic phase dried over MgSO4, and

the solvent removed to yield a clear Oil (609 mg, 100%).
The crude product was analyzed by GLC (column B, 180°C,

60 mL/min), and exhibited two barely resolved peaks at

6.4 min and 7.0 min. The relative areas Of these two com-
ponents were determined to be 1:12, respectively, at ~45°C,
and 1:6 at -23-0°C. The spectral properties of this

material were identical to those Obtained for the product

of disiamylborane hydroboration/oxidation Of diene 62,

96

with the exception of shoulders on the doublet at 03.58

1

in the H NMR (60 MHZ).

Cis, endo-2,6-bis(hydroxymethyl)bicyclo[3.3.0]octane

 

Dimethanesulfonate 64. Diol 6% (1.7 g, 0.01 mol) was dis-

 

solved in dry pyridine (15 mL) in a 25 mL round bottom
flask, and cooled to 0°C. Methanesulfonyl chloride (7.7 mL,
0.10 mol, previously distilled (bp 54.5-55°C/m10 mm)),

was added to the stirred solution. The reaction was stirred
for 2 h, and then poured onto ice. Extraction with chloro-
form (2 x 20 mL) was followed by multiple extractions Of

the combined organic phases with cold 5% NaOH (3 x 20 mL),
cold 1 M HC1 (4 x 20 mL), saturated NaHCO3 (1 x 20 mL),

and saturated NaCl solution (1 x 20 mL). The extract was
dried over MgSO4 and the solvent removed to afford di-
mesylate 64 (2.96 g, 91%) as a brown Oil which solidified

on standing. Similar runs Of this reaction gave yields
ranging from 90-96%. A 1H NMR of this material (CDC13,
60 MHz) was as follows: 64.16 (4H, d, J=7 Hz), 2.96 (s,

6H), 2.76—0.96 (12H, m). This material was not characterized

or purified further, but was used directly in the next

step.

Cis,endo-2,6—bis(iodomethyl)bicyclo[3.3.0]octane, 65.

 

TO a solution Of crude dimesylate 6% (2.96 g, 9.0 mmol) in

acetone (40 mL) was added sodium iodide (4.08 g, 27 mmol).

97

The reaction was refluxed with stirring for 12 h, after
which time an additional portion of sodium iodide (1 g)

was added and refluxing continued for an additional 6 h.
The reaction was cooled, filtered free of precipitated
salt, and the solvent removed under reduced pressure. The
brown residue was taken up in ether and washed with satu-
rated sodium sulfite (2x) and saturated NaCl solution (1x).
The organic phase was dried and the solvent removed to give
crude diiodide 65 as a brown oil. Flash chromatography on
silica gel packed with hexane yielded diiodide 65 (1.94 g,
55%) as a clear, colorless Oil, which could be kugel-

rohr distilled (100°C (0.05 mm)) only when very pure (g_.__g_i
after chromatography). Attempted distillation Of the crude
product resulted in a diminished yield of 65, Obtained as

a dark purple Oil. The spectral characteristics were as
follows: IR (neat) 2950, 2899, 1458 (triplet), 1282, 1182

1 1

cm- ; H NMR (CDCl 60 MHz) 63.18 (2H, dd, part of an

3'
AMX system, J=10 Hz, J=8 Hz), 3.03 (2H, dd, part Of an AMX
system, J=10 Hz, J=8 Hz), 2.83-0.93 (12H, two m centered

at 2.46 and 1.56); MS m/e 391 (M+1, 33. 0.02%), 390 (M+,

0.8%), 263 (7%), 135 (100%), 107 (9%).

Cis,endo-2,6-bis(bromomethyl)bicyclo[3.3.0]octane 62.

 

Reaction Of Diol 6% with Carbon Tetrabromide/Triphenylphos-

 

 

phine. Diol 63 (2.01 g, 11.8 mmol) was dissolved in aceto—

 

nitrile (200 mL) in a 500 mL three necked flask equipped

98

with either a good magnetic stir bar or a mechanical stirrer,
thermometer and stOpper. Carbon tetrabromide (16.1 g, 48.6
mmol) was added, and the solution stirred until it was all
in solution. Triphenylphosphine (12.8 g, 48.9 mmol) was
then added all at once, causing the reaction to warm up to
40°C. The solution immediately became a dark red, and
began to turn brown as triphenylphosphine oxide precipi-
tated. After stirring at room temperature for 2 h, TLC
analysis (silica gel, ether) showed only one component.

One half of the solvent was removed under reduced pressure
(beware Of bumping!) and the remaining solvent and pre-
cipitate poured into an equal volume Of ether. The solution
was filtered through celite packed in a fritted funnel,

and the filtered precipitate washed thoroughly with cold
ether. The solvent was removed under reduced pressure, and
more triphenylphosphine oxide crystallized from cold 1:1
ether:petroleum ether. The Oil finally Obtained was placed
under high vacuum overnight to remove bromoform (bp 150°C).
The remaining product was flash chromatographed on silica
gel (Whatman) packed with hexane to remove contaminating
carbon tetrabromide. The dibromide (2.12 g, 61%) was
isolated as a clear, colorless Oil with the following
spectral prOperties: IR (neat) 2950, 2875, 1440, 1275,
1225 cm-l; 1H NMR (CDC13, 250 MHz) 63.44 (2H, dd, part of

an AMX system, JAM=12 Hz, =8 Hz), 3.37 (2H, dd, part

JAX

Of an AMX system, J =12 Hz, =9 Hz), 2.68 (2H, sextet,

AM JMX
J=6 Hz), 2.43 (2H, d of q, Jd=20 Hz, Jq=7 Hz), 1.84 (2H,

99

=4 Hz), 1.65 (2H, m), 1.52 (2H, m),
13

q Of d, Jq=7 Hz, Jd

1.22 (2H, t of t, J=12 Hz, J=8 Hz); C NMR (CDCl 62.8

3!
MHz) 47.04, 46.57, 35.05, 31.11, 24.05 ppm; MS m/e 297
(0.04%), 296 (0.01%), 295 (0.10%), 293 (0.05%), 215 (11%),

135 (100%).

Cis,endo-2,6-bis[(amidinothio)methyl]bicyc10[3.3.0]-

 

Octane Dibromide 71. The Addition Of Thiourea to Dibromide

 

 

69. NB: The success of this reaction depends on the purity
Of the dibromide; substantial amounts of other bromocarbons
(gig; carbon tetrabromide) will seriously interfere-with
the calculated stoichiometry. Pure dibromide 62 (890 mg,
3.00 mmol) and thiourea (Aldrich, 685 mg, 9 mmol, 50%
excess) were placed in a screw-capped reaction tube, and
95% ethanol added (10 mL, iigi, to make the solution 0.3

M in dibromide). The tube was tightly capped and heated

on a steam bath for 18 h. During this time, the product
crystallized from the reaction. The solution was allowed
to cool to room temperature, and then chilled to -15°C

for 4 h. The resultant solid was filtered, washed with
cold ethanol, and air dried to give a white solid (1.46 g,
108%), mp l62-164°C. This material was not purified
further, although it Obviously contained some of the excess

thiourea.

100

Cis,endo-2,6-bis(mercaptomethyl)bicyclo[3.3.0joctane

 

66. Hydrolysis Of Bis-thiouronium Bromide 62. The thio-

 

uronium salt from above (1.46 g m3 mmol) was placed in a
100 mL round bottom flask, and suspended, with stirring,
in aqueous NaOH (40 mL of a 1.5 N solution). A serum cap
was fitted to the flask and the suspension stirred as
nitrogen was bubbled through the solution with a syringe
needle (another needle served as a vent). After 5 min,

a nitrogen purged condenser was fitted to the flask, and
the whole system kept under nitrogen for the duration Of
the reaction. The solution was refluxed for 12 h, where—
upon the suspension dissolved. After cooling tO 0°C,

the reaction was acidified to pH 1-2 with cold HC1 (4N)
in the hood. The mixture was transferred to a separatory
funnel, and extracted with ether (2 x 75 mL). The organic
phase was washed with water (2 x 20 mL) and saturated
NaCl aqueous solution (1 x 20 mL), and dried over NaZSO4.
Removal of solvent (stench!) left a yellow Oil (710 mg).
This was flash chromatographed on silica gel (Whatman)
with 5% ether in hexane. Dithiol 66 was obtained as a
clear, colorless (but not odorless) oil (585 mg, 96%

from dibromide 66), and was best stored in the dark under
nitrogen to discourage formation of disulfides. Dithiol
66 had the following spectral properties: IR (neat) 2924,
2857, 2654 (w), 1449 (triplet), 1250, 758 cm—1; 1H NMR

(CDC1 250 MHz) 62.57 (2H, m), 2.55 (4H, t, J=8 Hz),

3'

101

2.12 (2H, m), 1.79 (2H, d of pentet, Jp=7 Hz, Jd=5 Hz),

1.64-1.32 (4H, m), 1.36 (2H, t, J=8 Hz, -sg), 1.13 (2H,

13

tt, J=11 HZ, J=9 HZ); C NMR (CDC1 62.8 MHZ) 47.93, 46.51,

3!
31.52, 26.40, 24.05 ppm; MS m/e 203 (n+1, 8%), 202 (M+,

62%), 168 (37%), 135 (40%), 121 (82%), 93 (100%), 79 (88%).

The Reaction Of Dithiol 66 with Dibromide 66; A l L

 

three necked flask fitted with an addition funnel, ther—
mometer, magnetic stir bar and provision for inert atmos-
phere was flame dried under vacuum, and a nitrogen atmos-
phere established. Cesium carbonate (1.83 g, 5.64 mmol,
Alpha) was suspended in dry DMF (320 mL), and the tempera-
ture brought to 95°C. Dithiol 66 (570 mg, 2.82 mmol) mixed
with dibromide 62 (834 mg, 2.82 mmol) in DMF (70 mL) was
slowly added over a period Of 3 days. The reaction was
cooled to room temperature, and excess cesium carbonate re-
moved by filtration. DMF was distilled at 25°C (0.10 mm).
The residue was flash chromatographed on silica gel (75 g)
packed with 20% methylene chloride in hexane. Only one
component eluted in this solvent system (10 mg, ml%).
Increasing the polarity of the solvent with 40% methylene
chloride eluted another component (30 mg, %3%). Elution
with straight methylene chloride gave a third fraction (280
mg, 30%). Further elution of the column led to the isola-
tion Of small amounts of polymer. 1H NMR spectra Obtained

on these materials did not unambiguously identify them.

102

The smallest fraction is tentatively identified as disul-
fide 1%: MS m/e (CI) 241 (2.5%), 229 (5%), 201 (86%),

200 (8%), 167 (25%), 135 (100%). The 3% fraction showed

MS m/e 537 (2%), 536 (9%), 371 (3%), 167 (100%). The
largest fraction was insufficiently volatile tO be analyzed

by MS.

Cerous Chloride Mediated Sodium Borohydride Reduction

 

Of Bicyclo[3.3.0JOcta-3,7-diene-2,6-dione 1%. Cis,endo-2,6-

 

 

 

dihydroxybicycloE3.3.0]octa-3,7-diene 61. A three necked

1 L round bottom flask was equipped with a mechanical
stirrer, thermometer and drying tube. Dienedione 16 (2.0 g,
0.15 mol) (prepared by the method Of Hagadorn and Farnum)38
in absolute methanol (600 mL) was added, and stirring com—
menced. Cerous chloride hexahydrate (Fluka, 116 g, 0.33
mol) was added, and the solution cooled to -20°C with a dry
ice-acetone bath (below m-25°C, the cerous chloride begins
tO precipitate from solution). Sodium borohydride (22.6 g,
0.60 mol, 100% excess) was added slowly portionwise, allow-
ing the foam to subside between additions and keeping

the temperature below -10°C by the judicious addition of
dry ice tO the cold acetone bath. The addition was com-
plete after 1 h, and the acetone bath was replaced with

a NaCl-ice bath. Stirring was continued at -15°C for 2 h,
at which time TLC analysis (20% ethyl acetate in ether)

showed the reaction was complete. Warming tO 0°C, water

103

(300 mL) was added and the reaction stirred for 15 min.
Methanol and water were removed under reduced pressure,

the last traces on a vacuum pump with a large capacity cold
trap. The solid was transferred to a 2 L erlenmeyer flask,
crushed and extracted with boiling chloroform (1 x 1 L,

2 x 500 mL). After each extraction, the solution was cooled
and filtered through an 11 cm Buchner funnel. The combined
filtrates were passed through a Drierite cone and the
chloroform distilled under reduced pressure to afford the
isomeric dienediols 61 as a waxy, white solid (22 g, 106%
due to residual chloroform). Recrystallization of the crude
product from acetone at -15°C, followed by a second re-
crystallization Of the mother liquors at -30°C yielded

gig, ggggfdiol 61 as hard, white granules (10.9 g, 53%),

mp 92-93°C. The Spectral properties Of this material were

identical to those reported earlier.

Temperature Effects in the Sodium Borohydride/Cerous

 

Chloride Reduction of DienedioneiZQ. The reduction was

 

carried out as previously described on a l-lO 9 scale at
one Of the following temperatures or temperature range:
0°C to 22°C; -10°C; -15°C. The crude product from each
reaction was analyzed by GLC (column A, 190°C, 60 mL/min),

and the relative ratio of cis:trans isomers determined

 

(retention times of 10.1 and 12.2 min, respectively). At
-15°C, this ratio was 8:1, at -10°C, 7.6:1, and at 0°C-22°C,

6:1.

104

Iodomethyltri-g-butyltin. The title compound was pre-

 

pared from methylene iodide and iodomethylzinc iodide fol-

71 9

lowing the procedures of Seyferth and Andrews, and Still.4

The reaction was run on 0.02 to 0.80 molar scales, with iso—
lated yields ranging from 78-85% after distillation (bp
118-119°C/0.65 mm), (literature lOO-llO°C/0.01 mm). TLC
(hexane Rf 0.51; IR (neat) 2865, 1456 cm'l; NMR (c0c13,
60 MHz) overlapping absorptions at 61.90 (2H, s, flanked
by two isotope satellite peaks, J=18 Hz, ICHZSn), 1.55

(m), 0.93 (m).

Cis,endo-2,6-bis[(tri-g—butylstannyl)methoxy]bicyclo—

 

 

[3.3.0jocta-3,7-diene 66. Potassium hydride (6.42 g of a

 

35% mineral Oil dispersion, 0.056 mol) was weighed out in

a dry 500 mL three-necked flask. A mechanical stirrer,
addition funnel,nitrogen inlet adapter and rubber serum

cap were fitted to the flask, and a nitrogen atmosphere
established. The potassium hydride was washed free of min-
eral Oil with dry ether (3 x 20 mL), and then suspended in
dry THF (100 mL). Dienediol 67 (2.58 g, 0.019 mol) in dry
THF (100 mL) was added dropwise to the stirred suspension
at room temperature. When gas evolution ceased (gg. 2 h),
iodomethyltri-g-butyltin (17.7 g, 0.041 mol) in dry THF

(75 mL) was added dropwise, and the reaction stirred for

24 h at room temperature. TLC analysis showed the reaction

to be complete (20% ethyl acetate/hexanes) Rf 66, 0.68;

105

Rf 61, 0.30. Methanol (10 mL) was then added slowly to
destroy excess hydride. The dark brown solution was added
to an equal volume of petroleum ether and extracted with
water (2 x 100 mL), and saturated aqueous sodium chloride
(1 x 25 mL). The organic phase was dried over sodium sul-
fate and the solvent evaporated to leave a light brown
liquid (19.3 g). The crude product was applied directly
to the top of a 6 cm diameter column containing silica gel
(250 g) packed with hexane. When residual nonpolar com-
pounds were no longer detected in the eluate (TLC, 1%
ethyl acetate/hexane, Rf 0.67) the solvent was changed to
20% ethyl acetate/hexane whereupon the title compound
eluted, and was isolated as a very pale yellow liquid
(12.8 g, 96%) with the following properties: TLC (1%
ethyl acetate/hexane) Rf 0.17; IR (neat) 3040, 2857, 1613,

1

1453 cm" ; NMR (CDC1 60 MHz) 65.70 (4H, condensed AB q,

3!
J=7 Hz), 4.26 (2H, m), 3.75 (4H, 5, two isotope satellite
peaks, J=18 Hz), 3.33 (2H, m), 2.16-0.53 (54H, two mul-

tiplets, including a broad distorted triplet at 0.88).

2-Oxa-tricyclo[5.2.l.0.4’lo]nona-5,8-diene 16. Reaction

 

of 66 with Two Eguivalents Of p-Butyllithium. A 100 mL
three necked flask fitted with a serum cap, nitrogen
inlet valve, magnetic stir bar and addition funnel was
flame dried under vacuum. A nitrogen atmosphere was es-

tablished, and g—butyllithium (6.6 mL of a 1.1 M solution in

106

hexane, 7.3 mmol) was transferred to the flask and cooled
to -78°C. Dry THF (14 mL) was added slowly to the cold
g-butyllithium gig syringe, followed by big-stannylmethyl
ether 66 (2.5 g, 3.3 mmol) in THF (25 mL). The reaction
was stirred at -78°C for 0.5 h, and then allowed to warm

to -30°C. The nitrogen inlet valve was closed, and the
reaction stored in a freezer at -30°C for 36 h. At the end
Of this time, the reaction solution had become quite a deep
reddish-orange. A nitrogen line was reattached to the ap—
paratus and the reaction mixture allowed tO stand at room
temperature for several hours. Water (6 mL) was added and
the whole solution transferred to a separatory funnel.

The organic phase was washed with saturated sodium chloride
(1 x 20 mL) and then dried over MgSO4. The dry organic
extracts were filtered into a 250 mL round bottom flask
containing a magnetic stir bar, and the solvents removed

at atmOSpheric pressure through a 28 cm Vigreaux. The red-
orange residue left from the distillation was applied to
the top of a 4 cm diameter column containing silica gel

(65 g) packed in 20% ether/pentane. Fractions averaging

20 mL were collected, and the chromatography followed by
TLC. Tetra-g—butyltin eluted first, followed by tricyclic
ether 76. The fractions containing 66 were combined,

and the bulk Of the solvent removed at atmospheric pressure
followed by brief evacuation (aspirator pressure) to remove

the last traces Of solvent. In this way, Z6 was Obtained

107

as a clear, colorless, volatile Oil. If further purifica-
tion is desired, 66 can be distilled through a short path
distillation unit (bp 80°C/10 mm) into a cold receiver.

TLC (10% ether/pentane) Rf 0.20; IR (neat) 3067, 2941,

1

1618, 1076 cm” ; NMR (CDC1 250 MHz) 65.98 (1H, dd, J=6

3'

Hz, J=2 Hz), 5.75 (1H,dt, J =6 Hz, Jt=2 Hz), 5.68 (1H, dt,

d
Jd=6 Hz, Jt=2 Hz), 5.50 (1H, dt, J=6 Hz, J=2 Hz), 4.92
(1H, dd, J=8 HZ, J=2 HZ), 3.93 (1H, dd, J=10 HZ, J=7 HZ),
3.68 (1H, dd, J=10 HZ, J=4 HZ), 3.64 (1H, m), 3.49 (1H,
13C

q, J=8 Hz, with some fine structure), 3.26 (1H, m);

NMR (CDC1 62.8 MHz) 138.76, 134.41, 131.29, 129.59, 87.95,

3,
74.95, 57.01, 50.75, 50.63 ppm; MS m/e 135 (M+1, 0.3%),
134 (M+, 4%), 104 (100%), 78 (17%).

Tricyclic ether Z6 should be stored under nitrogen, and
preferably kept cold, as it appears tO be rather sensitive

to oxygen and susceptible to polymerization over a period

of time.

Low Temperature Quenching Of Dianion i_Derived from

 

Bis-stannane 66 and g-Butyllithium. Cis,endo-2,6-dimethoxy—

 

 

 

bicyclo[3.3.0]0cta-3,7-diene Z6. Big-stannane 66 (267 mg,
0.36 mmol) in either dry ether or dry THF (2 mL) was slowly
added to a solution Of g-butyllithium (0.92 mL of a 1.11 M
solution) in either dry ether or THF at -78°C. The reaction
was stirred for 15 min at -78°C, and then quenched with

0.5 M HC1 (4 mL). The organic phase was separated from

108

the aqueous phase and washed with water, and then dried
over MgSO4. The 60 MHz 1H NMR showed only one product for
both solvents. The crude product from the reaction done in
ether was analyzed by GLC (column A, 140°C, 60 mL/min) and
consisted of 96% dimethyl ether 74 and 4% tricyclic ether
Zé- 1 3,

Hz), 4.46 (2H, br m), 3.43 (8H, a sharp s superimposed on

H NMR (CDC1 60 MHz) 5.80 (4H, condensed AB q, J=6

a m).

2-Oxa-tricyclo[5.2.l.04'101nona-5,8-diene, 16. Reaction

 

 

Of 66 with One Equivalent of g—Butyllithium. A 500 mL three

 

 

necked flask equipped with a magnetic stir bar, nitrogen
inlet adapter and a rubber serum cap was flame dried under
vacuum, and a nitrogen atmosphere established. gig-stan-
nylmethyl ether 66 (11.5 g, 15.5 mmol) in dry ether (175
mL) was introduced gig cannula to the flask. The solution
was cooled to -78°C, and g-butyllithium (16.3 mL of a 0.95
M solution in hexane, 15.5 mmol) added slowly gig syringe.
After stirring for 0.5 h at -78°C, an aliquot from the re-
action showed the presence Of at least six components

(TLC 20% ethyl acetate/hexanes Rf = 0.85, 0.68, 0.51, 0.33,
0.28, 0.17). These correspond in order of migratory ap-
petitude on TLC, to tetra-g—butyltin, gig-stannylmethyl
ether 66, methoxy stannylmethyl ether 66, hydroxymethyl-
tri-g—butyltin 66, tricyclic ether 66 and dimethyl ether

66. By allowing the reaction to warm slowly to 0°C, and

109

stirring at 0°C for 5 h, followed by warming to room tem-
perature overnight, TLC analysis showed only tricyclic
ether 66, hydroxymethyltri-g-butyltin, and tetra-g—butyltin.
Apparently, the various bicyclooctadiene compounds fun-
neled off to the tricyclic ether. Workup as before gave

pure Z6 (1.44 g, 72%).

Cis,endo-Z-hydroxy-6-[(tri-g—butylotannyl)methoxy]-

 

bicyclo[3.3.0]Octa-3,7-diene, 66. This reaction was run

 

under conditions identical to those used for the prepara-
tion of gig-stannylmethyl ether 66, save for the fact that
only one equivalent of iodomethyltri-g-butyltin was used.
Thus diene diol 66 (664 mg, 4.81 mmol) in dry THF (5 mL)
was added dropwise to a stirred suspension Of potassium
hydride (700 mg of a 35% Oil dispersion, 5.77 mmol) in THF;
when gas evolution ceased, iodomethyltri-g—butyltin (2.06
g, 4.81 mmol) in THF (5 mL) was added drOpwise. After
stirring at room temperature for 2.5 h the reaction was
worked up as before to give a viscous brown oil (2.16 g).
The products were separated by flash chromatography on 150 g
silica gel packed in 20% ether/hexanes. gigfstannyl-
methyl ether 16 eluted first (TLC 20% hexane Rf 0.68),
(1.11 g, 31%), followed by the title compound 66 (TLC 20%

ethyl acetate/hexane Rf 0.30) (0.50 g, 24%), which exhibited

110

the following properties: IR (neat) 3344 (br), 3077, 2907,

l

1634, 1471, 1070 cm’ ; NMR (CDC1 60 MHz) 65.96-5.53 (4H,

3,
s at 5.83 overlapping an AB q at 5.53, J=6 Hz), 4.71 (1H,
br m), 4.21 (1H, m), 3.75 (2H, t, J=7 Hz), 3.33 (1H, m),

1.85 (1H, s, OH), 1.73-0.67 (12H, m at 1.33, distorted t

at 0.90, J=6 Hz). Further elution of the column with 40%
ethyl acetate in hexane yielded recovered dienediol 47

(150 mg, 23%). Thus the total yield of chromatographed

material represents a 78% mass balance.

n-Butyllithium Exchange with Cis,endo-Z-hydroxy-6—(tri-i

 

n-butylstannyl)methoxybicyclo[3.3.0]octa-3,7-diene 87.

 

Cis,endo-2-hydroxy-6-hydroxymethy1bicyclo[3.3.0]octa-3,7-

 

 

diene 88, and 2-Endo—hydroxymethyl-6-butylbicyclo[3.3.0]-

 

octa-3,7-diene 8%. A 250 mL three necked flask equipped

 

with a magnetic stir bar, nitrogen inlet and a rubber serum
cap was flame dried under vacuum and a nitrogen atmosphere
established. Dry ether (100 mL) was transferred to the
flask via cannula, and cooled to -78°C. n—Butyllithium
(14.3 mL of a 0.95 M solution in hexane, 13.6 mmol) was
slowly syringed into the cold ether, followed by (stannyl)-
methoxy alcohol 87 (2.93 g, 6.65 mmol) in ether (20 mL).
Stirring was continued at -78°C for 0.5 h; the reaction
mixture then gradually warmed up to 0°C whereupon stirring
was discontinued. Within one hour of standing at 0°C,

a heavy white precipitate appeared. The reaction was

111

allowed to stand at 0°C for 6 h, and then at room tempera-
ture overnight (12 h). Water (10 mL) was added and the

layers separated. The aqueous phase was back-extracted

with chloroform (l x 50 mL). The organic extracts were com-

bined and washed with saturated, aqueous sodium chloride,
and then dried over sodium sulfate. Removal of the solvent
left an orange-red oil (3.49 9). Gas chromatographic
analysis of the crude material (column A, 190°C) showed
the presence of at least four major products (besides
tetra-n-butyltin), two of which could not be completely
resolved, in a 2:1:l:3.5 relative ratio. The entire crude
product was taken up in a minimum of ether and chromato—
graphed on silica gel (100 g) packed with 1% methanol/
ether. Following elution of tetra-n-butyltin, the solvent
was changed to 25% ethylacetate in ether containing 1%
methanol. Butylated product 82 was isolated as an oil
(150 mg, 12%), and had the following properties: TLC (20%
ethylacetate/hexane) Rf 0.50; IR (neat) 3356 (br), 3030,

l l

2890, 1453, 1370, 1068 cm“ ; H NMR (0001 250 MHz) 05.71

3:
(1H, d J=6 Hz), 5.62 (2H, AB q, J=6 Hz), 5.53 (1H, d of t,
J=6Hz, J=2 Hz), 3.72 (1H, dd (AMX), J=10 Hz, J=6 Hz), 3.68
(1H, dd (AMX), J=10 Hz, J=7 Hz), 3.45 (2H, m), 3.01 (1H,
m), 2.74 (1H, m), 1.66-0.80 (10H, m centered at 01.35, and
a broad t at 0.89); ms m/e 192 (M+, 15%), 161 (64%), 135
(6%), 117 (85%), 105 (66%), 91 (83%), 41 (100%).

The next product to elute from the column was methoxy

112

alcohol 9% (70 mg, 7%), easily identified by the sharp

singlet at 03.4 in its 1H NMR spectrum: TLC (20% ethyl

1

acetate/hexane) Rf 0.51, H NMR (CDC1 60 MHz) 05.76 (4H,

3’
m), 4.76 (1H, br m), 4.35 (1H, br m), 3.20-3.80 (5H, sharp

singlet at 03.40 superimposed on a multiplet), 2.68 (1H,

S, OH).

Biol 88 next eluted from the column, and was obtained
as a clear, colorless, viscous oil (270 mg, 40%). TLC

(20% ethyl acetate/ether) Rf 0.27; IR (neat) 3280 (br),

3021, 2874, 1639, 1020 cm’l; 1H NMR (c001

13

3, 250 MHz)

(See Table 2, page 61); C NMR (CDC1 62.8 MHz) 134.14,

3!
133.62, 132.68, 130.38, 77.98, 63.01, 53.52, 51.04, 49.22

ppm; MS m/e 153 (M+l, 0.7%), 152 (M+, 7%), 134 (12%),
121 (49%), 104 (100%), 91 (58%), 78 (59%).

Finally, as the last product to elute from the column,
diol 47 was isolated as an oil which solidified on stand-
ing. Its 1H NMR spectrum left no doubt as to its identity.

TLC (20% ethyl acetate/ether) R 0.12.

f

n—Butyllithium Cleavage of Tricyclic Ether 15. Endo-

 

 

 

2-(hydroxymethyl)-6-n—buty1bicyclo[3.3.0]octa-3,7-diene 8%.
Tricyclic ether 75 (61 mg, 0.45 mmol) in dry ether (5 mL)

was cooled to 0°C and n-butyllithium (1.0 mL of a 0.95 M
solution) added via syringe. After the reaction was stirred
for l h at 0°C, an aliquot was quenched and analyzed by

GLC (column A, 135°C-l900C, 60 mL/min). A EE- 1:1 ratio of

113

15 to product was observed. The product had a retention
time identical to that of butylated material 82 isolated
earlier. The relative ratio of 75:82 did not change over
the course of 2 h, so additional n—butyllithium was added
(2 mL). The reaction was stirred at 0°C for 1 h and a1-
1owed to warm to room temperature overnight, after which
time GLC analysis of an aliquot indicated the reaction was
complete. Careful quenching with methanol to destroy
excess alkylithium, extraction with water and saturated
aqueous NaCl was followed by dryingtflmaorganic phase over
MgSO4. Removal of solvent at reduced pressure left a
yellow oil (70 mg, 80%) which was identical in all respects

to 82 isolated earlier.

Sodium Borodeuteride/Cerous Chloride Reduction of Bi-

 

cyclo[3.3.0]octa-3,7-diene-2,6-dione 73. Cis,endo-2,6-

 

 

dihydroxybicyclo[3.3.0]octa-3,7—diene-2,6-d_2 77. The

 

reduction was carried out as described previously, using
sodium borohydride. Dienedione 7% (1.75 g, 13 mmol) and
cerric chloride hexahydrate (10.1 g, 28 mmol) in methanol—
d (35 mL) was stirred at -15°C, and sodium borodeuteride
(2.2 g, 52 mmol, Merck Sharp & Dohme) added portionwise.
Workup of the reaction followed the procedure previously
outlined. The title compound was obtained as a mixture of
stereoisomers (1.90 g, 100%). Recrystallization of the

crude material from acetone afforded labeled Cis,endo diol

114

1

77 (632 mg, 35%). H NMR (c0013, 60 MHz) 04.10 (4H, AB q,

40:14 Hz, J=6 Hz), 3.36 (2H, br s), 2.83 (2H, br s, 03)?

MS m/e 141 (M+1, 0.5%), 140 (M+, 4%), 122 (99%), 108 (3%),

92 (100%).

Alkylation of 2,6-d Diol 77 with One Equivalent of Iodo-

2
methyltri-n—butyltin. Diol 77 (604 mg, 4.31 mmol) was

 

alkylated with iodomethyltri-n—butyltin according to the
procedure described earlier for the protio analog, g7.
Labeled bis-stannane 78 was isolated as a clear liquid
(620 mg, 19%) which was identical to 88 with the following

1 1

exceptions: IR (neat) 2079 cm- ; H NMR (CDC1 60 MHz)

3,
5.68 (4H, AB q, J=6 Hz, 00:10 Hz), 3.75 (4H, s, with two
isotope satellite bands, J=8 Hz), 3.35 (2H, s), 2.15-0.60
(54H, 2 m centered at 01.38 and 00.90).

Labeled hydroxystannane 87 was obtained as a clear,
viscous oil (660 mg, 35%) identical to 87 with the follow-
ing exceptions: IR (neat) 2151 cm-1; 1H NMR (CDC13, 60
MHz) 05.93-5.53 (4H, s at 05.83 overlapping an AB q cen-
tered at 05.40, J=6 Hz), 3.70 (2H, s, two isotope satel-
lite peaks, J=8 Hz), 3.33 (2H, br s), 1.75-0.80 (10H,

2 m centered at 01.36 and 00.90); MS (CI) m/e 443 (0.2%),

442 (0.7%), 387 (2%), 291 (100%), 251 (38%).

115

2—Oxa-tricyclo[5.2.l.04’lOJdeca-S,8-diene-1,6-g2.

Bis-stannane 78 (505 mg, 0.67 mmol) was destannylated with

 

two equivalents of n—butyllithium in THF following the pro-
cedure outlined for the protio analog 88. Labeled tri-
cyclic ether 78 was obtained as an oil (45 mg, 50%) which

1

had the following spectral properties: H NMR (CDC1

3!

250 MHz) 05.96 (1H, dd, J=6 Hz, J=2 Hz), 5.67 (1H, dd,

J=6 Hz, J=1 Hz), 5.49 (1H, br s), 3.93 (1H, dd, J=9 Hz,
J=7 Hz), 3.67 (1H, dd, J=9 Hz, J=4 Hz), 3.63 (1H, m), 3.49

(1H, t, J=8 Hz), 3.24 (1H, m).

gi§,endo-2-hydroxy-6-(hydroxymethyl)bicyclo[3.3.0]octa-

 

3,7-diene-2,6-d2 88_and Endo-Z-(hydroxymethyl)-6-n—buty1-bi

 

 

cyclof3.3.0]octa-3,7-diene-4,8-g 107. Labeled hydroxy-

2—’L"u’\a

 

stannane 87 (456 mg, 1.03 mmol) was destannylated with n-
butyllithium following the procedure outlined for the protio

compound, 87. Labeled diol 706 was isolated as a clear,
mm mm

1

colorless oil. H NMR (CDC13/D20, 250 MHz) 05.96 (1H,

dd, J=6 Hz, J=2 Hz), 5.82 (1H, dd, J=6 Hz, J=2 Hz), 5.76
(1H, dd, J=6 Hz, J=2 Hz), 5.65 (1H, dd, J=6 Hz, J=2 Hz),
3.73 (1H, d, J=12 Hz), 3.64 (1H, d, J=12 Hz), 3.58 (1H,
br d, J=7 Hz), 3.42 (1H, br d, J=7 Hz).

Labeled butylated product 787 was obtained as an oil.

1H NMR (CDC1 250 MHz) 05.62 (1H, br s), 5.54 (1H, br s),

3!

3.72 (2H, ABX; HA centered at 03.74, J=11 Hz, J=7 Hz; H

centered at 03.68, J=11 Hz, J=7 Hz), 3.46 (2H, m), 3.01

B

116

(1H, m), 2.75 (1H, m).
Labeled methyl ether 788 was identical to the parent
compound 88, save for the absence of signals at 04.76 and

04.33 in its 60 MHz 1H NMR spectrum.

Cis,endo—2,6-bis(hydroxymethyl)bicyclo[3.3.0]octa-3,7-

 

diene 88. Nucleophilic Ring Opening of Tricyclic Ether 78

 

with Methanol Dianion. Tri-n-butyltin hydride was prepared

 

from bis(tributyltin)oxide (Pfaltz and Bauer) and poly-
methylhydrosiloxane (Aldrich) according to the procedure

72

of Hayashi, Iyoda and Shiihara. Hydroxymethyltri-n-

butyltin 87 was prepared by the method described by Still.49
Thus, a 250 mL three necked round bottom flask was fitted
with a nitrogen inlet, stir bar, ground glass stopper and
a serum cap, and then flame dried under vacuum. A nitrogen
atmosphere was established, and dry THF (75 mL) transferred
to the flask. Diisopropylamine (10.3 mL, 73.5 mmol) was
added, and the solution cooled to 0°C, whereupon n—butyl-
lithium (73.7 mL of a 0.95 M solution in hexane) was slowly
added. Stirring was continued at 0°C for 5 min, and then
tri-n—butyltin hydride (18.4 mL, 70 mmol) was added slowly
from a syringe. After complete addition, the reaction was
stirred 15 min at 0°C, and then paraformaldehyde (2.1 g,

70 mmol; dried overnight in a vacuum oven at 40°C) was

added all at once. The reaction was stirred at room

temperature for 3.5 h, and then poured into petroleum

117

ether (450 mL). This solution was washed with water (2 x
300 mL), saturated NaCl solution (1 x 50 mL), and dried
over NaZSO4. Removal of the solvent at reduced pressure
yielded a colorless oil (19 g). The crude product was
chromatographed on a 2 inch diameter column containing
silica gel (150 g) packed with hexane. One large fraction
was collected (450 mL) in which residual nonpolar tin com—
pounds resided. The solvent was changed to 10% ethyl
acetate in hexane, whereupon hydroxymethyltri-n—butyltin
87 eluted. Removal of the solvent left a colorless oil
(13.8 g, 61% (lit. 67%)). This material had the same Rf
on TLC as reported in the literature: (10% ethyl acetate/
hexane) Rf 0.31. The NMR Spectrum was consistent with the

1

reported structure of the adduct. H NMR (CDC1 60 MHz)

3,
03.63 (2H, 5, two isotope satellite peaks, J=6 Hz), 1.93-
0.46 (28H, two overlapping multiplets). This material
partially decomposes within a day, and therefore it should
be used directly in the next step.

A 250 mL three necked round bottom flask was prepared
as before, and hydroxymethyltri-n—butyltin 87 (13.8 g, 43.1
mmol) in dry ether (75 mL) was transferred to the flask
with a cannula. This solution was cooled to -78°C under
a nitrogen atmosphere, and n-butyllithium added slowly from
a syringe (68.1 mL of a 0.95 M solution in hexane, 64.5

mmol) while the solution was vigorously stirred. The re-

action was stirred at ~78°C for l h, then warmed to -40°C,

118

and stirred for 45 min at this temperature. After cooling
the reaction to -78°C, tricyclic ether 78 (1.01 g, 7.51
mmol) in dry ether (10 mL) was slowly added. After com-
plete addition, the reaction was allowed to warm to 0°C
and stirred at this temperature for 2 h. During this time
a very heavy precipitate formed (the dilithium alkoxide

of 88 is insoluble in ether). At this point, the reaction
is probably complete. In this run, however, the reaction
was stirred overnight at room temperature, and then cooled
to 0°C and cautiously quenched with methanol (exothermic!).
A minimum of water was added to dissolve the precipitate
(10 mL) and the layers separated. The aqueous phase was
back extracted with chloroform (2 x 50 mL) and the organic
extracts combined. Drying over NaZSO4 and removal of sol-
vent left two immicible oils, one very viscous, and the
other quite fluid. The free flowing oil was removed from
the flask with a pasteur pipette, and chromotographed on

a 6 cm diameter column containing silica gel (125 g)
packed in 20% ethylacetate/ether containing 1% methanol.
Dienediol 88 eluted as a viscous, slightly yellow oil (140
mg) .

The crude viscous oil was chromatographed on a 2 cm
diameter column containing silica gel (20 g) packed with
20% ethyl acetate/ether and 1% methanol. Elution of the
column led to dienediol 88 (430 mg). The combined chromato-
graphed yield, therefore was 770 mg, or 62%.

After isolation of the diol, it was clear from

119

miscellaneous absorptions in the NMR in the 1 ppm region
that a small amount of tin contaminant remained. Thus, the
product was rechromatographed on silica gel (25 g) with 3%
methanol in ether to afford pure dienediol 88 as a clear,
colorless oil (470 mg) which began to crystallize on stand-
ing. A less pure fraction 93. 90% in 88 was also isolated
as an oil (90 mg). The yield of twice chromatographed
material was then 44%. The actual yield is much higher, how—
ever, as chromatography of a diol generally only leads to
70% recovery. It would be advantageous to recrystallize
the product directly from the crude product mixture.

The spectral properties of diol 88 are as follows: IR
(neat) 3360 (br), 3045, 2895, 1740, 1020 cm—1; 1H NMR

(CDC1 250 MHz) 6.82 (2H, dt, J =6 Hz, Jt=2 Hz), 6.61

3!
(2H, dt, J

d

d=6 Hz, Jt=2 Hz), 3.73 (2H, dd (AMX), J=11 Hz,

J=7 Hz), 3.66 (2H, dd (AMX), J=ll Hz, J=8 Hz), 3.58 (2H,
t of m, Jt=8 Hz), 3.04 (2H, m), 1.65 (1H, s, OH), 1.49

(1H: S! OH); 13

C NMR (CDC13, 62.8 MHZ) 131.67, 131.35, 63.72,
50.48, 51.22 ppm; MS (CI) m/e 167 (M+1, 9%), 166 (M, 0.7%),

149 (37%), 131 (100%), 121 (30%).

Cis,endo-2,6-bis(acetoxymethy1)bicyclo[3.3.0]octa-3,7-

 

diene 777. Diol 88 (3 mg, 0.018 mmol) in 5 mL chloroform
was refluxed for l h with dry pyridine (0.10 mL) and acetic
anhydride (0.10 mL). After cooling the solution, ice and

more chloroform were added, and the reaction was extracted

120

with dilute HC1 and saturated sodium bicarbonate (2 x 20

mL). Removal of the solvent gave an oil which was chromato-
graphed on silica gel with ether. One compound was iso-

lated (4.5 mg, 100%) which was quite pure from TLC analysis.
IR (neat) 3045, 2950, 2880, 1735, 1230 cm’l; 18 NMR (00c13,
250 MHz) 5.69 (2H, dt, Jd=6 Hz, Jt=2 Hz), 5.55 (2H, dt, Jd=6
Hz, Jt=2 Hz), 4.10 (2H, dd (AMX), J=11 Hz, J=7 Hz), 3.99
(2H, dd (AMX), J=ll Hz, J=9 Hz), 3.56 (2H, t of m, Jt=8

Hz), 3.13 (2H, m), 2.08 (6H, s); 13c NMR (CDC1 250 MHz)

3’
170.99, 131.47, 130.82, 65.34, 50.52, 47.48, 20.99 ppm;
MS (c1) m/e 250 (M+, 0.5%), 249 (3%), 205 (4%), 177 (2%),

163 (10%), 129 (34%), 117 (4%), 61 (100%).

SUMMARY

The synthesis of gig, Egggfz,6-§i§(hydroxymethyl)bi-
cyclo[3.3.0]octa-3,7-diene, a key intermediate in our ap-
proach to the synthesis of dodecahedrane, has been achieved.
Some interesting chemistry has been learned en route to
diol 88. This includes the observation of an extraordinarily
facile SNi' attack of an organolithium species on an un-
strained double bond in the generation of tricyclic ether
78. It can also be concluded from this work that a [1,2]
Wittig rearrangement rather than an anionic [2,3] sigma-
tropic rearrangement is the preferred, if not exclusive,
reaction path in a methyl-metallated bicyclo[3.3.0]-3,7-
diene-2,6-diol methyl ether. Finally, eXploitation of the
geometrical constraints in tricyclic ether 78 led to a
chelation—controlled stereoselective SN2' ring opening of
78 with methanol dianion.

Concerning the use of diol 88 in the synthesis of do-
decahedrane, model studies have served to convert the
saturated analog of 88 to the dibromide and dithiol in ac-
ceptable yields. The successful application of these re-
actions to diol 88 thus seems assured. An element of un-
certainty is introduced into the overall synthetic plan

(summarized in Figure 10) by the uncharacterized macro-

molecular sulfide products obtained in the model dibromide-

121

122

 

/""'o

SH

Summary of the sulfide contraction route to

Figure 10.
hexaene 78.

123

dithiolate dimerization. The appearance of a major product
in this reaction bodes well for its eventual characteriza-
tion as the desired gig-sulfide 88. Completion of the
model studies thus requires the isolation and identifica-
tion of gig-sulfide 88, and its transformation to diene

88. The application of these reactions to the "real system"
and the synthesis of hexaene 78 must then be addressed.
Finally, the exciting possibility of an electrocyclic ring
closure of hexaene 78 to dodecahedrane, and the possibility
of trapping a metal cation inside the cavity, needs to be

investigated.

APPENDIX

124

 

125

a~HH.m

o.H.N.NQOHoonMuOXOIMImeINVanmfioAwmz ow

.m

.Nw Aamlvlcmflpum.mnmompcs
Hooov snuuommm mzz cououm

.H¢ musmflm

126

Hfl wusmflm
on o v

 

 

AIA

v.3“.- .

 

 

mumL-DJ J - -

 

 

127

Imo.m.mQOHU>UHchmH>£umEImHQIm.N mo Aux: omm

.m

.ww wcmuoo
HUQUV Esuuommm mzz cououm

.N< musmflm

128

o

m< mucosa

0 _

v

m

 

1 4 d 11 A! 1‘ 111‘

 

4w

 

14‘ 1‘ 1‘ 1 ‘ J‘ 1‘ 14 J

4 1‘ .1 ‘ Id 4 f 4 11 < J

4 4

fiJJ4+11¢JJJA 4 4

 

 

%
I

we.

 

O

.20

.30

\

 

 

 

 

 

\

 

 

129

 

 

-Aasnumasxouvsncman-m.~-oocm.mao mo Awmz 0mm

.m

.Ww mangoomo.m.mHoHo>oHn
HUQUV Esnpommm mzz cououm

.m< musoflm

130

 

AL
l1
0

L

 

 

 

 

 

 

 

 

 

 

Figure A3

 

 

131

 

IaaxcumEOEounvmHQIo.NIO©:m.mflU mo ANmE 0mm

.m

.ww mcmuooho.m.mloHo>ofln
HUQUV Esuuowmm mzz cowoum

.v¢ wusmflm

132

O P

fil 14 4 4‘ d d 1 l - I4 1 l4 1 1“ c C -

 

 

J

 

 

v< muomflm
N
IJIll—‘l-‘lll‘l 404.114

 

D

 

 

 

 

 

133

 

 

 

 

 

IAH>£umEouQmoumEVmflbno.mI0©cm.wflo mo Ammz 0mm

.m

.ww wcmuoomo.m.mgoHomofln
HUQUV Esuuommm mzz :Ououm

.m4 wusmflm

134

 

:m.:

”u"

0

m4

 

 

 

2J1;
é

 

 

 

135

.ww Heacufla acm ww wofleounflm mo cofluommu mcaamsoo
mcp CH Awom .mmv acupoum HohmE wnu mo ANmE omm .mauoov Esuuowmm mzz cououm

.mfl wusmflm

136

 

 

 

$4 madman

 

 

137

.ww mamaouh.mumuoomo.m.onHomoflnflsxoaumsAH>:

 

LcmumawvsnnclfluuvQmflnlm.NIOUcw.mHo mo Aux: on

.m

HUQUV Eduuowmm mzz cououm

.nd musmfim

138

n4 wusmflm

 

 

 

 

 

 

. _— __ —-—

 

 

 

 

 

: a. ‘ 5...; :r. :C :\
d 1 fl 4 J a d 4 a d J _ q d 1 q H a
1 d 1 l4 1 fl 4 1 d d d - . d

 

 

"anew

 

 

139

IGCOCfi

OHJV

o.a.m.m_oHo>ofluunmxoum mo ANmz om.

m

.ww mcmflcum.m
Hooov gunpowmm mzz cououm

.m< musmflm

140

m< wusmflm

 

 

I :_ :x. .: 2.. as 5:; 2.. so 3\
4 d‘ d _ q d T d —r 1 1 i 4 — d d d d _ d d d 1 1 d 1 d — 1 11 d I F q
- q . 1 4 a < 4 4 — 4 q 4 a a 4 d — . a q q - a a . J < a q — . d

 

 

 

 

 

 

 

 

 

 

 

 

 

141

Imcocw

oa.v

o.a.m.mgoHo>oHuunmxoum mo “um: omm

.m

.$ 658$ .m
HUQUV Esuuommm mzz coeoum

.md wusmflm

142

 

 

 

 

 

 

m< ousmflm
)3 5
g; k c _.
_l l
mm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

143

.ww wcmflpuh.mumuoomo.m.muoHo>o

 

Iflnazusnuolawsuwewxou©>£|0pcwlm mo Aux: ow

.m

Huoov Esuuommm mzz conoum

.oam musmflm

144

 

oac musmflm

.1.

Sr

 

 

 

 

 

 

 

 

 

 

 

145

.Nw wcmfipnh.mumuoofio.m.mgoHozoflnwxonumEAawccmwm

l'l'lIl‘IIll

IawusnlmlfluuvIol>xou©>£nm100cw.mflo mo ANmz on

.m

HUQUV Esuuommm mzz cououm

.HH¢ musmflm

146

Haa mucous

 

 

4:.

 

 

: a. .1. 2.. 2.. L. 2.: on so .1
1 1 1 J‘l 1 1 41 + T 1| 1 1 1 1 d 1 1
_ 4 1 a d — M 4 - _‘ 4 d M — d M 1 4 d M _ J4 _ J
m
u _
v .
u l .
_ .
l _
_

 

-~_—_’.._- ..-

._.__.._--.__ -—--— .--

0041’I'II

 

 

,.__——-...—- _-—--—~ _--— ——-

 

 

- on“. -—.—_——... “o- “cu-mm o“..-

. --—_~o~-—-¢-o-..-

 

 

 

 

14?

.WW mcmwplh.MImuoomo.m.mgoHo>oflQ

lullllI' I|

.Aacumexxoup>cuoI>x0HU>£Im10pcm.mflo mo Aux: 0mm

.m

Huoov Esuuowmm mzz cououm

.Nad wusmflm

r..-

w

1.44444444444

41!..-

148

 

 

@

mafi mnsmflm
v m
444444444 d 41444 4 4 <4444‘4I4'14J14lfi4l14’414l4ljsl14l14l1

\ift

 

 

 

 

 

 

149

.ww wcwflp|>.mumuoomo.m.onH0%oH£

 

 

Ifiamcqumxoupacvmflblm.m10©cw.mflo wo ANS: omm

.m

Hooov snuuommm mzz cououm

.mam musmflm

150

 

ma< ousmflm
o P N n
—y* P F > FL F FLr \PLL F LVL+LPL P LP PL \P P ?

v

@

Qm

P}Flrk—LVPFlrlrxrkrfihrPLrbeF*L*Lrer—VP**LFLIPrxPx—rL

_ m

.01

i,

“"\

hm

u.111111J—111111111d1141u4111—111141111‘1

is

\w'“

10.

 

 

 

 

 

151

.wew mamaonb.mumuoomo.m.onHosofln

ill!!!"

Iflaxnuwemxouwomvmanlo.mnoocw.mflo mo Ammz 0mm

.m

HUQUV Eduuowom mzz cououm

.QH< musmam

 

VHxN Ohsmflh

 

152

 

0.1

44441q1444d4444d

J a...

 

 

 

é

N
H
H

 

 

 

\.....

040

 

 

 

REFERENCES

10.

11.

12.

13.

14.

15.

REFERENCES

Maier, G.; Pfriem, S.; Schaffer, U., Matusch, R.,
Angew. Chem. Int. Ed. Eng. 1978, 11, 520.

 

 

 

 

 

Rauscher, G., Clark, T., Poppinger, D., Schleyer, P. v.R. f
Angew. Chem. Int. Ed. Eng. 1278, 11, 276. n

i
Eaton, P. B., Cole, T. W. Jr., J. Am. Chem. Soc. 1964, :
El 9620 .
Ternansky, R. J., Balogh, D. W., Paquette, L. A. 3
J. Am. Chem. Soc. 1282, 04, 4503. E

For a review on synthetic efforts directed towards
dodecahedrane, see Eaton, P. E. Tetrahedron 1312, 35,
2189.

 

Schulman, J. M., Venanzi, T., Disch, R. L. J. Am.
Chem. Soc. 1275, 91, 5335.

 

Gasteiger, J., Dammer, O. Tetrahedron 1978, 34, 2939.

 

Schulman, J. M., Disch, R. L. J. Am. Chem. Soc. 1978,
100, 5677.

 

Woodward, R. B., Fukunaga, T., Kelley, R. C. J. Am.
Chem. Soc. 196%, 86, 3162.

 

(a) Paquette, L. A., Kramer, J. D., Lavrik, P. B.,
Wyvratt, M. J. J. Org, Chem. 19 , 42, 503. (b) Bosse,
D., deMeijere, A. Angew. Chem. 7%, 86, 706.

 

 

Gladysz, J. A. Chem. Tech. 1272, 9, 372.

 

(a) Paquette, L. A., Itoh, I., Farnham, W. B. J. Am.
Chem. Soc. 1975, 97, 7280. (b) Paquette, L. A.,
Itoh, I., LipkowiEE, K. B. J. Org. Chem. 1916, 41, 3524.

 

 

Oljan, R. Ph.D. Thesis, Harvard University, 1976.

Roberts, W. P., Shoham, G. Tetrahedron Lett. 1981,
4895.

 

Eaton, P. B., Mueller, R. H., Carlson, G. R., Cullison,
D. A., Cooper, G. P., Chou, T.-C., Krebs, E.-P. J. Am.
Chem. Soc. 1911, 99, 2751.

 

153

16.

17.

18.
19.

20.

21.
22.
23.
24.
25.
26.
27.

28.

29.

30.

31.

154

Paquette, L. A., Snow, R. A., Muthard, J. L., Cynkow-
ski, T. J. Am. Chem. Soc. 1918, 100, 1600.

 

Woodward, R. B., Hoffmann, R. "The Conservation of
Orbital Symmetry", Verlag Chemie, GmbH, Weinham/Bergstr,
1970, p. 70.

Salomon, R. G., Kochi, J. K. J. Am. Chem. Soc. 1911, gg,
1137.

 

Salomon, R. G., Folting, K., Streib, W. B., Kochi, J.
K. J. Am. Chem. Soc. 1911, _9, 1145.

 

(a) Salomon, R. G., Ghosh, S., Zagorski, M. G., Reitz,
M. J. Org. Chem. 1, 41, 829. (b) McMurry, J. E.,
Choy, W. Tetrahedron Lett. 1980, 2477.

 

 

Rathke, M. W., Lindert, A. J. Am. Chem. Soc. ,
2;, 4605. 12%

 

 

Farnum, D. G., Hagedorn, A. A. III Tetrahedron Lett.

1915, 3987.

Hagadorn, A. A. III Ph.D. Thesis, Michigan State Uni-
veristy, 1974.

 

Jenkins, C. L., Kochi, J. K. J. Am. Chem. Soc. 1911,
99, 843.

 

Taken, in part, from Boeckelheide, V. Acc. Chem. Res.
12§QI lg, 65.

Mitchell, R. H., Bolkelheide, V. J. Am. Chem. Soc.
1911, 99, 1547.

Mitchell, R. H., Otsubo, T., Boekelheide, V. Tetra—
hedron Lett. 1915, 219.

 

 

 

 

(a) Otsubo, T., Boekelheide, V. Tetrahedron Lett. 1915,
3881, (b) Otsubo, T., Boekelheide, V. J. Org. Chem.

1311, 33, 1085.

For a review on the Ramberg-Backlund rearrangement, see
Paquette, L. A., Organic Reactions Vol. 25, John Wiley
and Sons: 1977, pp. 1-71.

 

 

Swepston, P. N., Lin, S.-T., Hawkins, A., Humphrey,
S., Siegel, S., Cordes, A. W. J. Org. Chem. 1981,
49, 3754. .

 

Buter, J., Kellogg, R. M. J. Org. Chem. 1981, 49, 4481.

 

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

155

Ochrymowycz, L. A., Mak, C., Michna, J. D. J. Org.
Chem. 1919, 9, 2079.

 

Farnum, D. G., Mohaddes, M., Monego, T. A. unpublished
results. For organo aluminum-mediated aliphatic
Claisen rearrangements, see Takai, K., Mori, T.
Oshima, K., Nozaki, H. Tetrahedron Lett. 1991, 3985.

 

Hassner, A., Reuss, R. H., Pinnick, H. W. J. Org. Chem.
1919, 99, 3427.

 

For example, see Baldwin, J. B., DeBernardis, J.,
Patrick, J. E. Tetrahedron Lett. 1919, 353.

 

Still, W. C. Mitra, A. J. Am. Chem. Soc. 1978, 00,
1927. “””” 7“

 

Still, W. C., McDonald, J. H. III, Collum, D. B.,
Mitra, A. Tetrahedron Lett. 1919, 593.

 

 

Hagadorn, A. A. III, Farnum, D. G. J. Org. Chem. 1911,
99, 3765.

Wittig, G., B611, W., Krfick, K.-L. Chem. Ber. 1991,
99, 2514.

 

(a) Conia, J.-M., Limasset, J.-C. Bull. Soc. Chim.
fig. 1997, 1936; (b) Greenwald, R. , Chaykovsky, M.,
Corey, E. J. J. Org. Chem. 12199933 9_, 1128, (c) Peter-
son, D. J. J. Org. Chem. 780; Chan, T. H.,

Chang, E., Vinokur, E. Tetrahedron Lett. 1910, 1137.

 

 

 

 

Schostarez, H., Paquette, L. A. J. Am. Chem. Soc.
1991, 103, 722.

 

Dauben, W. G., Walker, D. M. J. Org. Chem. 1991,
1103.

 

Brown, H. C. "Organic Synthesis Via Boranes", John
Wiley and Sons: New York, 1975.

Scouten, C. G., Brown, H. C. J. Org. Chem. 1913, _§,
4092.

 

 

Brown, H. C., Zweifel, G. J. Am. Chem. HSoc. 199,
1512; Zweifel, G., Nagase, K., Brown,

s93. 129%, g3, 190.

Hooz, J., Gilani, S. S. H. Can. J. Chem. 1999, 99, 86.

8Chem.

 

 

Urquhart, G. G., Gates, J. W. Jr., Connor, R. Org.
Syn., Coll. Vol. 9, 1999, 363.

 

 

48.
49.
50.
51.
52.
53.
54.

55.

56.

57.
58.

59.

60.

61.

62.

63.

64.

65.

156

Farnum, D. G., Gur, Y. T., unpublished results. See
also Luche, J. L. J. Am. Chem. Soc. 1919,100, 2226.

 

Still, W. C. J. Am. Chem. Soc. 1919,100, 1481.

 

Wakefield, B. J. "The Chemistry of Organolithium Com-

pounds", Pergamon Press Ltd.: Oxford, 1974, pp. 89-
108.

 

Broaddus, C. D. J. Org. Chem. 1999, 99, 4431.

Veefkind, A. H., Bickelhaupt, F., Klump, G. W. Rec.
Trav. Chim. Pays-Bas 1999, 99, 1058.

-. land.”

 

Mulvaney, J. B., Gardlund, Z. G. J. Org. Chem. 1999,
99, 917.

 

Veefkind, A. H., Schaaf, J. v.D., Bickelhaupt, P.,
Klumppr G. W. Chem. Comm. 1911, 722.

 

 

Moriarty, R. M., Yeh, C. —L. Tetrahedron Lett.1979
383.

 

(a) Baldwin, J. B., DeBernardis, J., Patrick, J. E.
Tetrahedron Lett. 1919, 353; (b) Rautenstrauch,v.
Chem. Comm. 1919,

 

 

Baldwin, J. B., Patrick, J. E. J. Am; Chem. Soc. 1911,
99, 3556.

 

Lansbury, P. T., Pattison, V. A. J. Org. Chem. 1999,
91, 1933.

 

Cazes, B., Julia, S. Tetrahedron Lett. 1919, 2077.

 

Nakai, T., Mikami, K., Taya, S., Kimura, Y., Mimura,
T. Tetrahedron Lett. 1991, 69.

 

Baldwin, J. B., Urban, F. J. Chem. Comm. 1919, 165.

 

Evans, D. A., Andrews, G. C. Acc. Chem. Res. 1919, Z,
147.

 

(a) Baldwin, J. B., Hackler, R. E., Kelleyl, D. P.
Chem. Comm. 1999, 1083; (b) Greico, P. A., Meyers, M.,
Finkelhor, J. . Ogg. Chem. 1919, 99, 119.

 

 

Warpehoski, M. A., Chabaud, B., Sharpless, K. B.
J. Org. Chem. 1999, 91, 2897.

 

Seyferth, D., Weiner, M. A. J. Am. Chem. Soc.
84, 361. 128%

 

66.

67.

68.

69.

70.

71.

72.

157

 

Seyferth, D., Vick, S. C. J. Organomet. Chem. 1978,
144, 1. “ ”m

(a) Crandall, J. K., Clark, A. C. J. Org. Chem. lgzg,
31, 4236; (b) Crandall, J. K., Clark, A. C. Tetra-
hedron Lett. lgég, 707.

 

 

Gordon, A. J., Ford, R. A. "The Chemists Companion",
John Wiley & Sons: New York, 1972, p. 379.

Still, W. C., Kahn, M., Mitra, A. J. Org. Chem. ,
33, 2923. $2Z§

 

Watson, 8. C., Eastham, J. F. J. Organomet. Chem. $967,
9, 165.

 

Seyferth, D., Andrews, S. B. J. Organomet. Chem.
we 3_o. 151-

Hayashi, K., Iyoda, J., Shiihara, I. J. Organomet.
Chem. lgéz, lg, 81.