PART I
ACID-CATALYZED REARRANGEMENTS 7
OF (IYCLOHEXADSENONE EPOXEDES ‘ '

PART II
Paomcnmsmv OF‘ . _
cvcwnmnmnon-E Homes V

PART m
MISCELLANEOUS

Sisserfiafion is? the Eegree cf Ph. ‘3
MECHEGAN STATE fi‘s‘éé‘iEfiSiW
ma MU 3%

197,5

 

This is f”

   
  
     
  

~'fy that the
*hesis edfitled

Acid-Catalyzéd Rearrangements
vaplohexadienone

ep ides 6
Part II.3 Photochemistry officlohexadieno
Part II‘I', Miscellaneous ‘

'XJ.de~‘-‘I7-v.
:* r. presentefly ' ‘
*5 ,3”). .

54

Part I.

Eng Mu Shih,

has been accepted towards fulfillment
of the requirements for

Ph D . degree in Chemistry

4M LLJ

Major professor
/
Date ' 2 ' ‘71 5

 
      

2*: mm mm mt.
LIBRARY {2' ‘ZTQS
‘ W1. ,. ;

M
as

 

in

at-
n

_nly

ABSTRACT

PART I

ACID-CATALYZED REARRANGEMENTS OF CYCLOHEXADIENONE EPOXIDES

PART II

PHOTOCHEMISTRY OF CYCLOHEXADIENONE EPOXIDES

PART III
MISCELLANEOUS
BY
Eng Mu Shih

The acid-catalyzed rearrangements of cyclohexadienone
epoxides, §§.and'§2, are described in Part I of this thesis.
The «,B-epoxy ketones Egg and g2; (gig and trans-2,3-epoxy-
2,3,4,5,6-pentamethy1-4-vinyl-5-cyclohexenones) rearrange
in acid exclusively by vinyl migration, in preference to
the acetyl and methyl migration observed previously in
analogous compounds with a methyl group in place of the
4-vinyl substituent. When the vinyl group and epoxide ring
were trans (§§2)I rearrangement was much faster than with
the cis isomer (gfig), owing to homoallylic participation

during the epoxide ring opening. However the rearrangement

2 Eng Mu Shih
products in both instances were identical, i.e., a 1:4
mixture of cis- and trans-Z-acetyl-S-vinyl-Z,3,4,5-tetramethyl-

3-cyc10pentenone3'(46c and 46t).

9 o o
J' as
O ————-5
|
gaff, 4—§‘C,t

Rearrangement of 4,5-epoxy-3,4,6,6—tetramethyl-2-
cyclohexenone (22) in trifluoroacetic acid (TFA) at room
temperature gave 2,3,6,6-tetramethyl-2-cyclohexen-l,5-dione
(E1, 53%) and 4-methylene-5-hydroxy-3,6,6-trimethyl-2-
cyclohexenone (fig, 47%). In contrast to permethylated
compoundgg (4-methy1ene-S-hydroxy-2,3,5,6,6-pentamethyl-2—
cyclohexenone) £5 underwent no further rearrangement on
treatment with TFA, and this was attributed to preferential
protonation at the carbonyl oxygen to give the highly

delocalized cation 3 instead of at the hydroxyl group.

 

C)
H / O 9
\ \ \ / H
\ H I \\ + CH
C) (3 H

H H

39 a (53%) g (47%)

3 Eng Mu Shih

H
O
H+ Fl ,/
2g ' OH
H
B

The photochemistry of cyclohexadienone epoxides, 3g and
22, is described in Part II of this thesis. The UV irradia-
tion of 4,5-epoxy-2,3,4,5,6,6-hexamethyl-2,4-cyclohexadienone
(33) in ether through a Pyrex filter gave 4-acetyl-2,3,4,5,S-
pentamethyl-Z-cyclopentenone (fig), which rearranged further
to bicyclo[2.1.0]pentan-2-one (22). On further irradiation
through a Corex filter, compound 22 rearranged to an enol
lactone 23 (80%) and a compound which was tentatively assigned

structure 21 (20%).

 

0 o
\\ l
M ‘Q: ._,
_, .
\\ C)
O Pyrex
.33. 29. 22
O \_/
“\
a; _h_"__, o + o
Corex - O
96 97

 

LL

4 Eng Mu Shih

BicyclOpentanone 25 was thermally labile. It reverted slowly
to 32 either in the solid state or in a non-polar solvent
such as carbon tetrachloride. In methanol, it was converted
rapidly to lactones 22 (25%) and £22 (75%). Deuterium-
labeling experiments were consistent with the prOposed mechanisms

for the photorearrangement of 22 and the thermal rearrangement

of 25.

\

95 ‘3' CC14 9.9
A, MeOH

9
03 /[ + 0/
n

99 00

_£)

Photolysis of 4,5-epoxy-3,4,6,6-tetramethyl-2-cyclohexenone
(32) in ether through Pyrex afforded anti-1,3,3,6-tetramethyl-
bicyclo[3.1.0]hexan-2,4-dione (£94, 44%), §X231,3,3,6-
tetramethylbicyclo[3.1.0]hexan-2,4-dione (£25, 34%) and
4,4,7,7-tetramethyl-2-oxa-cyclohepta-3,5-dien-1-one (199, 22%).
Compounds 124 and 12§_were shown to originate from the
diketone intermediate 115. A new reaction pathway leading to
lgé from 32 seems to be operating, via the cyclOpropanone
intermediate'lgg. Deuterium-labeling experiments support

the proposed mechanism for the photorearrangements of 32.

5 Eng Mu Shih

o c') o
' u
H '.< M‘ H C/
____) —__)

F! Fl H

O o O

39
A/H‘//

 

 

 

o ‘ ‘
H I H\ I - H\
o \ o ‘E 4 “K O
- H , . ( \\ ( .
H

115 F'
‘_ 1.29.
/A, or hv
C?
Id ,. I
C) " C)
\\ ti” \\ // “ti
cf “kl \

 

£21 (44%) 105 (34%) 106 (22%)

 

6 Eng Mu Shih
On further irradiation through a Corex filter, 104 rearranged

to antifl,6-dimethyl-4-iseprOpylidene-3-oxa-bicyclo[3.1.0]-
hexan-Z-one ($21), $25 rearranged to syn-1,6-dimethyl-4-
isopropylidene-3-oxa-bicyclo[3.1.0]hexan-2-one (lgg, 75%)
and gynfs,6-dimethy1—4-is0pr0pylidene-3-oxa-bicyclo[3.1.0]-
hexan-Z-one ($22, 25%), $22 gave 2,2,6,7-tetramethyl—4—oxa-

bicyclo[3.2.0]hept-6-en-3-one (110).

104 ._____,

 

 

 

 

 

191
C;l
Fl
///LL\ C)
//J!\\
105 -————~ \\
_ .+ \H
H
\
£23 (75%) £92 (25%)
\H0
192* —o
/'
H\

110

7 Eng Mu Shih
Some miscellaneous results are described in Part III
of this thesis. The dichlorocarbene adduct of hexamethyl-
dewarbenzene,'l§2_(3,4,4-trichloro-1,S,6,7,8-pentamethyl-2-

methylene-tricyclo[4.2.03'5

]oct-7-ene), rearranged to
3,4-dichloro-l,6,7,8-tetramethyl-2,5-dimethylene-bicyclo[4.2.0]-

octa-3,7-diene (136) on thermolysis.

as .1

 

\\| A A ‘7 , Cl
/ C1 -HC1 , '

l4

 

 

 

 

 

C1

134 gig
On irradiation through Pyrex, 1,4,5,6,7-pentamethyl—
bicyclo[3.2.1]octa-3,6-dien-2,8-dione (119) reached a
photostationary state with 1,3,4,5,8-pentamethylbicyclo[3.3.0]-
octa-3,7-dien-2,6-dione (I42). On further irradiation of
this equilibrating mixture through Corex, a new photOproduct
l4; (1,4,5,6,7-pentamethy1bicyclo[3.2.0]hepta-3,6-dien-2-one)

was obtained presumably through the decarbonylation of 139.

f;
j m ‘1,
12 =‘ 98

139 (73%) 140 (27%) 141

 

 

\\

 

 

8 Eng Mu Shih
Treatment of bicyclo[3.2.1]octa-3,6-dien-2-one (143)

with trifluoroacetic acid followed by quenching with sodium
bicarbonate solution gave 4-trif1uoroacetylbicyclo[3.2.1]octa-
6—en—2-one (144); treatment of 142 with fluorosulfonic acid
followed by quenching with sodium methoxide gave 4—methoxy-

bicyclo[3.2.1]octa-6-en-2-one (145).

 

 

‘ TFA
/ ” 2 N HCO > /
. a 3 \(
124—2 OCOCF3
14.1
1. FSO3H ‘ /
2. NaOMe \\\(/
OMe
14_5

Treatment of l,4-dihydro-1,l,2,3,4,4-hexamethylbenzo-
pentalene (lgg) with mfchloroperbenzoic acid gave 1,4-
dihydro-Z-hydroxyl-3-methylene-l,l,2,4,4-pentamethylbenzo-
pentalene (lgfi), presumably through the rearrangement of
monoepoxide 141 catalyzed by a trace of acid. Photosensitized
oxidation 0f.l£§ in methanol led to a methanol adduct, the

exact structure of which, represented either as 150a or lSOb

remained uncertain.

9 Eng Mu Shih

 

& 13.7. 15.53.
OMe
OMe
146 TILL or
O OMe
sens. OMe
150a 150b

 

Epoxidation of 3,4,4,5-tetramethy1-2,5-cyclohexadienone
(151) with alkali hydrogen peroxide gave monoepoxide 154 (7%)
and diepoxides, 153 (5%) and 155 (45%).

0 H202

NaOH

 

MeOH

   

151 153 (5%) 154 (7%) -155 (45%)

PART I

ACID-CATALYZED REARRANGEMENTS OF CYCLOHEXADIENONE EPOXIDES

PART II

PHOTOCHEMISTRY OF CYCLOHEXADIENONE EPOXIDES

PART III

MISCELLANEOUS

BY

Eng Mu Shih

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

ACKNOWLEDGMENTS

The author is deeply indebted to his advisor,
Dr. Harold Hart, for his assistance and encouragement
during the course of this research.

I gratefully acknowledge financial support from
the National Science Foundation, National Institute of
Health and Michigan State University in the form of

research and teaching assistnatship.

ii

TABLE OF CONTENTS

PART I
ACID-CATALYZED REARRANGEMENTS OF CYCLOHEXADIENONE EPOXIDES
Page
INTRODUCTION......................................... 2
RESULTS AND DISCUSSION............................... 13
1. Acid-Catalyzed Rearrangement of cis- and trans
2,3-epoxy-2,3,4,5,6-pentamethyl-4-vinyl-

5-CYC10hexenOne (38C and 38t) 00000000000000. 13

2. Acid-Catalyzed Rearrangement of 4,5-epoxy-
3,4,6,6-tetramethy1-2-cyclohexenone (32).... 26

EXPERIMENTALOOCOOOOOOC0.0.0.0...0.0.000...0.0.0.0.... 35
1. General ProcedureSOOOOOOOOOOOOOOOCOOOOOOOOOO 35

2. Epoxidation of 2,3,4,5,6-pentamethyl-4-
ViflYl-Z I 5-CYC10hexadien0ne (£2) 0 o o o o o o o o o o o o 36

3. Acid-Catalyzed Rearrangement of cis- and
trans-2,3-epoxy-2,3,4,5,6-pentamethy1-4-
vinyl-S-cyclohexenone (38c and 38t)......... 39

4. Cleavage of trans-Z-acetyl-S-vinyl-Z,3,4,5-
tetramethyl-3-cyclopentenone (46t) with Base 41

5. 5-Trideuteromethyl-2,3-epoxy-2,3,4,6- *
tetramethyl-4-vinyl-S-cyclohexenone (§§_)... 42

6. Acid-Catalyzed Rearrangement of 3gf......... 43
7. Acid-Catalyzed Rearrangement of 38c......... 43

8. Epoxidation of 3,4,6,6-tetramethyl-2,4-
cyclohexadienone (23)....................... 44

9. Acid-Catalyzed Rearrangement of 4,5-epoxy-
3,4,6,6—tetramethyl-2-cyclohexenone (22).... 45

iii

TABLE OF CONTENTS (continued)
Page

10. 2-Deutero-3-trideuteromethy1-4,5-epoxy-
4,6,6-trimethy1-2-cyclohexenone (22*)...... 46

11. Acid rearrangement of 22*.................. 47

12. Acid-Treatment of 4-methylene-5-hydroxy1-
3,6,6-trimethyl-2—cyclohexenone’(22)....... 47

PART II

PHOTOCHEMISTRY OF CYCLOHEXADIENONE EPOXIDES

INTRODUCTIONOOOOOOOOO 000000000000 OOOOOOOOOOOOOOOOOOO 49
RESULTS AND DISCUSSION...O...OOOOOOOOCOOOOOCOOOOOOOO 60

l. Photochemistry of 4,5-epoxy-2,3,4,5,6,6-
hexamethyl-Z,4-cyclohexadienone (22)....... 60

2. Photochemistry of 4,5—epoxy-3,4,6,6-
tetramethyl-Z-cyclohexenone (22)........... 74

EDERIMENTAL... O. ......... I OOOOOOOOOOOOOOOOOOO .0. O. O 96
1. General Procedures......................... 96

2. Photolysis of 4,5—epoxy-2,3,4,5,6,6-
2,4-cyclohexadienone (22).................. 96

3. Photolysis of 4,S-epoxy-3-trideuteriomethy1-
2,4,5,6,6-pentamethy1-2,4-cyclohexadienone

4. Photolysis of 4,5-epoxy-3,5-bis(trideuterio-
methyl)-2,4,6,6-tetramethy1-2,4-
cyclohexadienone (22*: )................... 98

5. Photolysis of Bicyclopentanone (22)........ 98

6. Photolysis of Bicyclopentanone (22*)....... 99

.1.
7. Photolysis of Bicyclopentanone (225' )..... 100
8. Thermal Reaction of Bicyclopentanone (22).. 100

*
9. Thermal Reaction of Bicyclopentanone (22 ). 102

iv

TABLE OF CONTENTS (continued)

10. Thermal Reaction of Bicyclopentanone (22f,+)
ll. Photolysis of 4,5-epoxy-3,4,6,6-tetramethyl-
Z'CYCIOhexenone (22.)oooooooooooooooooooooooo
12. photolysis of 32*...........................
13. Photolysis of anti-l,3,3,6-tetramethyl-
bicyclo[3.1.0]hexan-2,4-dione (104).........
*
14. PhOtOlYSiS Of 104 OOOOOOOOOOOOOO0.0.0.000...
15. Photolysis of syn-l,3,3,5-tetramethyl-
bicyclo[3.1.0]hexan-2,4-dione (105).........
16. photolysis of 105*..........................
17. Photolysis of 4,5,7,7-tetramethyl-2-oxa-
cyclohepta-3,5-dien-l-one (106).............
18. Photolysis of 106*..........................
19. Photolysis of 2,3,6,6-tetramethy1-2-
cyclohexen-1,5-dione (22)...................
20. Photolysis of 22*...........................
21. Photolysis of l,3,3,5-tetramethy1-
bicyclo[3.1.0]hexan-2,4-dione (113).........
22. PhOtOlYSiSOf113*...ooooooooooooooooooooooo
PART III
MISCELLANEOUS
RESULTSOOOOO ...... 0...... ........... OOOOOOOOOOOOOOOOO
1. Thermal Rearrangements of some of the
Dichlorocarbene Adducts of Hexamethyl-
dewarbenzene................................
2. Photoisomerization of l,4,5,6,7-pentamethy1-

bicyclo[3.2.1]octa-3,6-dien-2,8-dione (139).

Page
103

104

105

105
106

107

108

108

109

110
111

111

112

114

114

115

“J -._

TABLE OF CONTENTS (continued)
3. Acid Treatment Of Bicyclo[3.2.l]octa-
3,6-dien-2-One (143)ooooooooooooooooooooooo
4. Epoxidation and Photosensitized Oxidation
of l,4-dihydro-l,l,2,3,4,4-hexamethyl-

benzopentalene (146).......................

5. Epoxidation Of 3,4,4,5-tetramethyl-2,5-
cyclohexadienone (151).....................

EXPERIMENTAL...OOOOOOOOOOOOOIOOOOOOOCOCOOOOOOOOOOOOO

1. Vacuum Distillation of the Dichlorocarbene
Adducts of Hexamethyldewarbenzene..........

2. Thermal Rearrangement of l34...............

3. Allylic Oxidation of 1,2,5,6,7-pentamethy1-
bicyclo[3.2.llocta-2,6-dien-9-One (138)....

4. Irradiation of l,4,5,6,7-pentamethyl-
bicyclo[3.2.1]Octa-3,6-dien-2,8-dione (139)

5. Irradiation Of l,3,4,5,8-pentamethy1-
bicyclo[3.3.0]Octa-3,7-dien-2,6-dione (140)

6. Treatment of BicyclO[3.2.llocta-3,6-dien-
2-One (143) with Trifluoroacetic Acid......

7. Treatment Of 143 with Fluorosulfonic Acid..
8. Michael Addition of Methanol to l43........

hexamethylbenzopentalene (146).............

10. Photosensitized Oxidation of l46...........

11. Epoxidation of 3,4,4,5-tetramethy1-2,5-

Page

117

118

120

121

121

122

123

124

126

126
127

128

128
131

cyclohexadienone (151) with mfchloroperbenzoic

aCidOCOOIOOOCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

12. Epoxidation Of 3,4,4,5-tetramethyl-2,5-
cyclohexadienone (151) with alkali hydrogen
peroxide................. ....... ...........

BIBLImRAPHYOOOOOOOOOOOOO ....... O ....... .0. ..... 0...

vi

132

132

138

LIST OF TABLES
TABLE Page
I. Migratory Aptitudes Of the B-substituent in

the Conversion Of «,B-Epoxyketones to
8-DiketoneSOOOOOOIOOOOOOOOOOI0.0.0.0....0.... 52

vii

PART I

ACID-CATALYZED REARRANGEMENTS OF CYCLOHEXADIENONE EPOXIDES

INTRODUCTION
Epoxides can be isomerized to aldehydes, ketones or

a mixture of these under acid conditions. Thus etyhlene

oxide 2 gives acetaldehyde 2 over aluminum oxide.

0

 

O
/ \ A1203 t CH3-C</

CH ——CH
2 2 H

Il-‘
IN

These acid-catalyzed rearrangements Of epoxides are of

special interest, since they provide a simple means of

I

converting Olefins to carbonyl compounds. For example,

in the presence Of aluminum trichloride, trimethylethylene
oxide 2, which can be obtained by epoxidation Of trimethyl-

4
ethylene 2, rearranges tO methyl iSOprOpyl ketone 2 :

CH3 H CH3
/ O A1C13

O
//
c=c 31—» \CH-C
/ \\ H ‘// \\
CH3 CH 3 CH 3 CH 3

 

IO)
lb
IUl

3

Similarly, isobutylene Oxide 2 gives largely isobutyraldehyde

4
1, and prOpylene oxide 2 yields prOpionaldehyde 2.

H

H
9. .7.
O O
H Fl ’ HI§___AV
H H \
9. 9.

In general, Lewis acids (i.e., aluminum trichloride,
magnesium bromide or boron trifluoride etherate) are used as
catalysts, although aqueous mineral acids and non-nucleOphilic
carboxylic acids such as trifluoroacetic acid may also
serve. Rearrangement reactions of this kind become especially
useful if one can predict in any particular case what the
major product will be. The major product formed from the
rearrangement of an epoxide is governed by two main factors:
the direction Of ring Opening, and the relative migratory
aptitudes Of the different substituents.

The direction Of ring Opening in acid-catalyzed
rearrangements may in general be predicted simply on the
basis of the relative ease Of ionization Of the two carbon-

Oxygen bonds in question. By analogy with other situations

4

in organic chemistry, the sequence is expected to be tertiary
> secondary > primary for the aliphatic series. The
rearrangement of isobutylene oxide g to isobutyraldehyde 1
illustrates this rule.

Aryl substituents on epoxides are generally superior
to alkyl substituents in enhancing the ease of ionization
of the carbon-oxygen bond. The above rule may be generalized
to state that, in rearrangements, the carbon-oxygen bond
that tends to be broken is the one to the carbon atom
substituted to the greatest extent with groups promoting
ionization, namely alkyl and aryl. Thus, indene oxide l2_
is isomerized to 2-indanone ll on treatment with magnesium

bromide etherate.5

O MgBr2 .ether

Vinyl substitution, like phenyl substitution, weakens
the adjacent carbon-oxygen bond with respect to ionization,
and there are a number of reports of rearrangements of
ethylene oxides involving breakage of the bond to a vinyl-
substituted rather than a phenyl-substituted carbon atom.6’7
For example, phenyl cyclohexenyl ethylene oxide la is

rearranged to phenyl cyclohexenyl acetaldehyde £39 in

contrast to the cyclohexyl analog 13! which rearranges with

S

cleavage of the other carbon-oxygen bond to give benzyl

cyclohexyl ketone lg.

O H 9
MgBr2
-A //__\‘ H
95 H H O 0

¢
//

 

 

1i
D
5
NF?
I I/
(9%

Once the direction of ring opening has been decided,
in epoxide g, for example, by the preferential formation of
a tertiary carbonium ion. There remains a choice between
hydrogen migration or alkyl migration. This will normally
be decided by the relative migratory aptitudes of the
groups, and in general this order is aryl > acyl >H > ethyl
> methyl. No rationale for this order of migratory aptitudes
has been prOposed, since the mechanistic details for this
type of rearrangement, which is closely related to the
pinacol rearrangement, are not yet clear. Among the few
exceptions to this order of migratory aptitudes, are those

in which, for steric reasons, hydrogen migrates in preference

to phenyl.2

As with simple epoxides, epoxy ketones also undergo
acid-catalyzed ring cleavage with rearrangement, and have
served as useful precusors to dicarbonyl compounds. For
example, in glacial acetic acid in the presence of sulfuric
acid, benzalacetoPhenone oxide lg isomerizes to formyl-

desoxybenzoin l1.

0
II
o o o C-C6H
/ \ ll H2804 \\ 5
c HS- CH— CH-C-C H ———-) C-CH
5 5 5 CH3C00H ,/
H C6HS
16 17

Although the ring Opening of «,B-epoxyketone under
the influence of acids can proceed in two directions, it

generally favors the more stable carbonium ion A and the

 

OOH
II I +
R-c-c—c—
H / '
O O + 0 0+
ll /\ H II /\ A
R-c-c—c- ._._.. R-c-c—c-
I ' ' ' \ o OH
II + I
R-c-c—c—
| I

7

observed products depend only on the migratory aptitudes of
the groups attached to the epoxide ring. Thus, in the
presence of boron trifluoride, a-ethylbenzalacetophenone
oxide l§_rearranges, via the intermediate ion lg, to

1,2-diphenyl-l,3-pentanedione lg, the product expected from

 

10
the preferential benzoyl group migration. ,_
+,BF3
o o H ¢co o H
H /\ / BF3 \l +/
¢~C-C-—C A, C——-C
I \ / \
Et {2) Et (5
L2 L9.
benzoyl
O O
//
C- (IIH-C
\
Et (2) ¢
20

Similarly, 2-benzalcyclOpentanone oxide ll yields diketone

ll,11 and iSOphorone oxide 33 gives 2-formyl-2,4,4—trimethyl—

12
cyc10pentanone 35.

O

 

o-—<)
m
we
w
m
w
W

 

Although the acid-catalyzed rearrangement of epoxy-
ketones is generally initiated by protoation of the epoxy
oxygen atom,13 a competitive protonation at carbonyl oxygen
appeared to play a role in the acid-catalyzed rearrangement
of‘l_5_.l4 This competitive protonation on carbonyl oxygen
was necessary to account for the formation of ll, in addition

to the expected product l1 (Scheme 1)

 

Scheme 1
O (I? O
. H+ OH -H+J
A O "’ .A OH --’ O
+~HO
g
+ 21
OH
A60
1
- | (“)0
H+ ~acyl wk
0 __. OH——H;

 

 

9

In a recent study,15 it was shown that acyl and methyl
migration compete approximately equally in the acid-catalyzed
rearrangement of «,B-epoxyketone lg. Thus on treatment with
trifluoroacetic acid, l§_rearranged (Scheme 2) to a nearly

equal mixture of 32 and (39 + ll).

 

 

Scheme 2
O O
I O
+ II I
O H , OHM-"CY1 7, \
_H+
22 g _2_9_ (50%)
~Me
‘9 9
O
OH O
, OH +
/
_D. _3_(_)_ (10%) 31 (35%)

Protonation and ring opening in lg occurs in such a manner
as to place the positive charge remote from carbonyl group,
giving ion 9. Acyl migration and proton loss affords the
major product l2. Methyl migration competes effectively,

giving the allylic ion Q_which may either lose a proton

10

to give 39 or rearrange by a more complex process, eventually
to give ll.

The epoxyketone 23, which is a vinylog of an «,8-
epoxyketone, rearranged in acid initially to allylic cation E,
not §,15 Presumably, the conjugative effect is an overriding
factor in determining the stability of the resulting carbonium

ion.

(I?
\ /
O O
| / OH
+-
fi _1, \ ~3-
0 0+
H O
33 |
OH

E

The intermediate §_can collapse to 23 by losing a
proton (Scheme 3). In neat trifluoroacetic acid, the same
intermediate may rearrange by a 1,2-alkyl shift (ring contrac-
tion) to give lg or may, following nucleOphilic attack
a to the carbonyl group, undergo a ring contraction, leading

to 31. The major product 3E may arise from protonation of

l‘

11

1,2 acyl shift

\+
+
H
I
‘\
A

if
1
(1)“
‘H

 

 

 

 

 

t
f
m +
H .I
nu w .+ th W Aw
. E. .. Lo 1&3
/ x
s
A
w
+
m HH
T nWN
nu /

 

Scheme
C)
32
C)
l,
C)
ll

12

ll followed by a 1,2-acyl shift and deprotonation; product
_2 is formed by dealkylation of 25 (Scheme 3). These
mechanisms were supported by deuterium labeling experiments.

Only a few examples of dienone epoxide rearrangements
have been studied. Although one can, from previous experi—
ence, write plausible reaction paths, one's predictive
ability as to which path will predominate in any given
instance is still rather low. Since the reactions can be
synthetically useful, especially for making variously
substituted cyclopentenones, it would be helpful to be
able to make such predictions with greater accuracy. Therefore
it is the purpose of this part of the thesis to examine the
manner in which the various methyls groups or other
substituents in 23 and lg determine the mode of the acid-
catalyzed rearrangements of cyclohexadienone epoxides.

The two cases studied here were lg, which differs from
lg by having a vinyl group at the C-4 position instead of

a methyl group, and 22, which differs from ll by having

hydrogens instead of methyls at the C-2 and C-5 positions.
C)
\
O O
II
I
lg
C)
: ;C>

H

Id

RESULTS AND DISCUSS ION

l. Acid-Catalyzed Rearrangement of cis- and trans-2,3-epoxy-

24},4,5,6-pentamethyl-4-vinyl-Segyclohgxenone (38c and 38t)

Treatment of the vinyl cyclohexadienonegg21 with an
equimolar amount of mechloroperbenzoic acid.(erPBA) in
methylene chloride gave a mixture of two monoepoxides,
lfig and 222! in which the vinyl group and epoxide ring

are cis and trans to one another, respectively.

g-CPBA

 

VV

   

Although it was not possible to separate the two isomers, it
was clear from the nmr spectrum of the mixture that epoxi-
dation had occurred exclusively at the «,8-double bond. The
area ratio of vinyl protons (multiplet, 6 4.8-6.1) to aliphatic

protons (6 1.1-1.7) was 1:5. With the aid of europium shift

13

14

16 . .
reagent and as a consequence of isolating one of the

isomers in pure form (38c; vide infra), it was possible to

 

completely assign the nmr spectrum of each isomer. The
ratio of 2&2 to lgt, as determined by integration of the
europium-shifted nmr spectrum, was approximately 3:2.

When excess mrchloroperbenzoic acid was used, di- and
tri-epoxides of 39 were formed. The diepoxides were a
mixture of glg and £l£.in which the two epoxide rings are
cis to one another and either cis or trans to the vinyl

group. There was no detectable amount of the isomer with the

   
     

.-’1.35 (1.0)

 

0')
3 O
I’ “1.22 (22.5)
1.10 (18.7) 1.48 (41.1) H
4.8-6.4
41c 41t

two epoxide rings trans to one another. It is known that
the direction of the attack by the peracid in epoxidation
may be influenced by nearby polar substituents. Thus,
epoxidation of allylic alcohol £l_gave predominantly the

17
cis product 43c.

15

 

 

H
O 3
,Fl C FL.‘ ,hl
x H
OH 6 5C03H ‘ OH
# +
g 533 (91%) L312 (9%)

The directive effect of the hydroxyl group has been suggested
to arise because of hydrogen bonding between the hydroxyl
group and the attacking peracid. This type of directive
effect has also been observed in compound 3§_which on

epoxidation gave exclusively cis-diepoxide 12°

 

 

16

The configuration of the tri-epoxide g§_is not known, though
it seems probable that the two epoxide oxygens on the
cyclohexanone ring are cis. Spectral properties of ll

and 32 are given in the experimental section.

C)

>
o O
\

\CO

When a methylene chloride solution of the mixture of
Egg and §§E_was treated with trifluoroacetic acid for 1 hr
at room temperature, only one of the two isomers reacted.
The unreacted isomer was separated from the rearrangement
products by preparative vPc, and was assigned structure 2§E°
The chemical shifts and eurOpium shift slopes16 of the
five methyl groups are shown on the structural formula.
The most striking difference in the nmr spectra of 2§g_and
2&3 is the chemical shift of the methyl group at C-4; these

signals are 0.22 ppm apart in the two isomers, a result which
8.18

is consistent with the nmr spectrum of

17

     
  

C)
(3.04)) O
1.67 \\ ,z1d40 (3.30) (3.03) 1.67 .\\ ’,1.39 (3.31)
C)
(1.23) x O

1.67 , \\] ‘1.31 (1.07) (1.20) 1.67 “1.31 (1.07)
’ |

(1.00) 1.15 (1.00) 1.37 H

38c

’1.40 (3.93)

C)

, ‘\1.40 (1.26)
\

(3.66) 1.74

  

(1.60) 1.81//'

(1.00) 1.09 1.29 (1.16)

The methyl assignments at C-4 in ZE were based on the
assumptions that (a) the methyl cis to the electronegative
epoxide oxygen should appear at lowest field, and (b) the
methyl cis to the epoxide oxygen, a possible coordination
site for eurOpium shift reagent, should have a slightly
larger eurOpium-shift slope than the methyl trans to the

epoxide ring (the main coordination site for europium is

18
clearly, however, the carbonyl oxygen). On these grounds

one can assign structure 322 to the epoxyketone which is
recovered from mild treatment of lfig + 32E with trifluoro-
acetic acid, since its C-4 methyl appears at higher field
(thus trans to the epoxide ring) than the C-4 methyl in its
isomer. As will be seen, this assignment is consistent
with mechanistic rationalizations of the acid-catalyzed
rearrangement of £23

In addition to recovered 30, two rearrangement products
were isolated from the mild treatment of §§g_+ 3§E_with
trifluoroacetic acid. These products were separated by
preparative vpc and are assigned structures‘fiég and 46E
(ratio 1:4), in which the methyls at C-2 and C-5 of the
cyc10pentenone ring are cis and trans, respectively. Each
isomer showed two strong carbonyl bands in the infrared
(~ 1700, 1740 cm—1) for the acetyl and cyc10pentenone
carbonyl groups; the uv spectra also showed that neither
carbonyl group was conjugated with a carbon-carbon double
bond. The nmr spectrum of each isomer showed an acetyl methyl,
two allylic methyls, two aliphatic methyls, and three
vinyl protons. The mass spectra of £62 and 4§E_were nearly
identical; striking features were a very weak parent peak
(m/e 206), a base peak at m/e 164 corresponding to the
loss of ketene, and three additional intense peaks for the
further loss of 15, 28 and 43 amu. All of the spectra

resemble closely the published spectra of‘ll.15

19

1.23 (1.38)

\‘%%y

  
 

/

\ (4.52) 1.93
1.13 (1.29)

(3.98) 1.93
(2.86) 1.23’

   

(2.09) 1.67 1.67 (1.00)
(2.24) 1.60 1.67 (1.00)
9—62 16.9
0 C)
H I /1.05 (1.36)
(4.22) 1.98 .//\\\
(3.58) 1.27 1-08 (1-61)
\\
(2.09) 1.60 1.71 (1.00)
31

The distinction between 4§g_and 4§E_is based on different
chemical shifts and europium slopes of the C-5 methyl
(adjacent to the vinyl substituent). This signal appears
at lower field and is affectedgmore by shift reagent when
the methyl is cis to the acetyl group at C-2 (i.e., in 465).
Also noteworthy are the lower overall europium shift slopes in
463 compared with 46g, presumably because the large vinyl
group cis to the acetyl group in’l§g_diminished complexation

with the shift reagent.

20

Chemical evidence for the structure of‘l§_was obtained
by base cleavage. The major isomer (322) was treated with
sodium methoxide in methanol at room temperature (12 hrs)
to give a mixture of two stereoisomers of a cyclopentenone
assigned structure 11. The ir (vc=o 1700 cm-1) and uv
spectra (Afing 237 nm, a 8,800) support the presence of a

cyc10pentenone moiety, and the nmr spectrum showed two

 

P
|
NaOCH "\
3 1.63 \ fl
46t ;
CH OH
3 1.90 ,:\
Fl
2.3
4.2.
homoallylically coupled methyls (6 1.63, 1.90, l_= 1 Hz),
three vinyl protons (6 4.6-5.8), two aliphatic methyls
(approx. 6 1.0) and the methine proton (m, 62.3). The
structure of gl_was further supported by the observation
that treatment with NaOCH3/CH3OD gave 47-d4, in which the
nmr signals at 6 1.90 and 2.3 were absent, that at 6 1.63
sharpened to a singlet, and the aliphatic methyl signals
simplified to sharp singlets.
A plausible mechanism for the formation of 52 from ll

is shown in Scheme 4 (the question of stereochemistry is

deferred for later discussion; vide infra).

 

21

 

 

 

 

Scheme 4
O
H O 9|
+ \\
O H OH vinyl \
*/ /
/ x A‘
' ll \
9.8. 9. .1
ring HQ OK
contraction \
___., OH OH ///
19. 1 4.9.
0
HC) ' C) C)
+
——-“ \98
f
*—

IL)
4:.
0‘

22

Protonation of the epoxide oxygen and ring-Opening gives ion
§_(analogous to g_from 32, Scheme 2, page 9). Vinyl migration
would give the allylic cation g (analogous to Q_from 9,

Scheme 2). Proton loss could give fig (analogous to 39,

Scheme 2) which, however, was not observed. Reprotonation,
ring contraction (to give 42) and a 1,2-acetyl migration
account for the product. Alternatively g could suffer ring
Opening to §2 which, on protonation and ring closure could

lead directly to g, and then 4g. These schemes are analogous

C) C) C)
I \\__J/
OH - H+ \ \ H+
/ / / —
\
g so

to the mechanisms established for the formation of'gl from
22.15

A labeling experiment was performed to test the
reasonableness of Scheme 4. Treatment of §§_(a mixture of
cis and trans isomers) with DMSO-d6 and potassiumgfbutoxide
gave §§* (see Scheme 4) whose nmr spectrum was identical

with that of §§_except that the area of the peak at 6 1.67

was reduced by 50% (label at C-S). Treatment of 32f with

23

*
TFA at room temperature for 1 hr gave‘38c (signal at 6 1.67

reduced in area by 50%, and sharper than for‘éfig) andgétf
whose nmr spectrum lacked the signal at 6 1.60 and had a
sharpened singlet at 6 1.67. Insufficient gggf was isolated
for spectral examination. These results clearly establish
that a vinyl migration occurred during the rearrangement

of §§_to fig (the vinyl group and labeled methyl in ggf are on
adjacent carbons, whereas in 4Q* they are separated by

an additional carbon) as outlined in Scheme 4.

The observation that 2§E is rapidly converted by TFA
into fig under conditions where §§g_is recovered unchanged
strongly suggested that Scheme 4 is oversimplified, and that
homoallylic participation occurs in the ring opening of
epoxide ggt. That is, the initially formed ion from‘§§£_
is not the simple tertiary carbonium ion g but a cyclopropyl-
carbinyl cation, one contributor of which is shown as
structure 5. Vinyl migration should give E (corresponds to
g in Scheme 4) in which the vinyl and hydroxyl groups are
trans. Unfortunately further steps in the mechanistic scheme
allow for stereochemical ambiguity and the product is a

mixture of 46t and 46c.

24

 

38t

IN
lb

In isomer 222! where the vinyl and epoxide groups are
cis, homoallylic participation in the ionization step is
not possible. Consequently, 3&2 is recovered unchanged
under these reaction conditions. It seemed important to
determine whether 3E2 would also rearrange in acid under
more forcing conditions. It was found that §§g_was inert to
TFA at room temperature, even after 12 hrs. However, when
the temperature was raised to 60°, §§g did rearrange slowly,
reaction being 85% complete in 7.5 hrs. The products were
4&3 and iEE' identical with those obtained from gfig, It is
concluded that both 2&3 and gfig rearrange exclusively by
vinyl migration. They do so, however, by different mechanisms,
§§E rearranging with homoallylic participation (via E, E, etc)

and 38c rearranging without participation (via g, g, etc)

25
or with participation but at a later stage (via g, fi,‘£, etc).

Even when the reaction occurs without participation, however,
acyl or methyl migration are unable to compete with vinyl
migration. Thus products such as §l_(which could be formed
via acyl migration) or §2_(which could be formed by proton
loss from the highly delocalized intermediate cation‘fl) were
not observed. One can conclude that vinyl migration is
preferred over acyl or methyl migration even when homoallylic

participation in the ionization step is not possible.

 

 

O
‘~acyl H .
jl A
( I
\
g;
0
+ \ H
H -H+ V
———, .———. OH
/’
§
E H 52

 

~vinyl
-——————+ £1 .____+ 46c + 46t

26
2. ‘Acidecatalyzed;Reagrangement of'4,5—epoxy-3,4,6,6-

 

tetramethylHZHchlohexenone,;§2

 

4,5-Epoxy-3,4,6,6-tetramethyl—2-cyclohexenone (22)
was prepared in good yield from the corresponding dienone
'53 and mrchloroperbenzoic acid. The prOperties and structure

proof have already been presented.19

 

O O
|_'\ I 5.52 (q) f—I H 1.07 (S)
mePBA . 1.25 (s)
/’ \~ \\.
'i 2.08 (d) #4 3.03 (s)
1.48 (s)
53 39

Treatment of §2_with TFA at room temperature for 1/2

hr gave two isomers which were assigned structures 52 and 55,

o
H

/
3_9 TFA ; l + 0
/ \0 OH

'H

——O

 

g5 (53%) ‘ g2 (47%)

27

The product structures were established by their spectral
prOperties. The diketone 54 showed two carbonyl absorptions,
at 1720 and 1665 cm’l, and strong uv absorption at 245 nm,
indicating one conjugated and another non-conjugated carbonyl
group in a six-membered ring.20 The nmr spectrum, which showed
two mutually coupled vinyl methyl groups (6 1.77, 1.97), and

eurOpium shift data are consistent with the structure.

(2.05) 1.77 \
\ 1.20 (2.92)

(1.00) 1.97 / \O
H H

3.10 (3.97)

54

The compound assigned structure 55 showed a Vc=o at
l

1670 cm— consistent with a conjugated carbonyl group in

a six-membered ring; the presence of a hydroxyl group was
clear from the VOH at 3500 cm.1 and from the presence of

a one-proton peak in the nmr spectrum at 6 2.90 which was
removed on D20 exchange. The nmr spectrum showed three

vinyl protons as multiplets (6 5.43, 5.57, 5.70), one
allylically coupled vinyl methyl (6 2.05) as a doublet with

a coupling constant of 2 Hz, and two gem-dimethyls as singlets

(6 1.00, 1.10). The ir spectrum indicated that two of the

28

vinyl protons were on a terminal methylene group (930 cm-1)
and the uv maximum at 270 nm (5 16,430) suggested that both
double bonds were conjugated with the carbonyl group.
specific assignments within the structure are based on a

labeling experiment to be discussed below.

1.10 (3.22)

 
  

5.70 (4.56) H\
2.05 (1.04) /

“~~ 1.00 (2096)

5.43 (1.00)

55

Compound 55 is undoubtedly formed from 32 by proton
loss from the intermediate cation E (Scheme 5). The alter-
native ring-opening mode to give g_possibly followed by
1,2-acyl shift would lead to structure 56 which is inconsistent
with the observed nmr and uv spectra. Ion §|is preferred
over 9 because it is tertiary and allylic, and its fate
(formation of 55) is entirely analogous to the formation

of 22 from 33 via ion g (Scheme 3, page 11).

29

Scheme 5
9 O 9 o
H+ / -H+
——9 —+ —)
, / H OH

CD+ . C) r)
F4 (j

22 _N_ 55

 

fi ‘fi 7
. l
1.2 acyl m -H+ Q5
\‘Fi .
OH O”

OH

IO
m
m

Possible routes to 54 are shown in Scheme 6. The same
intermediate cation E can loss a proton to give 51 which,
however, was not observed. Its rapid keto-enol tautomerism
accounts for the formation of 54. Alternative processes
involve a hydride shift (leading to g), deprotonation (to 5g),
followed by keto-enol tautomerism (to 54 via £1)' The

formation of 54 rather than §§_is probably a matter of

*H

30

thermodynamic control. Compound 54 contains the more
stable (Most substituted) double bond; also, any possible

unfavorable 1,3-methy1-methy1 interactions in §§_are absent

 

 

 

 

 

in 54.
Scheme 6
f C) ‘ C)
_. (H '
+
“H " OH )6 O
O J J
.51 '\ 2::
OH _ p
H O ( O
~H' k
2‘. H - H+ *H\
_) —-——9
* OH ,2
O
H
. H J
g 58

A labeling experiment was performed to test the accuracy
of Schemes 5 and 6, and to establish unequivocally the nmr
assignments of 55. Treatment of 32 With DMSO-d6 and
potassium Efbutoxide gave 32? whose nmr spectrum was identical

with that of‘gg (see page 26) except that the peaks at

62.08 and 5.72 were absent (label at C-3 and C—2). Treatment

31
* 'k
of 32_ with TFA at room temperature for 1/2 hr gave‘ég

(the signal at 6 1.97 disappeared and the singlet at 6 1.77
sharpened) with one less deuterium labeled atom thanԤ9*,
and‘ééf whose nmr spectrum lacked the signals at 6 2.05
and 5.70 (see pages 27, 28). These results show that no
skeletal rearrangement occurred during acid rearrangement
of 32 and that‘gg was formed via its enol tautomer‘§1_where
hydrogen exchange with solvent was possible.

In contrast to'gg, which on protonation rearranged to
gi'via a 1,2-acyl shift (Scheme 3, page 11), 55 underwent

no further rearrangement on treatment with acid under similar

conditions.

C)

I O o
\\ TFA acyl + + /fl\
_.__.. -H
(DH .1 --—z

I

33- 9.4.

C)

 

><

H TFA
; No Reaction
OH
H

32
Scheme 7 shows three possible modes for protonation of 55,

Scheme 7

I”

Pathway a can be eliminated on the basis that if the
allylic cation E'is formed, it should give §4_and'§§_on
quenching. But 55 did not rearrange to 54 on treatment with
acid. Pathway b is quite unlikely too, since if Q_were
formed, it is difficult ot understand why it would not
undergo a 1,2-acyl shift analogous to that observed for 33.
The only other mode of protonation is pathway c involving the
highly delocalized intermediate cation'g. This highly
stabilized cation 3 may simply be incapable of rearranging

under the prescribed experimental conditions, and was

33
converted back to starting material 55 on quenching. The
nmr spectrum of 55 in neat TFA showed two non-equivalent
gem-dimethyls at 6 1.27 and 1.30 where each appeared as a
singlet, one allylically coupled vinyl methyl at 6 2.20
as a doublet, a one—proton multiplet at 6 4.47, and three
vinyl protons at 6 5.82 (m, 2H) and 6.03 (m, 1H). The
strongly deshielded absorption at 6 4.47 placed that C-H
adjacent to the oxygen atom. These data are consistent with
the structure of cation 5. Therefore, a new mode of proto-
nation, namely at the carbonyl oxygen, appears to operate
on the acid treatment of 55.

In summary, the dienone epoxide 55 rearranges to the
highly conjugated systems 55 (via 51) and 55 (via the allylic
cation E). In contrast to the permethylated allylic cation 5
(Scheme 3, page 11), a nucleophilic attack by solvent or a
1,2-alkyl shift cannot compete effectively with a simple
proton loss from E to give 51 and 55. Thus, products such
as 52 (which could be formed by nucleophilic attack a to the
carbonyl group followed by ring contraction) or 55 (which
could be formed by a 1,2-alkyl shift followed by a loss of

proton) were not observed.

I2

34

CF3COO/\ O O
__) H —+
H
TFA (3
H
>< (\9 H+ 59
0 o
~acyl
H Hi
OH '0
59
O
-H+ H
OH
H
55
0 o
H!" H
—)

 

EXPERIMENTAL

1. General Procedures

 

Analytical gas chromatography (vpc) was carried out
on a Varian Aerograph Model 1400 (flame ionization detector),
and preparative vpc was performed with a Varian Aerograph
AutOprep Model 700 instrument (thermal conductivity detector).
Except where otherwise noted, all nmr spectra were measured
in CDCl3 or CCl4 solutions using TMS as an internal standard.
The 60 MHz spectra were recorded on a Varian T—60 spectrometer
and the 100 MHz spectra were recorded on a Varian HA-lOO
spectrometer. The small number placed next to protons in
the structures in the discussion section are the nmr chemical
shifts of those protons relative to tetramethylsilane. The
numbers in parentheses beside the chemical shifts are the
normalized eurOpium shift numbers. These were obtained by
adding small increments of tris-(l,l,l,2,2,3,3-heptafluoro—
7,7-dimethyl-4,6-octanedione)Eu(III) to the CCl4 or CDCl3
solution of the compound being investigated. After each
addition the nmr spectrum was scanned and the new frequency
of each absorption was recorded. The shift for each absorp-

tion is the difference between the frequency of the shifted

absorption and the original one. The normalized shift

35

36

numbers are ratios obtained by dividing the shift of each
signal in the spectrum by the shift of the least shifted
signal.

Infrared spectra were recorded on a Perkin Elmer 237
grating spectrophotometer and were calibrated against a
polystyrene film. Ultraviolet spectra were obtained with
a Unicam SP-800 in methanol unless otherwise noted. Mass
spectra at 70 eV were obtained from a Hitachi-Perkin Elmer
RMU-6 Operated by Mrs. Ralph Guile. Melting points were
determined with a Thomas-Hoover Melting Point Apparatus and
are uncorrected. Analyses were performed by Spang Micro-
analytical Laboratories, Ann Arbor, Michigan, or Clark

Microanalytical Laboratories, Urbana, Illinois.

2. Epoxidation of 2,3,4,5,6-pentamethyl-4-vinyl-2,5-cyclo-

 

hexadienone (£2)

 

21
To a solution containing 100 mg (0.53 mmol) of 39

in 5 ml of methylene chloride was added, at 0Q, a solution

of mfchloroperbenzoic acid (93.5 mg, 0.54 mmol) in 3 ml

of methylene chloride. The reaction, which was followed

by nmr, was complete in about 2 hours, during which time
mfchlorobenzoic acid precipitated from solution. The solvent
was removed by rotary evaporation, petroleum ether (bp 30—
600) was added, and the mfchlorobenzoic acid was removed by

filtration. The filtrate was washed with aqueous sodium

37

bicarbonate, then with saturated sodium chloride, and dried
(Na2804). The solvent was rotary evaporated, and the
residue was chromatographed on Florisil (80-100 mesh) using
ethyl acetate-hexane (1:4) as eluent. The first fraction
was a mixture of the monoepoxides 383 and 382 (43 mg, 88%
based on unrecovered dienone). The second fraction was
unreacted 42,(55 mg, 55%). The mixture of §§g_and 382
had the following properties: ir (neat) 1660 (s), 1625 (w),
1385 (m), 1350 (w), 1030 (w), 885 (w), 685 (w) cm'l;
uv (MeOH) Amax 253 nm (a 12,700), 212 (6,800); nmr (CC14)
6 1.15 (s) and 1.31-1.40 (overlapping singlets), combined
area 9H, 1.67 (m, 6H), 4.9-6.1 (m, 3H). Europium shift
reagenth, Eu(fod)3, resolved the spectrum and at 100 MHz
separate peaks due to the five methyl groups in each stereo-
isomer were discernible. The area ratios for peaks assigned
to 283 and §§E_was 3:2. The mass spectrum (70 eV) m/e
(rel intensity) of the epoxide mixture was 206 (11), 191 (30),
190 (39), 175 (53), 174 (39), 164 (100), 163 (56), 159 (36),
149 (83), 147 (57), 136 (52), 135 (85), 121 (67), 120 (32),
119 (78), 108 (25), 107 (36), 105 (55), 93 (65), 77 (55),
65 (37).

Anal. Calcd. for C13H1802: C, 75.69; H, 8.80

Found: C, 75.73; H, 8.82

In an effort to decrease the amount of unreacted 40

in the above epoxidation, the ratio of peracid to dienone

was increased to 3:1. To a solution containing 100 mg

(0.53 mmol) of 49 in 5 ml of methylene chloride was added,

38

at 0°, a solution of 280 mg (1.62 mmol) of mfchloroper-
benzoic acid in 10 ml of methylene chloride. The mixture
was stirred at room temperature overnight, then worked up
as described above. Chromatography of the crude product
over Florisil (80-100 mesh) with ethyl acetate-hexane (1:4)
as eluent gave as the first fraction the diepoxides 413'
and 412 (58 mg, 50%) and as the second fraction the mono-
epoxides 3§E_and 382 (43 mg, 40%). There was no unreacted
.40. The diepoxide mixture (2,3;5,6-diepoxy-2,3,4,5,6-
pentamethyl-4-viny1-cyclohexanone) had the following properties:
ir (KBr) 1690 (s), 1380 (m), 1100 (m), 950 (m), 870 (w),
680 (w) cm-l; uv (MeOH) AmaX210 nm (8 2,900); nmr (CC14)
6 1.07 (s), 1.20 (overlapping singlets), 1.33 (s), 1.50 (5),
area from 61.07-1.50 (m, 15H), 4.8-6.4 (m, 3H); Eu(fod)3
shift reagent showed that there were two sets of methyl
singlets, each with area ratios 1:2:2; for the nmr assign-
ments and europium shift slopes, see structures; the two
isomers were present in a 3:2 ratio; mass spectrum (70 eV)
m/e (rel intensity) 222 (<1), 179 (18), 165 (13), 151 (49),
137 (70), 125 (23), 124 (24), 109 (100), 93 (34), 91 (30),
81 (34), 79 (37), 77 (31), 67 (35), 55 (29), 53 (54).

Anal. Calcd. for C13H1803: C, 70.24; H, 8.16

Found: C, 70.28; H, 8.21

Repetition of the epoxidation exactly as above but
with 374 mg (2.2 mmol) of mechlorOperbenzoic acid and 100 mg
(0.53 mmol) of 40 in a total of 20 m1 of methylene chloride

gave 99 mg (85%) of the diepoxide mixture 41c and 4lt and,

39
as the second fraction from Florisil chromatograph, 10 mg
(8%) of triepoxide 45: ir (CC14) 1695 (s), 1450 (w),
1375 (m), 1080 (w) cm‘l; uv (MeOH) Amales nm (a 6,000);
nmr (CC14) 6 0.83 (s), 1.4-l.6 (overlapping singlets; total
area from 6 0.8-1.7, 15 H), 2.60-3.32 (m, 3H), no vinyl
protons; mass spectrum (70 eV) m/e (rel intensity) 238 (1),
195 (13), 167 (77), 164 (33), 163 (28), 153 (50), 149 (100),
147 (33), 137 (73), 135 (65), 133 (43), 125 (86), 123 (86),
121 (36), 119 (44), 109 (70), 107 (94), 105 (55), 97 (28),
95 (26), 93 (52), 91 (97), 84 (20), 81 (45), 79 (80), 77 (64),
67 (60), 65 (39), 55 (63), 53 (75).

Anal. Calcd. for C 65.53; H, 7.61

13H1804‘ 0'
Found: C, 65.63; H, 7.59

3. Acid-Catalyzed Rearrangement of cis- and trans-2,3-epoxy-

2,3,4,5,6:pentamethy1-4-viny1-5-cyclohexenone (38c + 38t)

A solution of 382 and 382 (200 mg, 0.97 mmol) in 1 m1
of trifluoroacetic acid was stirred at room temperature
for 1 hour, then poured into a slurry of aqueous sodium
bicarbonate solution and methylene chloride. The organic
layer was separated, washed successively with aqueous
sodium bicarbonate and saturated aqueous sodium chloride,
and dried (NaZSO4). Evaporation of the solvent left 186 mg
of a light yellow oil which, when subjected to analytical
vpc (5' x 0.125 in column, 3% SE-30 on rarOporl 30, 100-

120 mesh, 125°), showed two peaks corresponding to 46c + 46t

40
(ret. time 5.5 min, 42%) and unreacted 383 (ret. time 9.5
min, 51%). Preparative vpc (5' x 0.25 in column, 10% SE-30
on chromosorb W, 80-100 mesh, 125°) gave pure 923:2,3-
epoxy-2,3,4,5,6-pentamethyl-4-viny1-5-cyclohexenone 383:
ir (CC14) 1660 (s), 1630 (w), 1390 (m), 1250 (m), 935 (w),
880 (s) cm'l; uv (MeOH) AmaXZSO nm (5 10,000), 215 (4,830);
nmr (CC14) see structure, all methyl peaks were sharp singlets
except that at 0 1.67 which was slightly broadened; the
three vinyl protons appeared as a multiplet, 6 4.8-6.0;
mass spectrum (70 eV) m/e (rel intensity) 206 (12), 191 (22),
164 (50), 163 (40), 149 (62), 147 (20), 137 (10), 136 (50),
135 (100), 121 (57), 120 (35), 119 (65), 107 (25), 105 (49),
93 (45), 91 (55), 77 (48), 65 (22). By comparing the nmr
spectrum of the mixture of 3§g_and 3§E_with that of pure
§§g_and by plotting the chemical shifts XE- europium shift
reagent concentration and extrapolating back to zero shift
reagent it was possible to assign the nmr spectrum of 38E
(see structure).

Vpc (5' x 0.125 in column, 5% TCEP (tetracyanoethylated
pentaerythritol) on chromosorb W, 80-100 mesh, 125°, FID)
resolved the chromatograph of 463 and 4§E_into two peaks,
46$ (80%, ret. time 75 min) and 4§£_(20%, ret. time 90 min).
Preparative Vpc (5' x 0.25 in column, 15% TCEP on chromosorb W,
80-100 mesh, 125°) gave each pure isomer: Erangfz-acetyl—
5-viny1-2,3,4,5-tetramethy1-3-cyclopentenone, 46E: ir (CC14)
1740 (s), 1710 (s), 1250 (s), 935 (w), 875 (s) cm‘l; uv (MeOH)

Amax219 nm (8 2,650), 285 (720); nmr (CC14) see structure;

13

77

at
20(
115

Ins

41
the peaks at 6 1.60 and 1.67 were mutually coupled quartets,

g = 1.5 Hz, other methyl peaks were sharp singlets, and the
vinyl protons appeared at 6 4.6-5.8 (m, 3H); mass spectrum

(70 eV) m/e (rel intensity) 206 (<1), 164 (100), 149 (63),

136 (54), 135 (29), 121 (73), 119 (28), 105 (30), 91 (26),

77 (19), 66 (27), 65 (23), 44 (36).

Anal. Calcd. for C H O

13 18 2‘
Found: c, 75.66; H, 9.03

C, 75.69; H, 8.80

Cisz-acetyl-S-vinyl-Z,3,4,5-tetramethy1-3-cyclopentenone, 46g:
ir (cc14) 1740 (m), 1700 (s), 1250 (s), 875 (s) cm'l; uv (MeOH)
Amax218 nm (6 1,150), 283 (30); nmr (CC14) see structure;

all peaks were sharp singlets except for the vinyl protons

at 6 4.7-5.4 (m, 3H); mass spectrum (70 eV) m/e (rel intensity)
206 (<1), 164 (100), 149 (69), 136 (52), 135 (24), 121 (76),
119 (26), 105 (30), 91 (28), 77 (23), 66 (77), 65 (56).

Insufficient 46c was isolated for elemental analysis.

4. Cleavage of 46t with Ease

 

A solution of 462 (26 mg) and sodium methoixde (20 mg)
in 3 ml of methanol was stirred at room temperature for 12
hours, then poured into ice-water and extracted with ether.
The combined ether layers were washed with saturated sodium

chloride and dried (Na S04). Evaporation of the solvent and

2
analysis of the residue by Vpc (5' x 0.125 in column, 3%
SE-30 on rar0porl 30, 100-120 mesh, 1200) showed that all
the 46t was converted to a single product, assigned the

structure 2,3,4,5-tetramethyl-5-vinyl-2—cyclopentenone, 41,

42

Pure 41_was collected by preparative Vpc (5' x 0.25 in column,
10% SE-30 on chromosorb w, 80-100 mesh, 120°): ir (cc14)
1700 (s), 1650 (w), 1250 (m), 885 (s) cm‘l; uv (MeOH) Amax
237 nm (a 8,820); nmr (CC14) 6 1.0-1.2 (m, 6H), 1.63 (q, 3H,
g = 1 Hz), 1.90 (q, 3H, g = 1 Hz), 2.3 (m, 1H), 4.6—5.8 (m,
3H); mass spectrum (70 eV) m/e (rel intensity) 164 (85),
149 (100), 135 (49), 121 (60), 119 (20), 105 (38), 93 (30),
91 (29), 79 (25), 77 (23), 67 (25), 65 (15), 53 (26).

Anal. Calcd. for C11H16O: C, 80.44; H, 9.83

Found: C, 80.47; H, 9.92

Treatment of 41 (10 mg) with excess sodium methoxide
in CH3OD for 5 hours at room temperature followed by workup
analogous to that used in the preparation 0f.£l fxm 46E.
gave 41:94 whose nmr spectrum consisted of two sharp singlets

at 6 1.00 and 1.13 (3 H), a sharp singlet at 6 1.63 (3H), and

a vinyl proton multiplet at 6 4.6-5.8 (3H).

5. 5-Trideuteromethyl-2,3-epoxyf2,3,4,6-tetramethy1-4—

 

vinyl-S-cyclohexenone, 38*

 

To a solution containing 145 mg (0.7 mmol) of a mixture
of 38c and 38t (as obtained from epoxidation of 40) in 5 m1

of dimethylsulfoxide-d was added with stirring and under

6
N2, 95 mg (0.85 mmol) of potassium Efbutoxide. The mixture
was stirred at room temperature for 4.5 hours, then quenched

with ice-water and extracted with ether. The combined

43

organic layers were dried (Na2S04) and the solvent evaporated
to give a nearly quantitative yield of 38*. The nmr spectrum
was identical to that of the starting material, except that

the peak at 6 1.67 was decreased in area by 50%.

6. Acid-Catalyzed Rearrangement of 38*

 

The procedure and workup were as described for the treatment
of 322 and 385 with trifluoroacetic acid. The recovered
unreacted gggf had an nmr spectrum identical with that of
pure 383 except that the signal at 6 1.67 had sharpened

and was reduced in area to only 3H. The major rearrangement
product £§£_had an nmr spectrum identical with that of

pure 46; except that the signal at 6 1.60 was absent, and
that at 6 1.67 had sharpened to a singlet. The amount of

46c* collected was insufficient for an nmr spectrum.

7. Acid-Catalyzed Rearrangement of BBC

 

A solution of pure 389 (22 mg; recovered from the treat-
ment of a mixture of 332 and 3§E_with trifluoroacetic acid
at room temperature for 1 hour) in 0.5 m1 of trifluoroacetic
acid was allowed to stand at room temperature for 12 hours.
There was no change in the nmr spectrum. The solution was
then heated at 60° and the nmr spectrum gradually changed.

After 7.5 hours the reaction was essentially complete and

44

the solution was poured into a slurry of aqueous sodium

bicarbonate and methylene chloride and worked up (vide supra).

 

Vpc (5' x 0.25 in column, 10% SE-30 on chromosorb W, 80-100
0

mesh, 135 ) gave 15% of recovered 38c and 85% of a mixture

(4:1) of 46t and 46c whose spectra (ir, nmr) were identical

with those described above.

8. Epoxidation of 3,4,6,6-tetramethy1-2,4-cyclohexadienone

 

§_3_

To a solution containing 5.0 g (0.033 mol) of 5322 in 25 m1
of methylene chloride was added, at 00, a solution of m:
chlorOperbenzoic acid (5.9 g, 0.034 mol) in 50 m1 of methylene
chloride. The reaction, which was followed by nmr, was
complete in about an hour, during which time mfchlorobenzoic
acid precipitated from solution. The precipitated mfchloro—
benzoic acid was removed by filtration. The solvent of the
filtrate was removed by rotary evaporation, and the residue,
which consisted essentially of compound 32 and a trace

amount of mechlorobenzoic acid as shown by an nmr spectrum,
was chromatographed on a short column of Florisil (80-100
mesh) using ethyl ether as eluent. Compound §2_was obtained
in nearly quantitative yield and was identified by comparison

of its ir and nmr spectra with those of an authentic sample.19

4S
9. Acid-Catalyzed Rearrangement of 4,5-epoxy-3,4,6,6-

 

tetramethyl-2-cyclohexenone, 33

 

A solution of 39 (200 mg, 1.21 mmol) in 1 ml of tri-
fluoroacetic acid was stirred at room temperature for 1/2
hour, then poured into a slurry of aqueous sodium bicarbonate
solution and methylene chloride. The organic layer was
separated, washed successively with aqueous sodium bicarbonate
and saturated aqueous sodium chloride, and dried (MgSO4).
Evaporation of the solvent left 192 mg of a light yellow
oil which, when subjected to analytical vpc (5' x 0.125 in
column, 10% FFAP on chromosorb W, AW-DMCS 80/100, 1750),
showed two peaks corresponding to 35 (retention time 4.5 min,
53%) and 33 (retention time 16 min, 47%). Preparative vpc
(5' x 0.25 in column, 10% FFAP on chromosorb W, 80-100 mesh,
180°) gave 2,3,6,6-tetramethyl-2-cyclohexen-l,S-dione (fl):
ir (CC14) 2950 (s), 1720 (s), 1665 (s), 1640 (s), 1390 (m),
1365 (w), 1250 (w), 1165 (w), 1045 (w), 875 (s) cm'l:
uv (MeOH) Amax245 nm (6 7,800); nmr (CC14) 6 1.20 (s, 6H),
1.77 (m, 3H), 1.97 (m, 3H), 3.10 (m, 2H); mass Spectrum
(70 eV) m/e (rel intensity) 167 (5), 166 (39), 123 (ll),

97 (8), 96 (100), 70 (9), 68 (37), 66 (28), 53 (16), 44 (9),
43 (13), 42 (15), 41 (36).
5223. Calcd. for C10H1402: C, 72.26; H, 8.49
Found: C, 72.13; H, 8.55
4-Methylene-5-hydroxy-3,6,6-trimethyl—2-cyclohexenone

(éé): ir (CC14) 3500 (br), 3000 (m), 1670 (S), 1595 (w),

46
1390 (m), 1250 (w), 1170 (w), 1160 (w), 930 (w), 875 (s) cm‘l;

uv (MeOH) km x270 nm (8 16,430); nmr (CDC13) 6 1.00 (s, 3H),

a
1.10 (s, 3H), 2.05 (d, 3H, 3 = 2 Hz), 2.90 (s, 1H, disappeared
in the presence of 020), 4.10 (m, 1H), 5.43 (m, 1H), 5.57(m,1H),5.70
(m, 1H); mass spectrum (70 eV) m/e (rel intensity) 167 (5),
166 (21), 123 (28), 122 (13), 121 (100), 107 (20), 105 (44),
95 (18), 91 (27), 79 (27), 77 (25), 67 (35), 66 (53), 65 (20),
55 (16), 53 (21), 51 (21).
533;. Calcd. for c H o - c, 72.26; H, 8.49

10 14 2'
Found: C, 72.17; H, 8.39

10. 2-Deutero-3-trideuteromethyl-4,5-epoxy-4,6,6-trimethyl-

2-cyclohexenone, 33*

 

To a solution containing 500 mg (3.01 mmol) of 33 in
15 m1 of dimethyl sulfoxide-d6 was added with stirring and

under N 370 mg (3.30 mmol) of potassium Efbutoxide. The

2,
mixture was stirred at room temperature for 3 hours, then

quenched with ice-water and extracted with ether. The combined
organic layers were dried (Na2 V
evaporated to give a nearly quantitative yield of 33: The

504) and the solution was

nmr spectrum was identical to that of the starting material,

except that the signals at 6 2.08 and 5.72 had disappeared.

47

ll. Acid Rearrangement of 33?

 

The procedure and workup were as described for the
treatment of 33 with trifluoroacetic acid. The rearrange-
ment product 33* had an nmr spectrum identical with that of
33 except that the signal at 6 1.97 disappeared and that
at 6 1.77 had sharpened to a singlet. The other product,
33*, had an nmr spectrum identical with that of 33 except

that the signals at 65.70 and 2.05 were absent.

12. Acid Treatment of 4-methy3ene-5-hydroxyf3,6,6-

 

trimethyl-Z-cyclohexenone, 33

 

A solution of 33 (100 mg, 0.61 mmol) in 0.5 m.of
trifluoroacetic acid was stirred at room temperature, the
reaction being monitored by nmr. The nmr spectrum of 33,
which remained constant with time, in neat TFA showed:

6 1.27 (s, 3H), 1.30 (s, 3H), 2.20 (d, 3H, 3 = 2 Hz),

4.47 (m, 1H), 5.82 (m, 2H), and 6.03 (m, 1H). After being
stirred at room temperature for 15 hours, the reacti-on
mixture was quenched by pouring it into ice and saturated
NaHCO3 solution. The mixture was extracted with methylene
chloride and worked up as described for the acid rearrange-
ment of 33. The only product (determined by nmr) after

15 hr was unrearranged 33.

PART II

PHOTOCHEMISTRY 0F CYCLOHEXADIENONE EPOXIDES

INTRODUCTION

The unusual ground-state reactivity of epoxides has
long been recognized and exploited. Only in recent years, has
the excited-state chemistry of these inherently strained
substrates been intensively explored and the potential
synthetic utility of the photoreactions received attention.
Irradiation of the parent compound, ethylene oxide,

results in decomposition of the molecule.23

hv
____3 + H + co + H CO + CH cno
CH4 + C2H6 2 2 3

Aryl substituted oxiranes undergo photofragmentation
in solution to give arylcarbenes and carbonyl compounds.
The arylcarbenes react in alkene solvents to give the
correSponding cycloprOpanes via addition to the double

4
bond.2

49

50

 

£5

Fl

&I
I:
W)
O

 

Photolysis of ¢,B-epoxy ketones generally gives
' 25
B-diketones. Reusch and co—workers observed that a- and
B-pulegone oxide, 33 and 33, are interconvertable; the

B-diketones 33 are formed at a slower rate photochemically.

 

51
A similar transformation on irradiation of 3,5,5-

trimethyl-Z,3-epoxycyclohexanone‘33_results in a 1:9 mixture
of 2,5,S-trimethyl-cyclohexane-l,3-dione 33 and 2-acety1-

4,4-dimethylcyc10pentanone“33.

 

C) C) o
m + o
C) c) C)
54 _65 .98
1:9

The rearrangement of «,8-epoxyketones involves a
characteristic preferential shift of a B-alkyl rather than a
B-aryl substituent (see Table I). The following order in
migratory aptitude of the B-substituent is based upon the
observations described in Table I:. benzhydryl and benzyl
> hydrogen > methylene > methyl > phenyl.

Zimmerman and co-workers28 suggested that these
rearrangements are initiated by excitation of an electron
from a non-bonding orbital situated on the carbonyl oxygen
to an antibonding n-orbital. Substituents « to the carbonyl
group should be expelled readily by elimination either as
anions or as free radical species from the n,n*-excited
state. Thus, Zimmerman postulated that the unusual migratory
aptitude observed in these systems is in accord with

homolytic carbon-oxygen bond fission. Using the "circle,

52

m2

m0

0 3m

 

 

km :6
mm 62
mm Nmo
on m

.mmm 06:00

NAHMHucwuwwmmm
mcflumumfla
musmsuflpmnsmlm

ms

 

 

mmcoumxwo

am

 

monoumxmxomm

mmcoumxflolm on monoumxmxommlm.8

may mo coflmuw>cou on» CH ucmsuflumndmnm on» no mopsuwumd MHoumeHE

.H wanna

dots, y" notation

53

where n electrons are depicted as solid

dots, heavily S-weighted electrons as circular dots and Py

electrons as small y's, the suggested mechanism is depicted

in the following scheme:

Me
I
Ph-C—’CH=C-Me
I I,
10' .07

94.6
Ph-C-CHLCFMe
I l'
:0: ~05

l. 00

 

Me /O
hV \ o
> \\C-—-CH-C-Me
/ l
Ph 207

00

 

 

Me
electron I
: Ph-C-CH=C-Me
demotion I
10' :01
’° 00
00
o 0;:
I
> Ph
Me
Me

54

Migration is presumed to occur either by methyl-radical
expulsion with intermolecular recombination or by recombination
within the solvent cage.

26

In an elaboration of the mechanism, Markos and Reusch

propose the scheme:

 

 

;) C)
9 \LR hu . [5|] ___, [T']
Q (a
‘\\\ j/
(3 . \R
a O(
.5_7 \
1‘ ? C:
(3/ \
R
1") / (3K
(3 ~’ R
. ,, 6.2
‘3 53

A rationalization for the abnormal migration aptitudes
has been advanced in which it is suggested that the migrating
group has radical character, such as is also observed in the
fragmentation mechanism involving the caged radical pair 33,
The position of hydrogen in the order of migratory aptitudes,
however, militates against a general fragmentation mechanism,
since hydrogen atoms are not generally produced in preference
to alkyl radicals.30 These authors, therefore, prefer a
single step or synchronous route for rearrangement from 31

to 33.

55

Whereas the photochemistry of «,8-epoxyketones has been
studied extensively,31 there has been very little work done
on the photochemistry of vinylogous epoxycarbonyl compound.
Most studies in this field were done by two Swiss workers,
0, Jeger and K. Schaffner on y,6—epoxy-a,B-unsaturated
ketosteroids.

32
Jeger and coworkers noted that 13 isomerized to

13 and 13 upon photolysis.

 

The photochemical rearrangement of the y,6-epoxyketones

Z3 and Z3_has also been reported.33

 

d)

 

1‘1 (75%) +

   

75 _7_e_s (25%)

56
Their results have been used to illustrate to what extent

stereoelectronic control due to conformational constraints
in alicyclic systems may provide for selective transformations
of such a,B-unsaturated-y,6-epoxyketones.

Reaction pathways other than 6 + y migration become
available when the enone group is aliphatic and thus is
geometrically less constrained than it is within a cyclic
framework. 0. Jeger and co-workers34 studied the conversion
of trans-B-ionone epoxide 11 to the furyl ketone 33. They
suggested that the trans-cis isomerization of the double
bond to Z3 and epoxide cleavage to Z3_was followed by

cyclization between the ketone oxygen and the 7 carbon, and

1,4 cleavage of the resulting biradical 33.

 

57
As a consequence of our research in this field, to be
described in this part of the thesis, it is important to
also give here a brief account of the photochemical rearrange—
ments of 4-acyl-cyclopentenones and related compounds. In
general these compounds rearrange upon irradiation to give

35
bicyclo[2.1.0]pentanones. Matsuura and Ogura and later

Plank and Floyd36 reported the photochemical conversion of
Egggng,5-di-3fbutyl-4-pivaloylcyclopent-Z-enone'33’to the
bicyclo[2.1.0]pentanone 33 which isomerized slowly at room
temperature to its isomer 33. This in turn rearranged to

butenolides 33 and 31 upon thermolysis, possibly via ketene

intermediate 33.

 

 

#3 H
H
82 333-
-—- [,room temperature
C) H
86 “H

 

 

be

yi

HO

th

di

ph

by

Ac

58

Similarly, photolysis of’trans-2,5-di-Efbutyl-4-
benzoylcycloPent-Z-enone'33_in methanol gave a quantitative

36
yield of the bicyclo[2.1.0]pentanone 33.

hv

 

   

However, sometimes a bicyclo[2.1.0]pentanone is too
thermolabile to be isolated and a butenolide is obtained
directly as the result of photolysis. One example is the
photochemical rearrangement of spirodiketone 33 to butenolide
33, in which the intermediacy of ketene 3l_was established

37
by trapping with methanol to give 33.

 

 

 

 

59

One interesting aspect of these reactions is that both
33 and 33 rearrange to bicyclo[2.1.0]pentanones of identical
configuration. This stereospecificity suggests that these
rearrangements must involve either a concerted [02a + 1r2a]
process or must proceed via an oxa-di-n-methane mechanism.

Our purpose of this part of the thesis is to explore
reaction pathways and to examine the manner in which the
various methyl groups and hydrogens influence the mode of
photo-induced rearrangements of cyclohexadienone epoxides
33 and 33. The other purpose is to draw any similarities
or dissimilarities between the acid-catalyzed rearrangements
as described in Part 1 of this thesis and the photoinduced

rearrangements of compounds 33 and 33.

—O

RESULTS AND DISCUSSION

1. Photochemistry of 4,3jgpoxyf233,435,6,653examethyl-

 

2,4-cyclohexadienone, 33

 

A. Product Structures

Photolysis of 4,5-epoxy-2,3,4,5,6,6—hexamethyl-2,4-
cyclohexadienone 3315 in ether through Pyrex was followed by
nmr. The nmr signals from 33 began to diminish as the
reaction proceeded, while those corresponding to 33 began to
rise. Compound 33 is known from the acid-catalyzed rearrange-
ment of33.15 As the signals from 33 reached an intensity
as high as that of 33, a new set of signals corresponding
to another product, 33, began to appear at the expense of
both 33 and 33. The total conversion of 33 to 33 required

about 16 hrs. Identical results were obtained using a

uranium glass filter (A > 3300 A) in the photolysis.

O o
L O *5
O O
32 36

60

61

Compound 33 was assigned the structure shown, on the basis

j

 

/

f

95

of the following spectral data, and on its subsequent
reactions. The absence of infrared bands in the region of

1500-1680 cm“1

together with the fact that all the methyl
signals in the nmr spectrum except the one at 6 2.15
(corresponding to acetyl group) appeared above 6 1.3

and were all singlets indicated that 33_must be a saturated
compound. Infrared bands at 1760 and 1710 cm.1 are
attributed to the cyclobutanone and acetyl absorptions
respectively.20 Infrared bands at 1762 and 1701 cm-1 were

repdrted for 5-pivaloyl-1,3-di-37butylbicyclo[2.1.0]pentan-

2-one 33 by Matsuura and Ogura.

 

The
pee
frc
dis
ane

gec

in

688

Pro;
the

Spec

Mir :

62

The mass spectrum showed a parent peak at m/e 194 and a base
peak at m/e 152 corresponding to a loss of CH2=C=O, presumably
from the acetyl group. Due to its thermal lability, to be
discussed below, no attempt was made to obtain an elemental
analysis of 33, The acetyl group at CS was assigned to endo
geometry. This is based on the subsequent reactivity of
33 which can be accounted for only if the acetyl group is
in the endo position.

Compound 33 was photolabile and rearranged to 33 and
31 on photolysis through a Corex filter. Compound 33 was

assigned the structure shown, on the basis of its spectral

(2.90)

1.37 (1.62)
1.70 (2.08) \ / O

C)
1.30 (3.72) ‘—

l.50 (1.56) 1.87 (1.00)

properties. The ir absorption at 1790 cm"1

is consistent with
20

the structure of five-membered enol lactone; the uv

spectrum showed no conjugation within the molecule. The

nmr spectrum revealed two homoallylically coupled methyls

63
at 6 1.50 and 1.87, and singlets at 6 1.30 (3H), 1.37 (broad,

3H), and 1.70 (6H). The resonance at 6 1.37 is at a rather
high field for a vinyl—methyl signal and is possibly

affected by the anisotroPic effect of the carbonyl group.

The mass spectrum showed a base peak at m/e 151 corresponding
to a loss of methyl and carbonyl moieties.

Compound 31 was tentatively assigned the structure:

The fact that its nmr spectrum showed six singlet methyl
signals at or above 6 1.20 and the lack of infrared bands
in the region of 1500-1680 cm"1 indicate that 33_must be a
saturated compound. The ir absorption band at 1760 cm-1
suggested a five-membered ring lactone.20 The uv spectrum
showed no conjugation within the molecule. The mass
spectrum showed a parent peak at m/e 194 and a base peak at

135 corresponding to the loss of CO and CH3 moieties.

2
While structure 33 is consistent with the spectral data, it
has to be further confirmed by other means such as X-ray
analysis or chemical transformations in order to differentiate

it from isomeric structures such as 98.

64

Compound 33 was thermally unstable. At room temperature
the ir spectrum of freshly prepared 33 showed a well-defined
ketene band at 2300 cm-1, besides the frequencies characteristic
of 33. The relative intensities of these bands remained
constant, even though they did slowly diminish with and
were replaced by a new set of signals corresponding mainly
to 33. In an attempt to trap this ketene, freshly prepared
33 was treated with a few drOps of methanol at room temperature,
whereupon it rearranged to a mixture of two new compounds,

33 and 333 (ca. 1:3). Compound 33 was assigned the structure

shown:

  

1.60 (4.71)

1.77 (1.24) I

\:
I/ \ ’4 1.40 (2.04)

1.60 (1.00)

99

65
The molecular formula C12H1802 was confirmed by its mass
spectrum (parent peak m/e 194) and elemental analysis.
Compound '93 showed infrared bands at. 1745 (c=c) and 1700
(C=C) cm_1 and uv maxima at 223 nm (8 2,600) and at 255 nm
(2: 1,070) which indicate that g; is a Al-butenolide with
extended conjugation. Infrared bands at 1740 (C=O) and
1669 (C=C) cm-1 were reported for lactone 33 by Matsuura
and Ogura.35 In spite of an extended conjugation of the
butenolide in 33, it exhibited a normal uv maximum at
215 nm (a 8,430) for a Al-butenolide and the reason for
this was attributed to the non-planarity between the butenolide

group and the extended double bond.

 

8_6

In its nmr spectrum, compound 33 showed two multiplets at
6 1.60 and 1.77 corresponding to six protons each and a
broad six-proton singlet at 6 1.40 which may coupled with
other methyls via long-range coupling. The eurOpium shift

data are also consistent with the structure.

66

Compound 100 was assigned the structure shown:

0
l

C) / 1.72 (4.23)

1.53 (1.83) 1.82 (1.86)
-——<§5> 1.72 (1.86)
1.62 (1.00) \\\.I
(1.72)
100

That it was isomeric with 33 was shown by mass spectrometry
(parent peak m/e 194) and elemental analysis. In the ir
region, it showed absorption at 1745 cm- ; its uv spectrum
showed Amax230nm (a 3,100) consistent with the presence of a
Al-butenolide moiety. The nmr spectrum showed two multiplets
at 6 1.62 (6H) and 1.72 (6H), a singlet at 6 1.53 (3H) and
a quartet at 6 1.82 which was shown later by a labeling
experiment to be coupled to one of the methyls at 6 1.72.
Compound 33 reverted slowly to 33 when allowed to
stand in the solid State at -15° for several weeks. When 33
was heated in CCl in a sealed tube at 180° for 3 hrs, it

4
was converted mainly to 33. A minor product 101 was also

67
isolated by vpc collection. Compound 101 was assigned the

structure shown:

2.38 (4.77) H I
1.02 (3.38) O

1.50 (1.00) \ 4.53 (1.53)

\/ V
1 60 (1 12) \/'/\ 4.07 (1.26)
. ' 1.70 (1.41)

That 333 was isomeric with 33 was again shown by mass
spectrometry (parent peak m/e 194) and elemental analysis.
The ir absorption at 1795 cm"1 is consistent with the enol
lactone structure;20 in its nmr spectrum the compound showed
the expected resonances: two vinyl protons appeared as
doublets at 6 4.07 and 4.53, and are assigned to the
exocyclic methylene group; a mutually coupled three proton
doublet at 6 1.02 and a methine proton quartet at 6 2.38,
the latter having a chemical shift which require that it be
5 to the carbonyl group;39 two singlets at 6 1.50 (3H),

1.70 (3H) and a multiplet at 6 1.60 (GB). The mass spectrum
showed a base peak at m/e 151 corresponding to the loss of

CO and CH3 moieties.

68

B. Mechanism

The primary product from the photolysis of 33 through a
Pyrex filter was 33, The formation of the 4-acetylcyclopente-
none 33 may occur by the cleavage of the C4-O bond and a
shift of C6 from C5 to C4. The preferential ring contrac-
tion to 33 rather than methyl migration (to give‘333) is
consistent with the usual order of migratory aptitudes

observed for the photochemical rearrangement of «,8-epoxy—

ketones to B-diketones.

9 - 9
«+- ———, o
‘\C) C)
O
39.2. .3_2 3‘5;

Compound 33 rearranged further on photolysis to 33.

This type of photorearrangement was not new and has been

35

reported before by Matsuura and Ogura on photorearrangement

of 4-pivaloylcyclopentenone 33_and by Plank and Floyd36 on
4-benzoy1cyclopentenone 33, The mechanism for such stereo-

specific rearrangement was proposed as either O2a + 1r2a

36 38

concerted process or a stepwise oxa-di-n-methane mechanism.
Compound 33 rearranged to 33 and 33 on photolysis
through a Corex filter. The mechanisms leading from 33

to 33 and 33 are quite obvious as shown in Scheme 8. Whether

69
33, formed by recombination process e, or 33, by process f,
is the actual photoproduct remains to be seen.

To test the plausibility of the mechanism outlined in
Scheme 8, 33 labeled with CD3 groups in the positions marked
* and + (called 33f’+) was synthesized and irradiated.

The nmr spectrum of the primary photOproduct‘33 lacked the

signal at 6 1.95. When 33 labeled with a CD group only at

3

the position marked * (called 33f) was rearranged, the

resulting 33 had an nmr spectrum identical with that of

unlabeled-3315 except that the signal at 6 1.95 was reduced

in area by 50% and the band at 6 1.75 sharpened to a singlet.
The final photOproduct 33, which was isolated when

333 was the reactant, lacked the signal at 6 1.23. Starting

with 32:”, the resulting _9__5_ lacked the 6 1.23 and 2.15

signals. While there was no way of knowing the exact methyl

assignments in 33 except for the acetyl group, the label at

6 1.23 was assigned to the C-5 position. This is based on the

35
previous observations '36

that bicyclo[2.1.0]pentanone
systems were formed stereospecifically from 4-acety1-
cyclOpentenone systems, and on the subsequent transformation
of 33 which would necessitate that the * label appear at the
C-5 position in 33.

Compound 33 rearranged further on photolysis to 33
and possibly 33. When 33* was the reactant, the product 33
lacked the signal at 6 1.50 and the quartet at 6 1.87 was

* I
sharpened to a singlet. Starting with 33 ' , the resulting

33 lacked the signals at both 6 1.50 and 1.87. These label

 

70

(J

fidi-n-methane M>§'Z

hv, Corex

 

 

 

m ’ 29.

71

results and the suggested mechanism for the conversion of
33 +33 are shown in Scheme 8.

Compound 33 was thermo-labile and rearranged back to
33 either in its solid state or in a non-polar solvent such

as CC14. At room temperature, the ir spectrum of 33 showed a

well-defined ketene band at 2300 cm-1. The relative intensity
of this band to the carbonyl absorptions at 1760 and 1710 cm'1
remained constant with time. This suggested that 33 might
exist in equilibrium with its ring-opened form‘333. In an
attempt to trap this ketene, 33 was treated with methanol.
However, no ketene was trapped; instead two isomeric products
33 and 333 were obtained. A possible mechanism for the

thermal transformation of 33 to‘33_and‘333 is shown in Scheme 9.
This type of rearrangement is not new and has been observed

before.35'37

To test the plausibility of the mechansim in
Scheme 9, 33 labeled with CD3 groups in the positions marked
* and + (called 33f'+) was rearranged. The nmr spectrum

of the resulting 33 lacked the signal at 6 1.40, and that

of 333 lacked the signals at 6 1.53 and 1.82 and the multiplet
at 6 1.72 was sharpened to a singlet. When 33 labeled with
a CD3 group only at the position marked * (called 33f) was
rearranged, the resulting 33 had an nmr spectrum identical
with that of unlabeled 33 except that the area of the

peak at 6 1.40 was reduced by 50%; the resulting 333 lacked
the signal at 6 1.82 and the multiplet at 6 1.72 was reduced

in area by 50% and was sharpened to a singlet. These

labeling results fully support the proposed mechanism outlined

72

Scheme 9

 

 

 

 

\ _-
I / \ \ I
22 103
polar solvent
neat
or non-polar
solvent ,
M Vinyl ~ Me
O
\ t \\/
— \ \ /\\
o O“ I O‘
C) C) »*
J, \

73

in Scheme 9. The incapability of methanol to trap ketene
is possibly due to the preferred intramolecular reaction over
intermolecular addition to methanol. The reasons for the
different effect of various solvents on the thermal
rearrangement of 33 remain to be determined.

While most of the rearrangements outlined in Schemes
8 and 9 are not new, this work does offer an easy pathway to

permethylated bicyclo[2.1.0]pentanones.

74
2. Photochemistry of 4,57gpogy-333,6,§-tetramethyle2-

gyclohexenone, 33

 

A. Product Structures

The irradiation of 4,5-epoxy-3,4,6,6—tetramethyl-2—
cyclohexenone3319 in ether through Pyrex led to three
photOproducts, 333, 333 and 333. Compounds 333 and 333
are photolabile. They rearranged slowly on further irradia-
tion through a Pyrex filter, but much faster through a
Corex filter. Compound 333 rearranged to 333, and compound
333_rearranged to 333 and 333, Compound 333 was also
photolabile and rearranged to 333_on irradiation through a
Corex filter (Scheme 10)

Compound 104 was assigned the structure shown, on the

1.03 (1.94) 1.38 (1.73)
\/H
P
0.97 (2.08)
\\\
1'10 (2'04) 1.40 (1.00)
/
(3 II 2.22 (2.30)

104

75

 

 

 

 

109 (25%)

 

 

 

 

Scheme 33
o
H /
O H
33
1 '1,
H H O
C? Q|\/ .
O.
\ .
H \ / H
O O H I
303 (34%)
395 (44%) l 30_6_ (22%)
Jr 0
O H o'
\
% }_0_8_ (75%) H_ H
+ _\
H
o” / .132
.121. A
o H
/
\

76

basis of the following spectral data, and on its further
photoreaction. The molecular formula C10H14O2 was confirmed
by the compound's mass spectrum (parent peak m/e 166) and
elemental analysis. The absence of infrared bands in the region
of 1500-1680 cm-1 and the absence of olefinic hydrogen
absorption in the nmr spectrum together with the fact

that all the methyl signals appeared at or above 6 1.40
indicated that 333_must be a saturated compound. The uv
spectrum possessed only end absorption and the infrared
spectrum had carbonyl peaks at 1745 and 1710 cm-1. These

two ir absorption bands possibly originate from coupling
between the two carbonyl groups. The average value of the
two peaks (1727 cm-1) is consistent with the carbonyl
absorption characteristic of the bicyclo[3.1.0]hexan-2-one
system.40 Infrared bands at 1688 and 1717 cm.1 were reported
for 2,2,5,5—tetramethylcyclohexanedione—1,3 333 by Eistert

, 41
and Geiss.

C) \\/’

111

77

The structural and stereochemical assignments were
based on the following nmr data (CC14): 6 0.97 (s, 3H),
1.03 (m, 3H), 1.10 (s, 3H), 1.38 (m, 1H), 1.40 (s, 3H) and
2.22 (d, 1H, 3 = 9 Hz). The magnitude of the coupling implies
g3g-cyclopropyl vicinal coupling42 and thus requires that
the C6-methyl group to be endo. The europium shift data
are also consistent with the structure. The mass spectrum
showed a base peak at m/e 96 corresponding to the loss of
a (CH3)é%C=O moiety.

Compound 333 had nearly identical mass, ir and uv
spectra with those of 333, strongly suggesting that compound
333 was structurally and stereochemically isomeric with
compound 333. Its structure and stereochemical assignments

shown were based on the following nmr (CC14) data:

1.45 (4.37)
1.25 (1.00)

F)
O
H

0.95 (4.50) /\
1.02 (4.00) \\\ 1.33 (2.11)

Cféyl \\\34 1.70 (7.06)

105

78
6 0.95 (s, 3H), 1.02 (s, 3H), 1.25 (m, 3H), 1.33 (s, 3H),

1.45 (m, 1H) and 1.70 (d, 1H, 3 = 2 Hz). The small coupling
constant for the methine protons impliesgggggfcyclopropyl
vicinal coupling42 and thus requires that the C6-methy1
group be exo. The eurOpium shift data are also consistent
with the structure.

Compound 107 was assigned the structure shown, on the

0.96 (3.17) 1) 1.33 (3.34)
/Og
O
\\\ 1.37 (4.80)

 

/ \
1.63 (1.27) ‘ F) 2.38 (3.22)
1.68 (1.00) \\.

basis of the following spectral data. The molecula C10H14O2
was confirmed by the compound's mass spectrum (parent peak
m/e 166) and elemental analysis. In the ir spectrum, it
showed absorptions at 1780 (C=O) and 1710 (C=C) cm-l, and
_its uv spectrum showed Amax230 nm (a 3,320) consistent with

20
the enol lactone structure. Infrared bands at 1783 and

1698 cm"1 were reported for enol lactone 112 by Gibson.

79

 

 

112

The slightly lower frequency of the carbonyl band in 331_
could be due to partial conjugation of the carbonyl group
with the adjacent cycloprOpane ring. The stereochemical
assignments were based on the following nmr data (CC14):

6 0.96 (m, 3H), 1.33 (m, 1H), 1.37 (s, 3H), 1.63 (s, 3H).
1.68 (s, 3H) and 2.38 (d, 1H, 3_= 7 Hz). The magnitude of
the coupling constant implies g3§fcyclopr0pyl vicinal

coupling42

and thus requires that the C6-methy1 group be
endo. The mass spectrum showed a base peak at m/e 96. It
is interesting that the mass spectra of 333 and 331 are
nearly identical, suggesting that a similar rearrangement
may occur on electron impact. Specific assignments within
the structure are based on the europium shift data and

comparison with compound 133_(vide infra).

 

Compounds 108 and 109 had nearly identical mass, ir
and uv spectra with those of 107 which strongly suggested

that they were structurally and stereochemically isomeric

80
with compound 107. Their structures and stereochemical

assignments shown were based on their nmr data (CC14) and

   

1.17 (5.10) 1.15 (2.19)
\\»”’ 1.25 (4.00) ti 1.26 (1.00)
///////ji:
0
1.33 .
(5 76) (D
. \ -
/ )4 1.93 (3.80) 1' 1.44
(6.15)

\\
I.65’().oo) 1.60 (1.63) t\\)
l 65 (I 43) 1.70 (1.11)

1.50 (1.22)

108 109

the results of labeling experiments to be discussed below.
Compound 333 had: 6 1.15 (broad singlet, 3H), 1.33 (s, 3H),
1.17 (m, 1H), 1.65 (s, 6H) and 1.93 (broad singlet, 1H);
compound 333 had: 6 1.25 (m, 1H), 1.26 (m, 3H), 1.44 (d, 1H,
3 = 3 Hz), 1.50 (s, 3H), 1.60 (s, 3H) and 1.70 (s, 3H). The
magnitude of the coupling constant between the hydrogen at
ring juncture and the C6—H implies gi_s-cyc10pr0py1 vicinal
coupling42 and thus requires that the C6-methy1 group be exo.
The positions of the methyl groups at the ring junctures

were assigned based on the eurOpium shift data and by

81

comparison with the model compound 114 (vide infra).
Since the diketone 115, which was not isolated, was
suspected to be the precusor of compounds 104 and 105 in

the photolysis of 33, compound 33, an analog of diketone 115,

o
I

/

/

15

was subjected to photolysis under similar conditions.
Compound §3_was obtained from the acid-catalyzed rearrangement

of epoxyketone 33_(see Part I). On photolysis through a

Pyrex filter, compound 33 rearranged to 113.

hv

O Pyrex
fllfl

 

82

Compound 113 was assigned the structure shown, on the

1.47 (1.37) H H 1.08 (1.00)
\

O
'\
0.95 (1.53) .
1.05 (1.32)
f 1.33 (1.18)
O

113

 

 

basis of the following spectral data, and on its subsequent
reaction. Molecular formula C10H1402 was confirmed by the
compound's mass spectrum (parent peak m/e 166) and elemental
analysis. It had nearly identical ir and uv spectra with
those of 333 and 333. The symmetrical structure of the
molecule was suggested by the presence of two equivalent
methyls at the ring junctures as revealed by its nmr spectrum
and was later further confirmed by a labeling experiment

(see experimental section). Besides the two equivalent
methyls at 6 1.33 (s, 6H), the nmr spectrum showed two
three-proton singlets at 6 0.95, 1.05 and two one-proton
doublets at 6 1.08, 1.47 with a coupling constant of 5 Hz.
The assignments for the methylene bridge protons were based
on the eurOpium shift data, since the proton on the same

face of the carbonyl moieties (endo-H) should be shifted

83
downfield at a faster rate than the exo-H. The downfield

chemical shift of the endo bridged methylene proton at 6 1.47
as compared with exo methylene proton at 6 1.08 is attributable
to the anisoptrOpic deshielding effect of the carbonyl
moieties. The mass spectrum showed a base peak at m/e 96
corresponding to the loss of a (CH3)j%C=O moiety.

Compound 333 was photolabile and rearranged to 333
on irradiation through a Corex filter. Compound 333 was

assigned the structure shown, on the basis of its spectral

1.15 (4.57) 0.77 (3.37)
O
/H\
O 1.28 (5.40)
\ 1.45 (2.19)

1.67 (1.00)K\\\\3
1.77 (1.27)

114

properties. The molecular formula C10H14O2 was again confirmed
by the mass spectrum (parent peak m/e 166) and elemental
analysis. Compound 333 had nearly identical ir and uv

spectra with those of 333, 333 and 333. Its nmr spectrum
showed two three-proton singlets at 6 1.28 and 1.45 and two

one-proton doublets at 6 0.77 and 1.15 with a coupling

84
constant of 4 Hz. The assignments for the methylene
protons were made by applying the same arguments as were
used with compound 333, The assignments for the ring-
junctured methyls were based on eurOpium shift data.
Assuming that the shift reagent coordinates at the carbonyl
group,44 the methyl group at 6 1.28 which has the larger
eurOpium shift is assigned to the position next to the
carbonyl group. The mass spectrum of 333 showed a base
peak at m/e 96 and had a fragmentation pattern very similar
to that of 333.

Compound 106 was assigned the structure shown, on the

O \

u \

C)

6.13 (2.36) H \ /\H 5.13 (3.26)

1.25 (4.42)

1.80 (1.00)
(1.10)

basis of the following spectral data, and on its subsequent
reaction. The molecular formula C10H14O2 was confirmed by
the mass spectrum (parent peak m/e 166) and elemental

analysis. The presence of two equivalent gem-dimethyls

85

as revealed by its nmr spectrum suggested that compound 333
possessed a plane of symmetry, or readily passed through
such a conformation. Besides the two equivalent gem-
dimethyls at 6 1.25, the nmr spectrum showed two vinyl-
methyls centered at 6 1.80 as a multiplet and two vinyl
hydrogens at 6 5.13 and 6.13 where each appeared as a
broad singlet. Homoallylic coupling between the C3-H and
the C4-methy1 was revealed by a labeling experiment (see
experimental section). The uv spectrum showed a maximum
at 240 nm (8 3,225) which indicates that compound 333 has
a diene moiety. A uv maximum at 248 nm (log a 3.87) was
reported for 1,3-cycloheptadiene by Pesch and Friess.45

1 showed that there is no

The ir absorption at 1745 cm-
conjugation between the carbonyl group and the diene moiety.
The mass spectrum showed a base peak at m/e 123 corresponding
to the loss of methyl and carbonyl moieties. The exact
disposition of vinyl hydrogens and methyls on the diene

moiety is based on their chemical shifts, europium shift

data, on the structure of the photOproduct 110 (vide infra)

 

from 333, and on the labeling experiment to be discussed
below.

Compound 333 was assigned the structure shown on the
basis of its spectral prOperties. That it was isomeric with

106 was shown by mass spectrometry (parent peak m/e 166)

86

4.67 (3.72)

H

‘xe C)

 

1'68 (1:28) 0

 

 

 

1.10 (5.28)
H V”? 1.17 (5.55)

2.82 (2.38)

110

and elemental analysis. In the ir region, it showed absorp-
tion at 1780 cm.1 consistent with a y-lactone structure.20
The uv spectrum showed no conjugation within the molecule.
In its nmr spectrum the compound showed two aliphatic
methyls at 6 1.10 and 1.17 where each appeared as a singlet,
two vinyl methyls at 6 1.68 which happened to have the same
chemical shift and appeared as a singlet, a one-proton
multiplet at 6 2.82 and a one-proton doublet at 6 4.67

with a coupling constant of 4 Hz. The homoallylic coupling
between the two vinyl methyls was revealed by spreading the
signal at 6 1.68 into two multiplets using europium shift
reagent. The strongly deshielded absorption at 6 4.67
placed that C-H adjacent to the oxygen atom.39 The mass

spectrum showed a base peak at m/e 123 corresponding to

the loss of methyl and carbonyl moieties.

87

B. Mechanism

The photolysis of compound 33 in anhydrous ether through

a Pyrex filter resulted in the formation of333, 333 and 333.
Since the photochemistry of either «,B-epoxyketones or
vinylogous epoxyketones generally leads to [3-diketones,42'43'44
compound 333 was suspected to be involved as an intermediate
in the photorearrangement of 33. Though no direct evidence
could be obtained for the formation of 333_during the course
of the irradiations, support for the presence of 333 was
provided by the observation that compound 33, a structural
analog of 333, rearranged on photolysis to 333 which is
analogous to 333_and 333, Compound 333 rearranged further
on photolysis to 333 which is analogous to 33ZJ'333 and 333.
The formation of 333 from the photorearrangement of 33 may

occur by the cleavage of the C -0 bond and a hydrogen

4
migration from C5 to C4. The preferential hydrogen migration

to give 115 rather than ring contraction to give 116 is

116 39 115

88

consistent with the usual order of migratory aptitude

observed for the photochemical rearrangement of «,B-epoxy-
ketones to B-diketones.26 Assuming that 333 is an intermediate,
three possible mechanisms can be envisioned for the rearrange-
ment of 333 + 333 + 333 as were proposed by Saboz46 for the

photorearrangement of 117 + 118 + 119.

   

119

These are: 1) a step by step process through diradicals 3;
2) a step by step process through diradical 3_(33§., an
oxa-di-n-methane rearrangement); 3) or a concerted o2a + Tr2a
cycloaddition reaction.

'104 + 105

 

89
Compound 333 rearranged further on photolysis to 333,
and 333 rearranged to 333_and 333, The photochemical
rearrangement of non-enolizable B-diketones is known to give
predominantly enol lactones with an exocyclic olefinic
double bond. For example, Nozaki and coworkers47 reported that
2,2-dimethy1-1,3-cyclohexanedione 333 afforded exclusively

the exocyclic enol 6-1actone 121 on photolysis.

 

 

o O O \
0 Z 4
8': ° (—-> b ‘_-_-_-‘.
\
O . O \J
120 2 .121

The mechanism for the formation of the enol lactone was
preposed as follows: photoexicitation of 120 results in
m-fission giving diradical 3, which recyclizes to the enol
lactone 121.
. 43 .

A recent paper by Gibson on the photochemistry of
(-)-trans-verbenone epoxide 122 reported a similar type of
rearrangement, even though the proposed intermediate‘123

for the formation of 112 was too photolabile to be isolated.

90

 

 

 

 

/ O ‘
0 Me / H
Av \ ——_’
O
22 . o J

lring contraction 123 112 (73%)
C)
H
\

124 (27%)

Compounds 104 and 105 which are unsymmetrical diketones
could, on photoexcitation, undergo m-fission in either direction

to generate diradicals 3, E and 3, 3_respectively. Our

—4 £91
, 109
__, 108

 

91
data showed that the preferential m-fission is in the

direction which would generate more stabilized diradicals
E and 1.
Possible routes to 104-$99, consistent with the

labeling results, are shown in Scheme 11. Experiments were

 

 

 

 

Scheme ll
*H I * o o'
H\ *H\
-—————+ —————+
* O H k/ 0 H * OH
2.9. 2:.
, V c) ‘
O
*H I J‘H *H
\ ¥- '____ * 4/KO
* \OJ ' Ix
\ H \ ( J
H H
.22 2.5.
H
o H ff 9
I
\ o
\ k
* k H '\\ // Id
(3’ ’4} C) I4?

104 105 106

92

done with tetradeuterio-epoxyketone 32f (see Part I). The
mechanisms leading from 32 t°.l2£' l2; are fairly obvious,
but the mechanism leading from 22 to l2§_is less obvious.
The possibility that the intermediate ll; was involved in
the primary process leading to l2§_was ruled out by our
control experiment starting with E1, since compound‘éi
afforded exclusively ll; on photolysis. It is found that four
processes suffice to rationalize the transformation of 32
to lgg: (l) n-w* excitation, (2) a bond scission process
involving the C4-O bond to give §,(3) some rebonding processes
followed by n-electron demotion to give the cyclOpropanone
intermediate lag, and (4) ground state (or possibly excited
state) transformation of $32 to the observed product l2§°

In an effort to detect the transient formation of the
cyclopropanone intermediate lag, the irradiation of vinyl
epoxyketone 22 was carried out using absolute methanol as
solvent. Previous work on the photochemistry of tetra-
methylcyclobutanedione lag has shown that the corresponding
cyc10pr0panone intermediate £21 can be trapped by reacting with

ethanol to form tetramethylcycloPropanone ethyl hemiketal 128.48

93

O o
_L. _h_v__. \ /
-co -co / \
Q 127 129
126 EtOH
H OEt
128

However, we could obtain no evidence for the formation of
£3; under these conditions, and identical results were
obtained as when ether was used as the solvent. The
inability to trap the cyclopropanone intermediate with
methanol shows that, if lag is indeed an intermediate in the
process, its intramolecular reaction tolgé must be very
rapid with respect to the intermolecular addition of methanol.
Another mechanism which, although reasonable, is
definitely excluded by the labeling experiment, is outlined
in Scheme 12. Intermediate g (from Scheme 11) may either
under a hydrogen shift to give ll: or may undergo some
rebonding processes to give the intermediate bicyclic ketone

130. Excited state transformation of 130 could lead to 106.

 

 

 

 

Scheme 12 C?
~ H' ‘ * H
. * o
O H
at H\
— .12
*/
Q
E- * H
L_,
*
C) C)
U . Jl
C)
H
* f4 ‘\ /, H * \ /' Fl
\
*— r
106

However, had this mechanism been Operating, the product lgg
would have contained labels at the C3 and C4 positions
instead of at the C5 and C6 positions as observed.

The alternative ring-opening mode from the intermediate
E to give if would lead to structure 12$ (Scheme 13), which
is also reasonably consistent with the observed nmr spectrum

and the labeling results, but which is inconsistent with the

uv and ir data.

95

Scheme l3

 

In summary, the dienone epoxide‘gg rearranges mainly
via the diketone intermediate 115. A new reaction pathway
leading to 106 seems to be operating via the cyclopropanone

intermediate 125.

EXPERIMENTAL

1. general Procedures

 

Except where otherwise noted, all nmr spectra were
measured in CDCl3 or CCl4 solutions using TMS as an internal
standard. The 60 MHz spectra were recorded on a Varian T-6O
spectrometer and the 100 MHz spectra were recorded on a
Varian HA-lOO Spectrometer. Infrared spectra were recorded
on a Perkin Elmer 237 grating Spectrophotometer and were
calibrated against a polystyrene film. Ultraviolet spectra
were obtained with a Unicam SP-800 in 95% ethanol, unless
otherwise noted. Mass spectra were obtained from a Hitachi-
Perkin Elmer RMU-G Operated by Mrs. Ralph Guile. Melting
points were determined with a Thomas-Hoover Melting Point
Apparatus and are uncorrected. Varian Aerograph gas chro-

matographs were used. Analyses were performed by Spang

Microanalytical Laboratories, Ann Arbor, Michigan.

2. Photolysis of 4,55epoxyf2,3,4,5,6,6-hexamethyl-2,4—

cyclohexadienone, 32

A degassed solution of 50 mg (0.26 mmol) of g3 in

25 ml of anhydrous ether was irradiated through Pyrex with

96

97

a 450 W Hanovia lamp at 0°. The photolysis was followed
by nmr spectroscopy. The nmr signals from 33 began to diminish
as the reaction proceeded, while those corresponding to 2215
began to rise. As the signals from gg reached a intensity
as high as that of 32, a new set of signals from'gg began
to rise at the expense of both 32 and 39. The total conver-
sion from 33 to 22 took about 16 hours.

Recrystallization from petroleum ether (bp 30-600)
gave a solid mass of bicyclo[2.1.0]pentan-2-one, 25 (45 mg,
90%): ir (CC14) 3000 (m), 1760 (s), 1710 (s), 1470 (w),
1400 (w), 1370 (w), 1230 (w).cm '1; uv (MeOH) §ax 225 nm
(6 2,590); nmr (CC14) 6 0.77 (s, 3H), 1.07 (s, 3H), 1.18
(s, 3H), 1.23 (s, 3H), 2.15 (s, 3H); mass spectrum (70 eV)
m/e (rel intensity) 195 (3), 194 (20), 179 (9), 153 (13),
152 (100), 151 (31), 150 (9), 137 (56), 123 (48), 108 (54).
93 (28), 91 (28), 91 (18), 81 (34), 79 (13). Due to its

thermal instability, no attempt was made to obtain its

elemental analysis.

3. Photolysis of 4,S-epoxy-3-trideuteriomethy1-2,4,5,6,6-

pentamethyl-Z,4-gyclohexadienone, 32*

The procedure and workup were as described for the
irradiation of 32. The initial rearrangement product 3§*
had an nmr spectrum identical with that of ggls except that
the signal at 6 1.95 disappeared and the quartet at 6 1.75

*
sharpened to a singlet. The final product 25 had an nmr

98
spectrum identical with that of 25 except that the signal

at 6 1.23 disappeared.

4. Photolysis of 4,5-epoxy-3,5—bis(trideuteriomethyl)-

* +
2,445,6-tetramethy1—2,4-cyclohexadienone, 32 '

The procedure and workup were as described for the
.1.

irradiation of 32. The initial rearranged product 3g ’

*
had an nmr spectrum identical with that of 35 except
that the signal at 6 1.95 was absent. The final product

*

.9_5

1+

was identical with that of 95* except that the signal

at 6 2.15 disappeared.

5. Photolysis of Bicyclopentanone, 95

 

A degassed solution of 100 mg (0.52 mmol) of 22
in 30 m1 of anhydrous ether was irradiated through Corex
with a 450 W Hanovia lamp at 00. The photolysis was followed
by nmr. The reaction went to completion in about 12 hrs.
Analytical Vpc (5' x 0.125 in column, 10% FFAP on chromosorb W,
AW-DMCS 80/100, 1650) showed two components: 22 (79.6%,
ret. time 5 min) and 21 (20.4%, 7 min). Preparative Vpc
(5' x 0.25 in column, 10% FFAP on chromosorb W, 80/100, 1450)
gave lactone 2g: ir (CC14) 2960 (w), 2910 (m), 2850 (w),
1790 (s), 1700 (w), 1450 (m), 1385 (m), 1370 (w), 1365 (w),

1290 (w), 1275 (w), 1225 (m), 1030 (s), 980 (m), 930 (w) cm'l;

99

uv (BtOH) only end absorption; nmr (CC14) 6 1.30 (s, 3H),
1.37 (broad singlet, 3H), 1.50 (q, 3H), 1.70 (s, 6H),
1.87 (q, 3H); mass spectrum (70 eV) m/e (rel intensity)
194 (13), 179 (6), 151 (100), 126 (21), 123 (35), 81 (26),
67 (14), 55 (15), 53 (19).

Anal. Calcd. for C12H12D602: C, 71.94

Found: C, 72.00

Lactone 21; mp 104-106°; ir (c014) 2950 (m), 2920 (m),

2850 (m), 1760 (s), 1460 (w), 1380 (m), 1300 (m), 1050 (m),

950 (m) cm'l; uv (EtOH) xma 220 nm (e 890); nmr (cc14)

x
6 0.70 (s, 3H), 1.03 (s, 3H), 1.12 (s, 6H), 1.17 (s, 3H),
1.20 (s, 3H); mass Spectrum (70 eV) m/e (rel intensity)
194 (2), 179 (2), 155 (55), 135 (100), 120 (16), 119 (42),
107 (19), 104 (22), 93 (21), 91 (27), 44 (60), 43 (25),

41 (24), 39 (20).

Anal. Calcd. for C 02: C, 74.19; H, 9.34

12H18
Found: C, 74.22; H, 9.36

*
6. Photolysis of BicycloPentanone, 99

 

The procedure and workup were as described for the
irradiation of 99. The rearranged product 99* had an nmr
spectrum identical with that of 99 except that the signal
at 6 1.50 disappeared and the quartet at 6 1.87 sharpened
to a singlet; 916 was identical with that of 91 except

that the signal at 6 1.17 disappeared.

100

'k
7. Photolysis of Bigyclopentanone, 2§-,+

The procedure and workup were as described for the
* +
irradiation of 99. The rearranged product 99 ' had an

nmr spectrum identical with that of 99 except that the
* 1‘

' was identical

signals at 6 1.87 and 1.50 disappeared; 91
with that of 91 except that the signals at 6 1.17 and 1.20

were absent.

8. Thermal Reaction of Bicyclopentanone, 99

a). in Methanol:

When compound 99 was treated with a few dr0ps of
methanol, it rearranged to a mixture of 99 and 199_(ca. 1:3
as determined by nmr spectrum). Preparative Vpc (5' x 0.25
in column, 10% FFAP on chromosorb W, 80-100 mesh, 1700)
gave lactone 99 (ret time 25 min): ir (CC14) 2950 (w),
1745 (s), 1700 (m), 1460 (w), 1380 (w), 1280 (m), 1080 (m),
980 (m) cm’l; uv (MeOH) xmax223 nm (a 2,600), 255 (1,070);
nmr (CC14) 6 1.77 (m, 6H), 1.60 (m, 6H), 1.40 (broad singlet,
6H); mass spectrum (70 eV) m/e (rel intensity) 195 (11),
194 (78), 179 (14), 151 (20), 136 (23), 123 (12), 109 (36),
108 (78), 107 (14), 93 (100), 91 (28), 77 (25), 65 (13),

53 (20).
9231. Calcd. for C H O - C, 74.19; H, 9.34

12 18 2'
Found: C, 74.27; H, 9.50

101
Lactone 100 (ret time 45 min): ir (CC14) 2950 (w),
1745 (s), 1450 (w), 1390 (w), 1330 (w), 1260 (w), 1130 (w)

1090 (w) cm'l; uv (MeOH) Am x230 nm (8 3,100); nmr (CCl4)

a
6 1.53 (S, 3H), 1.62 (m, 6H), 1.72 (m, 6H), 1.82 (q, 3H);
mass Spectrum (70 eV) m/e (rel intensity) 194 (25), 179 (32),
151 (36), 150 (28), 149 (100), 137 (10), 136 (7), 135 (29),
134 (14), 133 (18), 126 (53), 125 (48), 123 (43), 119 (15),
109 (11), 107 (10), 97 (16), 93 (10), 91 (14), 81 (20).

9231. Calcd. for C H O : C, 74.19; H, 9.34

12 18 2
Found: C, 74.20; H, 9.47

b).. in CCl4

When compound 99 (40 mg, 0.21 mmol) in 0.5 ml CC14
was heated in a sealed tube at 180° for 3 hrs, it transformed
into a mixture of 191 and 99 (ca. 1:9 as determined by
nmr Spectrum). Preparative Vpc (5' x 0.25 in column, 10%
FFAP on chromosorb W, 80-100 mesh, 160°) gave diketone
9§}5 (ret time 45 min) and lactone 121 (ret time 35 min).
Lactone 191 had: ir (CC14) 2950 (w), 1795 (S), 1695 (w),
1660 (m), 1460 (m), 1390 (w), 1250 (w), 1200 (w), 1130 (w).
1080 (w), 1050 (m), 990 (w), 860 (w) cm-l; uv (MeOH) Amax
225 nm (a 1,310); nmr (CC14) 6 1.02 (d, 3H, 1_= 7 cps),
1.50 (s, 3H), 1.60 (m, 6H), 1.70 (S, 3H), 2.38 (q, 1H,
1 = 7 cps), 4.07 (d, 1H, 1,: 2 cps), 4.53 (d, 1H, 1_= 2 cps);
mass Spectrum (70 eV) m/e (rel intensity) 194 (32), 179 (32),
152 (23), 151 (100), 137 (29), 136 (15), 133 (25), 123 (13),

121 (10), 109 (20), 107 (10), 91 (14), 81 (12), 79 (12), 77 (14).

102

Anal. Calcd. for C12H1802: C, 74.19; H, 9.34

Found: C, 74.15; H, 9.34

c). Neat
When compound 99 was allowed to stand in the cryStalline

state at -15° for several eeeks, it reverted slowly to 99.
*
9. Thermal Reaction of Bigyclgpentanone 95

a). in Methanol:

The procedure and workup were as described for the
unlabeled 99. The rearranged product 99? had an nmr
Spectrum identical with that of 99 except that the area of
peak at 6 1.40 was reduced by 50%; 199} was identical with
that of 199 except that the signal at 6 1.82 disappeared

and the multiplet at 6 1.72 sharpened to a Singlet.

b). in CCl4
The procedure and workup were described for unlabeled

99. The rearranged product 191* had an nmr Spectrum identical

with that of 191 except that the Signal at 6 1.50 disappeared;

99? was identical with that of 99}5 except that the Signal

at 6 1.95 disappeared and that at 6 1.75 sharpened to a

singlet.

 

 

103
*1"
10. Thermal Reaction of Bicyclopentanone 99 '

 

a). in Methanol:

The procedure and workup were as described for unlabeled
99. The rearranged product99*'+ had an nmr identical with
that of 99 except that the signal at 6 1.40 disappeared;
199*'+ was identical with that of 199f except that the

signal at 6 1.53 disappeared.

b). in CC14:
The procedure and workup were as described for the

unlabeled 99, The rearranged product 191*Thad an nmr

identical with that of 191* except that the signals at

*1‘
at 6 4.07 and 4.53 nearly disappeared; 99 ’ was identical

*
with that of 99} except that the signal at 6 1.95 disappeared.

 

 

104
11. Photolysis of 4,5-epogy-3,4,6,6-tetramethy1-2-

gyclohexenone, 99

 

A degassed solution containing 300 mg (1.81 mmol)
of 99 in 30 m1 of anhydrous ether was irradiated through
Pyrex with a 450 W Hanovia lamp. The photolysis was
followed by analytical vpc (5' x 0.125 in column, 10% FFAP
on chromosorb W, AW-DMCS 80/100, 140°). The reaction was
complete in about 8 hrs. Vpc Showed three compounds:
199 (44%, ret time 8.5 min), 199 (34%, 12.5 min), and 199
(22%, 16.5 min). Preparative Vpc (5' x 0.25 in column,
10% FFAP on chromosorb W, 80/100, 150°) gave 9291fl,3,3,6-
tetramethyl-bicyclol3.1.0]hexan-2,4-dione 199: Ir (CC14)
3000 (m), 1745 (w), 1710 (s), 1465 (w), 1385 (w), 1300 (w),
1140 (w) cm'l; uv (MeOH) Amax215 nm (a 2,590); nmr (cc14)
6 0.97 (s, 3H), 1.03 (m, 3H), 1.10 (s, 3H), 1.40 (s, 3H),
1.38 (m, 1H), 2.22 (d, 1H, 9 = 9 Hz); mass spectrum (70 eV)
m/e (rel intensity) 167 (9), 166 (77), 151 (29), 138 (14),
124 (15), 123 (55), 107 (20), 105 (10), 96 (100), 95 (27),
91 (17), 81 (19), 70 (23), 68 (64), 67 (65), 53 (38).

Anal. Calcd. for CloH C, 72.26; H, 8.49

O :

l4 2
Found: C, 72.28; H, 8.49

syn-1,3,3,6-Tetramethy1bicyclo[3.1.0]hexan-2,4-dione

105: ir (CC14) 3000 (m), 1740 (m), 1705 (S), 1465 (w),

1390 (w), 1280 (m), 1130 (w), 1095 (m) cm‘l; uv (MeOH)

AmaXZZS nm (8 1,240); nmr (CC14) 60.95 (s, 3H), 1.02 (s, 3H),

105
1.25 (m, 3H), 1.33 (8, 3H), 1.45 (m, 1H), 1.70 (d, 1H,

9 = 2 Hz); mass Spectrum (70 eV) m/e (rel intensity) 167 (9),
166 (58), 151 (19), 149 (16), 138 (18), 124 (20), 123 (59),
107 (28), 105 (15), 97 (10), 96 (100), 95 (35), 91 (20),

81 (20), 78 (17), 76 (11), 70 (20), 68 (78), 67 (83), 55 (21),
53 (58), 51 (15).

Anal. Calcd. for C10H C, 72.26; H, 8.49

1402‘
Found: C, 72.25; H, 8.37

12. Photolysis of 99

 

The procedure and workup was as described for the

irradiation of 99. The rearranged product 199i had an

nmr spectrum identical with that of 199_except that the
Signals at 6 1.40 and 2.22 disappeared; 199? was identical
with that of 199 except that the signals at 6 1.33 and 1.70
disappeared; 199* was identical with that of 199 except
that the Signal at 6 5.13 disappeared and the area of the
peak at 6 1.80 was reduced by 50% and was Simplified to a

doublet.

13. Photolysis of anti-1,3,3,6-§etramethylbicyclo[3.1.0]-

 

ggxan-2,4-dione, 104

 

A degassed solution containing 100 mg (0.60 mmol) of

104 in 10 m1 of anhydrous ether was irradiated through

 

106
Corex with a 450 W Hanovia lamp. The photolysis was followed

by analytical Vpc (5' x 0.125 in column, 10% FFAP on
chromosorb W, AW-DMCS 80/100, 1400). As the reaction
proceeded, the peak with a retantion time of 8.5 min
(corresponding to 199) decreased in area and a product
peak appeared at 10.5 min. After 1.5 hr, the reaction
was complete and the product, 9291f1,6-dimethy1-4-isopr0pyli-
dene-3-oxa-bicyclo[3.l.0]hexan-2-one 191 was collected by
preparative Vpc (5' x 0.25 in column, 10% FFAP on chromosorb
w, 80/100, 170°): ir (cc14) 3000 (m), 1780 (s), 1710 (m),
1450 (w), 1305 (w), 1140 (w), 1100 (m), 1050 (m), 970 (w),
870 (m) cm‘l; uv (MeOH) xmax230 nm (5 3,320); nmr (c014)
6 0.96 (m, 3H), 1.33 (m, 1H), 1.37 (s, 3H), 1.63 (s, 3H),
1.68 (s, 3H), 2.38 (d, 1H, 9 = 7 Hz); mass spectrum (70 eV)
m/e (rel intensity) 167 (5), 166 (46), 151 (29), 148 (13),
138 (19), 124 (22), 123 (76), 107 (20), 105 (15), 96 (100),
95 (29), 91 (21), 68 (41), 66 (50), 55 (22), 53 (21).

9991. Calcd. for C H O - C, 72.26; H, 8.49

10 14 2'
Found: C, 72.27; H, 8.46

14. photolysis of 104*

 

The procedure and work-up was aS described for the
irradiation of 104. The rearranged product 107* had an nmr
Spectrum identical with that of 107 except that the signals

at 6 1.37 and 2.38 were absent.

 

107

15. Photolysis of syn-1,3,5,6-tegramethylbicyglo[3.1.0]-

hexan-2,4-dione, 105

 

A degassed solution containing 100 mg (0.60 mmol) of
199 in 10 m1 of anhydrous ether was irradiated through Corex
with a 450 W Hanovia lamp. The photolysis was followed
by analytical Vpc (5' x 0.125 in column, 10% FFAP on
chromosorb W, AW-DMCS 80/100, 140°). As the reaction
proceeded, the peak with a retention time of 12.5 min
(corresponding to 199) decreased in area and two product
peaks appeared with retention times of 17.0 and 28.5 min
in the ratio of 3:1. After 1.5 hr the reaction was complete.
Preparative vpc (5' x 0.25 in column, 10% FFAP on chromosorb W,
80/100, 170°) gave a major component, 929:1,6-dimethyl-
4-isopropy1idene-3-oxa-bicyclo[3.1.0]hexan-2-one 199:
ir (CC14) 3000 (m), 1780 (s), 1715 (m), 1450 (w), 1290 (w),
1140 (w), 1060 (m) cm‘l; uv (MeOH) Amax235 nm (8 2,400);
nmr (CC14) 6 1.15 (broad Singlet, 3H), 1.33 (s, 3H),
1.17 (m, 1H), 1.65 (s, 6H), 1.93 (broad Singlet, 1H);
mass Spectrum (70 eV) m/e (rel intensity) 167 (5), 166 (39),
151 (20), 138 (14), 124 (22), 123 (68), 107 (25), 96 (100),
95 (30), 91 (26), 79 (20), 68 (52), 67 (71), 55 (31), 53 (55).

9991. Calcd. for C10H1402: C, 72.26; H, 8.49

Found: C, 72.24; H, 8.57
The minor component was s n-5,6-dimethy1-4-isopro-

pylidene-B-oxa—bicyclo[3.1.0]hexan-2-one 109: ir (CC14)

108

2980 (w), 2940 (w), 1785 (S), 1700 (m), 1460 (W), 1280 (m),
1250 (w), 1180 (m), 1140 (w), 1080 (w), 975 (w), 890 (w) cm‘l;
uv (MeOH) AmaX235 nm (8 6,150); nmr (CC14) 6 1.25 (m, 1H),
1.26 (m, 3H), 1.44 (d, 1H, 9 = 3 HZ), 1.50 (S, 3H), 1.60

(s, 3H), 1.70 (S, 3H); mass spectrum (70 eV) m/e (rel
intensity) 167 (12), 166 (100), 151 (40), 138 (13), 124 (15),
123 (49), 107 (27), 97 (14), 96 (52), 95 (27), 91 (27),

81 (16), 79 (19), 70 (30), 69 (20), 68 (48), 67 (45), 55 (16),
53 (27), 51 (8). Insufficient 199 was isolated for elemental

analysis.

16. Photo1ysis of 105*

 

The procedure and workup was as described for the
irradiation of 199. The rearranged product 199* had an
nmr spectrum identical with that of 199 except that the
signals at 6 1.33 and 1.93 were absent. The amount isolated
for the other rearranged product 199f was not enough for

nmr spectral measurement.

17. Ehotolysis of 4,5,7,7-tetramethy1-2-oxa-cyclohepta-

 

3,5-dien-l-one, 106

 

A degassed solution containing 100 mg (0.60 mmol)
of 106 in 10 m1 of anhydrous ether was irradiated through

Corex with a 450 W Hanovia lamp. The photolysis was

 

109
followed by analytical Vpc (5' x 0.125 in column, 10% FFAP
on chromosorb W, Aw-DMCS 80/100, 135°). As the reaction
proceeded, the peak with a retention time of 12.5 min
(corresponding to 199) decreased in area and a product
peak appeared at 6 min. After 2 hr the reaction was complete
and the product, 2,2,6,7-tetramethyl-4-oxa—bicyclo[3.2.0]-
hept-6-en-3-one 119 was collected by preparative Vpc
(5' x 0.25 in column, 10% FFAP on chromosorb W, 80/100,
180°): ir (c014) 2960 (m), 2920 (w), 1780 (s), 1385 (w),
1330 (w), 1250 (w), 1160 (m), 1100 (s), 1060 (m), 1050 (m),
875 (s) cm’l; uv (MeOH) Amax210 nm (e 830); nmr (0014)
6 1.10 (S, 3H), 1.17 (s, 3H), 1.68 (s, 6H), 2.82 (m, 1H),
4.67 (d, 1H, 9 = 4 Hz); mass spectrum (70 eV) m/e (rel
intensity) 166 (4), 148 (40), 123 (100), 109 (18), 107 (32),
91 (28), 79 (22), 77 (13), 67 (22), 55 (19), 53 (16).

Insufficient 110 was isolated for elemental analysis.

18. Photolysis of 106*

 

The procedure and workup was as described for the
irradiation of 199. The rearranged product 1195 had an nmr
spectrum identical with that of 119 except that the signal
at 6 2.82 disappeared and the area of the peak at 6 1.68

was reduced in area by 50%.

110
19. Photolysis of 2,3,6,6-tetrameghy1-Zficyclohexen-l,5-

dione, 99

A degassed solution containing 200 mg of 99 in 15 m1
of anhydrous ether was irradiated through Pyrex with a 450 W
Hanovia lamp. The photolysis was followed by analytical
vpc (5' x 0.125 in column, 10% FFAP in chromosorb W, AW-DMCS
80/100, 160°). As the reaction proceeded, the peak with a
retention time of 7.0 min (corresponding to 99) decreased
in area and a product peak appeared at 2.0 min. After 3 hr
the reaction was complete and the product, 1,3,3,5-tetra-
methyl-bicyclo[3.1.0]hexan-2,4-dione 119 was collected by
preparative vpc (5' x 0.25 in column, 10% FFAP on chromosorb W,
80/100, 170°): ir (0014) 3000 (m), 1750 (m), 1710 (s),
1470 (w), 1390 (w), 1290 (m), 1070 (m) cm‘l; uv (MeOH)
Amax225 nm (8 1,190); nmr (CC14) 6 0.95 (s, 3H), 1.05 (s, 3H),
1.33 (S, 6H), 1.08 (d, 1H, 9 = 5 Hz), 1.47 (d, 1H, 9 = 5 Hz);
mass Spectrum (70 eV) m/e (rel intensity) 167 (4), 166 (38),
151 (10), 124 (20), 123 (53), 97 (7), 96 (100), 95 (14),
68 (38), 67 (35), 53 (15).

9991. Calcd. for C H 0 ° C, 72.26; H, 8.49

10 14 2'
Found: C, 72.23; H, 8.57

 

 

111
20. photolysis of 99*

 

The procedure and workup was as described for the
*
irradiation of 99. The rearrangement puoduct 113 had an
nmr spectrum identical with that of 113 except that the

area of the peak at 6 1.33 was reduced in area by 50%.

21. 99oto1ysis of 1,3,3,5-tetgame9hy1bicyclo[3.1.0]hexan-

2,4-dione, 113

 

A degassed solution containing 100 mg (0.60 mmol) of
119 in 10 ml of anhydrous ether was irradiated through Corex
with a 450 W Hanovia lamp. The photolysis was followed by
analytical Vpc (5' x 0.125 in column, 10% FFAP on chromosorb W,
AW-DMCS 80/100, 1600). As the reaction proceeded, the peak
with a retention time of 7.0 min (corresponding to 119)
decreased in area and a product peak appeared at 16 min.
After 1 hr the reaction was complete and the product,
l,5-dimethyl-4-isoprOpylidene-3-oxa-bicyclo[3.1.0]hexan-
2—one 119, was collected by preparative vpc (5' x 0.25 in
column, 10% FFAP on chromosorb w, 80/100, 170°): ir (c014)
3000 (m), 1780 (S), 1700 (m), 1350 (w), 1300 (w), 1135 (w),
1150 (w), 1070 (m), 1030 (m) cm'l; uv (MeOH) )maxz35 nm
(a 3,780); nmr (CC14) 6 0.77 (d, 1H, 9 = 4 Hz), 1.15 (d, 1H,
9 = 4 Hz), 1.28 (s, 3H), 1.45 (s, 3H), 1.67 (s, 3H),

1.77 (s, 3H); mass spectrum (70 eV) m/e (rel intensity)

112
167 (4), 166 (34), 151 (10), 124 (21), 123 (58), 97 (7),
96 (100), 95 (16), 69 (10), 68 (35), 67 (33), 53 (15).
9991. Calcd. for C1 H O - C, 72.26; H, 8.49

0 14 2°
Found: C, 72.04; H, 8.34

22. photolysis of 113*

 

The procedure and work-up was asdescribed for the

*
irradiation of 113. The rearrangement product 114 had an

nmr Spectrum identical with that of 114 except that the
areas of the peaks at 6 1.28 and 1.45 were reduced in area

by 50%.

PART III

MISCELLANEOUS

RESULTS

1. Thermal Rearrangements of some of the Dichlorocarbene

Adducts of Hexamethyldewarbenzene

49
It was reported that dichlorocarbene gives three
adducts with hexamethyldewarbenzene 132; the structures

were formulated as 133, 134 and 135.

 

 

 

 

 

 

 

 

 

 

/ / C1
\T—T/ :cc1 \‘ \ I ’ \ I c1
L___———- Cl 1 C1 C1
/ \ // C1 I
C1
C1 C1
_1_§_2_ 199 134 135

In the process of separating 199 from 199 and'199 by
vacuum distillation, compounds 199 and 191 (in the ratio 1:3)
were obtained as the high boiling fraction (120°-130° at
0.1 mm), besides 199 as the low boiling fraction (80-870
at 0.1 mm). Compound 199 was Shown to arise from the thermal

rearrangement of 134 by losing a molecule of HCl. The

114

 

115

origin of 137 remains unknown.

C1“) (C1

 

 

 

 

 

 

 

 

‘\
-HC1 C1
| c1 "“_’ I I
A
K. C1
H
134 136

137
2. Photoisomerization of 1,4,5,6,7-pentamet9ylbicyclol3.2.1]-

 

octa-3,6-dien—2,8-di0ne, l39

 

Allylic oxidation of 1,2,5,6,7-pentamethy1bicyclo[3.2.l]-
octa-2,6-dien-8-one 19921 with chromium.trioxide-pyridine
complex gave the corresponding dione 199_in moderate yield
(ca. 40%). On irradiation through Pyrex, 199_reached a
photostationary state with 199. On further irradiation of
this equilibrating mixture through Corex, a new phot0product
191 was obtained,53 presumably through the decarbonylation

of 139. Compound 141 was photolabile and rearranged slowly

116

to a ketene which was tentatively assigned the structure 142.

.43

/’ Pyrex

H

139 (73%) 40 (27%)

‘

LI

 

 

 

 

H
II
o

142

 

117

3. Acid‘Treatment'of‘Bicyclol3.2.1IOCta-3,6-dienh2-one, 199

Treatment of'19950 with trifluoroacetic acid (TFA) at
room temperature gave 199; treatment of 199 with fluorosulfonic
acid (FSOBH) at room temperature followed by quenching with
sodium methoxide gave199. The structure of 199 was confirmed
by the fact that it reverted to 199 at high temperature on
an SE-30 column by eliminating a molecule of trifluoroacetic
acid. The structure of 199 was confirmed by independent

synthesis via a Michael addition of methanol to 143.

 

 

C)
( \f TFA
pk? A r
143
“-‘T OCOCF
144

 

 

‘r

//O
\

2. NaOMe

 

l . FSOqH ¢b\
\

OMe

145

 

 

118

1,1,2,3,4,4-hexamethy1benzopentalene, 146

Treatment of19951 with 9fchloroperbenzoic acid (9fCPBA)
gave 199, presumbly through the rearrangement of monoepoxide
191 catalyzed by a trace of acid. When the reaction mixture
was chromatographed over neutral alumina, a minor product

149 was isolated besides 148. The origin of 149 remains unknown.

 

a . ‘ V
00 13.-rpm... OJ" _1MOH

146 147 ‘148 (85%)

/ [’0
\ I \ \”
149 (5%)

Photosensitized oxidation of 146 in methanol led to a
compound, the exact structure of which, represented either as

150a or 150b, remains uncertain. Possible routes to 150a

 

and 150b are shown in Scheme 14.

 

 

119

 

 

 

Me
e
146 1W: 02 _.
methylene 0
blue, MeOH
150a 150b
Scheme 19

x)

3‘6 :en302 \ /0.

 

 

. OMe J

OMe

 

 

J,

150a 150

 

 

120

5. EpOxidation of 3,4,4197te1ramet9y1é2,Sbcyc1ghexadienone, 151

The reaction of191’54 with gfchloroperbenzoic acid led
to no epoxidation of the double bond. Epoxidation with
alkaline hydrogen peroxide gave 191 (43%), 199 (5%), 199 (7%)
and 199 (45%). Compound 199 decomposed on preparative Vpc

column, and its structure remains to be determined.

   

151 53 (5%) 154 (7%) 155 (45%)

 

EXPERIMENTAL

1. Vacuum Distillation of the Digh1orocarbene Adducts of

Hexamet9y1dewarbenzene

 

A mixture of compounds 199 and 191 was obtained in the
vacuum distillation of dichlorocarbene adducts of hexa-
methyldewarbenzene49 at 120-1300 (0.1 mm), besides the
expected 19; at 80-870 (0.1 mm). Analytical Vpc (5' x
0.125 in column, 3% SE-30 on chromosorb w, 80/100, 150°)
of the mixture Showed a major component 191 (75%, retention
time 2.5 min) and a minor component 119 (25%, retention
time 3.5 min). Preparative Vpc (5' x 0.25 in column,

10% SE-30 on chromosorb W, 80-100 mesh, 165°) gave 191:

ir (CC14) 2960 (m), 1580 (w), 1450 (w), 1380 (w), 1250 (m),
880 (s) cm‘l; uv (MeOH) Amax265 nm (5 8,500); nmr (cc14)

6 1.15 (s, 3H), 1.18 (s, 3H), 1.50 (q, 3H), 1.60 (q, 3H),
4.53 (s, 1H), 5.17 (s, 1H), 5.40 (d, 1H, 9_= 1 Hz),

5.53 (d, 1H, 9': 1 Hz); mass spectrum (70 eV) m/e (rel
intensity) 258 (6), 256 (31), 254 (47), 241 (16), 239 (24),
221 (36), 220 (20), 219 (100), 206 (19), 205 (18), 204 (60),
203 (22), 202 (33), 200 (50), 189 (22), 184 (50), 183 (38),
169 (60), 167 (44), 164 (90), 156 (35), 155 (30), 154 (98).
153 (56), 152 (46), 141 (35), 139 (33), 129 (71), 128 (90),

121

 

122

127 (42), 115 (80), 91 (35), 77 (60), 75 (37), 65 (36),
63 (48), 53 (35), 51 (68), 39 (76); Cmr (CDC13)52 151.57 (t),
146.23 (m), 144.92 (m), 144.66 (s), 140.93 (s), 132.71 (s),
118.94 (t), 106.64 (t), 59.91 (S), 54.49 (s), 16.44 (q),
13.81 (g), 11.00 (g), 7.05 (q) ppm.

9991. Calcd. for C14H16C12: c, 65.94; H, 6.32

Found: C, 65.63; H, 6.50

The minor component 199 had: ir (CC14) 2960 (m),
1440 (w), 1250 (m), 950 (w), 880 (s) cm‘l; uv (MeOH)
Amax273 nm (a 15,900); nmr (CC14) 6 1.23 (S, 6H), 1.37 (s, 6H),
5.10 (s, 2H), 5.53 (s, 2H); mass Spectrum (70 eV) m/e
(rel intensity) 258 (1), 256 (7), 254 (11), 241 (41),
239 (64), 226 (20), 224 (30), 221 (24), 220 (15), 219 (70).
218 (13), 206 (35), 205 (30), 204 (100), 203 (40), 191 (10),
189 (25), 169 (34), 167 (12), 153 (30), 152 (30), 141 (13),
129 (22), 128 (35), 127 (19), 115 (36), 77 (29), 63 (22),

51 (29), 39 (34).

2. Thermal Re9rrangement of 134

 

A solution of 199 (50 mg, 0.17 mmol) in 0.5 m1 carbon
tetrachloride was sealed in an nmr tube and heated in an
oil bath maintained at 120°. The reaction, monitored by
nmr, went to completion in about an hour. Analytical vpc
(5' x 0.125 in column, 3% SE-30 on chromosorb w, 80/100,

150°) Showed nearly quantitative conversion of 134 to 136.

 

123

3. A11y1ic Oxidation of 1,2,5,6,7-pentamethy1bigyclo[3.2.1]-

 

octa-2,6-dien-9-one, 13821

 

To a slightly cooled solution of 4.75 g (0.06 mol)
of anhydrous pyridine in 80 m1 of methylene chloride was
added 3.0 g (0.03 mol) of chromium trioxide, and the
mixture was allowed to stir at room temperature for 45 min
under nitrogen. A solution of 510 mg (2.7 mmol) of 199
in a small amount of methylene chloride was added and the
mixture was allowed to stir for 24 hrs at room temperature.
The solution was decanted, the residue was rinsed with
petroleum ether (bp 30-600), and the combined organic
phase was washed in succession with saturated sodium bicar-
bonate solution, 2 N hydrochloric acid, saturated sodium
bicarbonate, and saturated salt solution. The organic
phase was dried (M9804), filtered, and concentrated under
reduced pressure, and the residual oil, when subjected to
analytical Vpc (5' x 0.125 in column, 10% FFAP on chromosorb W,
Aw-DMCS 80/100, 180°) Showed a major peak corresponding to
199 (retention time 9.0 min) and several not very well
resolved peaks with shorter retention times. Preparative
vpc (5' x 0.25 in column, 10% FFAP on chromosorb W, 80-100
mesh, 200°) gave 139; 224 mg (40% yield); ir (cc14) 3000 (m),
1770 (s), 1660 (m), 1450 (w), 1390 (w), 1250 (m), 880 (s) cm‘l:
uv (MeOH) Amax235 nm (6 2,540), 262 (2,500); nmr (CC14)
61.20 (s, 3H), 1.27 (S, 3H), 1.58 (q, 3H, 9 = 1 Hz),

 

124
1.68 (q, 3H, 9 = 1 Hz), 1.97 (d, 3H, 9_= 1 Hz), 5.53 (m, 1H);
mass spectrum (70 eV) m/e (rel intensity) 204 (44), 189 (33),
176 (28), 161 (37), 134 (13), 133 (100), 105 (23), 91 (33),
79 (16), 78 (8), 77 (28).
9991. Calcd. for C H O : C, 76.44; H, 7.90

13 16 2
Found: C, 76.31; H, 8.04

4. Irradiation of 1,4,5,6,7:pentamethy1bicyclo[3.2.1]-

 

octa-3,6-dien-2,8-dione, 139

 

A degassed solution containing 100 mg (0.49 mmol)
of 199 in 20 m1 of anhydrous ether was irradiated through
Pyrex with a 450 W Hanovia lamp. The photolysis was followed
by analytical Vpc (5' x 0.125 in column, 10% FFAP on chromo-
sorb W, Aw-DMCS 80/100, 190°). The reaction reached a
photostationary state in about 2 hr. Vpc showed 2 peaks
corresponding to 199 (retention time 6.2 min, 73%) and 199
(retention time 4.5 min, 27%). Preparative Vpc (5' x 0.25 in
column, 10% FFAP on chromosorb W, 80-100 mesh, 200°) gave
1,3,4,5,8—pentamethy1bicyclo[3.3.0]octa-3,7-dien-2,6-dione,
199; ir (CC14) 1700 (s), 1650 (w), 1620 (w), 1445 (w),
1390 (w), 1250 (m), 880 (s) cm‘l: uv (MeOH) lmax235 nm
(a 13,400), 335 (730); nmr (CC14) 6 1.20 (m, 6H), 1.57
(q, 3H, 9 = 1 Hz), 1.97 (q, 3H, 9 = 1 Hz), 2.07 (d, 3H,

9 = 2 Hz), 5.37 (m, 1H); eurOpium Shift data (see structure);

 

 

125

mass Spectrum (70 eV) m/e (rel intensity) 205 (15), 204 (100),
189 (68), 176 (55), 161 (48), 123 (85), 105 (16), 91 (27),
79 (12), 77 (22), 65 (12).

9991. Calcd for C H O : C, 76.44; H, 7.90

13 16 2
Found: C, 76.34; H, 7.90

1.20 (1.25)

 

Q 2.07 (1.64)
\
1.57 (1.00) \ \ H 5,37 (2.70)
1.97 (1.03) ' O

1.20 (1.88)

140

If the photolysis was continued through Corex after
reaching photostationary state, the peaks corresponding to
199 (retention time 6.2 min) and 199 (retention time 4.5
min) began to decrease in areas and a new product peak
appeared at 1.5 min when followed by analytical vpc
(5' x 0.125 in column, 10% FFAP on chromosorb W, AW—DMCS
80/100, 190°). After 1.5 hr, the reaction was nearly
complete and the product 191 was collected by preparative
Vpc (5' x 0.25 in column, 10% FFAP on chromosorb W,
80/100, 1600). Compound 191 had spectra data identical with

53

that of the known compound.

Compound 141 was photolabile and rearranged Slowly

f

 

126

on further irradiation through Corex presumbly to compound
142 as indicated by the appearance of the ir absorption
band at 2102 cm-1. No attempts were made to isolate this

compound.

5. Irradiation of l,3,4,5,8:pentame§hy1bicyg1o[3.3.0]-

 

octa-3,7-dien-2,6-dione, 140

 

The procedure and workup were as described for the
irradiation of 139. Identical results were obtained as

in the irradiation of 139.

6. Treatment of Bicyclo[3.2.1]octa-3,6-dien-2-one (143)

 

with Trifluoroacetic Acid

 

A solution of 50 mg (0.42 mmol) of 199?0 in 0.3 m1
of ice-cold trifluoroacetic acid was stirred at room
temperature, the reaction being monitored by nmr. It
went to completion in about 4 hr and was quenched by
pouring the mixture into ice and saturated NaHCO3 solution.
The product was extracted with methylene chloride and
washed with saturated NaCl solution and dried (NaZSO4).
An nmr spectrum of the crude product showed no starting
material. However, Vpc analysis (5' x 0.125 in column,
3% SE-30 on chromosorb W, 80/100, 100°) Showed three

compounds: starting material 143 (15%, ret time 6.3 min),

127
1999 (21%, ret time 7.5 min), and 1999_(64%, ret time 8.0 min).
The ratio of these three compounds depended on the column
temperature. At higher column temperatures, the peaks
corresponding to 1999 and 1999_decreased in areas, while
that corresponding to 199_increased in its area. The mixture
of 1999 and 1999, which could not be separated easily
without much decomposition on Vpc, was collected by prepara-
tive vpc (5' x 0.25 in column, 10% Se-30 on chromosorb W,
80/100, 135°): ir (neat) 2960 (w), 1780 (s), 1720 (s),
1460 (w), 1420 (w), 1350 (w), 1220 (s), 1160 (s) cm‘l:
nmr (CDC13) 6 2.3-3.4 (m, 6H), 5.35 (m, 1H), 6.2—6.4
(m, 2H); mass Spectrum (70 eV) m/e (rel intensity) 234 (3),
196 (3), 153 (4), 138 (8), 120 (51), 95 (14), 93 (12),
92 (41), 91 (94), 79 (25), 78 (38), 77 (15), 69 (30), 67 (28),
66 (100), 65 (28), 55 (23).

7. Treatment of Bicyclo[3.2.l]octa-3,6-dien-2-one (143)

with_F1uorosu1fonic Acid

 

50

To a solution of 50 mg (0.42 mmol) of 143 in CH C1

2 2
(0.4 ml) was added fluorosulfonic acid (0.1 m1) at -78°.

The mixture was slowly warmed to room temperature and stirred
for 1 hr. The reaction was quenched by pouring the mixture
into methanol and NaOMe solution at -78°. The methanol was

evaporated, water was added to the residue, and the product

was extracted with CH2C12 and washed with saturated NaCl

 

128

2
column, 10% FFAP on chromosorb W, AW-DMCS 80/100, 155°) of

solution and dried (Na 804). Vpc analysis (5' x 0.125 in

the crude product Showed nearly exclusive formation of

199 (ret time 24 min). Preparative vpc (5' x 0.25 in column,
10% FFAP on chromosorb W, 80/100, 1200) gave 199} ir (neat)
2960 (s), 1713 (S), 1460 (w), 1415 (w), 1170 (w), 1230 (m),
1100 (s), 1040 (w), 1010 (w), 980 (w), 940 (W), 920 (w),

750 (S) cm-l; uv (MeOH) Amax290 nm (e 150); nmr (CC14) E
6 1.6-3.3 (m, 6H), 3.33 (s, 3H), 3.60 (m, 1H), 6.20 (m, 2H);
mass spectrum (70 eV) m/e (rel intensity) 152 (2), 120 (14),

110 (9), 92 (15), 91 (33), 79 (16), 78 (12), 74 (39), 67 (15),

 

66 (100), 65 (15), 58 (11), 45 (14).

8. Michagl Addition of Methanol to 143

 

50

To a solution of 143 (50 mg, 0.42 mmol) in methanol

was added 23 mg (0.43 mmol) of NaOMe, and the mixture was
allowed to reflux for 14 hrs. The methanol was evaporated,

water was added to the residue, and the product was extracted

with CHZClz' and washed with saturated NaCl solution and

dried (Na $04). Evaporation of the solvent gave a quantita-

2
tive yield of 145.

\

9. Epoxidati09_of 1,4-dihydro-l,1,2,3,4,4-hexamethy1—

 

anzgpentalene, 146

 

To a solution of 14651 (190 mg, 0.80 mmol) in 5 ml of

 

129

methylene chloride was added, at 0°, a solution of 150 mg
(0.87 mmol) of 9fchloroperbenzoic acid in 2 m1 of methylene
chloride. The mixture was stirred for 1 hr at room temperature
during which time 9fchlorobenzoic acid precipitated from
solution. The solvent was evaporated, petroleum ether

(bp 30-600) was added to the residue, and gfchlorobenzoic
acid was removed by filtration. Evaporation of the solvent
from the filtrate left a brown oil; an nmr spectrum of the
crude material showed it to be > 90% 199. The crude product
was chromatographed on Alumina (80-200 mesh) using 20%
EtOAc/hexane as eluent, to give 173 mg (0.68 mmol, 85%) of
199 and 10 mg (0.04 mmol, 5%) of 199. Compound 199_had

mp 89-90°; ir (KBr) 3400 (s), 2950 (s), 1620 (m), 1460 (m),
1370 (m), 1300 (w), 1180 (m), 1140 (m), 1080 (S), 1000 (w),
950 (m), 860 (s), 790 (w), 750 (s) cm'l: uv (MeOH) Amax

320 nm (Shoulder, 6 8,180), 300 nm (6 10,200), 290 nm
(Shoulder, 6 8,500), 243 nm (Shoulder, 6 5,670), 235 nm

(a 6,820), 230 nm (shoulder, a 5,700); nmr (CDC13) 6 1.33

(m, 15H), 1.77 (S, 1H, disappeared with D20), 5.03 (d, 2H,

9 = 2 Hz), 7.26 (m, 4H); europium Shift data (see structure);
mass spectrum (70 eV) m/e (rel intensity) 255 (13), 154 (53),
240 (20), 239 (92), 238 (23), 237 (17), 236 (59), 224 (27),
223 (32), 222 (24), 221 (100), 211 (15), 209 (12), 208 (17),
207 (19), 206 (55), 197 (15), 191 (29), 190 (14), 189 (15),
181 (14), 179 (15), 178 (20), 169 (15), 165 (27), 152 (12),
103 (12), 89 (18).

 

130

Anal. Calcd. for C H O: C, 84.99; H, 8.72
-—-—- 18 22

Found: C, 85.05; H, 8.72

1.26 (1.41) 1.29 (1.73)
\

 

’ \\ (1 5.03 (3.07)

1.33 (1.00)
1.33 (1.02)

148

Compound 149 had mp 188-1890; ir (KBr) 2960 (m),
2860 (m), 1660 (s), 1540 (w), 1460 (s), 1400 (m), 1370 (m),

1; uv (ether)

1200 (w), 840 (m), 760 (s), 720 (m) cm-
Amax345 nm (6 4,280), 307 nm (Shoulder, 6 18,820), 295 nm
(6 11,100), 288 nm (Shoulder, 6 10,600); nmr (CDC13)
6 1.36 (s, 6H), 1.50 (S, 6H), 2.36 (s, 3H), 7.30 (m, 4H),
10.24 (S, 1H); europium Shift data (see structure); mass
spectrum (70 eV) m/e (rel intensity) 253 (26), 252 (100),
237 (45), 224 (12), 223 (29), 222 (22), 210 (14), 209 (67),
208 (17), 195 (10), 194 (30), 193 (22), 191 (12), 189 (10),
181 (11), 179 (31), 178 (39), 165 (15), 111 (11), 89 (13).
9991. Calcd for C1 H O: C, 85.67; H, 7.99

8 20
Found: C, 85.73; H, 7.96

 

131

1.36 (3.13)

(4 10.24
2.36 (1.64)

1.50 (1.00)

10. Photosensitized Oxidation of 1,4-919ydro-1,1,2,3,4,4-

hexametgylbenzgpentalene, 146

A solution of 19951 (147 mg, 0.62 mmol) in 25 m1 MeOH
through which oxygen was bubbling was irradiated with a
150 W Tungsten lamp in the presence of a catalytic amount
of methylene blue. The photolysis was stopped after 6 hr.
The methylene blue was removed on a short column of
Florisil. The crude product was recrystallized from
methanol to give a compound either as 1999 or 1999;
136 mg (70% yield); mp 150-151°: ir (KBr) 2990 (s), 1490 (m),
1460 (m), 1440 (m), 1380 (m), 1360 (w), 1320 (w), 1305 (m),
1260 (m), 1240 (w), 1200 (w), 1160 (m), 1150 (m), 1130 (m),
1100 (s), 1070 (m), 1050 (S), 990 (s), 970 (m), 930 (s),

 

132

900 (m), 770 (m) cm'l: uv (MeOH) Amaleo nm (6 4,740):
nmr (CC14) 6 0.93 (8, 3H), 1.37 (S, 3H), 1.40 (s, 3H),
1.49 (8, 3H), 1.52 (S, 3H), 1.87 (S, 3H), 2.97 (8, 3H),
3.30 (s, 3H), 7.00 (m, 4H); mass spectrum (70 eV) m/e
(rel intensity) 316 (25), 301 (33), 285 (28), 284 (42),
270 (23), 269 (100), 254 (19), 253 (44), 241 (13), 240 (12),
239 (49), 238 (23), 237 (24), 225 (20), 224 (45), 223 (35),
222 (15), 211 (13), 210 (14), 209 (31), 195 (16), 171 (12),
169 (15), 165 (25), 157 (18), 129 (14), 128 (17).

5991. Calcd. for c H 0 : c, 75.91, H, 8.92

20 28 3
Found: C, 75.84; H, 9.01

11. Epoxidation of 314,4,5-tetramgt9yl-2,5-cyclohexa-

 

hexadienone (151) with m-chlorgpegbenzoic acid

 

 

To a solution containing 200 mg (1.33 mmol) of 1915
in 10 m1 of methylene chloride was added, at 0°, a solution of
gfchloroperbenzoic acid (240 mg, 1.40 mmol) in 5 m1 of
methylene chloride. The reaction mixture showed no Sign
of reaction (nmr) after being stirred overnight at room
temperature. After the usual workup as in the epoxidation of

99 (see page 35), 151 was recovered in quantitative yield.

12. Epoxidation of 3,4,4,5-tetramet9y1-2,S-Eyclohexadienone

 

(151) 91th alkali 9ydrggen peroxide

 

To a solution of 1.86 g (0.012 mol) of 15154 and 12 ml

 

 

133
(0.12 mol) of 30% aqueous hydrogen peroxide in 30 ml of
methanol cooled to 150 was added dropwise and with stirring
7 ml (0.014 mol) of 2 N aqueous sodium hydroxide. After
being stirred at room temperature for 4 hrs, the reaction
mixture was diluted with water and extracted with ether.
The ether extracts were washed with saturated salt solution
and dried (M9504). Evaporation of the solvent left 0.975 g
of an oil which, when subjected to analytical vpc (5' x 0.125
in column, 10% FFAP on chromosorb W, AW-DMCS, 80/100, 1430)
Showed four components: 199 (43%, ret time 3.2 min),
199 (5%, 4.5 min), 199 (7%, 7.5 min), and 199 (45%, 9.5 min).
Preparative vpc (5' x 0.25 in column, 10% FFAP on chromosorb W,
80-100 mesh, 180°) gave 99999f2,3:5,6-diepoxy-3,4,4,5-
tetramethylcyclohexanone 199: ir (CC14) 3000 (m), 1720 (s),
1480 (w), 1400 (w), 1320 (w), 1295 (w), 905 (m) cm‘l:

uv (MeOH) Ama 215 nm (e 580); nmr (CC14) 6 1.20 (s, 6H),

x
1.28 (s, 6H), 2.83 (s, 2H); eurOpium shift data (see
structure); mass spectrum (70 eV) m/e (rel intensity)

182 (l), 167 (2), 153 (4), 150 (5), 139 (11), 135 (7),
126 (8), 125 (100), 123 (9), 110 (11), 107 (14), 97 (15),
91 (12), 83 (12), 79 (14), 77 (10), 69 (26), 67 (21),

55 (77), 53 (20).

 

 

 

134

/////F1 2.83 (3.90)
“o

I

  

1.20 (1.10)

5,6-Epoxy~3,4,4,S-tetramethyl-Z~cyclohexenone 154:

ir (CC14) 3020 (w), 2990 (w), 1670 (s), 1470 (w), 1430 (w),

1370 (w), 1250 (s), 870 (s) cm‘l; uv (MeOH) )max245 nm

(a 5,100); nmr (CC14) 6 1.17 (s, 3H), 1.27 (S, 3H), 1.40

(S, 3H), 1.83 (d, 3H, 9'= 2 Hz), 2.93 (S, 1H), 5.47

(d, 1H, 9 = 2 Hz); europium shift data (see structure);
mass Spectrum (70 eV) m/e (rel intensity) 166 (3), 151 (56),
150 (21), 149 (4), 138 (59), 137 (12), 135 (25), 123 (62),
109 (73), 107 (25), 105 (ll), 95 (18), 93 (ll), 91 (23),

81 (30), 79 (25), 77 (22), 69 (16), 67 (100), 65 (15),
55 (40), 53 (28), 51 (16).

 

 

135

5.47 (5.05) F'

1.83 (1.34)

 

1.17 (1.09)
1.27 (1.02)

‘\ 1.40 (1.00) F"

,zti 2.93 (5,22)

 

Cis-Z,3:5,6-diepoxy-3,4,4,5-tetramethylcyclohexanone 155:

ir (CC14) 3020 (m), 2980 (w), 1700 (S), 1470 (w), 1430 (W),

1405 (w), 1380 (w), 1305 (w), 1270 (w), 1075 (w) cm’l;

uv (MeOH) Amax215 nm (e 990), 233 nm (e 990); nmr-(CC14)

6 1.15 (s, 3H), 1.33 (s, 6H), 1.37 (S, 3H), 2.98 (s, 2H);

europium shift data (see structure); mass
m/e (rel intensity) 182 (3), 167 (4), 151
139 (8), 138 (8), 135 (13), 126 (10), 125

111 (13), 109 (13), 107 (27), 97 (15), 95

spectrum (70 eV)
(5), 150 (10),
(100), 123 (12),

(10), 91 (21),

83 (14), 81 (18), 79 (20), 77 (16), 69 (28), 67 (33),

65 (11), 57 (14), 55 (84), 53 (31).

 

jal:.i«c. a Pu Ii‘lll‘l
A... ‘

 

 

 

136

/,H 2.98 (2.02)

‘ 1.33 (1.00) g

1.15 (1.15) 1.37 (1.54)

 

 

55

Compound 152,which decomposed on the preparative Vpc

column, was not collected.

BIBLIOGRAPHY

 

BIBLIOGRAPHY

1. Ipatieff and Leontowitsch, Ber., 99, 2016 (1903)
2. (a). R. E. Parker and N. 8. Isaac, Chem. Rev., 59,
737 (1959). ‘-

(b). S. Winstein and R. B. Henderson in R. C. Elderfield,
ed., "Heterocyclic Compounds", Vol. 1, Wiley-
Interscience, New York, N. Y., 1950, pp 1-60

3. M. S. Malinovskii, "Epoxides and Their Derivatives“,

Daniel Davey, New York, N. Y., 1965; H. House,

"Modern Synthetic Reactions", 2nd ed., W. A. Benjamin,

Menlo Park, Calif., 1972, p. 318

4. E. L. Eliel and D. W. Delmonte, J. Amer. Chem. Soc.,
99, 1744 (1958).

5. B. Tchoubar, Compt. rend., 214, 117 (1942).
6. D. Abragam and Y. Deux, Compt. rend., 205, 285 (1937).

7. Y. Deux, Compt. rend., 211, 441 (1940); 208, 1090, 2002
(1939); 212, 795 (1941).

8. M. Tiffeneau and P. K. Kuriaki, Compt. rend., 209, 465
(1939).

9. E. Weitz, and A. Scheffer, Ber., 99, 2344 (1921).

10. H. 0. House and D. J. Reif, J. Amer. Chem. Soc.,
19, 6491 (1957).

11. H.O. House and R. L. Wasson, J. Amer. Chem. Soc.,
'19, 4394 (1956).

12. G. D. Ryerson, R. L. Wasson and H. 0. House, Org. Syn.,
Coll. Vol. 9, 957 (1963).

13. For an example, see H. 0. House, "Modern Synthetic Reactions",
2nd ed., W. A. Benjamin, Menlo Park, Calif., 1972, p. 320

14. H. Hart and I. Huang, J. Org. Chem., 99, 1005 (1974).

15. H. Hart, I. Huang and P. Lavrik, J. Org. Chem., 99, 999
(1974).

138

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

139

Chemical Shifts are in 6 units, with relative downfield
shifts in the presence of Eu(fod) given in parentheses;
see D. R. Kelsey, J. Amer. Chem. oc., 99, 1764 (1972).

a. H. B. Henbest, Proc. Chem. Soc., 159 (1963).
b. P. Chamberlain, M. L. Roberts and G. H. Whitham,
J. Chem. Soc., 9, 1374 (1970).

H. Hart, M. Verma and I. Wang, J. Org. Chem., 99, 3418
(1973).

H. Hart and P. Lavrik, J. Org. Chem.,‘99, 1793 (1974).

K. Nakanishi, "Infrared Absorption Spectroscopy,"
Holden-Day, San Francisco, Calif., 1962

H. Hart and M. Nitta, Tetrahedron Lett., 2113 (1974).

H. Hart and R. Lange, J. Amer. Chem. Soc., J. Org. Chem.,
91, 3776 (1966).

a. R. Gomer and W. A. Noyes, Jr., J. Amer. Chem. Soc.,
19, 101 (1950).

b. M. K. Phibbs, B. de B. Darwent and E. W. R. Steacie,
J. Chem. Phys., 16, 39 (1948).

c. R. J. Cvetanovic7_Can. J. Chem., 99, 1684 (1955).

W. Kirmse, "Carbene Chemistry," Academic, New York, 1964
p. 87

C. K. Johnson, B. Doming and W. Reusch, J. Amer. Chem. Soc.,
99, 3894 (1963).

C. S. Markos, and W. Reusch, J. Amer. Chem. Soc.,
99, 3363 (1967).

H. Wherlim C. Lehmann, K. Schaffner and O. Jeger,
Helv. Chim. Acta., 91, 1336 (1964).

H. E. Zimmerman, B. R. Crowley, C-Y. Tseng and J. W. Wilson,
J. Amer. Chem. Soc., 99, 947 (1964).

H. E. Zimmerman and D. I. Schuster, J. Amer. Chem. Soc.,
.99, 4527 (1962).

W. A. Pryor, "Free Radicals," McGraw-Hill, New York,
1966, p. 150

For general review, see R. Bertoniere and W. Griffin,
in "Organic Photochemistry," (O. L. Chapman, ed) Vol. 9
Dekker, New York, 1973

 

32.

33.

34.

35.

36.
37.

38.

39.

40.

41.

42.

43.
44.

45.

46.

47.

48.

140

J. A. Saboz, T. Iizuka, H. Wehrli, K. Schaffner and
O. Jeger, Helv. Chim. Acta., 91, 1362 (1968).

M. Debono, R. M. Malloy, D. Bauer, T. Iizuka, K. Schaffner
and O. Jeger, J. Amer. Chem. Soc., 99, 420 (1970).

B. R. von Wartburg, H. R. Wolf and O. Jeger,
Helv. Chim. Acta., 99, 852 (1972).

T. Matsuura and K. Ogura, J. Amer. Chem. Soc., 99, 3850
(1967).

D. A. Plank and J. C. Floyd, Tetrahedron Lett., 4811 (1971). f—“

G. F. Burkinshaw, B. R. Davis and P. D. Woodgate,
J. Chem. Soc., (C), 1607 (1970).

J. I. Seeman and H. Ziffer, Tetrahedron Lett., 4413 (1973).
R. M. Silverstein and G. C. Bassler, "Spectrometric

Identification of Organic Compounds," John Wiley & Sons,
Inc., New York, 1967, p. 137.

 

H. E. Zimmerman, R. D. Rieke and J. P. Scheffer,
J. Amer. Chem. Soc., 99 2033 (1967).

B. Eistert and F. Geiss, Tetrahedron., 1, l (1959).

a. W. G. Dauben and W. T. Wipke, J. Org. Chem.,
99, 2976 (1967).

b. M. P. Schneider and R. J. Crawford, Can. J. Chem.,
99, 629 (1970).

c. A. Padwa and S. Clough, J. Amer. Chem. Soc.,
91, 5803 (1970).

d. K. B. Wiberg and D. E. Barth, J. Amer. Chem. Soc.,
91, 5124 (1965).

T. Gibson, J. Org. Chem., 99, 845 (1974).

A. F. Cockerill, G. L. 0. Davies, R. C. Harden and
D. M. Rackham, Chem. Rev., 19, 553 (1973).

E. Pesch and S. L. Friess, J. Amer. Chem. Soc.,
19, 5756 (1950).

S. Domb, G. Bozzato, J. A. Saboz and K. Schaffner,
Helv. Chim. Acta., 91, 2436 (1969).

H. Nozaki, Z. Yamaguti, T. Okada, R. Noyori and
M. Kawanisi, Tetrahedron., 99, 3993 (1967).

H. G. Richey, J. M. Richey, and D. C. Clagett,
J. Amer. Chem. Soc., 99, 3907 (1964).

49.

50.

51.

52.

53.

54.

141

H. Hart and M. Nitta, Tetrahedron Lett.,

P. K. Freeman and D. G. Kuper, Chemistry
424 (1965).

A. C. Gripper Gray and H. Hart, J. Amer.
99, 2569 (1968).

Chemical Shifts are in parts per millipn
Splitting patterns from off-resonance H
experiment are Shown in brackets next to
Shifts.

2109 (1974).

and Industry.,

Chem. Soc.,

from TMS. The
decoupled
the chemical

H-N Junker, W. Schafer and H. Niefenbruck, Chem. Ber.,

100, 2508 (1967).

unpublished results by H. Hart and J. Griffiths.

 

M'TITI'IMITILTILIIMHMMIILEWQMMWES