MSU

LIBRARIES
M

b

 

 

RETURNING MATERIALS:
PIace in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

 

 

THE ACYLATION OF KETONE ENOLATES AND

THEIR TRIMETHYLSILYL DERIVATIVES

BY

Robin E. Tirpak

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

1982

ABSTRACT

THE ACYLATION OF KETONE ENOLATES AND

THEIR TRIMETHYLSILYL DERIVATIVES

BY

Robin E. Tirpak

The trimethylsilyl enol ethers of a variety of ketones
were acylated with acetyl chloride in the presence of zinc
chloride or antimony trichloride. The major product of
these reactions was the 1,3-diketone resulting from C-
acylation. Minor amounts of O-acylation were observed in
most cases. Yields of 1,3-diketones varied but were usually
good to excellent. Diethyl ketone trimethylsilyl enol
ether was acylated in an identical manner with various
acid chlorides. 1,3-Diketones were formed, along with
small amounts of enol esters resulting from O-acylation.
Yields for these reactions were also good to excellent.

Magnesium chloride and triethylamine were used to pro-
mote the acylation of several cyclic ketones. N-acyl
imidazoles were found to be the most effective acylating
agent under these conditions. Fair yields of 1,3-di-

ketones were obtained without the complication of products

resulting from O-acylation. Acyclic ketones failed to
react when treated with N-acetylimidazole, triethylamine
and magnesium chloride.

The carboxylation of a variety of symmetrical ketones
was accomplished with carbon dioxide in the presence of tri-
ethylamine and a magnesium halide. The rate of reaction
was fastest when mixtures of magnesium chloride and sodium
iodide were employed. It was assumed that this combination
generated magnesium iodide in situ. B-Keto acids were
isolated, in moderate yields, from these reactions by care-
ful acidification of the resulting magnesium chelated
B-keto carboxylate. A limited study utilizing unsymmetrical
ketones revealed that this carboxylation method is not

highly regioselective.

ACKNOWLEDGMENT

The author wishes to express his appreciation to Dr.
Michael W. Rathke for his invaluable guidance throughout
this investigation. In addition, the author considers
it a privilege to have been associated with him during the
past several years.

Thanks is extended to Dr. Eugene LeGoff for his service
as second reader.

The author further expresses his appreciation to his
fellow graduate students for making these past few years
enjoyable as well as educational.

Finally the author wishes to thank his wonderful wife

and family for their unending support and good wishes.

ii

TABLE OF CONTENTS

Chapter Page

LIST OF TABLES. . . . . . . . . . . . . . . . . . . Vi
LIST OF FIGURES . . . . . . . . . . . . . . . . . . Vii
CHAPTER I - AN INTRODUCTION . . . . . . . . . . . . 1

CHAPTER II - THE LEWIS ACID PROMOTED ACYLATION
OF TRIMETHYLSILYL ENOL ETHERS WITH

ACID CHLORIDES . . . . . . . . . . . . 8
Introduction. . . . . . . . . . . . . . . . . . 8
Results and Discussion. . . . . . . . . . . . . 11
Experimental. . . . . . . . . . . . . . . . . . 23

Materials . . . . . . . . . . . . . . . . . 23
Methods of Analysis . . . . . . . . . . . . 26

Synthesis of Trimethylsilyl Enol
Ethers. . . . . . . . . . . . . . . . . . . 26

General Acylation Procedure Used
to Survey Various Lewis Acids . . . . . . . 28

Optimum Procedures for the Acyla-
tion of Ketone Silyl Enol Ethers. . . . . . 29

Acylation of Ketone Silyl Enol
Ethers (Procedure A). . . . . . . . . . 29

Acylation of Ketone Silyl Enol
Ethers (Procedure B). . . . . . . . . . 30

Synthesis and Analysis of Some
1,3-Diketones . . . . . . . . . . . . . . . 31

CHAPTER III - A STUDY OF KETONE ACYLATION

PROMOTED BY MAGNESIUM CHLORIDE
AND TRIETHYLAMINE . . . . . . . . . . 37

iii

Chapter

Introduction. . . . . . . . . . . . . . . .
Results and Discussion. . . . . . . . . . .
Experimental. . . . . . . . . . . . . . . .

Materials . . . . . . . . . . . . . . .

Methods of Analysis . . . . . . . . . .
Preparation of N-Acylimidazoles . . . .
Preparation of N-Acetylpyrazole . . . .
Preparation of N-Acetylbenzotriazole.
Acylation of a Variety of Ketones
in the Presence of Magnesium Chloride
and Triethylamine . . . . . . . . . . .
CHAPTER IV - THE CARBOXYLATION OF KETONES WITH

CARBON DIOXIDE IN THE PRESENCE OF
MAGNESIUM HALIDES. . . . . . .

Introduction. . . . . . . . . . . . . . .

Results and Discussion. . . . . . . . . . .

Experimental. . . . . . . . . . . . . . . .
Materials . . . . . . . . . . . . . . .

Methods of Analysis . . . . . . . . .

Reaction of Cyclohexanone with CO2
and MgC12 in THF 0 O O O O O O O O O O 0

Reaction of Cyclohexanone with CO2
and MgBr2 . . . . . . . . . . . . . . .

Reaction of Cyclohexanone with CO2
and MgIz. . . . . . . . . . . . . . . .

Preparation and Analysis of Some
B-Keto Acids. . . . . . . . . . . . . .

Reaction of y-Butyrolactone with
C02 and M912. . . . . . . . . . . . . .

iv

Page

37
40
47
47
47
48
49

50

50

53
53
59
75
75

76

76

77

77

78

83

Chapter Page

Reaction of B-Dicarbonyl Compounds
with CO2 and M912 . . . . . . . . . . . . . 83

REFERENCES.....................84

 

Table

II

III

IV

VI

LIST OF TABLES

Reaction of MX with Cyclohexanone
Trimethylsilyl Enol Ether and Acetyl
Chloride. . . . . . . . . . . . . . .
Reaction of Trimethylsilyl Enol
Ethers with Acetyl Chloride . . . . .
Reaction of Various Acid Chlorides
with Diethyl Ketone Trimethylsilyl
Enol Ether. . . . . . . . . . . . . .
Reaction of g with Various Acylating
Agents. . . . . . . . . . . . . . .
Reaction of a Variety of Ketones
with N-Acylimidazoles . . . . . . . .

Carboxylation of a Variety of Ketones

vi

Page

14

20

24

41

45

69

 

LIST OF FIGURES

Figure Page
1 Acyl Chloride-Lewis Acid Complex. . . . . . l6
2 Solvent Study . . . . . . . . . . . . . . . 18
3 Concentration Study . . . . . . . . . . . . l9
4 Optimum Reaction Conditions . . . . . . . . 22
5 Mechanism of Carboxylation. . . . . . . . . 61
6 Rate of CO2 Absorption in a Series

of Solvents . . . . . . . . . . . . . . . . 63
7 Rate of CO2 Absorption for Three

Magnesium Halides . . . . . . . . . . . . . 65
8 MgI2 Stoichiometry Study. . . . . . . . . . 67
9 Et3 N Stoichiometry Study . . . . . . . . . 67

vii

CHAPTER I

AN INTRODUCTION

B-Diketones have long been important compounds in
organic chemistry. Their intrinsic reactivity and struc-
tural features make them useful synthetic building blocks
for larger and more complex molecules.

Classically, the synthesis of 1,3-diketones is accom-
plished using basic conditions by the Claisen condensation

1,2

of ketones with esters (Equation 1). Since esters are

relatively weak acylating agents, a strong base and high

0

 

CH SCH o u o NaH' Etzo ¢>
H + O H H
3 2 3 CHBCHZC c 2 3 “0-5500
(1)
o 0
CH 30H 30H CH + CH c " "
CH3 2 2 2 3 3 HzcoHCCH3
CH3

 

temperatures are usually required for satisfactory results.
Acylation at the methyl group of methyl alkyl ketones pre-
dominates in most cases. O-Acylation may compete kinetic-
ally with C-acylation, but due to the reversibility of the
reaction, only trace amounts of enol esters are obtained.3
Another method for the synthesis of B-diketones is the
irreversible acylation of alkali ketone enolates with acid
chlorides or anhydrides. Generally this reaction suffers

from several complications.l’3

Competing O-acylation

is often a serious side reaction. Also if the diketone
product is enolizable, proton transfer occurs between pro-
duct and starting enolate. For this reason it is normally
necessary to use at least two equivalents of enolate for
each equivalent of acylating agent to obtain high yields.
Recently, however, Seebach obtained high yields of 1,3-

diketones by a 1:1 reaction of lithium enolates with acid

chlorides4 (Equation 2). By carefully adding the enolate

Li 0 o o
__ n 4800 u u
R -—oH2 + RCCl —fiF—> RCCHZCR (2)

to the acylating agent at low temperatures the proton trans-
fer problem appears to be circumvented.
A similar transformation was reported by Howards.

A lithium enolate generated by use of LDA was acylated

with an acyl cyanide at 0° in THF solution (Equation 3).

O
H (3
$129.} + RCCN———-——9 R. )

This procedure gives good to excellent yields of diketone
without the complications of O-acylation or diacylation

products.

Kuwajima recently reported a l,3-diketone synthesis
which also utilizes lithium enolates of ketones6. o-
Chloroacyltrimethylsilanes are used as a-trimethylsilyl-

acyl equivalents and efficiently effect Specific C-acyla-

tion without competing O-acylation (Equation 4). Yields
" .. .3?‘
R'CH2C=CH2 + RCHCSiMe3 -———€> RCHZCCHZC-CHR
l
1 Me Si Cl
3
(4)

o
———9 R CH ZCCHZ C'CHR 031—3—9 R CH ZCCHZC|CH2R
SiMe

are good to excellent.
Another method which bypasses some of the complications
of reacting enolates with acyl halides makes use of the

anion of dimethylhydrazones7 (Equation 5).

 

o o o

1 2 II 1)M8NNH LP " u

R R CHCCH2I23 27mm gr > DR 0301 > RleoHcoHCRL’ (5)
LDA 2)H30 £3

Acidic conditions can also be utilized for construc-
tion of l,3—diketones. Several acidic catalysts have been
employed for the reaction of ketones with acid halides or
anhydrides. The most noteworthy of this type of trans-
formation uses boron trifloridez. This reaction is be—
lieved to involve electrophilic attack by the anhydride
Lewis acid complex on the enol derivative of the ketone
(Equation 6). The resulting l,3-dicarbonyl compound is
subsequently converted to the boron difluoride complex.
Treatment of the complex with sodium acetate yields the
free l,3-diketone. By manipulation of reaction conditions
selective acylation at the a or a' position of a methyl
alkyl ketone can be realized.

Noyori reported a synthesis of l,3-diketones which
does not utilize ketone enolates or enols. a, B-Epoxy-
ketones are converted to diones under catalysis by tetra-

kis (triphenylphosphine) palladium (O) and 1,2-bis-

 

(diphenylphosphine) ethanes. This method requires heating

—+

 

8 8 BF F333 3 95F3
H _ _
C 3CCH3 + (CHBC)20 -——63§+> . CHBC—O CCH3 + CHé-CCHB
+"’ [BFZO‘O
u REFS R
-———+> CH3CCH2CCH3 ——1§E€> CH3c=CH-CCH3 (6)
II II
NaOAC \ CH CCH CCH

2: i7 3 2 3

at high temperatures for prolonged periods of time. It
does, however, have the advantage of being accomplished in

good yields under neutral conditions (Equation 7).

0
ll /0\
CHBCCH-CH(CH

Pd(PPh3)u, dpe

 

O O
I I
) R’ CH SCH GCH(CH ) (7)
3 2 toluene, 140°C 7' 3 2 3 2
60 hours

Another palladium promoted process makes use of un-
saturated carbonyl compounds as starting materials. Tsuji

reacted NaZPdCl with the enones in the presence of t-

4
butyl peroxide and isolated l,3-diketones in acceptable

 

yields (Equation 8)9.

(8)

+oo+ % O

3COZH aq.

 

 

CH

Ishihara synthesized B-diketones via Se-acylmethyl
selenocarboxylateslo. Under the basic conditions of this

reaction good yields of diketones are observed (Equation

9).

P. P. 9. O
RCSeK + BrCHZCR' -——-f>t RCSeCHzgR'

(9)

+ t-CjHlloK R 8 '
«A; RCCH CR

 

Enamines derived from ketones can also be acylated
with acid chlorides or acid anhydridesll. Hydrolysis of
the intermediate provides l,3-diones in reasonable yield

(Equation 10).

 

CHAPTER II

THE LEWIS ACID PROMOTED ACYLATION OF TRIMETHYLSILYL

ENOL ETHERS WITH ACID CHLORIDES

Introduction

 

It is known that trialkylsilyl derivatives of esters

 

(silyl ketene acetals) can be acylated with acid chlorides

(Equation ll)12. This reaction provides a useful route to

N O O

H H
) 3+ s- RCCHZCOCZH5 (ll)
2 H o
3

/OSiMe2tBu 1)Et

H
HZC-C\OC H + R001
2 5

 

B-ketoesters. The reaction of trialkylsilyl derivatives of
ketones (silyl enol ethers) with acid chlorides is far

less satisfactory. Simple acyl halides fail to react with
silyl enol ethers in the absence of a catalystlB. In

the presence of mercuric chloride, trialkylsilyl enol
ethers give exclusive O-acylation, yielding enol esters
(Equation 12)14. Simple acid anhydrides give a similar

transformation with N-(4-pyridyl)pyrrolidine as catalyst

(Equation l3)l3.

 

O
u
OSiEt o 3
1|_ 3n 1)HgCl2 1 CR 2 (12)
R C--CHR2 + R 001 + 9 R C=CHR
2)H O
3
8
S' O O CCH
.9 1M83 u u I 3 (13)
R C=CHR + CHBCOCCHB ——> _R'C=CHR

In a special case, polyhalogenated acid chlorides will
react directly with trialkylsilyl enol ethers. B-Dike-

tones are observed after hydrolysis, but only in moderate

yields (Equation l4)15a'b. Polyhalogenated anhydrides

give parallel results (Equation 14)15a,b.

"
OSiMeB CXZYCCl ' '
R0=CH2 + or -———€> RCCHZCCXZY
(14)
N
(CXZYC)20 X2Y*C12H, C13, F3

KOpka found that silyl enol ethers can be effectively

acylated at carbon with acetyl tetrafluoroborate (Equation

'l’

Why.) '

10

15)l6. Moderate yields of l,3-diketones were reported

along with lesser amounts of enol acetates.

OS'
1' lmeB a 1)CH3N02O -3500
C=CHR2 + CH CBFu

R 3

 

”Name aq. % Ric'CHRZCCHB (15)

The reactions of silyl enol ethers described above
have several disadvantages for the synthesis of B-di-
ketones. The most important of these is the lack of
generality. The method of Murai (Equation 14)15 requires
the use of polyhalogenated acylating agents. This fact
severely limits the range of acylating agents which are
applicable and consequently the diversity of the B-di-
ketones which are obtained. Also the yields reported for
this work are usually poor. Kopka's acylation (Equation
15) requires the generation of acetyl tetrafluoroborate.

The fluoroborate in turn, is derived from acetyl fluoride
and BF3 gas. Clearly, the need for acid fluorides is a
drawback of this method. Finally the occurrence of O-acyla-
tion is an undesirable side reaction which should be avoided
to maximize the yield of B-diketones.

It is clear from the above discussion that the un-
catalyzed reaction of simple acid chlorides and silyl enol
ethers is not possible. If, however, the electrOphilicity
of the acylating agent is enhanced, as with alpha halo
acid halides or acyl tetrafluoroborates, B-diketones can

be obtained (see Equation 14). One alternate way of

ll

achieving this result would be the use of a Lewis acid as

a catalyst. Formation of a Lewis acid-acid chloride complex
should increase the electrophilic nature of the acyl
carbonyl carbon. This would possibly facilitate reaction
between the weakly nucleophilic silyl enol ether and an

acid chloride (Equation 16).

1?SiMe3 8 8 8
R o=CHR2 + RBCClu-LEWIS ACID ——9 R1CCHR2CR3(16)

We decided to study the reaction of trimethylsilyl
enol ethers with acyl chlorides in the presence of a
variety of Lewis acids. Our purpose was the discovery of
a general and productive route to a variety of l,3-di-

ketones.

Results and Discussion

 

We Chose cyclohexanone trimethylsilyl enol ether 1
and acetyl chloride 2 as our standard substrates. Our
initial investigation was a survey of a variety of Lewis
acids in an effort to find a candidate best suited to our
purpose.

Reaction of compound 1 and 2 in methylene chloride at
0°C with one equivalent of boron trifluoride etherate

failed to produce 2-acetylcyclohexanone (Equation 17).

 

12

OSiMe O O

H
+ CH CCl + BF -OEt -—€x&€>
3 3 2 (17)

[H
IN
Rb

This result was disappointing, since it had been shown that
BF3 etherate complex is active in the acylation of 1_with
acetyl fluoridel7. A reasonable explanation for its fail-
ure here may be that boron trifluoride has a relatively low
affinity for chloride, whereas its high affinity for fluoride
is well known. Therefore 2 may not be sufficiently acti-
vated in the presence of boron trifluoride. In accord

with this argument boron trichloride should facilitate the

reaction assuming it has a higher affinity for chloride.

This was found to be the case (Equation 18). A methylene

OSiMe

8 1)BCl ,CH2C12,O°C
+ CH 001 \

3 §)hydrolysis ’ (18)

[H

2 2

chloride solution of BCl3 was reacted with 1 and 2 and pro-
vided 27% of the diketone 3. No cyclohexanone enol acetate

4 arising from O-acylation was seen. Silyl enol ether 1

 

13

was not observed in our GC analysis. Also, cyclohexanone
§_which could have resulted from cleavage of the silicon-
oxygen bond of l by BCl3 or some protonic acid, was not
present. The complete absence of l and g suggests that the
Lewis acid is too potent and is responsible for the destruc-
tion of starting materials. Various other Lewis acid

gave similar results. For example, aluminum chloride and
ferric chloride gave 22% and 23% of 3 respectively. Little
or no enol acetate was detected in either reaction (Table
I, entry 7 and 8). Only small quantities of cyclohexanone,
presumably from protonation of starting material were ob—
served.

We concluded that the Lewis acidity of FeCl3, A1C13,
BCl3 and SbCl3 is more than sufficient to promote the
acylation of compound 1. In addition, they catalyze the
reaction of starting material along nonproductive paths.

It seemed that a compromise was needed. A Lewis acid of
intermediate strength was required in order to avoid destruc-
tion of reactants while still retaining the capability to
catalyze the C-acylation of the silyl enol ether.

Following this reasoning, tin tetrachloride was sub-
jected to our standard reaction (Reaction 19). GC analysis
indicated 34% of 3 in addition to 50% recovered starting
material, detected as cyclohexanone. None of the product
resulting from O-acylation was observed. The low yield
of diketone obtained in Equation 19 was unacceptable.

This was also true for several other Lewis acids which

14

Table I. Reaction of MX with Cyclohexanone Trimethylsilyl
Enol Ether and Acetyl Chloride.

SiMe 0
I 3 O

o C
II II II
+ CH 001 —MX——>
3 CH2012 . +

 

 

a

 

Entry MX C-acylated, % O-acylated, %a

1 ZnCl2 63 5

2 SbCl3 65 16

3 TiCl4 48 trace

4 SbCl5 35 7

5 SnCl4 34 0

6 BCl3 27 0

7 FeCl3 23 O

8 AlCl3 22 1

9 NiBr2 8 0
10 CoBr2 3 trace
11 SiCl4 trace 0
12 CuCl2 trace 6
l3 neBu3SnCl 0 O
14 CuCl 0 0
15 BF3oOEt2 0 O

 

 

aYields determined by GC analysis.

15

 

OSiMe3 0 0
o I H
H ))SnCl CH Cl 0°C
4’ 2 2’ CH
+ CHBCC1 2jnzo 9 3 (19)
1
_ g i
were investigated (Table I, entry 9-14). I

In contrast, acylation of 1 was accomplished in 63%
yield in the presence of zinc chloride. In addition, 5%

of enol acetate 4 was produced (Equation 20). When the r

 

 

o
. u
OSiMe o o CCCH
3 O u u 3
+ 2 1)Zn012,CH2C12,0 C\ + (20)
— 2)H20 ’
1 2 “

same conditions were applied to a reaction employing anti-
mony trichloride, 65% of 3 and 16% of 4 were present after

hydrolysis (Equation 21).

o
1)SbClB,CH2012,O C\
2)H20 '77 +

 

; + g (21)

HT

Accompanying the products (3 and 4) were 20% and 10% of
cyclohexanone in Equation 20 and 21, respectively.

The results described in the preceding paragraph were
the most encouraging of all those listed in Table I. It
appears that zinc chloride 6 and antimony trichloride l

approach the optimum Lewis acidity which allows them to

16

function as efficient acylating catalysts for our experi-
mental system. They provide reasonable yields and ade-
quately avoid the problem of O-acylation. We suggest
that the operative acylating agent is one of or some com-
bination of the three complex forms illustrated for ZnCl2

in Figure l.

 

R-CZEC+ ZnC12" g
u
RCC1 + ZnCl2
é 8.vZnClZ %
RCC1 RCC1u~ZnCl2
b
C

Figure l. Acyl Chloride-Lewis Acid Complex.

The overall reaction may be pictured as: 1) the acylation
of a trimethylsilyl enol ether by a Lewis acid activated

acyl chloride giving a B-diketone (Equation 22), 2) further

R1C:CHR2 + R3CC1 ————+> RloCHRZCR3
C1
I
n
O O
1U 2H '2 lé ll 3 (23)
CCHR CR“ + ZnCl ———> R =C-CR + HCl

l7

reaction of the diketone to yield a B-diketonate metal
complex 8 (Equation 23), 3) hydrolysis of workup of 8 to
free the dione product.

The liberation of protonic acid during the formation of
the diketonate metal complex 8 is a potential problem.
The starting silyl enol ethers are susceptible to cleavage
at the silicon oxygen bond under acidic conditionsla.
Degradation of starting material in this manner leads to
ketones which are unreactive to acylation under our condi-
tions. We assume that this side reaction is the major
factor limiting the yields with ZnCl2 and SbC13. One
possible solution would be to include a tertiary amine as
an acid trap in the reaction mixture to prevent protona-
tion of the silyl enol ether.

Triethylamine and pyridine were utilized in this manner.
In both cases more than 90% of the silyl enol ether was
recovered unchanged. Apparently the amine forms a complex
with the Lewis acid in preference to the acid chloride.
The formation of a precipitate on addition of the amine
to the acid chloride Lewis acid mixture is in agreement
with this. Other attempts to trap the acid with an olefin
before it cleaved starting material also met with defeat.

Next we attempted to optimize the yield of 2-acety1-
cyclohexanone by systematic variation of other reaction

conditions. Several solvents were studied in reactions of

l and 2 in the presence of antimony trichloride at 0°C.

18

The outcome of these experiments is shown in Figure 2.

 

Conditions: 1M solution, 0°C, 1 hour

 

 

 

Solvent CH2C12 CCl4 CHCl3 THF EtZO CH3CN F
% C-acylation 65 16 46 0 trace 56
% O-acylation l6 3 5 0 0 19
Figure 2. Solvent Study. i

Methylene chloride gave the best results and therefore, was
used for all subsequent reactions

In a similar manner, the dependence of product yield
on reaction concentration was studied (Figure 3). Investi-
gation of this narrow range of concentrations pointed to
one molar as the optimum value to be used in all remaining
reactions.

Finally we examined the problem of zinc chlorideE low
solubility in methylene chloride. It appeared that an in-
crease in the amount of Lewis acid in solution would pro—
vide a corresponding increase in its reactivity. To this
end, a mixed solvent system consisting of 15% ether in
methylene chloride was employed in the reaction of 1 and

2 in the presence of ZnClZ. As hoped, the higher solubility

19

 

Conditions: SbCl 0°C, 2 hours

 

3'
Concentration 2M 1M 0.5M 0.25M
% C-acylation 44 65 60 56
% O-acylation 6 15 16 8

 

Figure 3. Concentration Study.

 

of the catalyst in the ether provided an increase in yield
of almost 10% for C-acylation without substantial change in
the fraction of O-acylation. SbCl3 is completely soluble
in methylene chloride and perhaps for this reason does

not benefit from a mixed solvent system.

The net result of our investigation up to this point
was the determination of the optimum reaction conditions
for the acylation of silyl enol ethers with acid chlorides
using Lewis acid catalysis (Figure 4). Since both zinc
chloride and antimony trichloride gave comparable amounts
of l,3-diones, we decided to carry the two catalysts
through the balance of our investigation.

It remained to prove sufficient generality for our
conditions listed in Figure 4. Acetyl chloride was re-
acted, using procedure A and B, with the silyl enol ethers

in Table II. It is clear from the values in Table II, that

20

[II a; 412.”.IIIEI

 

 

 

 

 

ozflmo
NH 3 m C
0 0H <
moEwmo
3 mm m
m E <
mmzflmo
aw .pwuoawomlo aw .pmumeocno wmuspmooum Hmcum Hocm waflm muucm
CowpmHhoMIO scepmamoMIU
m .30: H
mmmmoflwam + mommmmmoofim AWI.N N Hoommo + mm mono m
m = o o 8 mo ._ N Z
mooo o o o o m
= mzwmo

.mUHHoHLU Hmpw04 spas mnwspm Hocm Hmaflmaxzuwfiflue mo nowpommm .HH canoe

.mflmwamcm 00 wn UwCHEHmuwp mpamflxn

.wpfluoHnoflHu wcoEHpcm moms m .Hmnum Hwnumflp paw opHHoHno ocHN moms < mucpwooumm

 

21

 

OH 3 m .hflW\\£m
o o < mmsfimo
ma on m ll. cm
0 mm < m
mzwmo
NH H0 m \f%
3 am < m
ozflmo
aw pmumawomlo ow .pmpmHmownu monopwooum nocum Hocm axaflm wnpcm

 

 

.pmSCHpcoo .HH canoe

22

the reagent of choice to perform the acylations is zinc

 

 

 

chloride.
Procedure Optimum Reaction Conditions
ZnC12, 0°, 15% EtZO/CHZCl2 by vol. ’
A 1 hour rxn. followed by warming to
room temp., H20 quench ,
SbCl3, 0°, CH2C12, 1 hour reaction 5
B followed by warming to room temp.,

H20 quench

 

Figure 4. Optimum Reaction Conditions.

Procedure A provides good to excellent yields of 1,3-
diketones as well as a marked selectivity for C-acylation
in almost all examples. An interesting observation is the
nearly quantitative results with diisopropyl ketone and iso-
butyrophenone (Table II). The acylated derivatives of these
ketones cannot form metal diketonate complexes (see Equation
23). Therefore, protonation of starting material is not
a problem in these cases. This data lends credence to the
assumption that acid hydrolysis is an important factor
limiting yields for procedure A.

The generality of our procedure was further tested.

Diethyl ketone trimethylsilyl enol ether was treated with

23

a selection of acyl chlorides under conditions of pro-
cedure A and B (Table III). Good to excellent yields of
diketones are obtained with zinc chloride catalyst. Again,
there is a high C versus 0 selectivity. Results for pro-
cedure B are less satisfactory.

It appears that procedure A with zinc chloride as a
Lewis Acid catalyst for the acylation of trimethylsilyl
enol ethers is the best procedure uncovered by our investi-
gation. It combines convenience, generality in substrate
and acylating agent, and a high selectivity for reaction
at carbon. It will probably complement other methods

(see Chapter I) reported for the C-acylation of ketones.

Experimental

 

Materials

 

Reagent grade carbon tetrachloride, chloroform, and
methylene chloride were dried over 4 A molecular sieves.
Tetrahydrofuran was distilled from sodium and benZOphenone.
Anhydrous diethyl ether was obtained from Mallinkrodt and
was used without further treatment. Acetonitrile was
distilled from calcium hydride as were all the commercially
available ketones used for the synthesis of the trimethyl-
silyl enol ethers. Acid chlorides, obtained from Aldrich
Chemical Co., were purified by simple distillation. Alum-

inum chloride,and Copper(II)chloride were obtained as

 

 

24

 

om Ho m Ho///\\. xc;mo
m mm m u HHH
m mm m Hull/\\\:d
=
v mm q 0 HH
mm Hm m HO <MIO
=
m be < o H
aw .pwHMH>OMIO aw .pmumH>OMIU mmuspmooum H0///\\\m . wuucm
=
o

 

 

CoprHmoMIO :oflpmHhoMID
\
Cay—MW
momonwpm + momoofim Aw Home + mmomouopm

: . .m m __ _
mwo o a 000 Ho mo 0 mmzflmo

.Hwnum Hocm
Hwaflmamcuwfiflue mcopox axsumflo SDHB mmpfluoHLU Uflod msoflum> mo :Ofluommm .HHH wHQMB

m

950: H

 

25

 

.00 mm conflfihmump mpamflwn

.prHOHsoHHu >coeflpcm mom: m .umnpm axnumflp ppm opflHoHso OGHN mom: m muscwoonmm

 

 

 

AH mm m H0///\\\omxmmov
= >
o om < O
«N mo m Ho momfimmov
lll\\\
m Hm fl ._ >H
O
aw .pmumHmomno ow .UoumH%OMIU mwuspwooum H0///\\\m muucm

 

 

.pwscHuCOU .HHH canoe

26

anhydrous reagents from Fisher Scientific Co. Boron tri-
fluoride etherate, acquired from Eastman Chemical, was dis-
tilled from calcium hydride prior to use. Antimony penta-
chloride was obtained from Matheson, Coleman, and Bell.
Cobalt dibromide, and copper(I)chloride were supplied by
Alfa Inorganics. All other Lewis acids were purchased from

Aldrich Chemical Co.

Methods of Analysis

 

1H NMR data were obtained on a Varian T-60 spectrometer

at 60 MHz. Chemical shifts are reported in parts per mil-
lion on the delta scale relative to TMS internal standard.
Infrared spectra were recorded on a Perkin Elmer 23-B
spectrometer with a polystyrene standard. Mass spectral
data were acquired with a Finnigan Model 4000 electron im-
pact GC/Mass spectrometer. Gas chromatographic analyses
were performed with Varian 920 and 90-P chromatographs
equipped with 6 ft. by 0.25 in. stainless steel columns

packed with 15% SE-30 on Chromosorb W.

Synthesis of Trimethylsilyl Enol Ethers

 

Cyclohexanone trimethylsilyl enol ether 1_was prepared
by the following procedure which is adapted from that of
Duboudinlg.

Sodium iodide (550 mmol, 82.4 g) was flame dried

 

27

under vacuum in a one liter round bottom flask equipped
with a septum inlet and a magnetic stirrer. The flask was
flushed with argon and charged with 500 mL of dry aceto-
nitrile. After the NaI dissolved 500 mmol (49.1 g, 51.8
mL) of cyclohexanone were added followed by 550 mmol (55.7
g, 76.5 mL) of trimethylamine. Trimethylchlorosilane (550
mmol, 69.8 mL) was slowly introduced to the flask so as to
avoid a rapid reflux due to the exothermicity of the reac-
tion. Large amounts of precipitate formed during this
addition. Stirring continued for approximately one hour,
at which time the reaction was diluted with 300 mL of pen-
tane. This mixture was washed with 200 mL of cold water
followed by a rapid separation of layers. The aqueous
layer was washed with an additional 100 mL of pentane. The
pentane layers were combined and dried over sodium sulfate.
Removal of solvent in vacuo, followed by vacuum distilla-
tion through a short vigreux column provided 63 g (75%) of
pure cyclohexanone trimethylsilyl enol ether: bp 63-65°C/

1

10 mm Hg; H NMR (CDC13) 0.2 (9H, s), 1.57 (4H, m), 1.93

(4H, m), 4.75 (1H, m).

Cyclopentanone Trimethylsilyl Enol Ether was prepared
1

 

as above: H NMR (CDC13) 0.2 (9H, m), 1.6-2.4 (6H, m),

4.5 (1H, bs).

3-Pentanone Trimethylsilyl Enol Ether was prepared
1

 

as above: H NMR (CDC13) 0.2 (9H, s), 1.0 (3H,

 

28

5, J=7Hz), 1.45 (3H, d, J=6Hz), 1.95 (2H, m), 4.45 (1H,

9, J=5Hz).

2,4-Dimethyl-3-Pentanone Trimethylsilyl Enol Ether was
1

 

prepared as above: H NMR (CDCl3) 0.2 (9H, s), 0.95 (6H,

d, J=7Hz), 1.55 (3H, s), 1.6 (3H, s), 2.73 (1H, sep, J=7Hz).

Acetophenone Trimethylsilyl Enol Ether was prepared as
l

 

above: H NMR (CDC13) 0.2 (9H, s), 4.35 (1H, d, J=2Hz),

4.75 (1H, d, J=2Hz), 7.0-7.5 (5H, m).

Isobutyrgphenone Trimethylsilyl Enol Ether was prepared
1

 

as above: H NMR (CDC13) 0.2 (9H, s), 4.35 (1H, d, J=2Hz),

4.75 (1H, d, J=2Hz), 7.0-7.5 (5H, m).

General Acylation Procedure Used to Survey Various Lewis

 

Acids

A flame dried 50 mL round bottom flask equipped with a
magnetic stirrer was flushed with argon and charged with
10 mL of dry methylene chloride. The Lewis acid (10 mmol)
weighed in a glove bag was added and the flask was cooled
to 0°C. Acetyl chloride (10 mmol, 0.72 mL) was introduced
into the flask via the septum inlet. Cyclohexanone tri-
methylsilyl enol ether 1 (10 mmol, 1.92 mL) was then added
by syringe to the vessel with stirring. Finally 10 mmol

(1.06 mL) of undecane was added to serve as an internal GC

 

29

standard. After stirring two hours at 0°C, 10 mL of water
was added and the layers were separated. The aqueous layer
was washed with 10 mL of methylene chloride. The combined
organic layers were dried with sodium sulfate. Aliquots

of the dried solution were analyzed quantitatively by GC

for 2-acetylcyclohexanone and cyclohexanone enol acetate.
Since the enol acetate 4 and silyl enol ether 4 have similar
retention times on the GC column, it was often necessary

to shake the methylene chloride layer with 3 mL of 3 M

HCl. This effectively converted all remaining 4 to cyclo-

hexanone.

Optimum Procedures for the Acylation of Ketone Silyl Enol

 

Ethers

Acylation of Ketone Silyl Enol Ethers (Procedure A)

 

A 50 mL flask with a septum inlet and magnetic stir
bar was flame dried and flushed with argon. Zinc chloride
(0.79 g, 5.8 mmol) was transferred in a glove bag to the
reaction flask. The flask was removed from the glove bag
and connected to a mercury bubbler. This was followed by
the introduction of 5.8 mL of methylene chloride and 0.93
mL of diethyl ether. After cooling the flask to 0°C, the
acid chloride (5.8 mmol) was added, followed shortly by
the dropwise addition of trimethylsilyl enol ether (5.8

mmol). After stirring one hour at 0°C the reaction was

 

30

allowed to warm to room temperature and was quenched with
approximately 10 mL of water. The layers were separated

and the aqueous layer was washed with methylene chloride.

The methylene chloride layers were combined and washed with

a saturated sodium hydrogen carbonate solution. The organic
extract was dried over sodium sulfate and analyzed by GLC

for C and O-acylated ketone. Pure samples of product were

obtained for Spectral analysis by preparative GLC.

Zinc Chloride - Zinc chloride was dried by refluxing in

 

thionyl chloride until gas evolution ceased. After the
removal of the excess thionyl chloride and zinc chloride
was stored inaadesiccator over KOH for twelve hours. It
was then transferred to a dry bottle and stored in a desic-
cator over P205. All transfers of zinc chloride were done

in a glove bag under argon.

Acylation of Ketone Silyl Enol Ethers (Procedure B)

 

A 50 mL flask with septum inlet and magnetic stir bar
was flame dried and flushed with Argon. Antimony trichloride
(1.28 g, 5.6 mmol) was added to the flask in a glove bag.
The flask was removed from the glove bag and connected to a
mercury bubbler. Methylene chloride (5.6 mL) was added to
the flask and it was cooled to 0°C. The acid chloride
(5.6 mmol) was added followed by the drOpwise addition of

trimethylsilyl enol ether (5.6 mmol). After stirring one

31

hour at 0°C the flask was warmed to room temperature and the
reaction was quenched with approximately 10 mL of H20.
The method for work up was as described for Procedure A

above.

Antimony Trichloride - Antimony trichloride was sub-

 

limed (60°C at mo.1 mm Hg) to give white crystals which
were transferred to a dry bottle and stored in a desiccator
over P205. All transfers of antimony trichloride were done

in a glove bag under argon.

Synthesis and Analysis of Some 1,3-Diketones

 

The following compounds were synthesized and charac-

terized during our investigation of this acylation reaction.

2—Acetylgyclohexanone was prepared from cyclohexanone

 

trimethylsilyl enol ether and acetyl chloride by Procedure
A and B. GC analysis of the crude product after workup
showed four components with the following identities and
retention times: cyclohexanone (1.6 min), cyclohexanone
enol acetate (3.5 min), undecane (4.2 min), 2-acety1cyclo-
hexanone (6.6 min). Spectral analysis of 2-acetylcyclo-

l

hexanone gave: H NMR (CDCl3) 1.70 (4H, m), 2.05 (3H, s),

2.31 (4H, m), 15.2 (1H, 5); IR (neat) 1730, 1705 cm'1

(C=O keto form), 1610 (C=C enol form).

 

32

2-Acetylcyclopentanone was prepared by both Procedure
A and B from cyclopentanone trimethylsilyl enol ether and
acetyl chloride. GC analysis of the crude product shows
four components: component identity (retention time),
cyclopentanone (l min), cyclopentanone enol acetate (1.9
min), 2-acetylcyclopentanone (3.4 min), dodecane (6.6 min).
Spectral analysis of 2-acetylcyclopentanone gave: 1H
NMR (CDC13) 1.96 (s), 1.65-2.7 (m), 2.20 (s), 3.3 (m),
13.3 (s), spectrum shows keto and two enol forms: IR

1 l

(neat) 1710, 1745 cm- (C=O keto form), 1650 cm— (C=C

enol form).

Cyclopentanone Enol Acetate was isolated from the reac-
tion of cyclopentanone trimethylsilyl enol ether and acetyl

1

Chloride: H NMR (CDC13) 5.3 (1H, m), 2.1-2.6 (6H, m),

2.05 (3H, s).

3-Methyl-2,4-Hexanedione was prepared by method A and

 

B from diethyl ketone trimethylsilyl enol ether and acetyl
chloride. The crude product was subjected to GC analysis
which showed a four component mixture: identity (retention
time), diethyl ketone (0.6 min), diethyl ketone enol acetate
(1.5 min), 3-methyl-2,4-hexanedione (2.5 min), undecane

(5.4 min). Spectral analysis of 3-methyl-2,4-hexanedione
gave: 1H NMR (CDC13) 1.0 (3H, t), 2.15 and 2.05 (3H, both

5), 1.9 and 1.3 (3H, s + d), 2.3 (2H, q), 3.65 and 16.2

‘1' 1.

33

(1H, q + s), spectrum shows keto and two enol forms; IR

1 l

(neat) 1690, 1725 cm- (C=O keto form) 1600 cm“ br (C=C

enol form).

Diethyl Ketone Enol Acetate was prepared by methods A

l

and B: H NMR (CDC13) 5.0 (1H, q), 1.4 (3H, d), 1.0 (3H,

t), 2.0-2.3 (5H, overlapping multiplet and singlet).

3,3,S-Trimethyl-Z,4-Hexanedione was prepared by methods
A and B from diisopropyl ketone trimethylsilyl enol ether
and acetyl chloride. GC analysis of the crude product
showed a four component mixture: identity (retention time),
diisopropyl ketone (0.9 min), diisopropyl ketone enol ace-
tate (1.9 min), dodecane (2.8 min), 3,3,5-trimethyl-2,4-
hexanedione (6.1 min). Spectral analysis of 3,3,5-trimethyl~

2,4-hexanedione gave: 1

H NMR (CDCl3) 1.05 (6H, d), 1.3
(6H, s), 2.1 (3H, s), 2.85 (1H, sept); IR (neat) 1710,

1730 cm‘1 (C=O).

DiiSOpropyl Ketone Enol Acetate was prepared from di-

 

isopropyl ketone trimethylsilyl enol ether and acetyl

l

chloride using both procedure A and B: H NMR (CDC13)

2.05 (3H, s), 0.95 (6H, d), 1.45 (3H, s), 1.7 (3H, s).

l-Pheny1-1,3-Butanedione was prepared from acetophenone

 

trimethylsilyl enol ether and acetyl chloride using procedure

 

34

B. GC analysis of the crude product mixture showed five
components: identity (retention time), acetophenone (1.5
min), acetOphenone enol acetate (2.5 min), benzoylacetone
(4.9 min), hexadecane (8.8 min) and an unidentified high
boiling fraction. Spectral analysis of benzoylacetone gave:
1H NMR (coc13) 2.13 and 2.24 (3H, both 5), 4.03 (s), 6.12

(s), 7.2-7.5 (3H, m), 7.65 (2H, m); IR 1600 cm"1 br.

Acetophenone Enol Acetate was prepared by method B:
1H NMR (c0c13) 2.2 (3H, s), 4.9 (1H, d), 5.35 (1H, d),

7.1-7.5 (5H, m).

1-Pheny1-2,2-dimethyl-l,3-butanedione was prepared
from isobutyrophenone trimethylsilyl enol ether and acetyl
chloride using both procedure A and B. GC analysis of the
crude product identified four major components: identity
(retention time), isobutyrophenone (2.8 min), isobutyro-
phenone enol acetate (4.0 min), l-pheny1-2,2-dimethy1—l,3-
butanedione (5.8 min), pentadecane (7.9 min). Spectral
analysis of l-pheny1-2,2-dimethyl-l,3-butanedione gave:
1H NMR 1.45 (6H, s), 2.05 (3H, s), 7.1-7.4 (3H, m), 7.65
(2H, m); IR (neat) 1675, 1715 cm”1 (C=O), 1600, 1580 cm"1

(monosubs. benzene).

1-Pheny1-2-Methyl-1,3-Pentanedione was prepared from

diethyl ketone trimethylsilyl enol ether and benzoyl chloride

 

35

by methods A and B. GC analysis of the crude product showed
three major components: identity (retention time), diethyl
ketone enol benzoate (5.3 min), 1-phenyl-2-methy1-1,3-
pentanedione (7.6 min), hexadecane (11.3 min). Spectral
analysis of 1-pheny1-2-methy1-1,3-pentanedione gave: 1H

NMR (CDC13) 1.0 (3H, t), 1.4 (3H, d), 2.4 (2H, q), 4.4

(1H, q), 7.2-7.5 (3H, m), 7.85 (2H, m), no enol seen; IR

1

(neat) 1690, 1720 cm- (C=O).

2,4-Dimethy1-3,5-Heptanedione was prepared from iso-

 

butyryl chloride and diethyl ketone trimethylsilyl enol
ether by procedure A and B. GC analysis of the crude pro-
duct identified three components: identity (retention time),
diethyl ketone enol isobutyrate (2.2 min), 2,4-dimethy1-
3,5—heptanedione (3.7 min), dodecane (6.4 min). Spectral
analysis of 2,4-dimethyl-3,5-heptanedione gave: 1H NMR
(CDC13) 3.8 and 16.5 (1H, q + s), 2.2-3.0 (3H, overlapping
septet and quartet), 1.8 and 1.15 (3H, s + d), 0.9-1.1

(9H, overlapping triplet and doublet), spectrum shows keto

and enol forms; IR (neat) 1700, 1725 cm-1

1

(C=O keto form) I

1580 cm- (C=C enol form).

4-Methy1-3,5-Heptanedione was prepared by methods A

 

and B from diethyl ketone trimethylsilyl enol ether and
propionyl chloride. GC analysis of the crude product showed
four components: identity (retention time), diethyl ketone

(0.8 min), diethyl ketone enol prOpionate (3.9 min),

 

36

4-methy1-3,5-heptanedione (6.7 min), undecane (10 min).
Spectral analysis of 4-methy1-3,5-heptanedione gave: 1H
NMR (CDC13) 3.65 and 16.3 (1H, q + s), 2.5 (4H, g), 1.8
and 1.25 (3H, s + d), 1.05 (6H, t), spectrum shows keto

and enol forms; IR (neat) 1725, 1700 cm-1

1

(C=O k8t0 £011“):
1600 cm- (C=C enol form).

Diethyl Ketone Enol Prgpionate was prepared by both
1

 

methods A and B: H NMR (CDC13) 5.0 (1H, q), 1.9-2.6

(4H, m), 1.45 (3H, d), O.9-l.3 (6H, m).

 

2,2,4-Trimethy1-3,S-Heptanedione was prepared by both

 

procedure A and B from pivaloyl chloride and diethyl ketone
trimethylsilyl enol ether. GC analysis of the crude product
showed four major components: identity (retention time),
diethyl ketone (1.0 min), diethyl ketone enol trimethyl
acetate (3.0 min), 2,2,4-trimethyl-3,5-heptanedione (5.2
min), dodecane (6.9 min). Spectral analysis of 2,2,4-

trimethy1-3,5-heptanedione gave: 1

H NMR (CDCl3) 4.1
(1H, q), 2.5 (2H, q), 1.35 (3H, d), 0.9-1.1 (12H, over-
lapping triplet and singlet), no enol seen; IR (neat)
1700, 1725 cm-1; mass spec. m/e (intensity), 41 (13),
57 (100), 85 (15), 86 (34), 99 (6), 114 (4), 142 (.3),
170 (M+, 1). Anal. Calcd. for C10H1802‘ c, 70.55; H,
10.66. Found C, 70.64; H, 10.80.

CHAPTER III

A STUDY OF KETONE ACYLATION PROMOTED BY

MAGNESIUM CHLORIDE AND TRIETHYLAMINE

Introduction

 

The acylation of ketones with acid chlorides or an~
hydrides supplements the Claisen condensation as the

classical method of l,3-diketone synthesis under basic

 

 

conditions (Equations 24 and 25). These two procedures,
0
u 0 Li 9'
+ PthI
m, -1000 07 (24)
0 0 0
n n N +'CH gcu
a
j. 00H3 2 3
DMSO, 60°C
0 O

u u (25)

37

 

38

which were described in Chapter I, require the use of very
strong basesl'z. In addition, extremes in temperature are
often necessary to obtain synthetically useful resultsl.

A procedure which employs milder conditions and uses
the more convenient tertiary amine bases for the acylation
of ketones, would be a useful alternative to the methods
discussed above. In practice, the relatively low basicity
of tertiary amines is insufficient to promote the formation
of l,3-diketones. Preliminary experiments with cyclo-
hexanone, acetyl chloride, and triethylamine tend to support

this assertion (Equation 26). Presumably the acidity of

the ketone needs to be increased in order to effectively

0

II
+ CH 001

2 Eth, 015cm \ (26,
3 o 7
25 C, 2 hrs.

 

0%

generate the necessary quantity of enolate. Then in the
presence of an acylating agent of appropriate reactivity,
l,3—diketones might be produced.

Ligands that are coordinated to a metal ion often exhibit
enhanced acidityzo. In particular, kinetic studies of
various metal catalyzed reactions of ketoneslunmarevealed
the increased acidity of a-hydrogens bonded to an O-co-

20

ordinated acetyl group . Conceivably, this effect could

be employed in the tertiary amine promoted acylation of

 

39

ketones, if a suitable inorganic Lewis acid is used. Mag-
nesium halides are ideally fitted for this purpose since
magnesium readily coordinates to oxygen ligandSZI. Also,
magnesium has a relatively low affinity for nitrogen ligands

in solution21. This suggests that amine bases will not

compete efficiently with the ketone for the magnesium

 

:-
halide. The tendency of magnesium enolates to acylate at I
carbon rather than oxygen also argues for the potential
utility of magnesium salts as promoters of acylation re-
actions3. i

Acyl chlorides, anhydrides and N-acylimidazoles are
promising choices for acylating agents in this reaction.
Of the three possibilities, N-acylimidazoles may be the
most useful. Their reactivity with respect to nucleophilic
displacement is approximately that of acyl chlorides and
anhydrideszz. In addition, acylations using acylimidazoles
produce imidazoyl anion. The presence of this basic species

may help in avoiding the problem of proton transfer between

product and starting enolate (Equation 27).

3 <2. 3 a
R0=CH2 + RCCHRCR ——9 RCCH3 + RC-C'3=CR (27)

R

Following the suggestions of the preceding discussion,
we investigated the reaction of ketones with a variety of
acylating agents in the presence of triethylamine and mag—

nesium chloride.

40

Results and Discussion

 

Mixtures of cyclohexanone 5, triethylamine, and mag-

nesium chloride were reacted with a variety of acylating

 

 

agents (Equation 28). The results are disappointing with
u 0 3 O L.
u Et N, MgCl R i
+ BOX 3 L)
CHBCN, 25°C 5

5 1 hr' 2 R=CH3 (28)
29 R=H ‘

one notable exception (Table IV). N-acetylimidazole 9 gave

a 67% yield of 2-acety1cyclohexanone 3. In contrast, acetyl
chloride performed poorly even though it is comparable in
reactivity to 2. This is probably due to the rapid forma-
tion of ketene from the reaction of triethylamine and acetyl
chloride. The destruction of acylating agent in this manner
limits the amount of 3 produced.

It is interesting to note that cyclohexanone enol
acetate was not formed in any of these acylation reactions.
One possible explanation may be that O-acylation does occur,
but due to the reversibility of the process, only the more
stable C-acylation product is observed. It has also been
suggested that Mg+2 enolates exist as tight ion pairs or as

covalently bonded species in many cases, and this fact

41

Table IV. Reaction of 5 with Various Acylating Agents.

 

 

 

 

Acylating Agent Product % Yieldb
O
u a
l) CH3CC1 3 28
a O
2) CH3 OQCH3 3 O
O
I R
3) CH3C-N N ,3 67
\cfiJ
0
ll INT~
4) CH3C-N 3 trace
/
C) N::
u I N
5) CH3C- % 26
O
u 2
6) CH3CH %% 5

 

 

a2 molar equivalents of Et3N were used in these cases.

ineld determined by GC.

42

tends to favor C rather than O-acylation.

Our initial results (Table IV, entry 3) for the acyla-
tion of cyclohexanone with g prompted further investigation
of this reaction. Treatment of 5 with N-acetylimidazole

in refluxing acetonitrile did not increase the yield of 2-

 

acetylcyclohexanone (Equation 29). The same procedure
Cl
0 I
u 0,Mg~.
+. 2 EtBN, MgCl2 S. “
CHBCN, ref. 7
i 1 hour (29)
O O

2

performed at room temperature for 43 hours gave 63% of 3.
Also, a systematic variation in the stoichiometry of the re-
agents failed to improve on the initial yield of 67%.
Apparently, one or more reactions restrict the yields
of l,3-diketone. Self condensation of cyclohexanone was
ruled out since the amounts of unreacted cyclohexanone in
all cases corresponded to 95-100% material balance. This
fact also eliminates the occurrance of polyacylation (Equa-
tion 30). A third possibility is the reaction of N-acetyl-
imidazole in a manner similar to the Claisen ester con-
densation (Equation 31). Staab has observed such a reaction
in an attempted synthesis of t-butyl acetate from 9 and t-

butanol in the presence of potassium Efbutoxidezz.

 

43

 

 

 

 

 

0 0
u u u
Et N M C1
+ 2 J . g 2 >
(30)
Et3N .
_/\ J
0 “Mg ”\N 0 0 |
u /::N \cgJ \ I. u -N (31)
2 CHBC QV/J /' CH OCH 0"
” 3 2

Self condensation, however, should be minimized by slow
addition of 9 to the acylation reaction mixture. Repetition
of the cyclohexanone acylation under these conditions sup-
plied 57% of 3. Use of less reactive heterocyclic amides
(Table IV, entry 4 and 5), which are probably less acidic
than 9, also failed to improve the yield of acylated product.
At this time it occurred to us that a second side reaction,

shown in Equation 32, could be functioning simultaneously.

H §::JJ
“ + t N ————+>
MgC12 + \g‘J E 3

(32)

RN
Et N HCl + ClMg N\/|
3 /

44

Slow addition of N-acetylimidazole would maximize this re-
action. Both difficulties (Equations 31 and 32) should be
circumvented by dropwise addition of 9 to a reaction mix-
ture containing a 100% excess of both triethylamine and
magnesium chloride (Equation 33). Unfortunately, this pro-

cedure resulted in only 62% yield of 2-acety1cyclohexanone

3.
"\3
0
II
+ 2 M501 + 2 Et N slow
2 3 add 'na'
01
‘ (33)
O/Mg'-.
ll 4601

Following these abortive attempts to increase the amount
of 3 produced in the acylation of cyclohexanone we reacted
several ketones with N-acetylimidazole in order to examine
the generality of this method (Table V). The conditions
employed were identical to those of Equation 28. The values
in Table V indicated that cyclic ketones are the only sub-
strates which give reasonable yields of acylated product
under our conditions. Even in these cases yields are often
poor. Acyclic ketones fail to react even when extended re-
action intervals were employed.

In conclusion, this method of acylation has the potential

of being a mild and convenient preparation of l,3-diketones.

 

 

 

 

 

Table V. Reaction of a Variety of Ketones with N-Acyl-
imidazoles.
Ketone Acylating Agent Product % Yield
8 o O 0
~\
CH 3F 6
3 / 7
5 9.
Pr 9 O
l \‘
i ((313)20ch 58
O O
l g 8
C7 2 GA .0
O O 0
II II I
8 O
R R
5 t-C H C-N N trace
1+ 9 /

 

 

 

46

 

 

 

Table V. Continued.
Ketone Acylating Agent Product % Yield
0
\\v/n\v/’ 2 \\v/E\T,Jl\ 0
O O O
‘\T/n\T/’ 2 \\r/g>x<¥\\ O
P: 8 9
Ph 2 HIM 0
O
1 8 8
P 2 PM 0
0 9 0
z/n\\ 2 I/L\v/flk\ 0

47

Additional study is required, however, to increase the
generality and productivity of the procedure. Perhaps the
use of an acylating agent less susceptible to self condensa-

tion will solve these problems.

Experimental

 

Materials

 

Dry tetrahydrofuran was obtained by distillation from
sodium and benzophenone. Reagent grade methylene chloride
was dried over 4 A molecular sieves. Acetonitrile, tri-
ethylamine, and all the ketones used for this work were
distilled from calcium hydride before use. All acid
chlorides were purchased from Aldrich Chemical Co. and
purified by simple distillation. Anhydrous magnesium
chloride, Aldrich Chemical Co., was stored and handled in
a glove bag under argon atmosphere. Methyl formate, imi-
dazole, pyrazole, and 1,2,3-benzotriazole were also ac-
quired from Aldrich Chemical Co. and were used without

further treatment.

Methods of Analysis

 

All products and intermediates were analyzed by 1H

NMR and mass spectroscopy. Infrared spectra were obtained

when necessary. GC analysis was performed where appropriate.

 

48

Descriptions of the techniques and instrumentation used
are provided in Chapter II. Melting points were determined
on a Thomas-Hoover capillary melting point apparatus and

are uncorrected.

Preparation of N-Acylimidazoles

 

N-Acylimidazole was prepared by the following procedure

adopted from the method described by Staab22. It is repre-

 

sentative of the preparation of all N-acylimidazoles used
in this work.

Imidazole (500 mmol, 34 g) was dissolved in approxi-
mately 300 mL of dry methylene chloride under an argon
atmosphere. Acetyl chloride (250 mmol, 17.8 mL) was added
slowly with stirring. The reaction was exothermic and
care must be taken to avoid rapid refluxing of solvent on
this scale. Precipitation of imidazole hydrochloride oc-
curred before the entire portion of acetyl chloride was
introduced. The reaction was allowed to stir for 3-4 hours.
The insoluble salt was removed by suction filtration and
washed once with dry methylene chloride. Concentration of
the filtrate provided a pure crop of white crystals which
were isolated by filtration. Evaporation of the remaining
solvent supplied a second crop of crystals which were re-
crystallized once from methylene chloride. Combination of
the two portions of crystals gave 26.1 g (95%) of a white

solid identified as N-acetylimidazole: mp loo-101°C

49

22 1

(lit. mp. 103°C );

br. s), 7.4 (1H, m), 8.1 (1H, br. s).

H NMR (CDC13) 2.55 (3H, s), 7.05 (1H,

N-Isobutyrylimidazole was prepared from isobutyryl

 

chloride and imidazole in a manner similar to that des-

cribed above: 83% yield; bp 85°/ 100 mm Hg (lit bp 92/18

mm ngz); 1H NMR (c0c13) 1.4 (6H, d), 3.2 (1H, septet),

7.05 (1H, br. s), 7.4 (1H, m), 8.1 (1H, br. s).

N-pivaloylimidazole was prepared from pivaloyl chloride

 

and imidazole by a procedure similar to that described

22 1

above: 75% yield; mp 55-57°C (lit mp 560 ); H NMR (CDC13)

1.4 (9H, s), 7.0 (1H, m), 7.5 (1H, m), 8.2 (1H, m).

Preparation of N-Acetylpyrazole23

 

Acetyl chloride (150 mmol, 10.7 mL) was added dropwise
to a mixture of pyrazole (150 mmol 10.2 g) and dry pyridine
(200 mmol, 12.1 mL) in approximately 200 mL of dry methylene
chloride at 0°C (145 mmol, 10.35 mL). The resulting solution
was stirred for one hour at 0°C and then 3 hours at room
temperature. The reaction mixture was washed with water,
aqueous Na2C03, and saturated sodium chloride solutions.
The organic layer was then dried over magnesium sulfate.
Evaporation of the solvent in vacuo gave an oily residue
which was distilled under reduced pressure. N-Acetylpyrazole

(12.5 g, 76%) was isolated as a clear colorless liquid:

Y'"

a, no

50

1H NMR (c0c13) 2.6 (3H, s), 6.35 (1H, m), 7.6 (1H, m),

8.15 (1H, m).

Preparation of N-Acetylbenzotriazole

 

This procedure was adapted from that described by
Staab for the preparation of N-acetylbenzotriazole24.
Acetyl chloride was added to a solution of benzotriazole
(11.91 g, 100 mmol) in 100 mL of dry methylene chloride.
The resulting mixture was stirred for approximately 4 hours.
The insoluble benzotriazole hydrochloride was removed by
suction filtration and the solvent was evaporated at reduced
pressure. The resulting white crystals were identified as
N-acetylbenzotriazole (7.7 g, 95%); mp 49-51°C (lit. mp

l

51°C); H NMR (CDC13), 2.95 (3H, s), 7.3—7.6 (2H, m),

7.9-8.3 (2H, m).

Acylation of a Variety of Ketones in the Presence of

 

Magnesium Chloride and Triethylamine

 

2-Acetylcyclohexanone was prepared from cyclohexanone

 

and a variety of acylating agents. The following procedure
employing N—acetylimidazole is representative of these.

A 50 mL round bottom flask equipped with a septum inlet
and magnetic stirrer was flame dried under vacuum, flushed
with argon and charged with 10 mmol (0.95 g) of anhydrous

magnesium chloride.

 

51

Dry acetonitrile (20 mL) was added to the flask. To
the resulting heterogeneous mixture were added 10 mmol
(1.4 mL) of triethylamine, 10 mmol (1.04 mL) of cyclo-
hexanone and 10 mmol (1.1 g) of N-acetylimidazole. The
reaction mixture was stirred for one hour at room tempera-
ture. After cooling to 0°C the reaction was quenched with
10 mL of 3 M hydrochloric acid. The resulting solution was
washed twice with anhydrous diethyl ether and dried over
magnesium sulfate. GC analysis (undecane as internal stan-
dard) of the ether solution indicated a 67% yield of 2-

acetylcyclohexanone: bp 111-1120/18 mm Hg; 1

H NMR (CDC13) ,
1.70 (4H, m), 2.05 (3H, s), 2.31 (4H, m), 15.2 (1H, s).
The synthesis of several 1,3-diketones were attempted

during this investigation. The general procedure used is

similar to that described above.

2-Acety1gyclopentanone was prepared from cyclopentanone
l

 

and N-acetylimidazole: 60% yield H NMR (CDC13), 1.96
(s), 1.65-2.70 (6H, m), 2.20 (s), 3.3 (m), 13.3 (1H, s),

spectrum shows keto and two enol forms.

2-Acety1cyclooctanone was prepared from cyclooctane and
1

 

N-acetylimidazole: 48% yield H NMR (CDC13), 1.4-1.8 (8H,
m), 2.1 (3H, s), 2.2-2.6 (4H, m), 16.3 (1H, 5), Spectrum is
predominately enol; mass spec (m/e) 168 (M+), 153 (M+-CH3),

125 (M+-CH3CO) 97, 84, 71, 55, 43.

52

2-Isobutyrylcyclohexanone was prepared from cyclo-
l

 

hexanone and N-isobutyrylimidazole: 58% yield; H NMR
(CDC13), 1.1 (6H, d), 1.5-1.9 (4H, m), 2.0-2.5 (4H, m),
2.85 (1H, septet), 3.6 and 15.9 (1H, m + s), spectrum shows
keto and enol forms; mass spec (m/e), 168 (M+), 125 (M+-

+
(CH3)2CH), 97 (M -(CH3)2CHCO, 71, 55, 43.

2—Pivaloy1cyclohexanone was prepared from cyclohexanone

 

and N-pivaloylimidazole. The very low yield was un-
determined and the product was identified by its GC/MS:

(m/e), 182 (M+), 125 (M+-C4H 98, 83, 70, 57, 41.

9)'
The acylations of the following ketones with N-acetyl-
imidazole were unsuccessful: diethyl ketone, diisopropyl

ketone, isobutyrophenone, benzophenone, acetone, 2,6-di-

methyl-cyclohexanone.

CHAPTER IV

THE CARBOXYLATION OF KETONES WITH CARBON DIOXIDE

IN THE PRESENCE OF MAGNESIUM HALIDES

Introduction

 

Biochemical investigationscxfthe mechanism of enzyme

induced decarboxylation frequently require synthetically

produced B-keto acidszs. These acids are also useful for

the preparation of ketonesZ6. Furthermore, B-keto acids
can serve as valuable intermediates in the synthesis of
natural product527_3l.

A large proportion of the literature pertaining to B-
keto acids deals with their decarboxylation. There are,
however, several reports of effective methods for the syn-
thesis of these compounds.

Alkaline hydrolysis of the corresponding esters has

been used successfully to prepare several aliphatic32'33

and aromatic B-keto acids (Equation 34)34. This method,

however, has proven to be unreliable in terms of its gen-

erality35'36.
Metz, Axelrod, and Hoffman utilized acid hydrolysis

in their preparation of some long chain B—keto acids

53

54

0 0 0
u u . -OH u u _ H* u u
RCCHZCOR ———-—) RCCHZCO ——9 RCCHZCOH (34)

(Equation 35)35. More recently, Logue treated tert-butyl-

B-benzoylisobutyrates with trifluoracetic acid36. The

O

RgCH 80R. -—l§a;_€). RSCH 8 (35)
2 HOAC 2 OH

 

result is a quantitative yield of the corresponding acids

(Equation 36).

0’1‘ CFECOZH (:> .H

15 min. 25°C

Ainsworth and Kuo developed a different type of ap-
proach to B-keto acid526. The dianion of a carboxylic
acid, generated with lithium diiSOpropyl amide, was acylated
with an ester and the intermediate was trapped with chloro-
trimethylsilane (TMCS). Solvolysis of the trimethylsilyl

ester under neutral conditions in methanol led to high

yields of the desired product (Equation 37).

55

 

 

20
1 -“ _
R RZCCO + R3C00H 0'5 hrs'%> R3CCR1 R 2Co
3 0°C
TMCS \_ 38 2"
R I
0.5 hrs. 2500fi7 CCR1R2C051M63 (37)
0 0
CH OH .3 [1
3 3; RBCCRlRZCOH

 

0.5 hrs. 2500

Van der Baan acylated bistrimethylsilyl malonate with

acid chlorides37. Hydrolysis under mild conditions followed

by decarboxylation gave B-keto acids (Equation 38). This

 

Et 0
. -.+
(Me381020)ZCH L1 + RdCl *5 2 I}, (MeBSiOZC)2CHJR
(38)
H20 g -Co2 g g
-_I;;;€> (HOZC)ZCH R -—————-€> R CH2 0H

mild procedure also gives excellent yields.

Direct reaction of carbon dioxide with enolate anions
also can be used to generate B-keto acids. Deprotonation
of appropriate ketones with sodium, potassium, or lithium
phenoxides followed by treatment with CO2 has led to a
variety of aliphatic and aromatic 3-oxoacids (Equations

38-40

39 and 40) Similarly, the use of strong bases

56

 

 

 

 

0'M+
0 0’
ll 0 '
PhCCH3 + O 25 C 5 PhCZCH2
(39)
0 O
002 \ u u
0 .7, PhCCHZCOH
Et20, 25 C
a 0'Li+ 0‘Li+
+. (:> EtZO \
:7
25°C
(40)
0 0
II II
C°2A OH
Et20I7

such as sodium amide or sodium triphenylmethide to generate
enolates has given B-keto acids after subsequent treatment
with carbon dioxide (Equation 4l)4l’42. Reaction of 1,8-

diazabicyclo (5,4,0)-7-undecane (DBU) with ketones in the

presence of CO2 results in the corresponding acids in

57

 

O
u NH 00
RCCH.a + NaNH ——}-> RE=CH ——2—->
9 2 2 EtZO
., (41) F
H II II
——-9 RCCHZC OH
43 i
moderate yields (Equation 42) .
. 0 0 0
u N l
002’ DMSO OH

 

>
N ‘\\ 3 hrs, 2500
61%

A useful method developed by Stiles and Finkbeiner
utilizes the reagent methyl magnesium carbonate (MMC)44.

MMC in dimethyl formamide solution carboxylates ketones that

have a hydrogens (Equation 43).

 

0 ,Mg\
II
u . CH OMgOCOCH
RCCH R 3 3 5 \ §
2 o R 0
0MP, 120 C
(43)
0 0

u m
——fi%—> RCCHRCOH
2

58

The magnesium chelate 19_can be alkylated or acylated
in situ. It is believed that the formation of the mag-
nesium chelate is a major driving force in this reaction.
This method provides good yields of B-keto acids but the
reaction is inhibited in some cases by steric factors. For
example, the reaction fails when the ketone used has only
one a hydrogen.

Sakurai recently reported a method for carboxylating ke-
tones using a bromomagnesium ureidecarbon dioxide adduct45.
This procedure gives fair yields of B-keto acids. The
reaction appears to be sensitive to steric factors and

carboxylation of the less hindered side of unsymmetrical

ketones occurs preferentially (Equation 44).

 

0
ll 8 o
1
R CHZCRZ + BngOZCN/ \NCOZMgBr QDMF; 110 C)
2)H 0
(CH2 n 3
n=293
(44)
P. 8
2 1
R CCHR COH 81:11, R2=Ph 51%

cyclohexanone 74%

In 1977, Matsumura reported a carboxylation reaction of
ketones and others active methylene compounds with CO2
which was promoted by a mixture of magnesium chloride and

46

amine base . This procedure requires long reaction times

and high carbon dioxide gas pressures to obtain good yields

 

59

of B-keto acids (Equation 45).

0

PthH MgClz, Eth, 17 hrs.

0
H01 l l
3 002(5 atm{) DMF ‘3; ‘————‘€> PhéCHZCOH (45)
76%

Matsumura's results were interesting to us, since our
initial investigation in the area of ketone carboxylations
indicated that this transformation could be accomplished
under much milder conditions and in much shorter times in
THF solution rather than in DMF. Armed with this initial
result and the improvement it seemed to offer over the pro-
cedure of Matsumura we decided to investigate the car-
boxylation of ketones with carbon dioxide in the presence

of triethylamine and magnesium halides.
Results and Discussion

Cyclohexanone was reacted with carbon dioxide at at-
mospheric pressure in tetrahydrofuran (THF) solution. Two
molar equivalents of triethylamine along with one molar
equivalent of magnesium chloride were used to promote the
carboxylation (Equation 46). This mixture remained hetero-
geneous at all times and as the reaction progressed became
extremely turbid. A rapid uptake of gas was observed and
after one hour 80% of the theoretical amount of carbon

dioxide was absorbed. At the end of a two hour period, a

60

full molar equivalent of CO2 had been absorbed. Due to the
susceptibility of B-keto acids to decarboxylation it was
decided not to attempt isolation of the 2-oxocyclohexane
carboxylic acid at this point. Instead we continued our
investigation by monitoring the rate of carbon dioxide ab-

sorption under various reaction conditions.

Mg
./
0 0 \0
u
MgCl2 Q:0
+ (302 + 2 EtBN W (46)

A simplified description of how this carboxylation
proceeds is offered in Figure 5. Initial deprotonation of
the ketones is accomplished with triethylamine. The slow
formation of enolate is assisted by coordination of mag-
nesium chloride with the ketone carbonyl oxygen. In this
manner the acidity of the alpha hydrogen is enhanced and
the proton abstraction with triethylamine is facilitated
(Equation 47). Subsequently, carbon dioxide, possibly in
the form of a carbamate salt, reacts with the enolate to
produce a magnesium keto carboxylate (Equation 48)46.

This species will quickly be converted to a magnesium
chelated dianion 10 by excess triethylamine. Stabilization
of the dianion in this way is reminiscent of what is believed
to occur in the reaction of Stiles' reagent (MMC) with

ketones44. Careful acid hydrolysis of 19 yields the B-keto

 

61

 

 

 

ngCl2 _+
a MgCl2 8' Et3N . MgCl
R CHZR > R CHZR 7‘ Rcz:CHR (47)
+ - 0
ul \ - (48)
RGZCHR + ,c=0 -———> RCCHR 0 MgCl
EtQN+ 11
5 __
M8
/ \
1— ’ \ Q —3——) RCCHRCOH
0

(49)

Figure 5. Mechanism of Carboxylation.

acid (Equation 49). Clearly, the magnesium halide plays
an important role since in its absence the carboxylation
reaction does not proceed at a measurable rate.

In contrast to our results with cyclohexanone, Matsumuro
reported that seventeen hours were required to obtain the

carboxylation of ketones in dimethylformamide (DMF) at a

62
carbon dioxide pressure of five atmospheres46. By repeating
the carboxylation of cyclohexanone (Equation 46) in DMF
we obtained a similar result. The reaction mixture re-

quired more than 8 hours to absorb one molar equivalent of

 

carbon dioxide (Figure 6). A possible explanation of the

slow rate of reaction in DMF may be that the magnesium

5]

chloride is extensively coordinated to solvent molecules.
Therefore, the magnesium chloride cannot efficiently aid

in enolate formation. THF, which is a weaker Lewis base

 

relative to DMF, enables the magnesium chloride to per- '
form more effectively in the deprotonation of the ketone.
Consequently, the rate of carboxylation is increased.
Following this reasoning, employment of a polar aprotic
solvent with negligible affinity for magnesium, could result
in an increase in the rate of reaction. A likely candidate
would be acetonitrile, since magnesium complexes with nitro-
gen are weak and tend to disassociate in solution46.
To test this idea, cyclohexanone was reacted with carbon

dioxide at atmospheric pressure in the presence of magnesium

chloride, triethylamine, and acetonitrile (Equation 50).

63

N

mpCo>Hom mo mofipom m CH coapmnomnm 00 do opmm .0 opsmflm

1.:fleo mane

om ma on mo om mm om m: 0: mm on mm om ms OH m 1

1

.OH
0N

.om
man

3. 0‘:

.om

om
.om

.om
mze

.00
20 mo

 

.OOH

moo a

amonompa

64

Thirty minutes were required for the mixture to absorb

80% of the theoretical amount of C02. This was a substan-
tial increase in the rate of absorption relative to reac-
tion in THF (Figure 6). Methylene chloride and dimethoxy-
ethane (DME) were also investigated for use as solvents.
Under conditions similar to those used for THF they provided
slower rates of reaction.

A logical extension of our investigation involved a
study of the effect of magnesium halides, other than the
chloride, on the rate of carbonation of ketones. The reac-
tion of cyclohexanone with carbon dioxide was repeated using
magnesium bromide. In a similar carboxylation experiment,

a two to one mixture of sodium iodide and magnesium chloride
was employed. It is assumed that this mixture reacts to
generate magnesium iodide in situ. Pronounced increases in
the rate of carbon dioxide absorption, relative to that ob-
served for magnesium chloride, were noted (Figure 7).

These results are not surprising when the Lewis acidity

of magnesium bromide and magnesium iodide are considered.
Both of these reagents are more acidic than magnesium
chloride and consequently would be more effective at
promoting the deprotonation of cyclohexanone (Equation 47).
Since enolate formation is probably the rate determining
step, the overall result is an increase in the rate of CO2
incorporation.

It was found that mixtures of magnesium iodide and

triethylamine absorbed carbon dioxide in the absence of

65

N00 mo mpmm .m opswflm

mooflam: Esflmocmme monzp how Coflfimhompw
A.cnev mafia

me as as m” .a mm mm mm 4m IN ms ms NH 6 m

I

*m

OH
ON

1on

 

.o:
.om
,ow
on
.om

Naomz .om

Hm:

 

M m
m E. _ ooH

uwppowpa moo a

66

ketone. The uptake of gas, however, was negligible and

did not complicate the measurement of reaction rates. A
possible explanation of this observation may be a slow
carboxylation of acetonitrile. Attempts to isolate the
expected product from this reaction were unsuccessful (Equa—

tion 51).

1)2 Mgl2
9 HO CCH

CHCN + CO + 4EtN 2 2

3 2 3 2)H30+ CN

(51)

The stoichiometry of the carboxylation requires a unit
ratio of ketone and magnesium salt. It was conceivable,
however, that use of an excess of magnesium iodide could
have a beneficial effect on the reaction. Results were
obtained, by the treatment of isobutyrophenone with CO2
in the presence of excess MgIz, which confirm this idea
(Figure 8). Use of exorbitant amounts of M912 (3 equiva-
lents or more) furnished a rate acceleration but provided
inordinate quantity of solids suspended in solution. For
this reason, two molar equivalents of catalyst were con-
sidered optimum. It was determined, in a similar fashion,
that the use of four molar equivalents of triethylamine
provided the highest yields of benzoyl acetic acid, through
carboxylation of acetophenone (Figure 9). By combining these
two observations we defined our standard carboxylation condi-

tions (Equation 52).

67

 

 

O O O
u n M812 1; u
PhCCH(CH3)2 + CD2 + 3 Et3N -———-—-)' PhCC(CHq)ZC-OMgI
J
M019 % C02 1 Equivalent 2 Equivalent 3 Equivalent
Absorbed MgI MgI2 MgI2
25% 46 min. 13 min. 11 min.
50% 99 min. 28 min. 22 min.
75% 185 min. 55 min. 39 min.
100% 313 min. 96 min. 79 min.

 

 

Figure 8. M912 Stoichiometry Study.

 

O
2 Mgl2 O

H
PhCCH +
3 C02 + n Et3N CHBCNZ fl-;L-€> PhCCH 2COH

 

Molar Equivalents of % of Yield of
Triethylamine Benzoylacetic Acid
2 52%
4 75%
6 67%

 

Figure 9. Et3N Stoichiometry Study.

68

 

 

.fi ' 2 MgI2
R CHZR + xs 002 + u EtBN i7
CHBCN, 25°C
H30+ ,
\
EtZO 7. RCCHRCOH <52)

The generality of the method was demonstrated by react-
ing a variety of symmetrical ketones under the standard
conditions. In all cases, incorporation of carbon dioxide
was fast. Typically, an equimolar amount of CO2 was ab-
sorbed in less than one hour. Stirring for an additional
1.5 hours while slow gas absorption continued, provided the
best yields of products (Table VI). The B-keto acids are
labile with respect to decarboxylation and therefore must
be carefully handled during isolation. Slow quenching,
of the aqueous solution containing the chelated dianion
19, with cold dilute hydrochloric acid furnished the most
satisfactory results.

An attempt was made to determine if this technique ex-
hibited any selectivity for either alpha position of an
unsymmetrical ketone. Treatment of 2—butanone with C02,
triethylamine and magnesium iodide gave a mixture of products.

1H NMR indicated that both pos-

Analysis of the residue by
sible B-keto acids were present (Equation 53). Further
treatment of the resulting mixture with bistrimethylsilyl

acetamide (BSA) yielded the silylated products which could

 

.J-fl

69

 

   
 

 

access . .
on ooom-aa mo _
_
o .
mm SomoTSi x. Q Q
oomoHIHoH
o 0
A0 mmnmmv
. 0 © Q
m OOQNIOB IO/\F\. \/\.
a a a
om oommlmm 30% g
. m o
mpamww w A.m.z “flay .m.z pontoum mcoumm mcfluumum

 

 

.mmcoumx mo >vmaum> a mo coflumawxonumo .H> manme

70

 

 

mat/“\hflw\)ll\\ Il&\)lm\)/(\\
O

 

mm o.
lav/f)
mm D = = %
O 0
m0 A003NV SDI/WVXA(\FII \\FIH\FII
comm v m 0
mm ll. mOI/\\AHHV AHHV
m a a
3 008.? :o . \‘l
= _
a o o
mpamww w A.m.2 uHHV .m.z uosooum wcouwx mafiuumum

 

 

.Umdcflucoo .H> manna

71

 

 

coouom

oms-mm

 

mcamflw w

A.m.z

pflav

.m.2

wcoumm mcflunmum

 

 

.6mscfiucoo .H> magma

72

 

 

 

 

 

.mumEOmfi mo onsuxfle ov\om m mm pmumHOme
.mumEOmH mo mmsuxflfi om\om m mm UmumHOmHo
.wuflamummuo ou omaflmm pompoumn
.mpawflw UmumHomHm
mmo
o l 38 gm pmoigm
mmo pmo
o I pmo /._\/_m\
a o
o I mo<Q «r. v
= =
o o a
mpawflw w A.m.z uHHV .m.z pospoum mcoumx mcfluumum

 

 

.Uwscwucou .H> magma

73

be examined by GC. Conversion in the silylation step was
low but the results indicated a nearly equal ratio of the

two possible products (Equation 53). In a similar procedure

 

O O
\/"\/u\
\
7
N 2>H30". 0°C +
0 O
u u
’/’\\T//~\\DH
M S'
::1;:E\\r/fi\\‘
\\ OSiMe3
(53)
BSA 3 +_
Me 810 0
3 ll
’/l\T/\‘031Me3
3-methyl—2—butanone was carboxylated. 1H NMR analysis of

the oily residue showed a complex mixture containing both

B-keto acids (Equation 54). It appeared that reaction at

 

0 0
II a 12
0
" ‘\r’/\\~/’A\\OH
1)co ,Et N,MgI
+ o
_ 2)H30 , O C O 14
13 u 8 ‘—
Me SiO 0 1 ’//::><:f\\‘on
3 u "‘5

\
+
o lé
u _
//f:><:¥1\‘OSiMe3 (54)

74

the methyl group of 12_predominated, providing a 60/40
ratio of 13/14. Silylation of I; and 14 followed by GC
analysis indicated the presence of 15. The silylation of
l§_was not detected since it is not GC stable. These re-
sults seem to denote a slight preference for carboxylation
at the less sterically hindered alpha position of an un-
symmetrical ketone. More extensive examination of this
problem is needed before a definite trend can be established.
However, it seems clear that the reaction is not highly
regioselective.

In addition to various ketones, y butyrolacetone l,3-
cyclohexanedione, ethyl acetoacetate, and diethyl malonate
were subjected to our standard carboxylation conditions
(Table VI). Reaction of y-butyolactone provided a complex
mixture in a very low yield even though some of its deriva-
tives have been successfully carboxylated using MMC47.
Ethyl acetoacetate and l,3-cyclohexanedione reacted very
sluggishly requiring 24 hours to absorb one equivalent of
carbon dioxide. Hydrolysis of the residues in both cases
yielded recovered starting material. Reaction of di-
ethyl malonate rapidly absorbed carbon dioxide. Hydrolysis,
however, failed to provide product. The absence of
carboxylation in the last three cases may be due to Very
facile decarboxylation of the desired species upon hy-
drolysis. Another possibility could be that the products

possess a high solubility in water and are lost during work-

up.

 

75

In conclusion, this procedure offers a useful route to
B-keto acids by direct carboxylation of symmetrical ketones.
The conditions are mildly basic, require inexpensive re-
agents, relatively short reaction times, and ambient tem-
peratures. Further study is needed to determine if this
method is capable of any positional selectivity when em-

ployed for the carbonation of unsymmetrical ketones.

Experimental

 

Materials

 

Tetrahydrofuran was distilled from sodium and benzo-
phenone. Acetonitrile, dimethylformamide, triethylamine
and pyridine were dried by distillation from calcium
hydride. All ketones used in this investigation were com-
mercially available and purified by fractionation over
calcium hydride. All the above mentioned reagents were
stored under an atmosphere of argon. Anhydrous diethyl
ether was obtained from Mallinckrodt and used without
further treatment. Sodium iodide, purchased from J. T.
Baker Chemical Co., was dried by heating under vacuum
prior to its use. Magnesium chloride, acquired
as the anhydrous reagent from Aldrich Chemical Co., was
stored in a dry bottle in a glove bag under argon. Bis-
trimethylsilyl acetamide was prepared from acetamide, tri-

ethylamine and chlorostrimethylsilane by the method of

 

 

76

Klebe, Finkbeiner, and White48. Solid carbon dioxide
was used as a source of CO2 gas which was dried by passage

through a tube containing anhydrous calcium sulfate.

Methods of Analysis

 

All products were analyzed by 1H NMR, IR, and mass

spectrometry. Where appropriate, GC analysis was also per—
formed. The descriptions of the instruments used were

provided in Chapter I.

Reaction of Cyclohexanone with CO2 and MgCl2 in THF

 

A flame dried 50 mL round bottom flask equipped with a
septum, gas inlet and magnetic stirrer was charged with 20
mL of dry tetrahydrofuran. Magnesium chloride (10 mmol,
0.95 g) and triethylamine (20 mmol, 28 mL) were added to
the flask and stirring was initiated. The flask was
flushed with dry carbon dioxide and connected to a gas

burette filled with dry CO After the fluid level in the

2.
burette had stabilized the gas volume was adjusted to 250
mL. Cyclohexanone (10 mmol, 1.04 mL) was added via syringe.

Absorption of CO was then recorded by observing the change

2
of the fluid level in the burette. 224 mL was considered
to be 10 mmol of carbon dioxide gas.

Carboxylation of cyclohexanone with CO2 in the presence

of magnesium chloride were also performed in DMF, aceto-

nitrile, DME, and methylene Chloride. The procedure is

 

77

identical to that described above with the substitution

of the appropriate solvent.

 

Reaction of Cyclohexanone with C92 and MgBr2

 

Magnesium bromide was prepared from magnesium and
ethylene dibromide in the following mannersz. Into a flame
dried flask equipped with a septum inlet and a reflux con-
densor were added 12 mmol (0.29 g) of magnesium turnings.
The flask was flushed with argon gas and charged with 20
mL of dry THF. Ethylene dibromide (10 mmol; 0.86 mL) was
added slowly with stirring. Reaction was exothermic and
refluxed spontaneously with concomitant evolution of
ethylene gas. When gas evolution slowed the flask was
gently warmed until ethylene gas formation is no longer
observed. After the flask cooled the condensor was re-
moved and replaced with a gas inlet. This was followed
by the addition of 20 mmol of triethylamine (2.8 mL). At
this time the flask was flushed with dry carbon dioxide
gas during which rapid stirring was maintained. A gas
burette filled with dry CO2 was connected to the reaction
flask and the gas volume allowed to stabilize. Injection
of cyclohexanone (10 mmol, 1.04 mL) initiated rapid CO2

absorption. The rate of gas uptake was recorded.

 

Reaction of Cyclohexanone with CO2 and MgI2

Sodium iodide (20 mmol, 3 g) was heated by flame in a

 

78

flask under vacuum (0.1 mm Hg) to drive off any moisture
present. The flask was flushed with argon and allowed to
cool. Introduction of 20 mL of acetonitrile was followed
by the addition of 10 mmol (0.95 g) of magnesium chloride.
The resulting mixture, which was assumed to generate mag-
nesium iodide, was stirred for 30 minutes at room tempera-
ture. Vigorous stirring was maintained while 20 mmol of
triethylamine (2.8 mL) were added and the flask was flushed
with dry C02. The flask was then connected to a gas bur-
ette filled with C02. After the gas volume had stabilized,
cyclohexanone was introduced into the heterogeneous solu-
tion via syringe. Carbon dioxide absorption began immedi—
ately and was monitored by observing the change in fluid
level in the gas burette.

Several reactions were done in which the number of
equivalents of magnesium iodide and triethylamine were
systematically varied in order to determine the optimum
conditions for carboxylation. These experiments were con-
ducted in a manner identical to the procedure described
above with the exception of the absolute amount of reagents

used.

Preparation and Analysis of Some B-Keto Acids

 

2,2-Dimethyl-3-Phenyl-3-Oxopropanoic Acid was prepared
from isobutyrophenone and CO2 by the following procedure

which is representative of the synthesis of all the B-keto

WI

 

79

acids in this study.
Anhydrous magnesium chloride (20 mmol, 1.9 g) and dry
sodium iodide (40 mmol, 6.0 g) were reacted for one half

hour in 30 mL of acetonitrile at room temperature under

 

argon. At that time triethylamine (40 mmol, 5.6 mL) was

added to the reaction. With vigorous stirring the flask

 

was flushed with carbon dioxide and connected to a gas bur- k
ette. After the gas volume in the burette had stabilized,
isobutyrophenone (10 mmol, 1.5 mL) was introduced into the
mixture. Absorption of CO2 ensued immediately. The reac- I

tion mixture became excessively turbid making stirring dif-
ficult. In some cases during this investigation small
amounts of solvent in addition to the original 30 mL were
added to maintain effective stirring. Allowances must then
be made for absorbtion of CO2 by the extra solvent. The
resulting mixture was stirred for 2.5 hours at which time

approximately 300 mL of CO had been absorbed. The mixture

2
was diluted with 50 mL of ice water and extracted with
ether. The aqueous layer was cooled to 0°C in an ice bath
and acidified to pH 3-4 with 0.7 M aqueous HCl at 0°C.
Vigorous stirring was maintained throughout this process.
The aqueous solution was extracted with two portions of
ether (30 mL each). The resulting organic layer was dried
with magnesium sulfate. Removal of the solvent in vacuo

provided 1.73 g (90%) of a white solid identified as 2,2-

dimethyl-3-phenyl-3-oxopropanic acid: melting point 93-950;

80

1H NMR (cac13) 1.58 (1H, s), 7.3—7.5 (3H, m), 7.65-7.9

(2H, m); IR (CHCl solution) 3400-2500 (br), 1710 (s),

3
1635 (s) cm-1; mass spec (m/e) 192 (M+), 148 (M+-C02),

105, 77.

2-Methy1-3-Pheny1-3-Oxopropanoic Acid was prepared from

propiophenone and CO2 by the procedure described above:
27). l

1.5 (3H3 d), 4.2 (1H, q), 7.3-7.55 (3H, m), 7.8-8.05 (2H,

85% yield; mp 76-78°C (lit mp 77-78°C H NMR (CDC13)

m); 10.1 (1H, 5); IR (CDCl3 solution) 3500-2400 (br), 1720
(s), 1680 (s) cm-l; mass spec (m/e) 178 (M+), 134 (M+-

C02), 105, 77.

Benzoylacetic Acid was prepared from acetophenone and

 

CO2 by the procedure described above: 73% yield; mp 101-
27). 1

5.7 (vinyl H from end, s), 7.25-7.6 (3H, m), 7.7-8.05 (2H,

1020c (lit mp 101-102°C H NMR (c3060) 4.05 (2H, s),

m); mass spec (El) 164 (M+), 120 (M+-C02), 105, 77.

 

2-Oxocyclohexane-l-Carboxylic Acid was prepared from

cyclohexanone and CO by the procedure described above:

C28); 1

2
70% yield; mp 79-80°C, (lit mp 78-800 H NMR (CDCl3)
1.4-2.0 (4H, m), 2.0-2.6 (4H, m), 3.2 (1H, m) 10.3 (1H,
br. 3), 11.65 (1H br. s) from enol; mass spec (m/e) 142

(M+), 124 (M+-H20), 98 (M+-C02), 68, 55.

 

 

81

1,5-Dimethy1-2-Oxocyclohexane-l-Carboxylic Acid was pre-

 

pared from 2,6-dimethy1cyclohexanone and CO2 by the pro-
cedure described above: 65% yield; mp. 58-61°C; 1H NMR
(CDC13) 1.05 (3H, d), 1.35 (3H, s), 1.4-2.7 (7H, m), 10.65
(1H, 5); IR (CHCl3 solution) 3500-2500 (br), 1720 shoulder
1700 (s) cm‘l. Mass spec. (m/e) 170 (M+), 152 (M+-H20),

126 (M+-C02), 111, 97, 87. i

2,2,4-Trimethy1-3-Oxopentanoic Acid was prepared from

 

diisoprOpyl ketone and CO2 by the procedure described
29), 1

 

above: 65% yield; mp 25°C (lit 24° H NMR (CDC13),
1.15 (6H, d), 1.45 (6H, s), 3.0 (1H, septet), 11.2 (1H, s);

IR (CHCl solution) 3400-2700 (br), 1690 (s) cm'l. Mass

3
Spec (m/e) 158 (M+), 14o (M+-H20), 114 (M+-COZ), 99 (M+-

COZMe) 71, 43.

2-Methy1-3-Oxopentanoic Acid was prepared from di-

 

ethyl ketone and CO2 using the general procedure described

above: 56% yield obtained as a viscous oil; 1

H NMR (CDC13),
1.1 (3H, tr), 1.4 (3H, d), 2.6 (2H, g), 3.6 (1H, q), 9185

+ + + +
(1H, 5); mass spec (m/e) 130 (M ), 101 (M -E ), 86 (M -

C02) ' 57o

2-Ethy1e3-Oxohexanoic Acid was prepared from 4-heptanone

 

and CO2 by the procedure outlined above: 75% yield iso-

l

lated as a viscous oil; H NMR (CDC13), 0.95 (6H, overlapping

triplets), 1.75 (4H, m), 2.55 (2H, tr), 3.45 (1H, tr),

82

11.1 (1H, 8); IR (CHC13 solution) 3500-2500 (br), 1725-

1700 cm‘l, mass spec (m/e) 158 (M+), 115 (M+-C3H 114

7),
(M+-C02), 71, 43.

2-Oxocyglgpentane-1-Carboxy1ic Acid was prepared from

 

cyclopentanone and CO2 by the procedure described above:

1

38% yield obtained as an oil; H NMR (CDC13), 1.5-1.9 (2H,

m), 2.1-2.6 (4H, m), 8.9 (1H, br. s).

3-Oxopentanoic Acid and 2-Methy1-3-Oxobutanoic Acid were

 

prepared and analyzed in the following manner.

2—Butanone (10 mmol 0.9 mL) was carboxylated by the
previously described method. A yellow oil (0.55 g, 47%
conversion) was isolated and appeared, by 1H NMR, to be a
mixture of the two possible -keto acids. This residue
was dissolved in 5 mL of diethyl ether and treated with
bis, triethylsilylacetamide according to the procedure of
Kleve, Finkbeiner, and White48. The resulting solution,
containing the two bis-silylated acid derivatives, was
subjected to CC analysis employing tetradecane as an internal
standard. The silylation yield was 26%. The two possible
products (A and B) were present in approximately equal
amounts with compound A eluting first followed closely by
8 : g; 1H NMR (c0c13) 0.27 (9H, 5), 0.29 (9H, s), 1.06
(3H, tr), 2.69 (2H, q), 5.02 (1H, 5); mass spec (m/e) 260
(M+), 245 (M+-CH3), 171, 147, 75; 8; 1H NMR (cnc13) 0.20

(9H, s), 0.35 (9H, s), 1.9 (3H, m), 2.30 (3H, 5); mass spec

 

 

 

83
(m/e) 260 (M+, 245 (M+-CH3), 171, 147, 73.

Me 810 0

MeBSiO 0 3 u
u
\\\//駧V//\\‘OSiMeQ ’/é:?T//\\‘081Me3
J
A B

Reaction of I:Butyrolactone with CO2 and M912

 

y—Butyrolactone (10 mmol, 0.77 mL) was subjected to the
standard carboxylation conditions. Workup as usual yielded
0.10 g of an orange oil which appeared, by NMR, to be a

complicated mixture which did not merit further study.

Reaction of B-Dicarbonyl Compounds with C09 and MgI2

 

Carboxylation of 1,3-cyclohexanedione was attempted by
the usual procedure. Absorbtion of CO2 was slow. Work—
up of the final mixture furnished 0.23 g of l,3-cyclohexane-
dione as the only product.

Ethyl acetoacetate was also subjected to the carboxyla-
tion procedure outlined above. The reaction required 24
hours to absorb 281 mL of C02. Workup and analysis yielded
0.20 g of recovered keto ester with no accompanying car-
boxylated product.

Diethyl malonate, when reacted as above, quickly ab-
sorbed over 300 mL of carbon dioxide. Analysis of the
residue after workup indicated that 0.27 g of starting

diester was recovered with no carboxylated product seen.

REFERENCES

 

10.

11.

12.

13.

14.

15.

References

House, H. 0. "Modern Synthetic Reactions"; W. A.
Benjamin, Inc.: Menlo Park, CA, 1972; 2nd Ed., pp.
747—765.

Hauser, C. R.; Swamer, F. W.; Adams, J. T., Org. Reac-
tions 8888, 8, 59.

 

Augustine, R. L. "Carbon-Carbon Bond Formation";
Marcel Decker Inc.: New York, NY, 1979; Vol. 1, pp.
250-263.

Seeback, D.; Weller, T.; Protschuk, G.; Beck, A. K.;
Hoekstra, M. S. Helv. Chim. Acta. 8888, 83, 716.

 

Howard, A. S.; Meerholz, C. A.; Michael, J. P. Tetra-
hedron Lett. 8888, 1339.

 

Kuwajima, 1.; Matsumoto, K. Tetrahedron Lett. 1878,
4095. m %

 

 

Enders, D.; Weuster, P. Tetrahedron Lett. 8888, 2853.

Suzuki, M.; Watanobe, A.; Noyori, R. J. Am. Chem. Soc.
8888, 102, 2095.

Tsuji, J.; Nagashima, H.; Hori, K. Chem. Lett. 8888,
257.

 

 

 

Ishihara, H.; Hirabayashi, Y. Chem. Lett. 8888, 1007.

House, H. 0. "Modern Synthetic Reactions"; W. A. Benja-
min Inc.: Menlo Park, CA, 1972; 2nd Ed., PP. 766-772.

Rathke, M. W.; Sullivan, D. F. Tetrahedron Lett. 8888,
1297.

 

 

Rasmussen, J. K. Synthesis 8888, 91.

Kramarova, E. N.; Baukov, Y. I.; Lutsenko, I. F.
Zhurn. Obsch. Khim. 8888, 58, 1857.

 

(a) Murai, S.; Kuroki, Y.; Hasagawa, K.; Tsutsumi,
S. J. Chem. Soc. Chem. Commun. 8888, 946.

 

84

 

15.

16.
17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

85

(b) Kramarova, E. P.; Baukov, Y. I.; Lutsenko, I. F.
Zhurn. Obsch. Khim. 8888, 32, 478.

 

Kopka, I. E.; Rathke, M. W. J. Org. Chem. 8888, 38, 3771.

 

Rathke, M. W.; Tirpak, R. E. unpublished results.

(a) Stork, G.; Hurdlik, P. F. J. Am. Chem. Soc. 8888,
89, 4462.

 

(b) Tidwell, et al. J. Am. Chem. Soc. 8888, 102, 5835.

 

Duboudin, et al. J. Organomet. Chem. 8888, 201, C9.

 

Houghten, R. P. "Metal Complexes in Organic Chemistry";
Cambridge Univ. Press: Cambridge, G.B., 1979; p. 106.

Cotton, F. A.; Wilkinson, G. "Advanced Inorganic Chem-
istry"; Interscience Pub., 1972; 3rd Ed., pp. 212-218.

Staab, H. A. Angew. Chem. Int. Ed. Engl. 8888, 8, 351.

 

 

Reid, W.; Kénigsten, F. J. LiebigTs Ann. Chem. 8888,
622, 37.

Staab, H. A. Chem. Ber. 8888, 89, 1321.

 

(a) Taguchi, K.; Westheimer, F. H. J. Am. Chem. Soc.
8888, 8;, 7413, and references therein.

 

(b) Logue, M. W.; Pollack, R. M.; Vihullo, V. P. g;
Am. Chem. Soc. 8888, 81, 6868.

 

(c) Spetnagel, W. J.; Klotz, I. M. J. Am. Chem. Soc.
8888, _9_8, 8199.

(a) Kuo, Y. N.; Yahner, J. A.; Ainsworth, C. J. Am.
Chem. Soc. 8888, 88, 6321.

 

 

(b) Krapcho, A. P.; Kashdan, D. S.; Jahngen, E. G. B.;
Lovey, A. J. J. Org. Chem. 8888, fig, 1189, and refer-
ences therein.

 

 

Miyano, M.; Dorn, C. R. J. Org. Chem. 8888, 81, 1818.

Herbert, I. C.; Jackson, F. B.; Nicolson, I. T. 8;
Chem. Soc. Chem. Commun. 8888, 450.

 

Van der Baan, J. L.; Barnick, W. P.; Bickelhaupt, F.
Tetrahedron 8888, 34, 223.

 

Crombie, L.; Hemesley, P.; Pattenden, G. Tetrahedron
Lett. 8888, 3021.

 

 

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.
41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

Kaishi 8888, 9, 1344.

86

Martin, J.; Watts, P. C.; Johnson, F. J. Org. Chem.
8888, 99, 1676.

 

Breusch, F. L.; Kreskin, H. Rev. Faculte Sci. Univ.
Istanbul 8888, 11a, 24.

 

Locquin, R. Bull. Soc. Chim. France 8888, 88, 595.

 

Swain, C. G.; Bader, R. P.; Estreve, R. M.; Griffen,
R. N. J. Am. Chem. Soc. 8888, 88, 1951.

 

Metz, M. A.; Axelrod, A. B.; Hoffman, K. J. Am. Chem.
Soc. 8888, 18, 1231.

 

Logue, M. W. J. Org. Chem. 8888, 99, 3455.

 

Barnick, J. W. F. K.; Van der Baan, J. L.; Bickelhaupt,
F. Synthesis 8888, 89, 787.

 

Bottaccio, G.; Chiusoli, G. P.; Felicioli, M. G.
Gaz. Chim. Ital. 8888, 103, 105.

 

Bottaccio, G.; Chiusoli, G. P. J. Chem. Soc. Chem.
Commun. 8888, 618.

 

, ;§, 4086.

 

 

Levine, R.; Hauser, G. J. Am. Chem. Soc. 88 8, 6, 862.

Baumgarten, B.; Levine, R.; Hauser, G. J. Am. Chem.
Soc. 8888, 99, 862.

 

Haruki, B.; Arakawa, M.; Matsumura, N. Otsuji, Y.;
Imoto, E. Chem. Lett. 8888, 427.

 

 

Stiles, M. J. Am. Chem. Soc. 8888, 81, 2598.

Sakurai, H.; Shirahata, A.; Hosomi, A. Tetrahedron
Lett. 8888, 88, 1967.

 

Matsumura, N. Yagyu, Y.; Imoto, E. Nippon Kaguku

 

Martin, J.; Watts, P. C.; Johnson, F. J. Chem. Soc.
Chem. Commun. 8888, 27.

 

 

Klebe, J. P.; Finkbeiner, H.; White, D. M. J. Am.
Chem. Soc. 8888, 89, 3390.

 

Straub, T. S.; Bender, M. L. J. Am. Chem. Soc. 8888,
98, 8881.

 

Linstead, R. P.; Wang, A. B. J. Chem. Soc. 8837, 807.

 

 

87

Ainsworth, C.; Kuo, Y. J. Organomet. Chem. kgzg, £9,
75.

Ashby, E. C.; Arnott, R. C. J. Organomet. Chem. k2é§'
£3, 1.

 

WM

1293 03177 8669

umummumhnfluit'u

3

 

un‘fiu’fifinum