'RARTI _
111E PHOTOCHEMTSIRY’ DH 3 4,: 5 E,

E 11753331333350ch {3 1 01119131232 ONE

PARTII‘

THE ACID CATALYZED REARRANEEMENTS 0E)1,3, 4,3 3 i *

¥HEXAMETHYLBICYCLD [3. 1. D] HEX-3 EN 2 ONE AND 2, 3, 45 '7

8, 6-HEXAMETHYL- 2, 4 CYCLOHEXADIENONE

T116313 for the Degree of Ph D
MICHIGAN STATE UNIVERSITY
DAVID WILLIAM SWATTON

1967 '

 

 
  
 

  

LIB RA R Y
Michigan State

Universé
1.!

H E313

 
  

This is to certify that the

thesis entitled
Part I - THE PHOTOCHEMISTRY OF l,3,4,5,6,6—HEXAMETHYLBICYCLO—
[3.1.0]HEX-3—EN—2-ONE
Part II — THE ACID—CATALYZED REARRANGEMENTS OF 1,3,4,5,6,6-
HEXAMETHYLBICYCLO [3 . l . 0 ] HEX-3-EN—2-ONE AND
2,3,4,5,6,6-HEXAMETHYL-2,4—CYCLOHEXADIENONE

presented by

David William Swatton

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Chemistry

 

QC’LA/C‘" ‘ 63 5’5 5'3.an

‘ Major professor

 

Date August 4; 1967

0469

RY
State

 

ABSTRACT

PART I

THE PHOTOCHEMISTRY OF i,5,4,5,6,6-HEXAMETHYL-
BICYCLO[5.1.0]HEX-5-EN-2-ONE

PART II
THE ACID-CATALYZED REARRANGEMENTS OF
1,5,4,5,6,6-HEXAMETHYLBICYCLO[5.1.0]HEX-5-

EN-Z-ONE AND 2,3,4,5,6,6-HEXAMETHYL-
2,4-CYCLOHEXADIENONE

by David William Swatton

The purpose of the first part of this thesis was to
investigate the photochemical behavior of 1,5,4,5,6,6—hexa—
methylbicyclo[5.1.0]hex—3—en—2—one [£31‘

Photolysis of §§ in methanol at 00 through Pyrex gave
the crystalline methoxyenol §§J which is thought to result
from reaction of a dipolar intermediate §;_with a molecule
of methanol. The photolysis could be sensitized by benzo—

phenone. Compound 56, on warming or treatment with dilute

O 0 OH
CH30
hv
3 CHQOH:
34 61 56

acid, loses methanol to form the enolic triene $91 which,

on further treatment with acid, gives dienone 55.

David W. Swatton

The mechanistic course of the photochemical and acid-

OH O

+
l ‘ ___Ji______>

4O 55
catalyzed rearrangements was established by using 34 labeled
with CD3 groups at various positions.

The photoproduct 56 was converted by basic methanol

to its keto form éé, which underwent thermal elimination of

methanol to Ogive dienone 46. Enol 40 could be thermally

isomerized to its keto form 41, isomeric with dienone 4gp
Irradiation of bicyclic ketone éé in ether afforded the
enolic triene $9: This is also proposed to result from
dipolar intermediate ,6; . Photolysis of §4 in acetic acid
gave dienone_§§, considered to result from the protonated
form of intermediate §$' The general significance of these
results with respect to the observed photochemical reactions

of bicyclo[5.1.0]hexenones is discussed.

David W. Swatton

The second part of this thesis is concerned with the
acid—catalyzed and thermal reactions of bicyclic ketone éé,
and the behavior of fully substituted cyclohexadienones in
acid media.

Treatment of g4 with 97% sulfuric acid gave dienone_§§
in excellent yield. The course of the acid-catalyzed re-
arrangement was established using éé variously labeled with
CD3 groups. The labeling results are in agreement with a
mechanism in which éé undergoes a cycloprOpylcarbinyl re-
arrangement to ion §§, which equilibrates rapidly with its

enantiomer 68a prior to ring opening to dienone éé; The

/ 6: 0

AZ; 68 68a

0

results also show that ion Qgican revert to bicyclic ketone
‘Qg, This is a novel type of rearrangement, not previously
obtained from bicyclo[5.1.0]hexenones. The implications of
these results for the acid-catalyzed reactions of bicyclo—
[5.1.0]hexenones are discussed.

Pyrolysis of bicyclic ketone ééialso led to dienone éé.
Labeling studies showed that a methyl migration was not in—

volved, but rather a mechanism which consisted of breakage

David W. Swatton

 

54

of bond 2, formation of a bond between the carbonyl carbon
and the quaternary carbon at C—6, and subsequent fission of
bond'g.

Treatment of dienone éfi with fuming sulfuric acid gave
dienone §§ in good yield. Dienone §§ is stable in concen-
trated sulfuric acid, but treatment of hexadeutero-éé,
labeled with CD3 in the C-3 and C-5 positions, with fuming
sulfuric acid led to complete scrambling of the label due

to a series of methyl migrations.

PART I

THE PHOTOCHEMISTRY OF l,5,4,5,6,6-HEXAMETHYL-
BICYCLO[3.1.0]HEX-3-EN-2-ONE

PART II

THE ACID-CATALYZED REARRANGEMENTS OF
1,5,4,5,6,6-HEXAMETHYLBICYCLO[5.1.0]HEX-3-
EN-Z-ONE AND 2,5,4,5,6,5-HEXAMETHYL-
2,4-CYCLOHEXADIENONE

BY

David William Swatton

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

1967

 

Dedication

This thesis is dedicated to my wife,
Tina, for her patience and invaluable assist—
ance in perfecting this manuscript.

ii

ACKNOWLEDGMENT

The author wishes to express his sincere appreciation
to Professor Harold Hart for his invaluable guidance and
encouragement during the course of this study.

Appreciation is due to Doctors Peter M. Collins and
Richard M. Lange for many fruitful discussions and sugges-
tions.

Appreciation is extended to the National Science
Foundation for financial support from June, 1965 to June,
1966 and from January, 1967 to August, 1967.

Appreciation is also extended to the National Institutes
of Health for financial support from September, 1966 to

December, 1966.

iii

 

TABLE OF CONTENTS

Page
PART I
THE PHOTOCHEMISTRY OF 1.5.4.5,6.6-HEXAMETHYL-
BICYCLO[5.1.0]HEX-5-EN-2-ONE
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . 2

RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . 11

A. Photolysis of Hexamethylbicyclo[5.1.0]hex-3—
en-2-one {£4) in Methanol. . . . . . . . . . . 11

1. The Keto Forms of 2,5,3,4,5,6—Hexamethy1-
6-methoxycyclohexa-1,4-dien-1-ol (fig) and
2,5,5,4,5-pentamethy1—6-methylenecyclo—
hexa-1,4-dien—1—ol (49). . . . . . . . . . 17

2. Photolysis of Labeled Bicyclo[5. 1. O]-
hexenones. . . . . . . . . . . . . . . . 21

5. The Photosensitized Conversion of Hexa-
methylbicyclo[5.1.0]hex—5-en-2—one (Q4) to
2,5,5,4,5,6-Hexamethyl-6-methoxycyclohexa—
1,4-dien-1-ol (Q6) . . . . . . . . . . . . 24

4. The Mechanistic Significance of the Photo-
chemical Rearrangement of Hexamethylbi-
cyclo[5.1.0]hex—5-en-2-one (Q4). . . . . . 25

B. The Photochemistry of Hexamethylbicyclo-
[5.1.0]hex—5—en-2-one (Q4) in Diethyl Ether. . 54

C. The Photochemistry of Hexamethylbicyclo—
[5.1.0]hex-5—en-2—one (Q4) in Acetic Acid. . . 57

EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . . 41
A. General Procedures . . . . . . . . . . . . . . 41

B. General Photolysis Procedures. . . . . . . . . 41

iv

TABLE OF CONTENTS - Continued

C. Photolysis of 1,5,4,5,6,6-Hexamethylbicyclo-

D.

[5.1.0]hex-S-en-2-one £Q4) in Methanol. . .

The Photosensitized Reaction of Hexamethyl-
bicyclo[5.1.0]hex-5-en-2-one (54) . . . . .

The Dark Reaction of Hexamethylbicyclo-
[3.1.0]hex—5-en-2-one (éé) in Methanol. . .

2,5,3,4,5-Pentamethyl-6—methylenecyclohexa-
1,4-dien-1-ol (g’g). . . . . . . . . . . . .

Acid-Catalyzed Conversion of 2,5,5,4,5,6-
Hexamethyl-6-methoxycyclohexa-1,4-dien-1-ol
(Q6) and 2,5,5,4,5—pentamethyl-6-methylene-

cyclohexa-1,4-dien-1—ol (49) to 2,3,4,4,5,6-

Hexamethyl-Z,5-cyclohexadienone (§§).

Preparation of 2,4,4,6-Tetramethyl-5,5-bis-

(trideuteromethyl)-2,5-cyclohexadienone (44)

2—Methoxy-2,3,4,5,5,6-hexamethylcyclohex-5-
en-1-one (45) . . . . . . . . . . . . . . .

2,5,5,5,6-Pentamethyl-4-methylenecyclohex-Z-

en—1-one (46) . . . . . . . . . . . . . . .

5,4,5,5,6-Pentamethyl—2-methylenecyclohex-S-

en—1—one (4J9 . . . . . . . . . . . . . . .

Preparation of 2,4,6,6-Tetramethy1-3,5~bis-

(trideuteromethyl)—2,4—cyclohexadienone (4&0.

Photolysis of 3,5,6,6-Tetramethy1-1,4-bis—
(trideuteromethyl)bicyclo[5.1.0]hex-3-en-2-
one (Egg. . . . . . . . . . . . . . . . . .

Photolysis of 2,5,4,4,5,6-Hexamethyl—2,5—
cyclohexadienone (ég) in Ether. . . . .

Photolysis of 2,4,4, 6-—Tetramethyl-—5, 5—bis-

(trideuteromethyl) -2, 5— —cyclohexadienone (44)

in Ether. . . . . . . . . . . . . . . . .

. Degradation of 1, 5, 6, 6— ~Tetramethyl- -4, 5-bis-

(trideuteromethyl)bicyclo[3. 1. O]hex-5- en- -2-
one (54).. . . . . . . . . . . . .

Page

42

44

45

45

47

49

49

50

55

55

55

56

56

TABLE OF CONTENTS - Continued Page

Q. Photolysis of 1,5,6,6-Tetramethyl-4,5-bis-
(trideuteromethyl)bicyclo[5.1.0]hex—5—en-
2-one (54) in Methanol. . . . . . . . . . . .

R. Photolysis of 1,5,4,5,6,6-Hexamethylbicyclo-
[5.1.0]hex—5-en-2-one (égj in Diethyl Ether

S. Photolysis of 1,5,4,5,6,6-Hexamethylbicyclo-
[5.1.0]hex-5-en—2—one (353) in 45% Acetic
Acid. . . . . . . . . . . . . . . . . . . .

SUMMARY . . . . . . . . . . . . . . . . . . . . . . .

PART II
THE ACID-CATALYZED REARRANGEMENTS OF
1,5,4,5,6,6-HEXAMETHYLBICYCLO[5.1.0]HEX-5-
EN-Z—ONE AND 2,5,4,5,6,6-HEXAMETHYL-
2,4-CYCLOHEXADIENONE
INTRODUCTION. . . . . . . . . . . . . . . . .
RESULTS AND DISCUSSION. . . . . . . . . . . . . . . .

A. Acid-Catalyzed Rearrangement of Labeled
Bicyclo[5.1.0]hexenones . . . . . . . . . . .

B. Discussion of the Rearrangement Mechanism . .
C. Miscellaneous Experiments . . . . . . . . . .

1. Acid-Catalyzed Rearrangements of Fully
Substituted Cyclohexadienones . . . . . .

a. The Acid-Catalyzed Rearrangement of
2,5,4,5,6,6-Hexamethyl-2,4—cyclo-
hexadienone (Qé) . . . . . . . . . .

b. The Ultraviolet Spectrum of 2,5,4,5,
6,6-Hexamethyl—2,4-cyclohexadienone
(QQ) in Sulfuric Acid. . . . . . . .

c. The Behavior<3f 2,5,4,4h5,6-Hexamethyl-

2,5—cyclohexadienone (gg) in Sulfuric
acid . . . . . . . . . . . . . . . .

vi

57

57

58
6O

65
68

7O
81

88

88

9O

92

95

TABLE OF CONTENTS - Continued

d. Acid-Catalyzed Rearrangement of
2,4,4,6-Tetramethy1-5,5-bis(tri-
deuteromethyl)—2,5—cyclohexa-
dienone (44). . . . . . . . . . .

2. Pyrolysis of 1,5,4,5,6,6-Hexamethylbi—
cyclo[5.1.0]hex-5-en-2-one géé)’ . . .

EXPERIMENTAL . . . . . . . . . . . . . . . . . . .

A.

Acid-Catalyzed Rearrangement of 1,5,4,5,6,6-

Hexamethylbicyclo[5.1.0]hex—5-en-2-one (54).

1,5,5,4,7,8—Hexamethyl—5,6-dicarbomethoxy-
bicyclo[2.2.2]octa-5,7—dien—2-one (98) . .

Acid-Catalyzed Rearrangement of Labeled

Bicyclo[5.1.0]hexenones. General Procedure.

Acid—Catalyzed Rearrangement of 5,5,6,6-

Tetramethyl-1,4-bis(trideuteromethyl)bicyclo-

[5.1.0]hex-5-en-2-one (59) . . . . . . .

Treatment of 2, 4, 6, 6- -Tetramethyl- -5, 5- di-

methyl— ds- -2, 4- -cyclohexadienone (49) with 97%

Sulfuric Acid. . . . . . . . . . . .

Preparation of 1,5,5,6,6-Pentamethy1-4—

methyl-ds-bicyclo[5.1.0]hex-5-en—2-one (102)

Acid-Catalyzed Rearrangement of 1,5,5,6,6-
Pentamethyl-4-methyl-d3—bicyclo[5.1.0]hex-
5—en—2-one (102) . . . . . . . . . . . . .

2, 5, 4, 4, 5, 6- -Hexamethyl- -2, 5- -cyclohexadienone

(55) . . . . . . . . . . . . . . . . . . .

Kinetics of the Acid- -Catalyzed Rearrangement

of 2, 5, 4, 5, 6, 6- -Hexamethyl- 2,4-cyclohexadi-
enone (55) . . . . . . . . . . . . . .

Measurement of the pKa of 2,5,4,4,S,6—Hexa—

methyl-2,5—cyclohexadienone £55) . . . . .
Rearrangement of 2,4,4,6-Tetramethyl-5,5-

bis(trideuteromethy1)-2,5-cyclohexadienone
(44) in Fuming Sulfuric Acid . . . . .

vii

Page

94

100

104

104

104

105

105

106

109

109

111

111

115

114

TABLE OF CONTENTS - Continued Page

L. NMR Study of 2, 4, 4, 6— —Tetramethyl- -3, 5-bis—
(trideuteromethy1)-2, 5-cyclohexadienone (44)
in 70% and 98% Sulfuric Acid. . . . . . . . . 116
1. 70% Sulfuric Acid . . . . . . . . . . . . 116
2. 98% Sulfuric Acid . . . . . . . . . . . . 116

M. Pyrolysis of 1, 3, 4, 5, 6, 6- -Hexamethylbicyclo-
[3.1. 0]hex-3-—en-—2-one (34). . . . . . . . . 116

N. Pyrolysis of 3, 5, 6, 6- -Tetramethyl- -1, 4- bis-
(trideuteromethyl)bicyclo[3. 1. 0]hex- 3- -en— -2—
one (50).. . . . . . . . . . . . . . . . 117
SUMMARY . . . . . . . . . . . . . . . . . . . . . . . 118

LITERATURE CITED. . . . . . . . . . . . . . . . . . . 120

viii

TABLE

II.

III.

IV.

VI.

VII.

VIII.

LIST OF TABLES

Page
NMR Spectra. . . . . . . . . . . . . . . . . . . 13
NMR Spectra. . . . . . . . . . . . . . . . . . . 72

Ultraviolet Spectrum of 2,3,4,5,6,6-Hexamethyl-
2,4-cyclohexadienone (33) in Sulfuric Acid . . . 92

Ultraviolet Spectrum of 2,3,4,4,5,6—Hexamethy1—
2,5-cyclohexadienone (35) in Sulfuric Acid . . . 94

NMR Spectrum of 2,4,4,6-Tetramethyl—3,5-bis-
(trideuteromethyl)-2,5-cyclohexadienone:(4&3 in
70% Sulfuric Acid at 250 . . . . . . . . . . . . 96

NMR Spectrum of 2,4,4,6-Tetramethyl-3,5-bis-
(trideuteromethyl)-2,5—cyclohexadienone (239 in
98% Sulfuric Acid at 250 . . . . . . . . . . . . 97

Extinction Coefficients of 2,3,4,4,5,6—Hexamethy1-
2,5-cyclohexadienone K35) in H2804 + H20
Mixtures . . . . . . . . . . . . . . . . . . . . 115

Extinction Coefficients and pKa Values for

2,3,4,4,5,6—Hexamethy1—2,5-cyclohexadienone (35)
in H2504 + H20 Mixtures. . . . . . . . . . . . . 115

ix

LIST OF FIGURES

FIGURE Page
1. Principles modes of bicyclohexenone isomeriza-
tion. . . . . . . . . . . . . . . . . . . . . . 4
2. Nmr spectrum of 2,3,3,4,5,6-hexamethyl-6-
methoxycyclohexa—l,4-dien—1-ol (36) . . . . . . 43
3. Nmr spectrum of 2,3,3,4,5-pentamethy1-6-
methylenecyclohexa-1,4-dien-1-ol (4£Q . . . . . 46
4. Nmr spectrum of 2,3,4,4,5,6-hexamethyl-2,5—
cyclohexadienone (égg . . . . . . . . . . . . . 48
5. Nmr spectrum of 2-methoxy-2,3,4,5,5,6-hexa—
methylcyclohex~3-en-1-one (45). . . . . . . . . 51
6. Nmr spectrum of 2,3,5,5,6-pentamethyl-4-
methylenecyclohex—2—en-1-one (46) . . . . . . . 52
7. Nmr spectrum of 3,4,5,5,6-pentamethyl-2-
methylenecyclohex—3-en-1—one (47) . . . . . . . 54

10.

11.

Nmr spectrum of hexadeuterated dimethyl
acetylenedicarboxylate adduct 100 . . . . . . . 107

Nmr spectrum of hexadeuterated bicyclic ketone
108 . . . . . . . . . . . . . . . . . . . . . . 108

Nmr spectrum of trideuterated bicyclic ketone
112 . . . . . . . . . . . . . . . . . . . . . . 110

Nmr spectrum of trideuterated dimethyl acetyl-
enedicarboxylate adduct 114 . . . . . . . . . . 112

PART I

THE PHOTOCHEMISTRY OF 1,3,4,5,6,6-HEXAMETHYL-
BICYCLO[3.1.0]HEX-3-EN-2-ONE

INTRODUCTION

The photochemical transformations of bicyclo[3.1.0]-
hex-B-ene-Z-ones have received considerable attention within
the past decade, and have been extensively reviewed (1,2).
The most commonly involved rearrangement is C-1, C—5 bond
fissiOn followed by skeletal transformations which furnish
isomeric phenols or dienones. A number of representative

examples follow:

hV
O 1 Neat H0

v

Wheeler and Eastman (3) (1959)

 

O OH
6
hV >_
¢ Hgo—Dioxane‘r .
14¢
fl
9.. A.

Zimmerman and Schuster (4)_(1961)

hv __
Ether

 

Chapman (5) and Fisch (6) (1963)

hv \_0
45% HOAC 7

 

   

Schuster and Fabian (7) (1966)

The mechanistic scheme presented in Figure 1 has been
proposed to account for the principle modes of bicyClohex-
enone isomerization: (1) n—->W* excitation(§~-—e>1£Q;

(2) 1,5 bond fission (1Q_—-e»11); (3) electron demotion to
a mesoionic ground state (4$.“€’i§)7 and (4) skeletal re—

arrangement (1§'-é>1§)-

Representation of rearranging intermediates as mesoionic

ground state species (12) has gained general acceptance in
interpretation of bicyclohexenone photochemistry. The simi-
larity of photochemical to ground state carbonium ion pro—

cesses suggest that such intermediates are involved.

.mm .m ..w .z .xuom 3mz .omma
..UGH .mcom paw mmHHB GSOb ..Um ..Hb .muuflm .z .b Ucm .Ucofifimm .m .w
..Hb .mmwoz .< .3 .w .HO> :.>HumHEm£Uou0£m GH mmocm>p¢c .Hmcwmmsom .M

Eoum UmOSUOHmmmV .coaumNHHmEOmH mcocmxmsoHoonQ mo mmUoE mamflocflum

 

 

 

.fi musmflm
.mfi \mm =ma
:m :m
um um
\ /
m I \
AEHMV ...m
u.-
0 mo
ma .mm 5%
:m 3m :m
m m m-
cm - um
ATIIIII AIIIIII
>3
ucom .-.m

on -..m

 

Jeger and co-workers have shown (8) that 1-dehydro-
testosterone acetate photoisomer 14, on treatment with either
acid or ultraviolet light gives a mixture of dienones 16 and
11: These authors suggest that a similar intermediate, 1gb

is involved in both reactions.

\ 0"}i\\x
\

  

 

14 15

W

. 0“

 

16 17

M M

The intermediacy of ionic species has also been sug—
gested by Kropp, who showed that irradiation of ketone 18
in acetic acid gave phenol Eivas the only product (9).
The proposed sequence 1§’——>xggr-—)»21’has ample analogy
in the series of rearrangements 23I-a>3§4,—%>_§§~pr0posed by
Davis and Halsall to intervene in the acid-catalyzed trans-
formation of hydroxyketone gg’to phenol 2§_(10). It was
therefore suggested that similar intermediates are involved

in both of the above transformations.

 

HO

20

OH

22
HO

, \

24

M

 

l
HO '
hv >: -———e»
HOAc I
I
l

19
H0

21

W

 

V

HO WH\\
H+
———%>

23

m

 

\/

110 X
D.

25

M

Bicyclohexenone rearrangements depend markedly on the

solvent employed.

Kropp proposed the mechanistic scheme

presented below to account for the photochemical behavior of

ketone 26 in neutral and acidic media (11).

 

 

O. h"

0' O
331 27 28

 

H0

29 30

M M

Irradiation of géiin methanol gave the isomeric dienone 28.
in 57% yield, whereas photolysis in aqueous acetic acid
furnished 47% of 2§,and 33% of phenol 39; The author states
that in neutral media there is minimum charge separation in
ion 21; hence localization of positive charge adjacent to
the electron rich oxygen causes preferential rearrangement
to 28: In acidic media, however, the ionic species may be
protonated on oxygen thereby reducing this electronic stabi—
lization factor. Ion §§_may therefore form products gévand
39—with approximately equal facility.

Zimmerman and co—workers have invoked a similar argument
to account for the solvent dependence observed in photochemi-

cal transformations of ketone_3_(12).

In addition to being dependent on solvent, further re—
arrangements of species 12 (Figure 1) are influenced by the
relative migratory aptitudes of R and R"in the order of

C (1). Examples of this are the re—

CH>CH2>CH3>C
ported conversions Z~——9:§_and ;§.__+R§$3 The hypothetical
dipolar species 31 derived from lumisantonin (5), however,‘
undergoes methyl migration to give dienone §J rather than
rearrangement through the spiro intermediate as is normally
preferred. Kropp has suggested that this is due to the
presence of the trans-v—lactone ring, since formation of the
spiro intermediate_§2’would require the highly strained

trans fusion of two five—membered rings (13).

——->

 

It was discovered earlier in our laboratories that
irradiation of dienone 33 in ethanol gave bicyclohexenone 34

in excellent yield (14,15). Further irradiation of 34 in

_N~

this solvent, however, gave dienone 35 in good yield.

0 O O
Ethanol
33 34 35

M M M

Although a few examples of photochemical conversions of bi—
cyclohexenones to 2,5-cyclohexadienones are known (7,8), this
result was somewhat surprising, since the reverse reaction

is much more common. Indeed, it was observed that irradia-
tion of 35 in ether gave 34’(14).

The possibility of an isolable intermediate in the photo-
isomerization of 34 to_3§ led us to carry out a careful study
of the photolysis of this ketone in methanol (chosen in
preference to ethanol in order to simplify the detection of
intermediate species by nmr).

The results of this study, presented in Part I of this
thesis, show that the transformation of ééuto‘§§_is not
direct, but involves the enolic ether 36! which is converted
to §§_yia_a series of acid-catalyzed rearrangements. It is
proposed that_§6 is formed by reaction of a dipolar inter—
mediate analogous to 1§_(Figure 1) with a molecule of methanol.
The photochemical behavior of 34 in ethyl ether and aqueous

acetic acid is also described. The mechanistic implications

10

of the above reactions are discussed.

CH30

36

RESULTS AND DISCUSSION
A. Photolysis of Hexamethylbicyclo[3.1.0]-
hex-3een—2-one (34) in Methanol
A solution of 1,3,4,5,6,6-hexamethylbicyclo[3.1.0]—
hex—3-en—2-one (34) (15) in methanol which was freshly dis—
tilled from sodium methoxide to remove any trace of acid was
irradiated at 00 with light of A > 280 mu. The reaction was
monitored by the change in.the ultraviolet spectrum of the
solution. As the photolysis proceeded, bands due to 34 at
234 and 274 mu decreased in intensity and the ultraviolet

region became nearly transparent. Filtration of the solu-

tion afforded a 76% yield of a white crystalline photoproduct

which was assigned structure 36 (14). The product could be
0 OH
CH30
hv E
CH30H
32.4: 2:5.

stored for extended periods of time at ~800, but formed a
viscous yellow oil on standing overnight at room temperature.
Freshly prepared samples had a melting point range of 58—650
and gave a satisfactory microanalysis.

The structure of 36Ifollows from its analysis, spectral

properties, and chemical transformations. Low wavelength

11

 

 

12

ultraviolet absorption at 202 mu (6 4,300) agrees with the
value of AfiggH 195 mu (6 7,600) reported for 1,2-dimethyl-
1,4-cyclohexadiene (16). A structurally similar compound
which has also been reported to have intense end absorption
is 3,3,6,6-tetramethyl-1,4—cyclohexadiene (17). The two
mechanistically feasible isomers, élvand 38 can therefore

be excluded since these compounds should exhibit intense

end absorption at about 278 mu (18). The infrared spectrum

OH OH
OCH3
OCH8
:22. 58

of 36 exhibited no carbonyl band but did contain a hydroxyl

1 consistent with its enolic

stretching frequency at 3555 cm"
structure. The nmr spectrum of the photoproduct also sup-
ports the assigned structure (see Table I). That the
spectrum contains three allylic methyls eliminates the hemi-
ketal structure galwhich could also be postulated for the
photoproduct. The assignment for the three allylic methyls
presented in Table I is based on labeling studies described

OH
OCH3

39

13

Table I. NMR Spectraa

 

 

 

Compound Chemical Shift (J)b AssignmentsC
7.97 (q, 0.8) C-3 methyl
8.14 (m) c- -2,c- -4, c- 5 methyls
8.89(s) C—6 methyls
o 8.12 (q, 0.9) c-4 methyl
2 8.45 (q, 0.9) C-3 methyl
1 8.78 (s) C-5 methyl
3 6 8.90 (s) c-1 methyl
4 8.90 (s) C-6 exomethyl
9.08 (s) C—6 endo methyl
34
8.05 (q, 0.8) C- 3, -5 methyls
8.17 (q, 0.8) c- 2, -6 methyls
8.79 (s) C- 4 methyls
5.75 (s) d Hydroxyl proton
7.13 (s) 7.11 (s)d Methoxyl
8.33 (s) 8.30 (s) dC—2 methyl

8.33 (q, 0.9) 8.30(q,0.9)dC-4 methyl
8.58 (q, 0.91 8.37(q,0a9) c-s methyl
C-

8.75 (s) 8. 70 (s)d 6 methyl (?)
8.92 (s) 8. 90 (s) C-3 methyls(?)
8.96 (s) 8. 95 (s)d

5.05 (5) Vinyl proton

(cis to hydroxyl?)

5.20 (8) Vinyl proton
8.18 (s) 8.18 (s)§ c-4,c—5 methyls
8.23 (s) 8.22 (s)d C-2 methyl

8.89 (s) 8.88 (s) C-3 methyls

 

continued

1'

 

 

14

 

 

 

Table I -— continued

Compound Chemical Shift (J)b AssignmentsC
7.10 (s) Methoxyl
7.28 (q, 7.0) C-6 proton
8.27 (q, 0.9) C—3,C—4 methyls
8.57 (q, 0.9)
8.73 (s) C—2 methyl (?)
8.85 (s) C—5 methyl (2)
9.01 (d, 7.0) C—6 methyl
9.22 (s) C-5 methyl (?)
4.74 (5) Vinyl protons
4.90 (s)
7.85 (q, 7.0) C—6 proton
7.98 (s) C-3 methyl
8.20 (s) C-2 methyl
8.90 (s) C—5 methyls
8.97 (s)
9.07 (d, 7.0) C-6 methyl
4.28 (8) Vinyl proton

(cis to hydroxyl?)

4.97 (8) Vinyl proton
8.13 (s) C—3,C—4 methyls
8.89 (s) C—5 methyls
9.07 (S)
8.96 (d, 7.0) c—e methyle

 

 

aAll spectra are in CCl4 except as noted.

Shifts are reported as T values, with TMS as an internal
reference. J's are in Cps. All spectra were run at 60 Mc.
All areas are consistent with the assignments. Multiplicity
of peaks shown in brackets: s, singlet; d, doublet; q,
quartet; m, multiplet.

CA question mark indicates that the assignment is tentative.
Chemical shifts observed in methanol.

eThe C-6 proton responsible for the splitting of this methyl
was lost in noise and could not be located with certainty.

 

15

below. Evidence that the methyl groups at T 8.30 and 8.37
are adjacent is the observed homoallylic coupling of 0.9
cycles between them. Further proof for the enolic structure
is the band for the hydroxyl proton of 3§Jobserved in CCl4
solution and the absence of a methyl group split by an
d—hydrogen (compare with spectrum of ketone 45). A definite
assignment for the three aliphatic methyls cannot be made,
although it is likely that the one at lowest field is bonded
to the carbon bearing the methoxyl group.

When a solution of 36/in CCl4 was warmed to 350 for a
few minutes, a band appeared in the nmr spectrum at T 6.62
due to formation of methanol. This was accompanied by other
changes in the spectrum; resonance bands due to éé‘decreased
in intensity and a new series of bands was observed. These
changes are attributed to 1,2—elimination of methanol from
36, leading to enol 40.
mm, c-

OH

40

M

This compound could be isolated as a yellow oil with a lfi::H

252 mu (€ 13,400). This is in good agreement with the uv

spectrum of the related triene 41 (15,19) which has a

xgzgoctane at 256 mu (8 21,400). Isomers 42 and 45, which

 

 

16

OH OH

41 42 V 45

M M M

might result by the 1,4-elimination of methanol from 36! can
therefore be excluded since they should exhibit intense
ultraviolet absorption at 313 mu (18). The nmr spectrum of
49_(Table I) is also quite similar to that of triene 4;;
which consists of three singlets at T 5.23, 8.22, and 8.91
with relative areas 1:6:3 (15). The observation that one of
the vinyl protons in 49 exhibits a chemical shift close to
that observed for those of 41’suggests that this proton is
trans to the hydroxyl group. The assignment for the allylic
methyls presented in Table I is based on labeling experi-
ments described below. An infrared spectrum of égvexhibited
a strong hydroxyl stretching bandzuz3450 cm-1 and no carbonyl
absorption, thus confirming the enolic structure.

Treatment of either §§/or égdwith dilute hydrochloric

acid gave dienone 35 (20) in excellent yield. The structure

of this compound is based on its spectral data and micro—

0

'55

M

analysis. It exhibited infrared bands at 1653 and 1624 cm‘1

 

 

 

17

(C = O and C = C) and lfiggH at 246 mu (6 14,800), consistent
with the 2,5-cyclohexadienone structure. The nmr spectrum

of 3§_(Table I) is also in agreement with the symmetrical
structure. The low-field band is assigned to the B-allylic
methyls, which are observed to undergo coupling of about
0.8 cycles with the d-allylic methyl protons, and can be

easily exchanged for deuterium (35 ——*'44) by refluxing the

 

dienone in basic deuteromethanol. In agreement with the

0 0

NaOCHa §_

CH30D

Ds
CD3

35 44
M M

above assignment, 44’has no nmr band due to the B-methyls
and the band attributed to the d—methyls is a sharp singlet.

Experimental evidence that éé’is not converted directly
to 35 by acid, but first forms‘49i was obtained by following
the change in ultraviolet spectrum when a trace of acid was
added to a solution of 3§din methanol. The absorption
maximum shifted rapidly from 202 mu (36) to 252 mu (49) then
changed to 246 mu (35) on adding more acid.

1. The Keto Forms of.2,3,3,4,5,6-Hexamethy1-6-
methoxycyclohexa-l,4-dien-1-ol(36) and
2,3,3,4,5-pentamethyl-6-methylenecyclohexa—

1, 4-dien-1-ol (4.0)
In order to provide further chemical support for the

structures of the photOproduct 36, and its elimination

 

18

product 42; an attempt was made to convert each enol to its
keto form.

A solution of é§,in methanol was treated at 00 for 0.5
hr with 0.2 N sodium methoxide in methanol. Extraction with
methylene chloride gave a colorless oil, assigned structure
3&9 in excellent yield. This structure is based on the
spectral and chemical properties of ééh

O
CHSO

45

M

1 corresponding

It exhibited a carbonyl frequency of 1716 cm“
to an unconjugated sinmembered ring ketone. Further evidence
for the absence of conjugation is the ultraviolet spectrum in
ethanol which showed only end absorption (kmax 202 mu,

6 5,350). The nmr spectrum (Table I) is consistent with
structure 45; significant features are the presence of two
allylic methyls which show homoallylic coupling of about 0.9
cycles, and the splitting pattern for the proton and methyl
group on C—6. The low—field aliphatic methyl at T 8.73 is
probably attached to the carbon which bears the methoxyl
(compare with spectrum of_§§). Also notable is the fact that
only one of the two possible stereoisomers of_4§ is obtained.

Further support of structure 45 was obtained by its

thermal 1,4—elimination of methanol when passed through a gas

 

19

chromatograph at 2000. The product was a colorless oil which
was assigned the structure éfidon the basis of its analysis
and spectral properties. Its infrared spectrum showed a
conjugated carbonyl and carbon—carbon double bond (1667, 1600

cm‘l) and a terminal methylene group (905 cm‘l).

O 0 Ii 0 D
CD3
46 47 48
EtOH

The uv spectrum exhibited a kmax at 280 mu (€ 13,800),
reasonable for the assigned structure. The nmr spectrum
(Table I) is also consistent with the structure. Particularly
important is the low-field position (T 7.98) of one of the
allylic methyls. This is expected for the B-methyl of an
d,8-unsaturated cyclohexenone (compare with the spectrum of
gfip and thus rules out isomer Ellwhich would result from
1,2—elimination of methanol from 45;

Confirmation of structure éévwas obtained by the change
in its nmr spectrum on treatment with deuteromethanol contain—
ing a catalytic amount of sodium hydroxide (ééL-—+'g£9. The
bands due to the C—3 methyl and C-6 proton disappeared, and
the doublet assigned to the C-6 methyl became a singlet.

The keto form of 40 was also prepared. Injection of an

ether solution of 42_on the gas chromatograph at 1800 afforded

 

20

a product assigned structure $1.0“ the basis of its spectral
data and method of synthesis. The infrared spectrum showed

“1 and a terminal methylene at 920

a carbonyl band at 1700 cm
cm'l. Although the carbonyl frequency appears to be rather
high for an d,8-unsaturated ketone (usually found at 1675 cm‘l),

Erskine and Waight (21) have presented a number of examples

of cisoid d,8-unsaturated ketones with similarly high carbonyl

0
H
47
M
frequencies. The nmr spectrum of 41'(Table I) is consistent
with the proposed structure. In particular, the absence of a

low-field allylic methyl is significant (compare with 46).
Also important is the similarity of the chemical shift of
the C-6 methyl of élvto those of_4§ and 4§’(Table I). The
relatively large chemical shift difference between the
methylene protons in 4Z’(compare with 46) suggests the marked
difference in environment expected if the methylene group is
adjacent to the carbonyl.

The above experiments establish the following sequence

of reactions:

21

 

OH OH 0
CH30 I
A H5
_fi —>.
orH
36 4O , 55
M M A M
NaOCH3
CH30H
o o
CH30 H H H
A
—————>
45 46 47
M M M

2. Photolysis of Labeled Bicyclo-
|3.1.0|hexenones

In order to confirm the structures which have been
described, and to assign as many of the nmr positions of the
various methyl groups as possible, the photolysis of two
labeled bicyclo[3.1.0]hex-3-en—2—ones was carried out in'

methanol.

 

22

Dienone 49, prepared as previously described (15),
gave ketone §9_on photolysis in ethanol. Ketone 59‘used
in the remainder of the above sequence had only 50% of the
label at C-1, but was completely labeled at C—4. The nmr
spectrum of églwas similar to that of 34’(Table I) except
that the band at T 8.12 (due to the C—4 methyl) was missing
and the band at T 8.45 for the other allylic methyl was a
sharp singlet, rather than a quartet. The area of the band
at T 8.90 (due to the C-1 and one of C-6 methyls) was con—
sistent with 50% label at C—1. Irradiation of ég’in methanol
gave 51, whose nmr spectrum in methanol was similar to 36
(Table I) except that the quartet at T 8.37 was missing and
the peak at T 8.30 had decreased in intensity to 4.5 protons.
This permits the assignment of the C—2 and C—4 allylic
methyls of 51 to the band‘at T 8.30.

When a solution of 51 in methanol was warmed, it gave
§§J whose nmr spectrum in methanol differed from 49 (Table I)
in that the peak at T 8.22 was reduced to a shoulder on the
peak at T 8.18, which had an area of 4.6 protons. This per-
mits the assignment of the high-field allylic methyl in £2,
to C—2.

Treatment of éguor 51 with dilute hydrochloric acid_
afforded dienone 53: whose nmr spectrum showed only one low-
field allylic methyl at T 8.05; the area of the high—field
allylic methyl at T 8.17 demonstrated that it contained

slightly more than one methyl. The observation that both of

 

list—F

 

23

these bands were singlets, rather than quartets, establishes
that the two CD3 groups are on opposite sides of the ring.
The other series of experiments with labeled ketones is

shown in the following scheme.

 

CD3

 

 

Dienone éélwas prepared by refluxing the unlabeled com—
pound (55) in deuteromethanol containing sodium methoxide.
Irradiation of 44 in ether afforded ketone 54, whose nmr
spectrum differed from 34'in that it lacked a band at T 8.12
(due to C-4 methyl) and the quartet at T 8.45 became a sharp
singlet. The band at T 8.78 was also missing; this is there-
fore assigned to the C—5 bridgehead methyl of 34; In order
to verify that the CD3 group in ééiwas on the bridgehead

position, this compound was subjected to ozonolysis with

24

oxidative work-up, affording Egas the final product. The nmr
spectrum of the anhydride (see Experimental) showed the
absence of one of the bridgehead methyls.

The photoisomerization 44d——>j§4_has ample precedent in
other 2,5—cyclohexadienone systems (22). Irradiation of 5%,
in methanol gave the crystalline photoproduct 55; The nmr
spectrum of 55/in methanol was similar to that of 36, except that
the quartet at T 8.37 was missing and the band at T 8L30 was
a sharp three-proton singlet. This confirms the assignment
of the latter band to the C—2 allylic methyl.

When a solution of 55 in methanol was allowed to remain
in a refrigerator overnight it was converted to 56, whose nmr
spectrum in methanol exhibited a 3-proton singlet at T 8.22
and a small peak (perhaps one proton) for the low-field
allylic methyls at T 8.15. This confirms the assignment of
the high-field allylic methyl of égflto C—2.

In one particular experiment, 54 was irradiated in
methanol which contained a trace of acid, and workup gave 51
as the major product. Its nmr spectrum showed three singlets
at the same positions as for 35 (Table I) but with relative

areas of 3:6:3.

3. The Photosensitized Conversion of Hexamethyl—
bicyclo[3.1.0]hex-3—en-2—one (34) to 2,3,3,4,5,6-
Hexamethyl—6—methoxycyclohexa-1,4—dien-1-ol (36)

Previous reports of triplet intermediates in ultraviolet
light induced bicyclohexenone rearrangements (6,12) led us

to irradiate 34.in the presence of benzophenone, a known

triplet energy donor (Et = 69.2 Kcal. mole"l (23)).

25

A solution of 0.55 mmoles 34’in 10 ml of methanol was
placed in a Pyrex test tube. In a similar tube was placed
0.55 mmoles of 34 and 27.6 mmoles of benzophenone. Both
tubes were irradiated with a 200-watt Hanovia mercury vapor
lamp cooled by a stream of air (this had the effect of
decreasing the photolysis rate of_34). The disappearance
of 34 and the appearance of the thermal decomposition product
of gédwere followed by vpc. After 225 minutes, photolysis
had proceeded to the extent of 13%-and 54%, respectively, in
the two samples.

The ability of benzophenone to enhance the rate of
photolysis of_§4, even though it is present in sufficient
quantity to act as a filter, together with the observation
that the same product was observed in both runs, strongly
suggests that a triplet species is involved in the conversion
of ééito 363‘

4. The Mechanistic Significance of the Photochemical
Rearrangement of Hexamethylbicyclo[3.1.0]hex—3—
en—2-one (34)

Several aspects of the photochemical properties of 34
which are mechanistically significant deserve comment. The
first of these is the predominant formation of_3§’in methanol.

The majority of light induced bicyclo[3.1.0]hexenone re-
arrangements can be explained by n ——>'w* excitation of the
ketone (59) to a dipolar species 69 followed by migration of

one of the groups at C—6 to C—1 or C—5 to give isomeric

26

 

 

O
3 hV c
n —-+ ”*7, R
4
R
332 .629.

dienones, or, if R = H, loss of a proton from C—6 with subse-
quent phenol formation (see Introduction).

The dipolar species 61 derived from bicyclohexenone 34}
could undergo a 1,2—methy1 migration to give dienones 33 or
35, or, if generated in a nucleophilic solvent such as

methanol, could react with the solvent to form 36.

o o” o
t j; —» fl
372.4, E}, M55
CH30H
o CH30 OH
931. 9.9..

This is the interpretation placed on the isolation of 36 as
the major photolysis product of bicyclohexenone 34 in

methanol.

27

The structure of photoproduct 36 strongly suggests that
it is produced by addition of methanol to an ionic species
such as 6;; A radical species 62 isoelectronic with gl'would
probably lead to 64 as the major product since the C—H hydrogen

atom of methanol is much more easily removed than the O-H (24).

O’ 7 OH HOCH OH
CHBOH E 'CHaOHa
92, 55, 84.

The ultraviolet light—induced conversion of_34 to the
dipolar intermediate 61 may proceed via the series of steps
presented below: (1) n-<>-W* eXCitation followed by inter-
system crossing to the triplet (34 —<>-§§); (2) bond alter—

ation t§§:-%*‘§2); and (3) intersystem crossing to a ground

state singlet (62-—%>_61).

0 0o 0
—h—V——> —>
(1) (2)
54 65 62

66

 

 

28

The above rearrangement is proposed to result from excitation
of the n -—**w* band of 34 (A 320 mu, shoulder, e 605), since
the intense W —>W* bands at 235 mu (6 6270) and 274 mu

(6 3240) appear at much shorter wavelength and the conversion
of §4_to.3§ occurs in Pyrex where little light below 300 mu
is available. Evidence for a triplet intermediate in the
above scheme is the observation that the reaction could be

photosensitized by benzophenone.

There is also the possibility that 65 undergoes inter-
system crossing to the ground state singlet 66, which then
undergoes ring opening to_61. The acid-catalyzed reaction
of 34 (described in Part II of this thesis) suggests that the
latter process does not occur. Treatment of 34,with acid gave
dienone éé'in excellent yield. Labeling studies demonstrated
that 33 is not formed by a methyl migration (involving the
protonated form of 61) but gig a cyclopropylcarbinyl rearrange-

ment (QZd-+§§’—>§§_).

+
O OH OH O
H+
—>’ ———>' -——9'
54 15.7. .628 93:.

It is therefore concluded that to the extent that the acid—
catalyzed reaction of 34 defines the behavior of the dipolar
species 66, such an intermediate is not involved in the

photochemical conversion of 34 to 36.

29

Special comment should be made concerning the formation
of enol 36, rather than its keto form. Since treatment of

36 with basic methanol furnishes ketone 45, this tautomer

OH
CH30 CH30

NaOCH3\_
CH30H

 

56 48.

M

is assumed to be the more stable of the two. Thus, 3§_is
probably formed yi§_a kinetically controlled process, perhaps
involving a rapid protonation of 61 on oxygen.

Exclusive formation of 36; rather than isomers 31 or 38;
in spite of the fact that the two latter structures have con—
jugated double bonds, may be due to the charge distribution
in 61: Preferential attack of methanol at the C-3 position
of 61’suggests that the charge density at this position is
greater than at C-1 or C-5. This is consistent with recent
observations in similar systems. In a study of nmr chemical
shifts of protonated alkylbenzenes, MacLean and Mackor (25)

have assigned the charge density distribution presented below

H Ii

30

to ions of type 69. The nmr spectrum of heptamethylbenzenon-

ium ion 70 is also in agreement with greater charge density

 

 

 

/ V >
‘ l‘
11 19. 41

in the C—4 position than in the C-2 and C-6 position (19).

The formation of 1,4— rather than 1,3-conjugated products

(36 rather than 31 or 38) also has ample precedent. Doering

has shown, for example, that ion 29 loses a proton to give

41 as the only product (19): similarly, treatment of.§§ with

sulfuric acid converted it exclusively to dienone 35 (14).

O . OH O
H+
——%>- ———%r
\\
55 7:4 5.5.

This reaction is considered to proceed via ion 72, which
may then undergo a methyl migration in either of two directions.

As expected, the product has the 1,4— rather than 1,3—diene

structure. Also notable is the 1,2-elimination of methanol

from.§§ to give 40 (compare with 1,4—elimination of methanol

from £2,

31

OH OH
CH3

-CH30H

36 40

M M

In view of the proposed mechanistic path which rational-
izes the photochemical conversion of 34 to 36, it is now of
interest to consider why this type of reaction has not been
observed previously. One reason is the dearth of examples
of bicyclohexenone photolyses in alcoholic solvents, presum—
ably because of the possibility of photoreduction. A few
examples of bicyclohexenone isomerizations in alcoholic
solvents have been reported, however. Dfirr has shown (26)
that irradiation of 13 in ethanol or dioxane with light of

l > 280 mu gives phenol Zévin excellent yield. The absence

0 O OH
8 8 M Q; g; 8 8
‘E;E§6§*’ ‘——2"
8 8 8
8 25 8
1.5.. 1.3.4 1.4,

of products resulting from reaction of dipolar_species
73a with ethanol is not surprising, as proton loss from this
species would be expected to be rapid in comparison to a

bimolecular reaction with the solvent.

32

KrOpp has reported that irradiation of bicyclohexenone
26 in methanol or dioxane leads to dienone 28 in good

yield (11).

 

- o
o o
hv
\ CHgOH) G. 5
5259. 27 28

No products resulting from reaction of species El'with

methanol were observed. This suggests that conversion of
21 to 28 is rapid compared to reaction with solvent. Com-
parison of the above reaction with the photolysis of 34 in
methanol leads to the conclusion that the rearrangement of

dipolar Species 61 (derived from 34) to dienones 33 or 35

O 0
0r
qt
0 - 9&0
55 55
W

I“ .
CH epic? OH
3011 y CH30

61

36

33

is quite slow with respect to the reaction of 61Jwith methanol.
The difference in rearrangement rates of 21_and 61 may be

due to the observed preference for methylene rather than
methyl migration to a positive carbon atom (see Introduction).
Stiles and Mayer, for example, have shown that the partial
rate factor ratio of ethyl vs. methyl migration in pinacol
rearrangements is 17:1 (27). The acid-catalyzed rearrange—
ment of dienone léuto phenol 252 was reporged by Miller and
Margulies (28). This reaction also demonstrates the favor-

able migration of a methylene group with respect to a methyl.

 

O OH
+
H > 90%
75’ 75a

There is also the possibility that these dipolar species
73a and Elvdo react with the alcoholic solvent to give
products similar to 36, but that these are then rapidly
converted by a trace of acid in the solvent to phenol lé’and
dienone 28! respectively. Evidence for this is the smooth
conversion of éélto 35 on treatment with a trace of acid.
Indeed, it was found that irradiation of 34-in ethanol or
methanol which had not been previously purified to free it
from acid (distillation from sodium methoxide) gave the dienone

§§_in good yield. Clearly, the photolysis of bicyclohexenones

34

léyand‘zfi in acid-free methanol would be worthwhile to
investigate.
B. The Photochemistry of Hexamethylbicyclo-

[3.1.0]hex-3—en-2-one (34) in Diethyl Ether

The observation that photolysis of 34’in methanol leads

to a single product 3§_which is apparently formed by reaction
of a dipolar intermediate 61’with a molecule of methanol led
us to study the photochemical behavior of 34 in diethyl ether,
a non-nucleophilic neutral solvent. It was thought that
photolysis of 34 would furnish the dipolar species 61! which,
in the absence of a nucleophile (such as methanol) might
undergo alkyl migration to give dienones 33 or 35, as has been

observed in other systems (see Introduction).

0

O...
:25
\ O

61
~33

A solution of bicyclohexenone 34 in anhydrous ether
was irradiated at 00 with light of"A > 280 mu. The reaction
was followed by ultraviolet spectroscopy. As the photolysis

proceeded, uv bands due to the starting ketone decreased in

35

intensity and a strong band appeared at 252 mu. When the
reaction was complete (as determined by the absence of any
change in the intensity of the 252 mu band on further irradi-
ation) the ether was evaporated in vacuo at room temperature.
The nmr spectrum of the resulting yellow oil in CCl4
showed it to contain a small amount of ether; the remaining
singlets at T 5.00, 5.20, 8.17, 8.23, and 8.90 were virtually
identical in intensity and chemical shift to the nmr spectrum
(Table I) of the enolic triene 491(prepared by warming 3&9
in CCl4 solution. Also consistent with structure 49’was the
strong ultraviolet absorption at 252 mu observed for the
photolysis solution.

OH

40

M

The structure of the photoproduct was further reinforced
by converting i.t;tQ 2 ,i3,,.'4,, 4, '5,, 6-hexamethyl-2 , 5—cyc lohexadienone
(35) with a trace of acid. Addition of 18 ul of 1% hydro-
chloric acid to 5 ul of the photolysis solution dissolved in
about 3 ml of 95% ethanol resulted in the disappearance of
the ultraviolet band at 252 mu (A = 1.42) due to 49_and
formation of a new band at A 244-245 mu (A = 0.96) corres—

ponding.to dienone 35:

36

A possible pathway for the conversion of éé’to 49’in

ether is outlined below. It is proposed that the enol 40 is

 

OH
O
40
OH
34 61
42

formed via_a rapid proton transfer from the C—3 methyl of 61
to the adjacent negatively charged oxygen. Loss of a proton
from the C—1 methyl would lead to fully conjugated product 42:
The fact that this product is not observed further demon*
strates the preferential formation of 1,4— rather than 1,3-
conjugated products in these systems. Formation of the enol
form of 49 rather than its keto tautomer £1, even though the
latter appears to be more stable under equilibrium conditions
(heat), suggests that éQ,iS formed yia a kinetically controlled
process such as the rapid protonation on oxygen depicted above.
There remains the question of why the intermediate Ea;
undergoes proton loss from the C—3 methyl, rather than alkyl
migration, as is normally observed (see Introduction). The
answer may lie in the unexpected stability of enol 49! perhaps

due to the double bonds being substituted with alkyl groups.

 

37

Formation of~49 as the only product requires that proton loss
from 61 be much more rapid than methyl migration to give 33'
or éfib

An example of this phenomenon in a similar system is
the acid—catalyzed dehydration of alcohol 16 reported by

Hart, Collins and Waring (15). These workers found that

OH
H
l _; + _4
H
// I //
16.3.. 199, 1.7,

treatment of Z§_with HCl in pyridine gave the triene Zl_in
greater than 95% yield. This was attributed to rapid proton
abstraction from 163 by the solvent before aromatization
could occur. In the case of 61, the negatively charged
oxygen may perform the same function as the basic solvent
described above.

C. The Photochemistry of Hexamethylbigyclo-
[3.1.0]hex-3—en—2-one (34) in Acetic Acid

 

In view of the interesting solvent effects observed in
bicyclohexenone photochemistry (see Introduction), a study
was made of the photochemical behavior of 34 in aqueous
acetic acid. A solution of 34 in 45% acetic acid at 00 was
irradiated with light of 7\ > 280 mu. The reaction was

monitored by vpc. As the photolysis proceeded, the vpc peak

38

due to 34,decreased in intensity and a new peak with a longer
retention time appeared. This peak increased in intensity at
the expense of 34 until all the bicyclohexenone had reacted.
There was no evidence for the formation of any intermediate
between ééiand the observed product. Bicyclic ketone 34 was
stable in the dark under the photolysis conditions.

An nmr spectrum of the crude material obtained by work—
up of the photolysis solution showed it to contain 7% unre-
acted 34 and 93% dienone_§5. Recrystallization from petroleum
ether afforded a white crystalline solid which had an nmr
spectrum and vpc retention time identical to that of an
authentic sample of 35 (14). The following sequence is there-

fore established:

hv
HOAc

:4 55

M

A possible pathway for the conversion of 34 to 35 in

acidic media is outlined below. One significant feature is

o 0" OH 0
h H+
——-l> ——>- ——>»
5.5.1}. -54 13, .53

39

the marked sensitivity of the photochemical behavior of

34 to the presence of acid. This constitutes yet another
example of a solvent effect in bicyclohexenone photochemistry.
The difference in behavior of 34 in acidic media is probably
due to protonation of species 61 on oxygen. Strong evidence
for ion 12 being a precursor of dienone 35 is the very prob-
able intermediacy of this species in the acid-catalyzed re—

arrangement of ggland 33 to 3E?

OH
\+ OH
.49, ”CH3
X
0 y
‘69
r10
73.. 55

flflma preferential. formation of dienone 35} rather than 33’
can be attributed to the low stability of linearly conju—
gated cyclohexadienones compared to cross—conjugated cyclo-
hexadienones.

Miller has shown, for example, that dienone Z§Vrearranges

smoothly to 53,0n heating for 2—3 hours at 100—1100 (29).

40

O O
____,.
78 79

Similarly, Denivelle and Fort have reported that 6—bromo—
cyclohexa-2,4-dienones rearrange completely to 4—bromocyclo~

hexa-2,5-dienones on standing (30).

EXPERIMENTAL

A. General Procedures

The nmr spectra were obtained with a Varian A-60
spectrometer, using CCl4 solutions (unless otherwise stated)
with tetramethylsilane as an internal reference. Infrared
spectra were taken on a Unicam SP-200 Spectrometer, in CCl;
solution unless otherwise stated. Varian-Aerograph gas
chromatographs were used. Ultraviolet spectra were taken on
a Beckman DB spectrophotometer. Mass spectra were carried
out by Harold Harris with a Consolidated Electrodynamics
Corporation 21-103C instrument. Melting points are uncor-
rected. Analyses were performed by Spang Microanalytical

Laboratories, Ann Arbor, Michigan.

B. General Photolysis Procedures

All irradiations were conducted using either a Type
S 200 or 450w Hanovia lamp. This was placed in a water—
cooled quartz jacket which was inserted in a glass vessel of
a slightly larger diameter. The resulting annular space
held about 450 ml of solution, which could be stirred by a
stream of nitrogen gas. Irradiation of small amounts (< 25 ml)
of solution was carried out in a Pyrex test tube taped to

the side of the quartz immersion well.

41

42

C. Photolysis of 1,3,4,5,6,6—Hexamethylbicyclo-
13.1.0]hex-3-en-2-one (34) in Methanol

 

The methanol should be freshly distilled from sodium
methoxide, and all glassware should be washed with alkali
before use. A solution of 1.81 g of éé.(15) in 400 ml of
methanol, stirred with a stream of dry nitrogen, was irradi-
ated at 150 with a 200—w Hanovia lamp through a Pyrex filter.
The reaction, which was complete in about 250 minutes, was
followed by vpc (disappearance of enone) or by uv (disappear—
ance of the peaks at 235 and 274 mu due to_34). Evaporation
of the solvent in_yaggg at 200 led to the isolation of
2,3,3,4,5,6-hexamethyl-6-methoxycyclohexa-1,4-dien-1-ol (égL
as white crystals. V

Anal. Calcd. for Cl3H2202: C, 74.24; H, 10.54.

Found: C, 74.03; H, 10.43.

MeOH

at 202 mu (6 4,300). Its nmr
max

Compound_§§ showed a A
spectrum was most easily measured in methanol (see Table I).

If one works quickly a similar spectrum is observed in CCl4
(Figure 2) (bands at T 7.13, 8.33, 8.38, 8.75, 8.92 and 8.96)
except for an additional one-proton singlet at T 5.75 (-OH);
at the usual probe temperature (about 350% this spectrum
changes rapidly (see below).

An infrared spectrum of 36 in CCl4 was obtained by cool-
ing the cells in a freezing chest; the cells were then filled
with a solution of 36, and the spectrum rGh quickly. Prominent
bands appeared at 3555 (O—H), 2980 and 2930 (C-H), 1618 (C=C).
1375, 1345, 1290, 1215, 1105 and 1072 cm'l; there was no

carbonyl absorption.

43

K3
.Ammv Houalsmflplw.fi

ImxonoaomohxonuofilmIamgumfimmeIm.m.¢.m.m.m mo Eduuowmm HEZ

q
‘3'

.m musmflm

‘

 

 

 

 

 

 

 

 

 

 

44

In a separate experiment to determine the isolable
yield and stability of 36} a solution of 0.327 g of_§4_(15)
in 2 ml of methanol (freshly distilled from sodium methoxide;
all glassware alkali-washed) in a Pyrex test tube sealed with
a serum cap was irradiated by attaching it to the side of a
water-cooled quartz well containing a 450 watt Hanovia lamp.
Photolysis was complete in 130 min (Vpc). The solution was
cooled to —700, and the photoproduct was filtered and rinsed
with cold (-700) methanol. The yield of white, crystalline
36 was 0.294 g (76%). The crystals could be stored indefi-
nitely at —700, but on standing overnight at room temperature
they were transformed to a viscous yellow oil. Starting at
room temperature, with a heating rate of 2O/min, 36 has a mp
of 58-650.

D. The Photosensitized Reaction of Hexamethylbicyclo-'
[3.1.0]hex-3—en-2—one (34)

The following experiment was performed to determine
whether the photolysis could be sensitized. Starting enone
‘34’and benzophenone were washed with 5% sodium hydroxide,
then water, before use; methanol was freshly distilled from
sodium methoxide, and all glassware was washed with alkali.

A solution of 0.098 g of 34 in 10 ml of methanol was placed

in a 25 ml Pyrex test tube, sealed with a serum cap. In a
similar tube was placed a solution of 0.098 g of‘ég’and 5.02

g of benZOphenone in 10 ml of methanol. Both tubes were taped

to the side of a water—cooled quartz well containing a 200

 

45

watt Hanovia mercury vapor lamp. A stream of air was passed
over the lamp; this had been observed to decrease the photoly—
sis rate of_§4. The photolysis was followed by injecting 10
ul samples on a gas chromatograph, 5' x 1/4" 20% SE-30
column, 1800. Disappearance of starting enone and appearance
of photoproduct (or vpc decomposition product thereof) were
both followed. After 225 min, photolysis had proceeded to
the extent of 13.2% and 53.9% respectively, in the two tubes.
E. The Dark Reaction of Hexamethylbicyclo-
i3.1.01hex-3-en-2-one :34) in Methanol

Enone 34 (0.066 g) was dissolved in 3 ml of methanol
and the resulting solution was stored in the dark at 00 for
70 hours, at which time the methanol was evaporated_iglyaggg,
yielding a clear oil (0.063 g). The nmr spectrum of the
product (Table I) showed it to consist entirely of 34.
F. 2L3]3y4y5-Pentamethyl-6-methylenecyclo-

hexa-1,4-dien-1-ol (49)

When an nmr tube containing a solution of_§§ in CCl4
was allowed to remain in the probe (about 350) for 5 min,
it was noted that a new band appeared at T 6.62, due to
methanol. After 10 min, this band reached maximum area, and
other spectral changes were complete. The product, which is
2,343,4,5-pentamethyl-6—methylenecyclohexa-1,4—dien—1-ol
(49), had an nmr spectrum as shown in Table I and Figure 3.

MeOH

Evaporation afforded a yellow oil with a kmax at 252 mu

(8 13,400). An ir spectrum of 49 showed a strong hydroxyl

46

. d3 6-756% J
ImxmnoHchomawamnumalmIHNSumEmucmmlm.¢.m.m.m mo Eduuuomm H82

.m musmflh

 

 

 

w m D Q
d

Eziééiié __

-

 

 

 

 

 

—1

—

 

47

band at 5450 cm-1 and other bands at 2945, 2900, 2840 (C—H),
1620, 1575, 1450, 1575, 1270, 1208, 1108, 1075, 1025 and

864 cm-1.

G. Acid—Catalyzed Conversion of 2,3,3,4,5,6-
Hexamethyl-6-methoxycyclohexa-1,4—dien-1—
01 (35) and 2,3,3,4,5—pentamethyle6-
methylenecyclohexa—l,4-dien—1-ol (40) to
2,3,4,4,5,6-Hexamethyl-2,5—cyclohexa—

dienone (35)

A few crystals of 36 were dissolved in 30 ml of methanol

 

 

 

and 2 drops of dilute HCl were added. The solution was
evaporated in vacuo at room temperature, and the nmr spectrumf
of the residue was identical with that of dienone 35 (14,20),

except for a small amount of an unknown impurity.

The nmr spectrum of dienone 35 is presented in Table I

and Figure 4. Its ir spectrum showed bands at 1653 cm-1
(C = 0) and 1624 cm-1 (C = C). Its uv spectrum had xfiggH at

246 mu (6 14,800).

A 10% solution containing a 50:50 mixture of égvand_3§
in methanol was acidified with 2 drops of dilute HCl. An
nmr spectrum run on the resulting solution showed peaks due
only to 35.

To 3 ml of a solution of EéLin methanol in a uv cell
(xmax 202 mu, A = 0.97) was added 4 01 of 1% HCl in methanol.
The solution now had a'strong absorbance at 252 mu (A = 0.88)
due to the formation of‘49; After 10 ul additional HCl, the
solution had a Kmax at 251 mu (A = 0.79), and after another

10 ul of HCl, (max at 247 mu (A = 0.68), due to conversion

to 35.

48

z.

42

.MMMW mnocmfipmxmfloHU%Ulm.mlamnumfimxmslm.m.¢.¢.m.m mo Ednuommm HEZ

\.

.d musmflm

 

 

 

 

 

 

 

 

 

 

..
q

 

 

49

H. Preparation of 2,4,4,6-Tetramethyl-3,5-
bis(trideuteromethy1)L275-cyclohexadi-
enone (44)

 

 

To a solution of §§’(7.8 g) in 63 ml of CH30D (31)
was added a 0.5 cm3 piece of sodium. The resulting solution
was refluxed for 70 hrs, after which time an nmr spectrum
showed that exchange was complete. Most of the methanol
was evaporated i§_yagug, and the remaining solution was poured
into 200 ml of methylene chloride. This was extracted with
cold (00) water to remove the remaining methanol, and the
solution was dried over anhydrous magnesium sulfate. Evapora—
tion of the solvent afforded 7.6 g (97%) of crystalline_44
which had an nmr spectrum with two equal singlets at T 8.18
and 8.78.

I. 2-Methoxy-2,3,4,5,5,6-hexamethylcyclohex-
3—en-1-one £45)

A solution of 34 (0.81 g) in 25 ml of methanol was
irradiated at 00 in a Pyrex test tube for 74 min with a 450
watt Hanovia lamp. After this time white cryStals of_§§~
could be seen suspended in-the methanol, and photolysis was
complete (vpc). The photolysis solution was poured into
25 ml of 0.2N sodium methoxide in methanol and stirred for
30 min at 00. Water (35 ml) was added, the solution was ex-
tracted with methylene chloride, and the latter solution was
washed with water and dried over anhydrous sodium sulfate.

Evaporation of the solvent in vacuo gave 2-methoxy72,3,4,5,5,6-

 

hexamethylcycloheXHS—en—1-one (45) as a colorless oil in

nearly quantitative yield. Its ir spectrum showed bands at

50

1716 cm-1 (C = 0) and 1115 cm‘1 (C-O—CHs). Its nmr spectrum

is shown in Table I and Figure 5. Its uv spectrum had a

EtOH
xmax at 202 mu (€ 5,350).

J. 2,3L5,5,6-Pentamethyl—4-methylenecyclohex—
2—en-1—one (46)

 

 

When 45 was injected on a gas chromatograph (5' SE—30
column, 2000) one major peak was observed; collection of the
product showed that 45'had lost a mole of methanol. -The new
product, which was a colorless oil, is assigned the structure
2,3,5,5,6-pentamethyl-4-methylenecyclohex¥2een¥1;one (46).

Anal. Calcd. for C12H180: C, 80.85; H, 10.18.

Found: C, 80.79; H, 10.09.
.46 had ir bands at 1667 cm71 (C = 0), 1600 cm‘1 (C = C) and

905 cm":L (terminal CH2). Its uv spectrum had a kfiggH

at 280
mu (6 13,800). Its nmr spectrum in CC14 is shown in Table I
and Figure 6.

The nmr spectrum of 46’in CD30D consisted of bands at
T 4.66 (1H,s), 4.88 (1H,s), 8.06 (1H, q, J = 7.0 cps), 8.20
(3H,s), 8.45 (3H,s), 9.22 (6H,s) and 9.37 (3H, d, J = 7.0
cps). Sodium hydroxide (0.0033 g) was added to a solution
of 4§_(0.075 g)=in 0.2 ml of CDSOD. After the solution had
remained at room temperature for 16 hrs, its nmr spectrum

showed that the quartet at T 8.06 and the singlet at T 8.20

were missing, and the peak at T 9.37 had become a singlet.

 

51

.Po.

.Nmeq.meo-e
Icmlmlxmsoaomoamnumfimxmnlm.m.m.¢.m.mlmxonumalm mo Ednuowmm H82

.m musmflm

 

 

 

m > u m .r
_ ul 1 q A W

-
‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

El»: t) 'P’IDIIDE Inf.» Db till 11):»? Dbl...) i I'DII'I 2|). IPI') b p I, [It'll 8)! )
7% 1411111 14‘11‘1 ‘11114111 4144114 11 14 [14111111144411.1111

 

52

.Awdv mgOIHIGm
ImuxosoHchomnmamnuwfilwlahnumfimucmmlm.m.m.m.m mo Ednuommm H82

.m musmfim

«m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

53
K. 3,4,545,6-Pentamethy1—2-methylenecyclo—
hex-3-en-1-one714ll
When an ether solution of‘gg’was injected on a gas
chromatograph (10' Apiezon-L column, 1800) one major peak

(other than solvent) was observed; collection gave a clear

 

oil which is assigned the structure 3,4,5,5,61pentamethyl-2-
methylenecyclohex-2—en41-one (41). Its ir spectrum had
prominent bands at 1700 cm"1 (C = 0) and 920 cm‘1 (terminal
CH2). Its nmr spectrum is reported in Table I and Figure 7..
L. Preparation of 2,4,6,6-Tetramethyl-3,5-
bis(trideuteromethyl)-2,4-cyclohexadi-
enone (49)

Dienone ég’was prepared by a modification of the method_
of Hart, Collins, and Waring (15). To a solution of un-
labeled dienone 33 (9.66 g) in 50 ml of CH30D was added a
0.8 cm3 piece of sodium. The resulting solution was refluxed
for 96 hours, at which time most of the methanol was evapor—
ated in yagug, and 25 ml of CH30D was added. The resulting
solution was refluxed for 49 hours, after which time an
nmr spectrum showed that exchange was complete. Most of the
methanol was evaporated in yagug, and the remaining solution
was poured into 200 ml of methylene chloride. This was ex—
tracted with cold (00) water, and the solution was dried over
anhydrous magnesium sulfate. Evaporation yielded an oil,
which on distillation gave 6.94 g (70%) of dienone 49 (bp
84-870 at 1.4 mm). Its nmr spectrum in CCl4 consisted of

three singlets at T 8.13 (3.3H), 8.18 (3.0H) and 8.90 (6.0H).

 

 

54

If."
.Ahwv occlalgmlm

IxmnoaUmomqwahguwfilNlahsuoamucmmlm.m.m.¢.m mo Esuuummm H82 .5 mnsmflm

«2

 

 

_

1?;

 

 

 

"7i:

 

m
_
1

 

 

(<34?

 

55

M. Photolysis of 3,5,6,6—Tetramethyl—1,4—bis—
(trideuteromethyl)bicyclo[3.1.0]hex—3-én-
2-one

Labeled enone §Q~was prepared as previously described
(15). Its nmr spectrum showed that label was essentially
complete at the C—4 methyl, and 50% complete at the C—1 methyl.
The enone was irradiated in methanol as described above for
the unlabeled compound. The nmr spectrum of the resulting
crystalline photoproduct (51) in methanol was identical with
that of unlabeled 36 (Table I) except that the peak at T 8.37
was almost absent, and the peak at T 8.30 was a very sharp
singlet, area 4.5H.

Thermal conversion of this labeled photoproduct (as
described above for unlabeled 34) gave a mixture of_§2 and
53. Disregarding bands due to the latter (see below) gave
an nmr spectrum for 52 in methanol which was identical with
that of unlabeled material, except that the allylic methyls
appeared as a sharp singlet at T 8.18 with a shoulder at T
8.22, total area 4.6H.

Treatment of labeled 5; or 52 with a little dilute HCl
led (as described above for unlabeled compounds) to labeled
éé’whose nmr spectrum had bands at T 8.05, 8.18 and 8.80 with
relative areas 2.6:3.4:6.0. The two low—field peaks were
singlets.

N. Photolysis of 2,3,4,4,5,6-Hexamethyl-2,5-

cyclohexadienone (35) in Ether

A solution of 2.0 g of_3§ in 400 ml of anhydrous ether

kept at 150 under a nitrogen atmosphere was irradiated with

 

56

a 200 watt Hanovia lamp through a Vycor filter with magnetic
stirring. After 81 min the photoproduct had reached a maxi-
mum concentration as shown by vpc and the reaction was termi-
nated. Preparative vpc afforded 0.39 g of hexamethylbicyclo-
J3.1.0]hex-3-en-2-one (34). The photolysis was not allowed
to proceed to complete disappearance of the dienone, because
the enone photoproduct undergoes further photorearrangement
under these conditions.
0. Photolysis of 2,4,4,6-Tetramethy1-3,5-bis-

(trideuteromethyl)-2,5-cyclohexadienone

(3&9 in Ether

The deuterated dienone was photolyzed as described above

for the unlabeled dienone. "The photoproduct 1,3,6L6-tetra-
methyl-4,5—bis(trideuteromethyl)bicyclo[3.1.0]hex-3-en-2-one
(54) had the following nmr spectrumzsinglets‘at T 8.45,
8.90 and 9.07 with relative areas 3:6:3.
P. Degradation of 1,3,6,6—Tetramethyl-4,5—bis—

(trideuteromethyl)bicyclo[3.1.0]hex-3—en-

2-one (ffifi

The hexadeuterated enone 54 was subject to ozonolytic

degradation, followed by an oxidative work-up. The procedure
used was similar to that employed by Hart, Collins and Waring
(15) to degrade the unlabeled enone #4, except that the esteri—
fication step was omitted. Ozonolytic degradation of_§4’led
to a tri-deuterated anhydride 58, whose nmr spectrum in
acetone-d5 differed from that of the unlabeled compound in

that the band at T 8.63 (due to the equivalent methyls on the

57

ring junctions) was reduced in area from 6H to 3H; in other
respects, the nmr spectrum was identical with that of the
unlabeled anhydride (15).
Q. Photolysis of 1,3,6,6-Tetramethyl-4,5-bis-
(trideuteromethyl)bicyclo[3.1.0]hex-3-en-
2-one (54) in Methanol

Labeled enone éé’was irradiated in methanol as described
above for the unlabeled compound. An nmr spectrum of the
product in MeOH showed it to be a mixture of_5§,and‘5§. If
peaks due to the hydroxytriene 56 were disregarded, a spectrum
due to labeled 55 resulted which lacked the band at T 8.37
and showed only a 3 proton singlet at T 8.30; in other re—
spects, the nmr spectrum was identical with that of unlabeled
Efib

The methanol solution of photoproduct was allowed to re-
main in the refrigerator overnight. It was then evaporated
in_yagug to 0.5 ml, giving a solution of labeled gnghose nmr
spectrum showed a small singlet (about lH) at T 8.15, and
singlets at T 8.22 (3H) and 8.88 (6.7H).

In one photolysis of 54¢ a trace of acid must have been
present, since the major reaction product was labeled hexa—
methyl-Z,5-cyclohexadienone,‘51. Its nmr spectrum showed
singlets at T 8.05 (3H), 8.18 (5.2H) and 8.80 (3.2H).

R. Photolysis of 1,3,4,5,6,6-Hexamethylbicyclo-
[3.1.01hex—3-en-2-one (34) in Diethyl Ether
A solution of 0.096 g of éévin 6 ml of anhydrous ether

(freshly distilled from lithium aluminum hydride), was :

58

irradiated at 00 with a 450 watt Hanovia lamp through a

Pyrex filter. The reaction, which was complete in 131 min,
was followed by ultraviolet spectroscopy (disappearance of
the bands at 235 and 274 mu due to 34). Evaporation of the

solvent in vacuo at 200 led to the isolation of 2,3,3,4#5—

 

ppentamethyl—6-methylenecyclohexa—1,4—dien—1-ol (49) as a
yellow oil (0.117 9). Compound 49 showed a kfiigH at 252 mu.
Its nmr spectrum in CCl4 was virtually identical with a
spectrum of an authentic sample of 42’(Table I), and con-
sisted of bands at T 5.00, 5.20, 8.17, 8.23, and 8.90. To

0

3 ml of a solution of photoproduct_4Q in ethanol in a uv

cell (Am 252 mu, A = 1.42) was added 18 ul of 1% aqueous

ax
HCl. The solution now had a strong absorbance at 244 mu
(A = 0.96) owing to the formation of 35,

S. Photolysis of 1,3,4,5,6,6-Hexamethylbicyclo—
j3,1.0]hex-3-en-2-one ggg) in 45%5Acetic Acid

A solution of 0.112 g of égiin 8 ml of 45% aqueous acetic
acid kept at 00 was irradiated with a 200 watt Hanovia lamp
through a Pyrex filter. After 150 min, the photoproduct had
reached a maximum concentration as shown by vpc, and the
reaction was terminated. Water (40 ml) was added, the solu-
tion was extracted with methylene chloride. The latter solu—
tion was washed with water and dried over anhydrous magnesium
sulfate. Evaporation of the solvent in_y§gug gave 0.119 g of

yellow crystals whose nmr spectrum in CCl4 showed the pres-

ence of 93% of gé'and 7% unreacted 34. Recrystallization from

59

petroleum ether gave 0.046 g of white crystals of §§’(14,

20), identified by its nmr spectrum and retention time.
After a solution of 0.099 g of 34 in 8 ml 45% acetic

acid had remained in the dark at 00 for 150 min, its vpc

trace showed that no 3§Lhad been formed.

 

SUMMARY

Photolysis of 1,3,4,5,6,6-hexamethylbicyclo[3.1.0]hex—
3-en-2-one (34) in methanol at 00 gave a crystalline
photoproduct, 2,3,3,4,5,6—hexamethyl—6—methoxycyclohexa—

1,4-dien-1-ol (36), in 76% yield.

The photoproduct (36) was quite unstable and lost methanol
on treatment with acid or on warming to give 2,3,3,4,5-
pentamethyl—6—methylenecyclohexa—1,4—dien—1-ol (40),.which
was converted to 2,3,4,4,5,6-hexamethyl-2,5-cyclohexadi4

enone (35) on further acidification.

The course of the photochemical and acid—catalyzed re-
arrangement was established by using enone éévvariously

labeled with CD3 groups.

The photoproduct éé’was converted by basic methanol to
its keto form, 2—methoxy-2,3,4,5,5,6-hexamethylcyclohex-
3-en—1—one (45), which thermally lost methanol to give

2,3,5,5,6—pentamethyl-4—methylenecyclohex—Z-en-1-one (4&9.

Photolysis of enone 34 in ether at 00 gave the enolic

triene 40 in excellent yield.

Photolysis of 34 in 45% acetic acid afforded dienone 35

as the only product.

60

61

7. A mechanism is presented which accounts for the observed

photochemical reactions of_§4.

 

PART II

THE ACID-CATALYZED REARRANGEMENTS OF
1,3,4,5,6,6-HEXAMETHYLBICYCLO[5.1.0]HEX-5-
EN-Z-ONE AND 2,3,4,5,6,6-HEXAMETHYL-
2,4-CYCLOHEXADIENONE

62

 

INTRODUCTION

Rearrangements of cyclopropylcarbinyl derivatives (89)
have been the subject of numerous investigations in a variety
of systems. These reactions have been extensively reviewed
by Breslow (32) and Richey (33). The most common modes of
reaction observed are the formation of either cyclopropyl-
carbinyl derivatives (81), cyclobutyl derivatives (813) (34),
or products resulting from cleavage of the cyclopropane ring

(82 and 83).

 

 

 

 

.82.

W

83

M

The unrearranged cyclopropyl products have the fastest rate
of formation, but the more stable ring—opened products pre—
dominate under thermodynamically controlled conditions. One
of the many examples of this is the initial formation of

the methyl ether §§/on treatment of 84dwith acid (35),
followed by the slower conversion to the ring-opened product

86 on prolonged refluxing.

63

64

 

OH + ?CH3 +
[:>—-é CH H 5‘ [:>‘C'CH H
I 3 CHSOH 7 I 3 CH O CHSO
H H 3 . CH3
,prolonged
84 ‘85 refluxing éfii

The ring-Opened products obtained from cyclopropyl-
carbinyl cations are usually those which correspond to the

formation of the most stable carbonium ion. Hence, 87, on

0H
+ +
H ,,,”47
——>
OH
\\*\:§ 88
55L
HOCHg
89

treatment with sulfuric acid, gives the cyclohexenol 88'
rather than the isomeric cyclopentenol ‘89 (36). Similarly,

treatment of 90 with p-toluenesulfonic acid gives the more

9 OH
n_¢ +
\H —5¥—+>- 02C=CHC(CH3)=CH¢
H3

99. 91

65

stable ring-opened product 91 in 95% yield (37).

A different mode of reaction has been observed for
cyclopropylcarbinyl derivatives incorporated in 6-membered
rings. In a study of cyclopropylcarbinyl rearrangements in
the thujopsene series, Dauben and Friedrich have shown (38)
that treatment of dideuterothujopsene (92) with refluxing

dioxane acidified to ..02N"with perchloric acid

 

H H H H
l l I
I , I I
*5. A A .
v— '7” ‘— .OH
K + e x
\\’
X
X X
22, X = D gé. 93a 21x X = D
92a, X = H 943., X = H

gives 2£,as the most rapidly formed product. Similar treat—
ment of 94a gives thujopsene (923) as the initial product
with the fastest rate of formation. To explain these results,
the authors proposed the intervention of two discrete cyclo-
propylcarbinyl cations 93 and 933, which are readily inter—
convertible. Proton loss from géJgives 92, while ring open-
ing of 93a, followed by reaction of the resulting cation with
water affords 94.

A similar cyclopropylcarbinyl rearrangement was proposed
by Tadanier (39) to account for the acid-catalyzed conversion

of the cyclosteroid §§lto the ring—opened product 96.

66

+
E+> -—+ 329
CHSO CH30 CH30 CH30 H
OH OH
95’ 95a 95b 98

A cyclopropylcarbinyl rearrangement of the above type has
also been proposed to account for the conversion of the

benzenesulfonate of 10-hydroxymethyl—Al(9)—octalin, under

9(j‘O)—octalin

acetolysis conditions, to 1—acetoxymethyl—A
reported by Hikino and de Mayo (40).
The possibility for cyclopropylcarbinyl rearrangements

also exists in acid—catalyzed reactions of bicyclo[3.1.0]-

hex—3—en—2—one systems. The cation 59a generated on

OH

 

/ Rt

97a R
I R 0H 0H 2.72 OH
+
R R
or or
59a + R
R
R
97c R 97d 97e
protonation of bicyclic ketone 59 can undergo isomerization

. to 97a or 97b (comparable with 9§~-—>'93a, 95a-—<> 95b) or

direct ring opening to 97c, 97 , or 97e.

MM

67

In view of the interesting possibilities for rearrange—
ment of this system, a study of the behavior of 1,3,4,5,6,6-
hexamethylbicyclo[3.1.0]hex-3-en-2-one (34) in acidic media
was undertaken. This ketone can be obtained in good yield

by ultraviolet irradiation of dienone §§Iin ether (15).

O 0
_D2__,.
322 .514,

The results of this study, and the mechanistic implications

derived from it, are presented in Part II of this thesis.

RESULTS AND DISCUSSION

Treatment of bicyclic ketone 34 (15) with 97% sulfuric
acid at 22.60 for 30 minutes gave the fully conjugated
dienone ééjin excellent yield. Its structure follows from

the nmr and infrared spectra, which are identical to those

0 0
H+ \
-—————e>
£4. :11

of an authentic sample (15). The nmr spectrum consists of
three bands at T 7.97, 8.14, and 8.89 (see Table I for assign-
ments) with relative areas of 1:3:2. The infrared spectrum
of 33 shows the expected carbonyl absorption at 1642 cm‘1
(ccli).

0n the basis of previously reported reactions of cyclo-
propylcarbinyl derivatives (see Introduction) two principal
modes of rearrangement might be expected for the conversion
of 62, the conjugate acid of bicyclic ketone 34; to dienone
5&5 (A) cleavage of the C—1, C-5 bond followed by methyl
migration to C-1, or (B) a cyclopropylcarbinyl rearrangement
to §§Iattended by consequent ring opening to give 33’

directly (Chart I).

68

 

 

 

69

H “Mono

1
mQU H m .md

\.))\.
emu n m Ham mo
mu
m m m m Q0
T , mmU
ml
+
o mo
800 u m .mmm
omo u m .mm. .MN
m m
HIHI.
m

 

 

70

The rearrangement of ion 72 to dienone 33 presented in
w M
Path A represents a deviation from the previously observed
tendency of similar systems to form cross-conjugated, rather
than fully conjugated products (see Part I of thesis). It

was found, for example, that treatment of 40 with a trace

0
0H OH I
+
.452. 1.2. .22

of HCl gave dienone éé'as the major product. The selective
rearrangement of ion 12, produced by protonation offiég, to
the cross-conjugated dienone 35, strongly suggests that
Path A, which requires conversion of this ion to dienone 33;
is not operative in the rearrangement of éé’to‘fié.
A. Acid-Catalyzed Rearrangement of Labeled
Bicyclo[3.1.0]hexenones

To determine which, if either, of the two mechanistic
paths A or B was responsible for the rearrangement, it was
decided to carry out the acid-catalyzed reaction using_§9’
labeled in the C—1 and C-4 positions with CD3 groups. This
compound was prepared by photolysis of labeled dienone 49’

O O

hv \‘ CD3

EtOH
CD3 CD3 cps

49.. 50

71

(synthesized by a variation of the method of Hart, Collins,
and Waring (15)) in 95% ethanol. The sample of énghich
was used in this study was completely labeled at the C-4
methyl, but only 91% labeled at the C—1 methyl, as determined
from its nmr spectrum. Thus, the nmr spectrum of éngas
similar to that of 34 (see Table I) except that it lacked
the quartet at T 8.12 (due to the c-4 allylic methyl) and
the band at T 8.45 for the C-3 allylic methyl was a singlet,
rather than a quartet. The singlet at T 8.90 was reduced
in area from 6.00H to 3.27H, consistent with 91% label at
C—1.

Rearrangement of bicyclic ketone 3Q'via Path A (see
Chart I) would give dienone 33g, labeled with CD3 groups in
the C—3 and C-6 positions, whereas Path B would furnish
dienone 49 with CD3 in the C-3 and C—5 positions. Although
dienones 33g and 49 could be differentiated on the basis of
their respective nmr spectra, the similarity in chemical
shift 0f the C—2, C-4, and C-5 allylic methyls of 33 (all
appear as a broad band at T 8.14) would make it difficult to
distinguish between égvand an isomer containing CD3 groups
at the C—3 and C-2, or C—3 and C—4 positions. Consequently,
it was decided to convert the dienone obtained on treatment
of égiwith acid to its dimethyl acetylenedicarboxylate
adduct (98) (41). The nmr bands of this compound which
correspond to the C-2, C—4, and C-5 allylic methyls of 33'

(see Table II) are well resolved; the band at T 6.34 due to

 

72

 

 

 

 

Table II. NMR Spectraa
Compound Chemical Shifts Assignments
(area)
0 6.34 (d, 6.0H) C-5,C-6 methyls
8.22 (m, 3.0H) C-8 methyl
‘3- 8.30 (m, 3.0H) C-7 methyl
8.57 (s, 3.0H) C-4 methyl
/ ‘ OOCH3 8.60 (SI 5.0H) C-i methyl
8 / 9.02 (s, 3.0H) c-5 methyls

 

 

 

 

9.08 (s, 3.0H)

6.34 (d, 6.00H)
8.22 (s, 1.67H)
8.50 (s, 1.54H)
8.57 (s, 1.45H)
8.60 (s, 1.73H)
9.02 (s, 2.88H)

9.08 (s. 3.02H)

8.12 (s, 1-53H)
8.45 (s, 1.64H)
8.78 (s, 1.46H)
8.90 (d, 4.70H)
9.08 (s, 2.94H)

8.10 (q.J=1,1.54H)

8.45 (m, 5.05H)
8.77 (s, 1.40H)
8.92 (s, 6.11H)
9.08 (s, 2.90H)

Same as for 98
M

C 4
C 3
C-5 methyl
C 1
C 6

methyl

methyl

continued

73

 

 

 

Table II —— continued
Compound Chemical Shifts Assignments
(area)
6.34 (d, 6.00H) Same as for 98

8.22 (s, 1.53H)
8.30 (s, 1.54H)
8.57 (s, 2.82H)
8.80 (s, 2.87H)
9.02 (s, 2.92H)
9.08 (s, 2.96H)

 

 

 

aAll spectra are in CCl4. Shifts are reported as T values,
with TMS as an internal reference. All spectra were run at
60 Mc. All areas are the average of three electronic inte—
gration measurements. Multiplicity of peaks shown in brac-
kets: s, singlet; d, doublet; q, quartet; m, multiplet.

bThe spectrum is normalized to a total of 12.27 protons.

CNormalized to a total of 15.00 protons.

 

74

the carbomethoxy protons of gg'can also serve as an internal
standard for measurement of the total amount of deuterium in
the molecule.

When a solution of dienone 33 was refluxed in xylene

with an equimolar amount of dimethyl acetylenedicarboxylate,

 

O O
\
R
R R R COOCH3
33,R=CH3 '98,,R-CH3
43, R = cos 98a, R = CD3

the adduct gngas obtained in good yield. Its nmr spectrum
is presented in Table II. Hexadeuterated dienone 49’was
also converted to its dimethyl acetylenedicarboxylate adduct
98a. The nmr spectrum of the adduct differed from that of
98 in that the bands at T 8.30 and 8.57 were missing and
the band at 8.22 was a sharp singlet. The high-field allylic
and bridgehead methyls of 9§Iare therefore adjacent to the
carbonyl group, a result which has also been found for the
maleic anhydride adduct of 33/(15).

Treatment of bicyclic ketone 39'with 97% sulfuric acid

at 22.60 for 30 minutes gave dienone 99, whose deuterium

content and distribution were determined by converting it to

- ....- ..____.___———. 4

75

 

0 O 5.90 0

03 1.75 \580 1.5{4W 1.73

+
. H
0 —‘>' 9' / COOCH3
CD3 1.54 .45 1.6 /
1.45 00CH3
1.87

(9.9. 39.2. .1198

the adduct 109. The nmr spectrum 0fޣ22,is presented in
Table II.

The number of protons in each band was calculated by
assigning the band at T 6.34 (due to the carbomethoxy methyls)
a value of 6.00 protons. Dienone 92 therefore has the hydro-
gen distribution shown above; since 99 contains 12.29 protons
(compare with 12.27 protons in 30) it is apparent that no
appreciable loss of deuterium occurred in the conversion of
30 to 99.

The label distribution of dienone 99 is not consistent
with its being formed by path A or B (see Chart I), but best
corresponds (within the experimental error of the nmr inte—
gration measurements of about i 8%) to a mixture of 52% 49'
and 48% 194 (see Chart II), which should have the hydrogen

distribution presented in structure 101, taking into account

0
1.89 \6.00

1.57

1.56
491-

 

76

HH

mmo H mm .moo u am . oa
moo u mm.am .mMfl
hmo n mm.am .mm
am
0
mac H mm .890 u am .8 a a
m IN.H \f)...
no u m m .mda mom
H Hm
IL
Alli.
o no

phone
8:0 H mm .800 n am
800 u mm.am
omo u mm.am
mm am
0

.ooo N am .moa
mm.am .mm.
omo n mm.am xwm
am
IIIV
luau...

77

the fact that the starting ketone §9,is only 91% labeled.

Dienones 49 and 194 may be formed gig the series of
reactions presented in Chart II: protonation of bicyclic
ketone 3Q_followed by a cyclopropylcarbinyl rearrangement
gives ion 63! which can either ring open to give‘49, or
undergo further rearrangement to 383, which can furnish
dienone 194. The almost equal mixture of 49_and 194'
(obtained on treatment of §Q_with acid) requires a rapid,
nearly complete interconversion of ions gé’and gga'prior to
ring opening to the respective dienones.

Dienone égvcould be recovered unchanged after treatment
with 97% sulfuric acid for 30 minutes at 25.50; the nmr
spectrum of the recovered dienone (which consisted of three
singlets at T 8.12, 8.17, and 8.90; the relative area of the
first two bands to the third was 1.00:0.97) was identical to
that of the starting material and therefore no scrambling of
the CD3 groups had occurred. This result demonstrates that
dienone 4§_does not revert to ion égiunder these acidic
conditions; if 33 was formed it would isomerize to gggdand
thus some dienone $94 would be produced.

In order to ascertain whether the conversion of 3Q_to
§§,(Chart II) is rapidly reversible under the isomerization
conditions, the acid-catalyzed reaction of 39’was carried
' out in such a way that the bicyclic ketone was only partially
converted to the dienone. Inspection of the nmr spectrum of

the recovered ketone would show whether or not any scrambling

 

78

of the CD3 groups had occurred. Treatment of 39’with 97%
sulfuric acid at 18.70 for 1 minute resulted in a 37% con—
version to the dienone. The nmr spectrum of the recovered
bicyclic ketone 19§_(Table II) shows it to have the label
distribution presented below, in good agreement with a 50:50

mixture of ketones 50 and 106 which would exhibit the

0 0 0
CD3 1.84 1.70 1.64 .84
+
_§L__,_
CD3 1.55 J25°94 1.50 J/8.00
1.50

1.46
50 108 109
M W M

distribution of label-presented in structure 193_(taking

into account that §Qvis 91% labeled at C-1). Further support
for the formation of a mixture of the two labeled ketones~39,
and $Q§.is the observation that the nmr bands due to the
allylic methyls are singlets, rather than quartets.

The formation of an equimolar mixture of 3Qvand 193 on
treatment of éngith acid demonstrates that there is a rapid
interconversion between 39: 68, 333, and 193’(see Chart II).

Since it has been established that in the presence of
strong acid ggiequilibrates rapidly with the isomeric ketone
193, an alternative mechanism for the conversion of bicyclic
ketone 3Q'to the dienone must be considered. Complete equi-

libration of 3Q_and 106 (see Chart II) would produce a

 

79

0 0 0
CD3 1,64 1.64 + 1.54 H \4.64 H
i, —H—->"
CD3 1.5 J/é.00 1.50 H 5-00 H
1.50 1.50 H
39.. 449 41,4

bicyclic ketone labeled as in 112: Ring opening of this ketone
followed by methyl migration to C-1 (see Chart I, Path A)

would give dienone 111. As the label distribution of 111’is
quite different from that of fig! the actual product obtained

on treatment of ég'with acid, it is concluded that the above
mechanism is not operative.

The mechanistic sequence proposed in Chart II for the
conversion of ég to §§_was further reinforced by treating
labeled bicyclic ketone 19g with acid. This ketone was pre—
pared by warming a solution of gé in deuteromethanol contain-
ing sodium methoxide. The nmr spectrum of 19? was similar

to that of 54 (Table I) except that the quartet at T 8.12

0 0
5.11 2.87 “\35.88
CH ONa +
CH30DD
1. 54 ./)5'90 1,54 2-82
1.40 1.55
‘54 102 112 .115

 

 

 

 

80

due to the C-4 methyl was missing (consistent with 100%

label at c—4) and the band at T 8.45 due to the c-s allylic
methyl was a sharp singlet. Treatment of 193 with 97%
sulfuric acid at 18.80 for 12 minutes resulted inan 85%
conversion to dienone ééé; The nmr spectrum of the recovered
bicyclic ketone 11; (see Table II) was in agreement with the
hydrogen distribution presented in 112, which corresponds,
within experimental error, to the expected equimolar mixture
of 19$ and 191 (see Chart II); this mixture would exhibit

the label distribution presented in structure 11g. The

deuterium content and distribution of the dienone (115) was

0
3.0 5.00
1.5 J/%.OO
1.50
115

determined from the nmr spectrum (Table II) of its dimethyl
acetylenedicarboxylate adduct 11$. The labeling pattern
observed corresponds to an equimolar mixture of labeled
dienones 1Q§ and 19g (Chart II) which would exhibit the label

distribution presented for structure 116.

 

81

116

M

This result further reinforces the mechanistic sequence pred
sented in Chart II; i.e., dienones 103 and 105 are formed

via ring opening of ions ééfiand 68a respectively.

B. Discussion of The Rearrangement Mechanism

The labeling experiments described above agree with
the proposal that the acid-catalyzed conversion of\§é_to:§i
proceeds yia the mechanistic scheme presented in Chart II;
§§vrearranges to 88! which equilibrates with its enantiomeric
form 888 prior to ring opening to furnish dienone éé: The
conversion of éé'to §§Vhas also been shown to be reversible;
ring Opening of §§~to §§~is not reversible, however. The
results obtained also demonstrate that the conversion of éé’
to éé does not involve the series of rearrangements presented

below.

H 3 ””CH3;'

3’3. .12., 52.1

 

82

The sequence éév-—>-§§v—+r‘§§’is analogous to the mechanis—
tic scheme proposed by Dauben and Friedrich (58) to account
for the acid—catalyzed rearrangement of dideuterothujopsene
(fig) to widdrol (2%) (see Introduction).

The conversion of bicycloTS.1.0]hexenone §é_to dienone §§J

via ion gglrequires that bond b (structure éfi)‘be broken and

54 68 53

M M M

 

that a new bond be formed between the carbonyl carbon and

the quaternary carbon at C-6. Two possible routes can be
written to describe this process: (1) After (or simultane—
ously with) fission of bond b_there is bond formation between
the C-2 and C-6 carbon atoms of 61, (2) Migration of C-6

to the carbonyl group, followed by a second migration to C—1,
to give the ionic species 68. Cleavage of the cyclopropane
ring gives dienone QQ. It should be noted that exo and endo
substituents on C—6 may invert, or retain their stereo-
chemistry, depending on whether or not there is rotation

about the C1-C6 bond in the rearrangement of 61 to 66,

 

83

OH

 

0 \ OH
\
1 0*

A decision between paths 1 and 2 may be based on the
following results. It was recently discovered in our labor-
atories (42) that treatment of 3% with methylmagnesium
iodide, followed by hydrolysis with saturated ammonium
chloride solution gave 139 as the only product. This reaction
may proceed yia the mechanistic scheme presented below. Ring

opening of 119 to give 119a is analogous to the conversion

of 34 to 117 proposed in mechanism 1.
MM

#44 W—Hiajif—i;

120
[‘3’4’ 119 119a

’W W

84

There is also precedent for ring closure of 111’to give 6g;
DeVries has shown (43) that the cyclopentadienyl derivative
121 undergoes ring closure to the bicyclo[3.1.0]hexene 122’
in pyridine at room temperature. The double bonds may

participate in aiding the departure of the brosylate anion.
CH2 -0B:s

. . +
Pyridine; O -H

121 122

M M

 

The failure of 1129 to undergo a similar ring closure is
probably due to the large amount of water present which can
act as a base and remove a proton, thereby affording 129.
It is clear that mechanism 1 agrees with these results inas-
much as there is a precedent for each step involved. On the
other hand, mechanism 2 requires the formation of the highly
strained intermediate 118. Since path 1 does offer a satis—
factory rationalization for the rearrangement of‘fil to §§,
there appears to be no need to invoke the unusual reactions
required for path 2. A mechanistic sequence similar to that
presented in path 1 can also account for the interconversion
of 68 and its enantiomeric form 6 a (see Chart II).

The conversion of ion‘gl to 111 requires that there be
considerable overlap of bond b_with the positive charge on

C-4. This is in agreement with recent reports (33) that

 

85

there is a substantial amount of positive charge on the
B—carbon atoms of cyclopropylcarbinyl cations. This can be
accounted for by resonance forms such as 123a. Evidence

has also been presented which demonstrates that cyclopropyl-

B H +
D.— c512+ <——> l/=
a
8
125 123a

carbinyl cations have the bisected geometry shown for struc—
ture 12§’(44). This geometry is most favorable for overlap
of bond a_with the positively charged carbon atom.

It is therefore clear why rearrangement of ion gl’to
111 is favored over ring opening to ion 12; Bond b_is ideally
situated for overlap with the positively charged C—4 carbon
of 61,‘whereas bond 3 is not.

It is of interest that a few reported examples of acid-
catalyzed bicyclo[3.1.0]hexenone reactions do not involve
a cyclopropylcarbinyl rearrangement of the type_§4’-9~§§L
but can be better explained by cleavage of the cyclopropane
ring to give products corresponding to formation of the most
stable cation.

Eistert and Langbein have shown (45), for example, that

treatment of bicyclic ketone Z§Jwith sulfuric acid gave the

isomeric phenol 74, rather than 126.

 

 

86

OH

’1

OH

 

Ring opening of 124 to give 1243 is therefore much more rapid
than the possible rearrangement to 12g; The fast rate of
ring opening is probably a consequence of the presence of
phenyl groups on the C—1 and C—5 positions of 13; The con-
version of 124 to igéa'must involve development of positive
charge on either the C—1 or C—5 position of ion 124J(depend-
ing on which way the C-1, C-5 bond cleaves); the presence of
phenyl groups at these positions stabilizes the developing
positive charge and thereby enhances the rate of the reaction.
Support for this argument is the report by Walborsky and
Plonsker (37) that the rate of ring opening Of cycloprepyl
ketones is greatly enhanced by the presence of phenyl groups

on the 3— or 4—position of the cyclOprOpane ring (see below).

4

 

87

A similar argument would account for the acid-catalyzed
conversion of bicyclic ketone 121,to 12§_(and not 129) re—
ported by Zimmerman and his co-workers (12). The phenyl
groups on the C-6 position of 12] stabilize the positive
charge developed on this position in the transition state

going to 127a, thereby favoring this process over rearrange—

ment to 127b.

OH O

 

 

a

128 0
W
OH

I,

d
129
W

In the case of bicyclic ketone 34! however, a cyclo-
propylcarbinyl rearrangement to 68,might be expected to com-

pete with ring opening to 22} as the methyl groups on 34

 

88

OH

\
. y g
0 \K .4

34

W

72
M
would not favor direct cleavage of the cyclopropane ring to

the extent that the phenyl groups on 1§_and 119 would; a
phenyl group may be more effective at stabilizing an adjacent

positive charge than a methyl group.

C. Miscellaneous Experiments

1. Acid-Catalyzed Rearrangements of Fully
Substituted Cyclohexadienones

The most commonly observed reaction of 2,4- or 2,5-cyclo-

hexadienones on treatment with acid is the dienone-phenol

rearrangement. This reaction may be repreSented by the
equations:
0 R OH
H+ R
R ,
R
130 131

 

 

89

3 OH
_H?;.
R R R
E2 444

The rearrangement involves protonation of the dienone
on oxygen followed by alkyl or aryl migration to afford the
isomeric phenol. Two representative examples are the con—
version of 1§Q_(R = CH3) to 131’on treatment with sulfuric
acid reported by Marvel and Magoon (46), and the formation
of phenol 1§§don treatment of 1321(R = 0) with hydrochloric
acid reported by Zimmerman and Schuster (47).

This reaction is not available to the isomeric dienones
§§,(15) and §§L(14), however, since these compounds are fully
substituted with methyl groups and therefore cannot form
phenols on treatment with acid. A possible mode of rearrange-

ment for these dienones is presented below.

O OH O
+ +
H : H 3 \
T— r—
33 72 35

It is seen that these two compounds should be readily inter—
convertible under acidic conditions via a series of methyl

migrations.

 

90

In view of the known stability of cross-conjugated versus
fully conjugated dienones (see Part I of this thesis), it was
thought that 33, on treatment with acid, would furnish
dienone 35; This hope was fulfilled. The acid-catalyzed
rearrangement of ééito_§5’and the interesting behavior of~3§’
in concentrated sulfuric acid is reported below.

a. The Acid-Catalyzed Rearrangement of 2,3,4,5,6,6-
Hexamethyl—2,4jgyclohexadienone (33L

Treatment of‘fiéawith fuming sulfuric acid for 10 minutes
at room temperature, followed by quenching on ice and work-up,
gave dienone §§vin good yield. This compound was isolated
as a white solid (mp 117.5-118.20). Its nmr (Table I) and
infrared spectra were identical to those of 35 prepared by

treatment of égpwith dilute acid.

OH O
+
CH3O i)
4.4, 55

The conversion of 33’to‘35 also occurs slowly in 97%
sulfuric acid. The first-order rate constant for the re-
action, measured in 97.0% sulfuric acid at 48.20, was found
to be 2f06 i..01 x 10‘2 min‘l. The percent conversion ofigi
to 35, as determined by the uv spectrum of the acid solution

at the end of the kinetic run, was essentially quantitative.

 

91

The conversion of 33 to 35 can be accounted for by the
MM

mechanistic scheme presented below: rapid, reversible

O OH OH
H+ ’VCH ”VCH
.__x .___4;> ____&fi>
‘r—' k1 k2
3‘73 0H 2,845 1,20
+
’ "H
——3
155 55

protonation of ééito giveééé, followed by a series of

methyl migrations to furnish dienone 35: The first—order
kinetics are in agreement with either k1 or k2 being the rate—
determining step. Evidence that k1 is the rate-determining
step is the observation (presented in Part I of this thesis)

that the acid-catalyzed conversion of £2,t0 35, which

OH OH
H+ N CH
——$h -:fi;39>

.49, ‘ 72 4:2,

 

92

presumably involves ion 72 as an intermediate, proceeds
rapidly under very mildly acidic conditions (methanol contain-
ing a few drops of 1% HCl).

b. The Ultraviolet Spectrum of 2,3,4,5,6,6-

Hexamethy1-2,4-cyclohexadienone (33) in
Sulfuric Acid

 

The uv spectrum of dienone éévwas studied as a function
of acid concentration. The results are presented in Table III.
The marked difference in the spectrum of 33 in 10.0 and 77.9%
sulfuric acid strongly suggests that, in the latter solvent,
éélexists as the protonated ion 134. Since there is no

Table III. Ultraviolet Spectrum of 2,3,4,5,6,6—Hexamethyl—
2,4-cyclohexadienone (33) in Sulfuric Acid

 

 

% H2304 xmax,mu (e) kmax,mu (6)
10.0 255(2,550) 552(5,750)
77.9 266(4,680) 402(9,750)
97.0 266(4,690) 402(9,750)

 

appreciable difference between the uv spectrum of éévin 77.9%

OH

 

93

and 97.0% sulfuric acid, it is further concluded that the
dienone is 100% protonated in this region. The uv spectrum

of 134 is quite similar to that of the structurally related
WV

Cl

ax at 287 mu

heptamethylbenzenonium ion (12) which has A:

70

M

(6 6,760) and 397 mu (6 8,500) (19).
c. The Behavior of 2,3,4,4,5,6-Hexamethyezzfi-
cyclohexadienone (35) in Sulfuric Acid

Dienone §§Iis quite stable in concentrated sulfuric
acid; the nmr spectrum of a solution of §§Iin 98% sulfuric
acid exhibited no new bands after 56 hours at room tempera-
ture. The ultraviolet spectrum of §§'in different concen-
trations of sulfuric acid is presented in Table IV. The
marked change in the spectrum with increasing acid concen-
tration is attributed to a reversible protonation of_§§’to
give ion 135. The pKa of §§Jwas calculated to be —2.6 (see
Experimental).

OH

94

Table IV. Ultraviolet Spectrum of 2,3,4,445,6—Hexamethyl—
2,5-cyclohexadienone (35) in Sulfuric Acid

% H2804

kmax,mu (e)

xmax,mu (e)

 

10.0
20.0
30.0
39.0
49.0
59.0
68.0
78.0
97.0

105.0

254(14,500)
254(14,100)
257(12,500)
265(11,700)
270(12,600)
271(12,600)
271(12,500)
271(12,500)
271(12,500)

271(11,300)

555(4,700)
555(7,100)
336(7,800)
338(7,9SO)
338(8,150)
340(8,200)

340(7,900)

 

The nmr spectrum of 35 in fuming sulfuric acid exhibited

three singlets at T 7.55,

7.80, and 8.57 (with respect to

internal N(CH3)4BF4 at T 6.87) with relative areas of 1:1:1.

These are assigned to the C-3, C—5 methyls, the C—2, C—6

methyls, and the C—4 methyls of 135, respectively. The

spectrum was quite clean and showed no other bands besides

those due to 135.

d. Acid-Catalyzed Rearrangement of 2,4,4,6—Tetra-
methyl-3,5-bis(trideuteromethyl)-2,5-cyclo-

hexadienone (44)

The intriguing possibility exists that 135 may equili—

brate with ions 72 and 134 (see Chart III) via a series of

 

95

O OH OH OH
+
EH5. __;x :23: R
T_”' xr—-
R
R R R ' ' R
E3 R = CH3 155 72 154
44, R = CD3
W
Chart III

methyl migrations. The existence of such an equilibrium

could be determined by treatment of dienone éédwith sulfuric
acid and inspecting the nmr spectrum of the recovered dienone.
If 1§§,(R = CD3) equilibrates with‘lg, some label would end
up in the C—4 position of the recovered dienone; similarly,
equilibration of 12_(R = CD3) with ion 134 would introduce
label into the C—2 and C-6 positions of the dienone.

Treatment of hexadeuterated dienone éguwith fuming sul—
furic acid for 5 minutes at room temperature gave dienone 136,
whose nmr spectrum in CCl4 consisted of three singlets at
T 8.03 (C-3, C-5 methyls), 8.15 (C—2, C—6 methyls), and 8.78
(C-4 methyls) with relative areas 1:1:1. The mass Spectra
of dienones 44/and_1§§l(see Experimental) were essentially
identical in the high mass/charge region; therefore no
deuterium was lost in the conversion offigg -f>-1§§’and the

recovered dienone has the hydrogen distribution presented below.

 

96

0
2.0 2.0
2 0 2.0
4.0
136

The complete scrambling of the CD3 groups in 136~demonstrates
that there is indeed a rapid equilibration between ions 135,
121 and 134 in fuming sulfuric acid.

This equilibrium also occurs in 70% and 98% sulfuric
The results obtained on treat-

acid, but much more slowly.

ment of 44 with these solvents are summarized in Tables V

 

 

 

and VI.
Table V. NMR Spectrum of 2,4,4,6—Tetramethyl—3,5—bis—
(trideuteromethyl)-2,5-cyclohexadienone (3&9
in 70% Sulfuric Acid at 25 a
Time 7.65(C-3,C-5) 7.92(c—2,c-5) 8.66(C-4) protons/
molecule
30 min. 2.18H 5.90H 6.08H 14.65
18 hrs. 2.34H 6.17H 5.74H 14.88
47 hrs. 2.24H 6.11H 5.80H 14.01
44 days 4.40H 5.79H 4.29H 14.30

 

aShifts are reported as T values, with N(CH3)4BF4(T = 6.87)
as an internal reference. All spectra were run at 60 Mc.
In each case, the number of protons per molecule was normal—
ized to a total of 14.16, in agreement with 2.16H at the C-3
and C-5 positions of 44;

bCalculated using a known weight of internal standard and
dienone.

97

Table VI. NMR Spectrum of 2,4,4,6-Tetramethy1-3,5-bis-
(trideuteromethyl)-2,5-cyc$ohexadienone (éflfl
in 98% Sulfuric Acid at 25 a

 

 

 

Time 7.61(C-3,C-5) 7.84(C-2,C-6) 8.59(C-4) protons/5
molecule

30 min. 2.11H 6.00H 6.03H 15.15

56 hrs. 2.22H 6.09H 5.84H 15.01

56 hrs.c+ 3.52H 6.12H 4.50H 14.64

20 mi8.

at 80

86 hrs.C+ 4.79H 4.98H 4.57H 15.01

810 min.

at 80

 

aShifts are reported as T values, with N(CH3)4BF4(T = 6.87) as
an internal reference. All spectra were run at 60Mc. In each
case, the number of protons per molecule was normalized to a
total of 14.16, in agreement with 2.16H at the C—3 and C-5
positions of 44;

bCalculated using a known weight of internal standard and
dienone.

. . o
CAmount of time sample remained at 25 .

The sample of éédused was only 64.0% labeled in the C—3 and
C-5 positions, as judged from its nmr spectrum. Thus the nmr
spectrum of 44 in CC14 was similar to that of 35’(Table I)
except that the band at T 8.05 (due to the C-3 and C—5 methyls)
was reduced in area from 6H to 2.16H, consistent with 64.0%
label at these positions.

The change in the nmr spectrum of éépin 70% sulfuric acid
(Table V) after 44 days can be rationalized by the mechanistic

scheme presented above for 44J(Chart III). The essentially

 

98

complete scrambling of the label between the C-3, C-5 posié
tions (T 7.65) and the C-4 position (T 8.66) observed after

44 days can be explained by equilibration of ions 135 (R = CD3)
and 12; The small change in area of the band at T 7.92 (due
to the C-2,C-6 methyls) demonstrates that only a small amount
of the label (CD3) is incorporated into these positions; thus
equilibration of ions 135 (R = CD3) and lgvis much more rapid
than interconversion of 72 and 134 in this solvent.

Similar results were obtained on heating a solution of
éépin 98% sulfuric acid (Table VI). Prolonged heating at 800
resulted in incorporation of a substantial amount of the
label into the C-2, C—6 positions (T 7.84) of the dienone; the
label was introduced much more rapidly into the C-4 positions
(T 8.59), however, in agreement with the results obtained
for 44’in 70% sulfuric acid.

It is of interest at this point to consider the relative
stabilities of ions 12, 134, and 135 (Chart III). The results
obtained on treatment of éé’with fuming sulfuric acid demon-
strate that these ions equilibrate rapidly in this solvent;
the nmr spectrum of §§’in fuming sulfuric acid, however, is
in agreement with its being completely converted to ion 135;
Similarly, the uncomplicated nmr spectrum of 44 in 70% and
98% sulfuric acid (Tables V, VI) is consistent with that of
the symmetrical ion 135, rather than the unsymmetrical species
zgvand 13%, even though the observed scrabbling of the label

suggests that the interconversion of ions 72¢ 134, and 135

99

is occurring. Hence, it is concluded that the ionic species
135 is more stable than either 72 or 134.
M MM

McLean and Mackor have calculated that the positive
charge in ions such as 135 is concentrated mainly on the C—1

position (25); the presence of hydroxyl at this position in

+
0H

 

135

M

135, compared with methyl at this position in ions lavand
‘134/ may account for the stability of éééiwith respect to
the two latter ions. The positive charge can be delocalized
onto the hydroxyl group of 135 as shown above, thereby en-
hancing the stability of this ion.

A brief comment seems in order regarding the extremely
rapid scrambling of the label observed on treatment of dienone
éévwith fuming sulfuric acid. One possible explanation is
that 44’may be doubly protonated in this strongly acidic

medium to give ion 137, which undergoes rapid equilibration

+
0H2

137

 

llllll'l! 1

100
with the diprotonated form of ions 12 and 134.

2. Pyrolysis of 1,3,4L5,6,64Hexamethylbi—
cycloi3.1.0]hex-3—en-2-one (34)
Wheeler and Eastman reported (3) that pyrolysis of
umbellulone 9&2 gave thymol (2) as the major product, along

with 5—10% sym—thymol (138). They proposed that the reaction

proceeded via the ionic pathway presented below.

i (Di-:1} 7‘

138

M

1

’W

The possibility that similar rearrangements might be
observed for bicyclic ketone 34’led us to carry out the
pyrolysis of this compound.

When gé’was injected on a gas chromatograph (injector
temp. = 3900, column temp. = 1700) two major peaks were ob—
served; collection of the product showed that the one with
the shortest retention time was due to recovered enone 34,
and the other was due to dienone 33. The dienone was identi-

fied by its infrared spectrum and retention time, which

 

101

33

M

were identical to those of an authentic sample (15). Injection
of éé’on a gas chromatograph with the injector temp at 2500
(column temp = 1700) gave no peak due to 33; this demonstrates
that the rearrangement of 34’t0.33’is occurring in the injector
block, and not on the VPC column.

The conversion of 34 to éé’can be formulated as a cyclo-
propylcarbinyl rearrangement analogous to the mechanism of
the acid-catalyzed conversion of 34Ito~§§l(see Chart II), in
which the zwitterionic species éé'rearranges to 139/ which
equilibrates completely with its enantiomer 1399 prior to ring
Opening to give 33; It is seen that, if the C-1 and C—4
positions of 34’are labeled with CD3, this mechanistic sequence
would give hexadeuter01§§,with the label equally distributed
among the C-2, C-3, C-4, and C—5 positions (1,§,, a 50—50

mixture of dienones ég/and 104).

 

102

 

 

O
R R
g 0
R R
139 139a
M ML
R \3 O O
I flF———— R
6 COOMe
. / D D
R R .
COOMe ,
98a, R = CD3 §Q§J R = CH3 R
_ 53 R = CH3
31% R '— CD3 ~ A“,
10 , R = CD3

Bicyclic ketone\59’was pyrolyzed as described above for
the unlabeled compound. This gave the hexadeuterated dienone
32/ whose deuterium content and distribution were determined
by converting it to the adduct 983, the nmr spectrum of which
lacked the bands at T 8.30 (due to the allylic methyl at C—7;
the other allylic methyl, T 8.22, was now a singlet rather
than a multiplet) and T 8.57 (due to the C-4 bridgehead
methyl). Comparison of the area of the band at T 6.34 (due
to the C—5, C-6 carbomethoxy protons) with the remaining bands
in the Spectrum showed that the dienone contained 12 protons
per molecule; thus, no deuterium was lost during the pyrolysis.
The nmr spectrum of recovered §Qlwas essentially identical to
that of the starting material. Therefore no scrambling of

the label occurs under the pyrolysis conditions.

 

103

The labeling results prove that the reaction does not
involve an equilibrium of the type 139 :fo1329, analogous
to that found for the acid-catalyzed rearrangement of\34 to
‘33, The results are, however, in agreement with the mechanis—
tic sequence 34’——>{§§’-%>'132"P*133, or a similar sequence
involving free radical, rather than ionic intermediates.
Finally, it should be pointed out that the rearrangement
of 340to 33’is analogous to the thermal conversion of
umbellulone (1) to sym—thymol (which may also proceed 113
a free radical path). Phenol formation in the latter case is
due to the presence of hydrogen atoms on the ring carbons

(compare with 34) which are easily removed.

 

EXPERIMENTAL

A. Acid-Catalyzed Rearrangement of 1,3L4,5,6,6-

Hexamethylbicycloj3.1.0lhex-3-en—2-one (341

A solution of 0.214 g of~34 (15) in 97% sulfuric acid

was kept at 22.60 for 30 min, and then poured on ice. The
resulting solution was extracted with methylene chloride,
and the latter solution was washed with water and dried over
anhydrous magnesium sulfate. Evaporation of the solvent
ig_y§ggg_gave 0.203 g of amber colored oil which was shown
by vpc to consist of 2% 34 and 98% 2,3,4,5,6,6-hexamethvl-

2,4-cyclohexadienone (33). The nmr Spectrum (Table I) and ir

 

spectrum of_33’were identical with those of an authentic

sample (15).

B. 1,3L3,4,7,8-Hexamethyl-5,6-dicarbomethoxy-
bicyclo[2.2.21octa—5L7-dien—2-one CQQL

 

 

The dienone 33’(0.140 g) in 4 ml of xylene was heated
under reflux with 0.113 g of dimethyl acetylenedicarboxylate:
for 22 hrs. Preparative vpc afforded 0.080 g of~9§’(41) as
a clear oil.

The nmr spectrum is presented in Table II. Hexadeutero-
33, labeled at C—3 and C—5 with CD3 groups, was also con—
verted to its dimethyl acetylenedicarboxylate adduct (983).
The nmr spectrum differed from that of unlabeled 9§’(Table II)

in the following ways: the bands at T 8.30 and 8.57 were

104

 

105

absent, and the band at 8.22 was a singlet. Thus, the high-

field allylic and bridgehead methyls are adjacent to the

carbonyl group.

C. Acid-Catalyzed Rearrangement of Labeled Bicyclo-
[5.1.0]hexenones, General Procedure

The acid was equilibrated to the reaction temperature
prior to each run. All reactions were run in a constant
temperature bath with occasional stirring. Control runs
showed that labeled dienone gg, enoneԤ9/ and Diels-Alder
adduct §§9 suffered no scrambling or loss of the label when
subject to the same vpc conditions (S-ft 20% SE-SO column at.
165-ZZOO) used to collect the products of the acid-catalyzed
rearrangements of the labeled bicyclo[5.1.0]hexenones.

D. Acid-Catalyzed Rearrangement of 5,5L6,6-Tetra-
methyl-1,4-bis(trideuteromethyl)bicycloi5.1.0]—
hex-5-en-2-one Q§QL

Labeled enone égflwas prepared as previously described
(15), except that 95% ethanol was used as the photolysis
solvent. The nmr spectrum of vpc purified material consisted
of four singlets at T 8.45 (C-S methyl), 8.78 (C-S methyl)
and 8.90, 9.08 (gem-methyls) with relative areas 1.00:0.91:
1.06:1.00, respectively. A comparison of the areas of the
bands at T 8.45, 8.78, and 9.08 with that of the band at

8.90 indicates that the C-1 methyl is 91% deuterated (0.27H

at C-1). The absence of the band at T 8.12 (due to the C-4

methyl) demonstrates that this position is 100% labeled.

Enone §Q (0.908 g) was dissolved in 25 ml of 97% sulfuric

 

106

acid and the resulting solution kept at 22.60 for 30 min.
Quenching on ice, followed by the usual work-up, gave 0.931
g of a yellow oil. Preparative vpc afforded 0.184 g of
hexadeuterated dienone 99, which was converted to its
dimethyl. acetylenedicarboxylate adduct 199; The adduct was
purified by vpc, its nmr spectrum is reported in Table II
and Figure~§s

A solution of 0.251 g of éQ/in 10 ml of 97% sulfuric
acid was kept at 18.70 for 1 min. Quenching on ice, followed
by the usual work-up yielded 0.216 g of a yellow oil which
was shown by vpc to consist of 63% hexadeuterated enone 198,
and 37% hexadeuterotfig, The nmr spectrum of_198’is reported
in Table II and Figure 9.
E. Treatment of 2,4,6,6~Tetramethyl-3,S-dimethyl-

d3—2L4—cvclohexadienone (49) with 97% Sulfuric
A_c_ig

The dienone 4§v(0.192 g) was dissolved in 5 ml of 97%
H2804 and the resulting solution kept at 22.50 for 34 min.
Quenching on ice, followed by the usual work-up gave 0.187 g
of a yellow oil. Preparative vpc afforded dienone 49., The
nmr spectrum was identical to that of the starting material
and consisted of three singlets at T 8.12, 8.17, and 8.90.
The relative area of the two bands at T 8.12 and 8.17 to the

band at 8.90 is 1.00:0.97.

 

107

.ooa uoDUpm mpmHhxonnmo
Iflpmcoahuoom ngumfiap pmuwumusmpmxmn mo Esuuommm HEZ .m musmflm

«8... use 3.0 P... 05 .35 $3

 

 

 

 

 

 

 

 

 

108

\i‘.

.moa wcouox UHHUNUfiQ Umumumpsmpmxmfi mo Esnuommm HEZ .m musmflm

L. 8.: 2.3 g.» 2.6 «Q
. 1

 

 

 

1 ‘
j n‘ _ 1 ‘

1
_.
I

-‘_.._.————-.

 

 

 

109
F. Preparation of 1,3,5,6,6-Pentamethyl-4-methyl—
darbicyclo[3.1.0]hex-3-en-2-oneL19§L

The enone 34’(2.09 g) was dissolved in 15 ml of methanol-
d (31) containing a 0.5—cm3 piece of sodium. The resulting
solution was refluxed for 65 min, after which time annmr
spectrum showed that exchange was not quite complete. Most
of the methanol was evaporated i§_y§gug, and 10 more ml of
methanol-d was added. The resulting solution was refluxed
for 70 min. Most of the methanol was evaporated i§_y§gug,
and the remaining solution was poured into methylene chloride.
This was extracted with cold (00) water to remove the remain-
ing methanol, and the solution was dried over anhydrous mag-
nesium sulfate. Evaporation of the solvent afforded an oil
(2.09 g), which on distillation at 1 mm gave 1.47 g (70%)
of 192 with hp 66-680. The nmr spectrum of_192 differed from
that of_34 (Table I) in that the band at T 8.12 was missing
and the band at T 8.45 was a sharp singlet-

G. Acid—Catalyzed Rearrangement of 1,3,5,6,6-
Pentamethyl-4—methy1-d3-bicyc1013.1.0]hex-
3-en—2-one (193)

A solution of enone 192 (0.937 g) in 25 ml of 97% sulfuric
acid was allowed to remain at 18.80 for 15'min, after which
time the solution was poured on ice.and worked-up. The product,
0.900 g of yellow oil, was shown by vpc to contain 17% tri-
deuterated enone 112 and 83% dienone 113; Preparative vpc
afforded 0.059 g 112 and 0.189 g 113; The nmr spectrum of

enone 112 is reported in Table II and Figure 10. Dienone 113

110

\‘n‘.
Nda maoumx Uflaohofln pmumumusmpfluu mo Esuuommm H82

«3 «5 b5 2;
2.»

.OH mnsmflm

 

 

 

 

 

111

was converted to its dimethyl acetylenedicarboxylate adduct
114 which was purified by vpc. Its nmr spectrum is reported

in Table II and Figure 11.

H . 2., 3,74 . 4 ,5 ,6;Hexamethyl-2 , 5-c1c lohexadienonefl
2,3,4,5,6,6-Hexamethyl-2,4-cyclohexadienone (fig), 15.2

g, was added slowly to 70 ml of 20% fuming sulfuric acid at

00. The resulting brown solution was stirred at room tempera-

ture for 10 min, then poured on ice. After dilution to 2

liters with ice water, and filtration, the resulting tan solid

vvas dissolved in ether. The ether solution was washed with

water and dried over magnesium sulfate. Recrystallization

from hexane gave 8.76 g of colorless crystals of 2,3,4,4,5,6-

 

hexamethyl-Z,5-cyclohexadienone (35), mp 117.5-118.2O.

 

Anal. Calcd. for C12H180: C, 80.85; H, 10.18.
Found: C, 80.91; H, 10.11.
. . ' -1 EtOH
§§Jhad prominent 1r bands at 1653 and 1624 cm , and a kmax
at 246 mu (e = 14,800). Its nmr spectrum is in Table I and
Figure 4.
I. Kinetics of the Acid-Catalyzed Rearrangement_
of 2,3,4,5,6,6-Hexamethy1—2,4-cyclohexadi-
enone 33)
The rearrangement of dienone éfilto 35 was followed
spectrophotometrically (ultraviolet), by measuring the de—
crease in absorbance of 33 at 402 mu. A solution of 33 in 95%

ethanol was added to a solution of 97% sulfuric acid at 48.20.

The concentration of the stock solution was such that the

112

\n‘fl.‘
.fifia
onwathQHmoHUwcmahuwom Hmnumfiflp Umumumusmpfluu mo Esuuom

uUDUpm
mm H52

.6 no... dog. cit hm.“ 04.6 dd .9

.fifi musmflm

 

 

_ _ a 4 _ _

 

 

 

 

 

 

 

 

 

113

initial absorbance was 0.90 and the acid concentration was

97.0%. The reaction was followed to at least 75% completion

and the first-order rate constant determined from a plot of

log A(% 402 mu) versus time, which was linear up to about

2 half lives. The pseudo first-order rate constant k was

0.02055 1.0.00005 min‘l. Comparison of the uv spectrum

before, and after, one kinetic run, showed that the percent

conversion of éélto 35 was 98%.

J. Measurement of thenga of 2,3,4,4,5,6-Hexa-
methyl-2,5-cyclohexadienone (35L *1

x
The spectra of various sulfuric acid solutions 1.068 x 10‘

4
molar in dienone were measured at A«250. The sample of_§§,
used was twice recrystallized from hexane (mp 117.2-118.00).
The solutions were prepared by adding 8 ul of a stock solution
of §§’in 95% ethanol to 5 ml of acid. The spectrum was
measured as soon as possible (/»'5 min) after preparation of
each solution. The Spectral data are summarized in Tables IV
and VII.

The first method of Hammett, Flexser, and Dingwall (48)
was used to calculate pKa. Values of Ka and €BH+ at various

wavelengths in solutions of known ho values (49) were obtained

by simultaneous solutions of the equation:

 

+ hQ _ he =
Ka+€BH ( €B_€ ) —Q——(€B_€) 0 (1)

The wavelengths were chosen in the region in which the
6B curve (the plot of A vs % in 10% sulfuric acid) was rela-

tively flat, and therefore the effect of a lateral shift in

 

 

114

the spectrum caused by a medium effect was small. The two
values of e, the extinction coefficient measured at a parti—
cular wavelength, used in each solution of equation (1),
were those measured in 30.0 and 39.0% sulfuric acid. The
value obtained for pKa of 35 was -2.57 i 0.11. In Table VII
are summarized the spectral measurements used in calculating
pKa.
K. Rearrangement of 2,4,4,6-Tetramethy1—3,5-bis-
(trideuteromethyl)-2,5-cyclohexadienone (44L
in Fuming Sulfuric Acid

Dienone 44’was prepared as described above. Its nmr
spectrum consisted of two singlets at T 8.17 and 8.79 with
relative areas 1:1. The band at T 8.05 due to the C-3, C-5
methyls was reduced in size to a small bump, consistent with
essentially complete label at these positions. The uncor-
rected mass spectrum of 44’showed it to contain 1.6%-d2,
7.15%-d3, 22.2%—d4, 36.5%-d5, 28.6%-d6, and 4.0%-d7.

Dienone 44’(.0457 g) was added to 1 ml of 20% fuming
sulfuric acid at 250. The resulting amber solution was kept
at room temperature for 5 min, then poured on ice, Work-up,
followed by preparative Vpc, afforded .0212 g of hexadeutero-
dienone 136. The uncorrected mass spectrum of 136’showed it
to contain 1.4%—d2, 7.6%-d3, 22.0%-d4, 36.5%-d5, 28.3%—d6,
and 4.1%-d7. The nmr spectrum of 136 (CCl4) consisted of
three singlets at T 8.03, 8.15, and 8.78 with relative areas
1:1:1. Dienone ééawas recovered unchanged (nmr) when subjected

to vpc conditions used to collect 136.

 

 

115

Table VII. Extinction Coefficients of 2,3,4,4,5,6-Hexamethy1-
2,5-cyclohexadienone (35) in H2SO4 + H20 Mixtures

 

%H2S04 i=527 mu x=550 mu i=555~mu 7x=358 mu 1:540 mu
10.0 188 281 94 94 94
20.0 488 488 575 575 281
50.0 1890 1690 1590 1500 1410
59.0 4500 4590 4880 4590 4210
49.0 8480 6930 7120 7120 8740
59.0 8880 7210 7590 7890 7500
88.0 8580 7210 7590 7880 7980
78.0 6560 7210 7590 7980 8180
97.0 8190 7120 7590 7980 8250

 

Table VIII. Extinction Coefficients and pKa Values for
2,3,4,4,5,6-Hexamethyl-2,S-cyclohexadienone (QQ)
in H2804 + H20 Mixtures

 

 

 

% H2804
. 10.0 30.0 39.0
%(mu) 5 e e pKa
327 188 1690 4500 -2.49
330 281 1690 4590 -2.38
333 94 1590 4680 -2.62
336 94 1500 4590 -2.70
340 94 1410 4210 -2.65
Avg. —2.57 1.0.11

 

 

 

‘ 116

L. NMR Study of 2,4,4,6-Tetramethyl-3,57
bis(trideuteromethyl)-2i5-cyclohexadi-
enone (44) in 70%and 98% Sulfuric Acid

 

Dienone 44’was prepared as described above. The sample
of 44/ used was 64.0% labeled in the c-s and c-5 positions;
thus, its nmr spectrum was similar to the undeuterated dienone
§§’(Table I) except that the band at T 8.05 (due to the C-3
and C-5 methyls) was reduced in area from 6H to 2.16H, con-

sistent with 64.0% label at these positions.

1. 70% Sulfuric Acid

Dienone 44 (0.0924 g) and N(CH3)4BF4 (0.0368 g) were
dissolved inv~/0.5 ml 70% sulfuric acid. The sample was
stored at room temperature in an nmr-tube and the nmr spectrum

run at various times (see Table V for nmr spectra).

2. 98% Sulfuric Acid

Dienone 44’(0.0761 g) and N(CH3)4BF4 (0.0293 g) were dis-
solved inz~r0.3 ml of 98% sulfuric acid. The sample was
stored at room temperature in an nmr—tube, except for when it
was heated to 800 in an oil bath. The nmr spectra are pre—
sented in Table VI.
M. Pyrolysis of 1,3,4,5,6,6-Hexamethy1bigyclo—

[3.1.01hex-3—en-2-one (34)

When égjwas injected on a gas chromatograph (5—ft SE-30
column, column temp = 1700, injector temp = 3900) two major
peaks were observed; collection of the products showed that

the peak with the shortest retention time was due to enone

 

117

‘34, while the other peak was due to 2,3,4,5,6,6—hexamethyl—

2,4-cyclohexadienone (33). The infrared spectrum of‘fié,

(V0014

max = 1642) and retention (5.8 min) were identical to

those of an authentic sample of‘§§'(15). Injection of‘34'on
a gas chromatograph (column temp = 170, injector temp = 2500)
gave only one peak due to 34; no peak for 33'was observed.

N. Pyrolysis of 3,5,6,6-Tetramethyl-1,4-bis-
(trideuteromethyl)bicyclo[3.1.0]hex-3-en—

2-one Q§92

Labeled enone 59’was prepared as previously described
(15). Its nmr spectrum showed that the label was complete
at the C—4 methyl, and 91% complete at the C-1 methyl. The
enone was pyrolyzed as described above for the unlabeled
compound. The vpc trace of the product showed it to consist
of 31% enone 59, and 69% dienone‘gg, labeled at the Ce3, C-5
positions with CD3. The nmr spectrum of unreacted §Q_was
essentially identical with that of the starting material.
The dienone was converted to its dimethyl acetylenedicar—.
boxylate adduct 98a as described above for the unlabeled
compound. The nmr spectrum of the adduct 383 differed from
that of unlabeled §§J(Table II) in the following ways: the
bands at T 8.30 and 8.57 were missing and the band at T 8.22
was a singlet. The ratio of the area of the nmr band at T 6.34

to the rest of the bands in the spectrum was 6:12.

SUMMARY

Treatment of 1,3,4,5,6,6-hexamethylbicyclo[3.1.0]hex-3-
en-2-one [34] with 97% sulfuric acid gave 2,3,4,5,6,6-

hexamethyl-—2,4—«cyclohexadienone (33) in excellent yield.

Labeling experiments (using.34 variously labeled with CD4)
show that a methyl migration is not involved; the results
are, however, in agreement with a mechanism in which.§g;
undergoes a cycloprOpylcarbinyl rearrangement to ion 68,
which rapidly equilibrates with its enantiomeric form 689
prior to ring opening to afford dienone 33; It is also

shown that ion 68 can revert to enone 34.

0H 0H

\ r
11“

68 ’ 68a

Pyrolysis of bicyclic ketone 34 gave dienone 33 as the
only product. The reaction course was determined using

enone éé’labeled with CD3 groups.

Treatment of dienone §§,with fuming sulfuric acid gave
2,3,4,4,5,6-hexamethyl-2,5—cyclohexadienone (35) in good

yield.

118

 

 

5.

119

The ultraviolet spectra of dienones ééland 35 in sulfuric
acid show that these compounds exist as their conjugate

acids in concentrated acid. The pKa of dienone 35 is -2.6.

Hexadeutero:§§, labeled at the C—3 and C—5 positions with
CD3, underwent rapid scrambling of the label on treatment
with fuming sulfuric acid. This reaction also occurred in

70% and 98% sulfuric acid, but much more slowly.

 

10.

11.
12.

13.

14.

15.

L ITERATURE C I TED

Schaffner, K. in "Advances in Photochemistry," edited by
W. A. Noyes, Jr., G. S. Hammond, and J. N. Pitts, Jr.,
Interscience Publishers, New York, Vol. 4, 1966, p. 81.

Kropp, P. J. in "Organic Photochemistry," edited by
0. L. Chapman, Marcel Dekker, Inc., New York, 1967, p. 1.

Wheeler, J. W. and R. H. Eastman, J. Am. Chem. Soc., 8;,
236 (1959).

Zimmerman, H. E., and D. I. Schuster, J. Am. Chem. Soc.,
§§, 4486 (1961).

Chapman, 0. L. and L. F. Englert, J. Am. Chem. Soc., 88,
3028 (1963).

Fisch, M. H. and J. H. Richards, J. Am. Chem. Soc., 88,
3029 (1963).

Schuster, D. I. and A. C. Fabian, Tetrahedron Letters,
No. 34, 4093 (1966). .

. Frei, J., C. Ganter, D. Kagi, K. Kocsis, M. Miljkovic,

A. Siewinski, R. Wenger, K. Schaffner, and 0. Jeger,
Helv‘. chi‘m. Acta., 4_9_, 1049 (1966).

Kropp, P. J., J. Am. Chem. Soc., 88, 3779 (1963).

Davis, B. R. and T. G. Halsall, J. Chem. Soc., 1833 (1962).

KrOpp, P. J., J. Am. Chem. Soc., 88, 4053 (1964).

Zimmerman, H. E., R. Keese, J. Nasielski, and J. S. Swenton,

J. Am. Chem. Soc., 88, 4895 (1966).

Kropp, P. J. and W. F. Erman, J. Am. Chem. Soc., 88,
2456 (1963).

Hart, H. and D. W. Swatton, J. Am. Chem. Soc., _8, 1874
(1967).

Hart, H., P. M. Collins, and A. J. Waring, J. Am. Chem.
Soc.,.ég, 1005 (1966).

120

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

121

Stich, K., G. Rotzler, and T. Reichstein, Helv. Chim.
Acta., 88, 1480 (1959).

Reusch, W., M. Russel, and C. Dzurella, J. Org. Chem.,
28. 2446 (1964).

Jaffe, H. H. and M. Orchin, "Theory and Applications
of Ultraviolet Spectroscopy," John Wiley and Sons, Inc.,
New York, N.Y., 1962, p. 200.

Doering, W. von E., M. Saunders, H. G. Boynton, H. W.
Earhart, E. F. Wadley, W. R. Edwards, and G. Laber,
Tetrahedron, 4, 178 (1958).

Shybin, V. G., V. P. Chzhu, A. I. Rezvukhin and V. A.
Koptyug, Bull. Acad. Sci. USSR, Div. Chem. Sci., 2056
(1966).

Erskine, R. L. and E. S. Waight, J. Chem. Soc., 3425
(1960). -

Chapman, 0. L., in "Advances in Photochemistry," 1, 330
(1963).

Herkstroeter, W., A. Lamola, and G. Hammond, J. Am.
Chem. Soc., 88, 4537 (1964).

Calvert, J. G. and J. N. Pitts, Jr., "Photochemistry,"
John Wiley and Sons, Inc., New York, 1966, p. 824.

Maclean, C. and E. L. Mackor, J. Chem. Phys., 88, 2208
(1961).

Dfirr, H., Tetrahedron Letters, 81, 5829 (1966).

Stiles, M. and R. P. Mayer, J. Am. Chem. Soc., 8;,
1497 (1959).

Miller, B. and H. Margulies, J. Am. Chem. Soc., 81,
5106 (1965).

Miller, B., J. Am. Chem. Soc., 81, 5115 (1965).

Denivelle, L. and R. Fort, Compt. rend., 240, 2423
(1955).

Streitweiser, A. Jr., L. Verbit, and P. Stang, J. Org.
Chem., 88, 3706 (1964).

Breslow, R., in "Molecular Rearrangements," Part I,
P. de Mayo, Ed., Interscience Publishers, Inc., New York,
N. Y., 1963, p. 233.

 

 

33.

34.

35.

36.

37.

38.

39.
40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

122

Richey, H. G. Jr., in "Carbonium Ions," G. A. Olah and
P. von R. Schleyer, Ed., to be published by John Wiley
and Sons, Inc., New York, N. Y.

Mazur, R. H., W. N. White, D. A. Semenow, C. C. Lee,

M. S. Silver, and J. D. Roberts, J. Am. Chem. Soc., 81,
4390 (1959).

Pearson, R. G. and S. H. Langer, J. Am. Chem. Soc., 18,
1065 (1953).

Wallach, 0., Ann., 360, 82 (1908).

Walborsky, H. M. and L. Plonsker, J. Am. Chem. Soc., 88,
2138 (1961).

Dauben, W. G. and L. E. Friedrich, Tetrahedron Letters,
no. 18, 1735 (1967).

Tadanier, J., J. Org. Chem., 81, 2124 (1966).
Hikino, Y. and P. de Mayo, Chem. Comm., 550 (1965).

Kakihana, T., M. S. Thesis, Michigan State University,
1966.

Hart, H., J. DeVrieze, R. M. Lange, Manuscript in
Preparation.

de Vries, L., J. Am. Chem. Soc., 82, 5242 (1960).

Pittman, C. U. Jr., and G. A. Olah, J. Am. Chem. Soc.,
.81, 2998 (1965).

Eistert, B. and G. Langbein, Ann. Chem., 678, 78 (1964).

Marvell, E. N. and E. Magoon, J. Am. Chem. Soc., 11,
2542 (1955).

Zimmerman, H. E. and D. I. Schuster, J. Am. Chem. Soc.,
84. 4527 (1962).

Flexser, L. A., L. P. Hammett, and A. Dingwall, J. Am.
Chem. Soc., 81, 2103 (1935).

Paul, M. A. and F. A. Long, Chem. Rev., 81, 1 (1957).

 

 

 

1 .

1
.