0 -Yt'.—1 1 O " v. . “ 'l . W??? 5:1
.. -- V ‘ -

- "\H‘w ‘i'ii

k r O , s. . f. . U >3:
a V ?-f_

. . o. u . ‘ . ' -':

.
V l
u; f‘JHI.t,‘,‘ ‘

' 5, >4
.fl) 0": ‘- .‘gi‘fizi'

" ' ‘ ""l.'.':""l'

.5

. 'I
"M
I

*

\' I." 1. H
l:“ t n' "I?!”
*0 . ’ ' ’ u"; 1".” "lo.

.I.‘ .. “‘ .‘ HI 9;..39:
' ' l "' I V! D

: 3-9023. .. l I“

I ' . '

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled
THE GENERATION AND REACTION OF ENOLATE ANIONS WITH
TRIETHYLAMINE IN THE PRESENCE OF MAGNESIUM 0R
LITHIUM HALIDES

presented by
Michael Anthony Nowak
has been accepted towards fulfillment

of the requirements for

Doctoral degree in Chemistry

CC)

Major professor

Date August 5: 1985

"Till. n- f o.’ I ' I!" In . I « . 0.12771

 

 

MSU

LIBRARIES

 

 

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

 

 

 

'flllGlllmflflflliflllIlflflflnltu'llfllfllIflflnlfllflfll
TIIIHHHJHDIIII'flllPilfllllloflflflllflflllal
Lrnunnlmunnns

BY

MICHAEL ANTHONY NOWAK

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of
DOCIIIOIIIIHIEPIY
Department of Chemistry

1985

 

Bis(trimethylsilyl) malonate, in the presence of
triethylamine and lithium or magnesium halides, is converted
to its enolste anion. Under these conditions, the enolate
anion is C-acylated in good yields with a variety of acid
chlorides or ethyl octanoyl carbonate. Important exceptions
are the hindered acid chloride pivaloyl chloride, which
gives a fifty percent yield of C-acylated product, and
crotonyl chloride, which gives no c-acylated product. Under
these conditions, acyl imidazoles give modest yields of C-
acylated product. Subsequent hydrolysis and decarboxylation
of the acylation product gives p-keto acids or methyl
ketones.

In the presence of triethylamine and lithium or
magnesium halides, triethylphosphonoacetate reacts with a
variety of aldehydes to give s,p—unsaturated esters in
excellent yields. Under the same conditions,
triethylphosphonoacetate is unreactive towards simple methyl
ketones.

Compared to other procedures which employ strong bases,
these procedures are inexpensive, safe and convenient,
especially for large-scale preparations. They are also
particularly attractive for use with base-sensitive

substrates.

 

In memory of my father, Edward Nowak.

 

The author wishes to thank Professor Michael W. Rathke
for his patience and his guidance throughout the course of
this work. I consider it a privilege having been associated
with him. I would like to express special thanks to
Professor Willian Reusch for serving as second reader.

I would like to thank my wife, Nanette, for always
being there and believing in me. I would like to thank my
Mom, Stan J. and my families in New York and Michigan for
their faith, interest and continuing support.

I would like to thank my fellow graduate students,
especially the members of the Rathke group and lost
especially Rob Tirpak, for sharing their friendship, advice
and chenistry. Thanks are extended to Professors Steven
Tanis and William Reusch for the stimulating discussions at
group meetings.

I would like to thank the C.T.’s of the MSU Chemistry
Department for their cheerful and invaluable technical and
administrative assistance. Appreciation is extended to
Michigan State University and SOHIO for financial support
during my graduate studies.

iii

 

‘flflfll QPCNIflflflE

List of Tables. . . . . . . . . . . . . . . . . .
Chapter I - Acylation of Bis(Trimethylsily1)
Malonate Using Triethylamine and
Magnesium or Lithium Halides. . . . .
Introduction. . . . . . . . . . . . . . . . . . .
Results and Discussion. . . . . . . . . . . . . .
Experimental. . . . . . . . . . . . . . . . . . .
Materials. . . . . . . . . . . . . . . . . .

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

General Procedure for the Acylation of
Bis(Trimethylsilyl) Malonate . . . . . . . . .

General Procedure for the Preparation of
Methyl Ketones . . . . . . . . . . . . . . . .

General Procedure for the Preparation of
fi-keto acids . . . . . . . . . . . . . . . . .

Isolation of 10 Using Lithium Bromide and Tri—
ethylamine . . . . . . . . . . . . . . . . . .

Characterization of 10 Prepared Using g-
Butyllithium . . . . . . . . . . . . ~.- . .

Attempted Acylation of 9 with Isobutyryl
Chloride Using n—Butyllithium. . . . . . . . .

Chapter II - The Horner-Wadsworth-Wittig
Reaction Using Triethylamine
and Magnesium Halides . . . . . .

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

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

10
21
21
22

22

23

23

24

25

25

27

28
38

 

Page

Experimental. . . . . . . . . . . . . . . . . . . . 48
Materials. . . . . . . . . . . . . . . . . . . . 48
Methods of Analysis. . . . . . . . . . . . . . . 48
General Procedure for the Phosphonate
Olefination Reaction . . . . . . . . . . . . . . 49

General Procedure Used to Survey Metal
Salts. . . . . . . . . . . . . . . . . . . . . . 49

General Procedure Used to Survey Solvent
Systems. . . . . . . . . . . . . . . . . . . . . 50

General Procedure for the Isolation of
Unsaturated Esters . . . . . . . . . . . . . . . 50

1H NMR Study of Triethylphosphonoacetate/
Lithium Bromide/Triethylamine Mixture. . . . . . 51

1H NMR Study of Triethylphosphonoacetate/
Magnesium Bromide/Triethylamine Mixture. . . . . 52

Isolation of the Magnesium Enolate of Triethyl-
phosphonoacetate . . . . . . . . . . . . . . . . 52

Bibliography. . . . . . . . . . . . . . . . . . . . 54

 

Table

Table

Table

Table

Table

Table

Table

Table

lflST¢fl'IflfllS

Acylation of Diethyl Malonate. . . .

Reaction of Bis(trimethylsilyl
Malonate with Acid Chlorides
Using MgClz. . . . . . . . . . . . .

Reaction of Bis(trimethylsilyl)
Malonate with Acid Chlorides Using
LiBr . . . . . . . . . . . . . .

Acylation of Bis(trimethylsilyl)
Malonate Using Lithium Bromide .

Reaction of Bis(trimethylsily1)
Malonate with Acyl Carbonates and
Imidazoles . . . . . . . . . . . . .

Reaction of Cyclohexanone with 19 in
the Presence of Metal Halides. . . .

Reaction of Carbonyl Compounds with
19 in a Variety of solvents. . . . .

Reaction of a Variety of Carbonyl
Compounds with 19 in the Presence of
Triethylamine and Metal Halides. .

12

14

16-17

19

39

41-42

44

 

(annals I

THE ACYLATION OF BIS(TRIMETHYLSILYL) MALONATE
USING TRIETHYLAMINE AND MAGNESIUM 0R LITHIUM HALIDES

 

Carbon-carbon bond forning reactions are the backbone
of synthetic organic chemistry. Base-pronoted removal of a
proton alpha to a carbonyl group gives a resonance-
stabilized enolate anion (eq. 1). Subsequent reaction with
an electrophile (eq. 2) constitutes one of the most
frequently used nethods of effecting such bond-forming

reactions.

0 o o-
" base N ‘ I
RCCH: R1 R2 '-—> R’C-CR)’ R2 H RC=CRI R2 (1)
0- 0
l 3’ ll
R-C=CHIR2 -—-—-—> R—C-CRIRZE (2)

For the past decade, synthetic applications of enolate
reactions have emphasized the use of very powerful bases,
primarily the lithium dialkylamides (pK. > 35).1

It is well known that complexation of a metal cation to
a ligand can enhance the acidity of the ligand.2 If metal
complexation with carbonyl compounds would enhance their
acidity sufficiently so that enolate formation could be
promoted by weak bases such as the tertiary amines (pK.~10),

then a new route to enolate chemistry would be available

 

(eq. 3). The synthetic advantages of such a route to

enolate species are nost apparent for acylation reactions.

M

I

o o”n1 0’“
II M’ II I EtaN |
RCCHR‘R2 ‘--->- R-C-C-R2 ----t- R-C=CR182 (3)
The standard procedure for acylation reactions involves
a two-step sequence (eq. 4). In the first step, an enolate
anion is generated using a strong base such as a metal
alkoxide, lithium alkyl or lithium dialkylanide. The second
step involves addition of an acylating agent. This
procedure is plagued by the fact that the product 3 is a
stronger acid than starting compound 1 and thus may
neutralize the starting enolate 2, often limiting yields to
fifty percent.3 One solution to this problem is to effect

the reaction in a single step using two equivalents of base

0
0 0- H O O
H strong I RZC—X | |
RCCII2R1 —* RC=CIIR1 ——> R1CCRR2CR3
base -
1 2 3 (4)
2 ._ ' 1
3 _. Ii ' R' +

 

to produce enolate 4 stoichiometrically. However, because
of side reactions with the acylating agent, such procedures
are not usually feasible with the strong bases used in
enolate chemistry. If the metal complexation route of
equation 3 were successful, the weak but tolerant tertiary
amine bases would accommodate the high acidity of the
acylation product but avoid reaction with the acylating
agent.

Initial investigation of this proposed enolate
chemistry began with the relatively acidic (pK.~15)‘ [-
dicarbonyl compounds. Formation of enolate complexes from
this class of compounds seems relatively straightforward,
not only because of their high intrinsic acidity, but also
because the chelating nature of the enolate should provide a
strong driving force for bonding with an appropriate metal
ion (eq. 5). Unfortunately, the same factors which make 5
easy to prepare also make 5 relatively inert towards
electrophiles. For example, many acetyl acetonate
complexes, including chromium acetyl acetonate 8, are weak

nucleophiles5 which undergo typical enolate reactions only

IM\ IM\
I \O of \o
R R ———————>./ji\v/fl\\ . J
,/”\\//~\\ 9 M. , ‘————-> ’ (5)
5
with powerful electrophiles, such as Friedel—Crafts

reagents, or at elevated temperatures.

 

The objective of this investigation was to identify
carbonyl substrates and metal ions for which complex 5 has
sufficient stability to be formed in useful concentrations
by weak bases but also is able to react with a variety of
electrophiles. The last step in equation 5 is a proton-
exchange reaction, and the overall driving force of the
reaction depends on the stability of the metal oxygen bond
in 5. The greatest chance for success would appear to be
with magnesium or zinc where the metal ions have appreciable
complexing ability for oxygen ligands and where their
enolates and known to possess sufficient reactivity towards
a number of electrophiles.2' Also, magnesium enolates are
known to have a low tendency to acylate at oxygen rather
than carbon°, and oxygen acylation is often a significant

problem in acylation reactions of ambident enolate anions

(eq. 6).
0
0 II
o- o u o—c—R1 o o
I ll R‘CX I II II
R-C=CH2<—DR-C-CH2' ——>RC=032 + llCCHzC-lll (6)

A member of our research group’ examined the acylation
of diethyl malonate in the presence of a number of metal
salts (Table l). The malonate esters are one of the least
acidic members of the p-dicarbonyl class of compounds

(pEa~l4). Nevertheless, diethyl malonate was acylated in

Till! I. Acylation of Diethyl Mhlonate.‘

 

 

l) MX, EtaN
EthCCRzCOzEt _——’ (EtOzC)2CHC—Ph
2) PhC-Cl H
II 0
0
Recovered Starting
Entry MX Yield Product, x Material, x
1 ___ 0 87
2 ZnClz 0 93
3 CuClz 0 99
4 FeCla 0 63
5 MgCl: 85 8
6 MgClz 0 98b
7 L101 0 —-—
8 LiBr 88 ---

 

8)

b)

All entries except entry 8 are taken from:

Cowan, Patrick, Ph.D. dissertation, Michigan State
University, 1983.

Reaction carried out in the absence of triethylamine.

 

good yields in the presence of magnesium chloride and
triethylamine (pKa~10).1' Control experiments demonstrated
the necessity of both the metal salt and EtaN for the
success of the reaction. The resulting acylation products
may be of interest to synthetic organic chemists, but more
importantly, the methodology should be applicable to similar
systems.

There are three general routes to fi-keto acids:
acid15-19 or base-pronoted hydrolysis17 of the corresponding
keto ester (eqs. 7,8), carboxylationm“21 of ketone enolate
anions (eq. 9) and acylation'-22 of carboxylic acid dianions

(eq. 10). A serious problem associated with the acid-

0
N H’, 320 N
RCCHzCOle — RCCH2C02H (7)
O 0 0
II ‘0“ ll 3* ll
RCCIIzCOzR1 ———-—. RCCHaCOz' —-> RCCIhCOzH (8)
0 0 0
II CO: H H’ II
RCCHz' —-’ RCCHzCOz' _’ 13008200le (9)
0
|| 0 0
BOX H 8* H
'CHzCOz‘ ——> RCCHzCOz‘ _—> RCCHzCOzH (10)

catalyzed hydrolysis of keto esters is that subsequent

decarboxylation of the keto acid (eq. 11) may occur.

 

0 O
H H‘ H
RCCHzCOzH ——> RCCHa + C02 (11)
A
Alkaline hydrolysis of p-keto esters avoids this

decarboxylation but is complicated by possible retro-Claisen
cleavage of the p-keto ester (eq. 12). An alternative

approach, the acylation of silyl acetates (eq. 13), gives

0 0
u -on u
RCCR‘zCOsz --—-> RC—OR + -CR12C02R2 (12)
° 0 O
1) Base | 830+
CHac-OSiRa ------a- Rlc-Cflac-OSiRa '---———>
2) RCX
H
O
0
H
nlccuacOza (13)

keto esters that are readily hydrolyzed under neutral

conditions.23 The carboxylation of ketones (eq. 14) using

"9
O
of ‘o

H
Rccnzn1

 

(14)
7

/
C)

magnesium methyl carbonate (MMC)17 in dimethylformamide
(DMF) solution gives keto acids, and formation of complex 8
appears to provide the thermodynamic driving force for this

reaction. Consequently, a necessary requirement for using

 

MMC is that starting ketone 7 contain at least two alpha hy-
drogens. A second drawback to using MMC is the large
excesses (5—20 fold) of the reagent required to obtain good
yields.

The synthesis of fl-keto acids via acylation of
bis(trimethylsilyl) malonate 9 (a synthon of trimethylsilyl
acetate) using two equivalents of the strong base n-
butyllithium (eq. 15) has been described.2‘ Two equivalents
of 9 in an ethereal solvent were cooled to -60°C and treated
with two equivalents of n—butyllithium. After warming to
0°C, the solution was treated with acylating agent and
stirred for 10 minutes. The acylating agents described are
a variety of aryl acid chlorides, pivaloyl chloride, and one
example each of n-alkyl acid chloride, a mixed anhydride and

acyl benzotriazole. Aqueous workup gave p—keto acids in 63-

938 yield.
0 0
N H l) 2 n-BuLi H
2 (CH:)3SiOCCHaCOSi(CH3)3 -—---—---> RCCHzCOzH +
2) RCX
9 ll
0
3) H30+
CR:(CO:H)2 (15)

Substrate 9, which is structurally similar to diethyl
malonate, is ideally suited for metal cation complexation,
and routes to enolate chemistry gig weak tertiary amine
bases should be applicable here. In addition, after

acylation, the readily hydrolyzable silicon esters should

10

provide an easy route to l-keto acids and to methyl ketones

by decarboxylation of the corresponding p-keto acid.

llflnflBIMIIBnlmBSRII

Because of the similarity of the substrates,
investigation of the acylation of 9 began with the
conditions found suitable for the acylation of diethyl
malonate.7 An initial set of experiments provided a survey
of solvent systems using magnesium chloride and
triethylamine as promoters of the reaction. Addition of 9
to a stirred slurry of magnesium chloride in the solvent
resulted in complete dissolution of the magnesium chloride.
The resulting homogeneous solution was then treated with
triethylamine and a precipitate formed immediately. After
stirring ten minutes, the heterogeneous mixture was mixed
with an acid chloride at room temperature. After an aqueous
work-up, the reaction mixture was analyzed by gas
chromatography (CC) for the methyl ketone obtained by

decarboxylation of the corresponding p—keto acid (eq. 16).

I) MgCIz, EtzN

MeaSiOzCCHzCOzSiMea V RCOCH3 "6)
2) RCOCI
3) HeO’

 

In this manner bis(trimethylsi1yl) nalonate was acylated
with benzoyl chloride in good yield in either diethyl ether
or acetonitrile solution. Since diethyl ether can be
purchased as anhydrous material, it was selected as the
solvent of choice. Attempts to acylate 9 with other acid

chlorides under the conditions described gave only poor to

ll

moderate yields of acylation product. The results of these
initial experiments are summarized in Table 2.

The ability of other metal salts to promote the
reaction was then examined. Lithium bromide proved to be an
effective alternative to magnesium chloride. Use of lithium
bromide offered two advantages: 1) Lithium bromide is easier
to obtain and handle in the anhydrous state. 2) The lithium
enolate should be more reactive than the corresponding
magnesium enolate.

We wished to verify that these acylation reactions were
proceeding via an enolate intermediate by isolating 10.
Addition of triethylamine to a solution of lithium bromide
and bis(trimethylsilyl) malonate in diethyl ether
immediately gave a precipitate. No precipitate formed in
the absence of lithium bromide. After stirring ten minutes,
the precipitate was isolated and dried under high vacuum;
the filtrate was concentrated under high vacuum. The
precipitate accounted for 99% of the theoretical weight of

triethylamine hydrobromide (eq. 17) and exhibited a 1H NMR

EtaN : 1
9 + LiBr ——>

+
Mo,Si . c; 05"“: EtaNHBr‘ (17)

spectrum identical to that of triethylamine hydrobromide.
Mass spectral analysis of the precipitate also gave a
spectrum corresponding to the relatively volatile
triethylamine hydrobromide (EtaNH’Br‘). The filtrate
residue gave a 1H NMR spectrum exhibiting silicon methyls

and a singlet 0.55 ppm upfield from the alpha proton signal

 

TABLE 2. Mia: of lis(tri-tlwlmilyl) manta
wiullcthEhmddulUmhuflflfla

1) MgClz, EtaN

 

 

 

MesSiOzCCRzCOzSiMea r» RCOCEa
2) RCOCl
3) 830*

Entry Reaction Time Solvent R Yield, 8
1 1 hour C83CN Ph 90
2 1 hour Etao Ph 90
3 1 hour C82Clz Ph 32
4 -1 hour THF Ph 50
5 1 hour Etzo 08:03:03: 0
6 1 hour Et20 n—Cvflis l5
7 1 hour Et20 i-Pr 6
8 18 hours EtaO i-Pr 15
9 18 hours 082012 i-Pr 52

 

Standard Reaction Conditions: 5 mmol scale: 1.0 equiv.
BSM, 1.05 equiv. Mgclz, 2.1 equiv. EtaN, 5.0 mL solvent.
After 10 min., 1.0 equiv. RCOCl. All yields are GC
yields.

 

13

of bis(trimethylsilyl) malonate. Comparison with the 18 NMR
spectrum of 10 prepared by reaction with grbutyllithium con—
firmed that enolate formation had taken place.

Acylation reactions of bis(trimetylsilyl) malonate,
using the lithiun bromide procedure, were examined under a
number of different reaction conditions (Table 3). Efforts
to improve the yields focused on reaction conditions which
were likely to promote enolate formation and minimize
possible side reactions. Diethyl ether remained the solvent
of choice. Coloration of the reaction mixture occurred when
either isobutyryl chloride or crotonyl chloride was added to
the reaction flask. This suggested that side reactions were
occurring, possibly ketene formation. Weaker basesl than
triethylamine proved ineffective in promoting the reaction.
The reaction temperature was an important factor. Best
yields were obtained when the malonate ester, lithium
bromide and triethylamine were stirred for several minutes
at room temperature and then cooled to 0°C before addition
of the acid chloride.

One problem frequently encountered in acylation
reactions of enolate species is competition of O-acylation
with C-acylation (eq. 6). The failure to detect any C-
acylated material derived from pivaloyl chloride suggested
that O-acylation might be the predominant reaction with
hindered acylating agents. Likewise, the modest C-acylation
yields obtained with isobutyryl chloride might be attributed
to competing O-acylation. Consequently, it was important to

determine if in fact O-acylation was responsible for our

Till! 3. limitiam aflflm(tnhlfliu1silyl)IiflommteldflhlAcid
ChlorhhmIUEBMILilr.

1) LiBr, EtaN
MeaSiOzCCHaCOzSiMea

 

RCOCH:CO:H or RCOCHs
2) RCOCl
3) H20

 

Entry Reaction Time Temperature Solvent R Yield, x

 

l 18 hours 25 Et20 Ph 57
2 18 hours 25 082C12 Ph 45
3 18‘hours 25 CHaCN Ph 55
4 18 hours 25 EtaO Ph 0‘
5 18 hours 25 03201: Ph 10°
6 18 hours 0 Et20 Ph 60b
7 1 hour 0 EtzO Ph 88c
8 1 hour 0 EtaO i-Pr 50d

 

Same reaction conditions as Table I, with LiBr in place of
MgClz.

a. Pyridine used as base.
b. Base added after RCOCl.
c. Isolated yield.

d. Base added at 25° and stirred for 10 minutes, then
cooled to 0°.

 

limited success with this procedure by isolating and
identifying possible O-acylation products. To this end,
isobutyryl chloride was added to a stirred reaction mixture
containing 9, lithium bromide and triethylamine in diethyl
ether at 0°C. The crude reaction mixture was analyzed by
gas chromatographic mass spectroscopy (GC/MS). No 0-
acylation product could be detected. 18 NMR analysis of the
crude reaction mixture did not reveal the presence of 0—
acylated product because: 1) The 18 NMR spectrum of the 0-
acylated product is expected to be nearly identical to that
of a mixture of 10 and isobutyryl chloride. 2) If the two
spectra were distinguishable, the CH: signal of the ammonium
salt would be superimposed on the enolate (vinyl) hydrogen.
Attempts to isolate O-acylated material from the crude
reaction mixture were unsuccessful. Thus, there is no
direct evidence of O-acylation.

In difficult cases, successful C—acylations can often
be accomplished by using excess starting enolatezs, which
then reacts with the O-acylation product to give the desired
C-acylation product together with the starting enolate (eq.

18). The acylation reactions were, therefore, re-examined

0
0 ll 1
‘C-R ‘ 0' O O
R-c=CH2 + 032:0'3 -——-——”R-C=CH2 + Rl-C-CHz-CR (18)

under a variety of conditions using a slight excess (58) of
bis(trimethylsilyl) malonate, lithium bromide, and

triethylamine (Table 4). For those acid chlorides which had

TABLE 4. Acylation of lis(tr1methw1sily1) Milan-ta
Ibingldthhlllnudde

0
l) LiBr, EtaN, Et20 H H
4‘ RCCH:COzR or RCCH:

 

CH:(COaSi(CHa)a)2
2) RCOCI
3) 830

 

 

Entry R Reaction Time Yield (X)
1 Ph 1 hour 85
2 g-Cvflis 1 hour 91
3 QfC7H15 1 hour 80f
4 g-Cvflis 1 hour 84hf
5 g—Cvflis 1 hour 65"-f
6 g-Cvfiis 1 hour 42°!,
7 9701815 1 hour 45".f
8 (CRa)2CH 1 hour 50
9 (CHa)zCH 1 hour 30c
10 (Cfls)2CH 1 hour 12°
11 (083)30 1 hour 0
12 (083)30 12 hours 35'
13 (083)30 18 hours 50I
l4 (CEa)aC 40 hours 10‘

 

TMHIIL oufldmld

 

 

Entry R Reaction Time Yield (3)
l5 083CH=082 1 hour 0
16 (083)208 0.25 hour 0h

 

a) Two full equivalents of lithium bromide used.

b) Malonate ester, base and lithium bromide were stilled
at 0°, not room temperature.

c) Ratio of reagents used is 1.1 BSM: 1.05 LiBr: 2.1
EtaN: 1.0 RCOCl.

d) Reaction mixture allowed to stir three hours before
addition of RCOCl.

e) (irPr)2NEt used in place of EtaN.

f) Isolated yield of keto acid.

g) Isolated yield of methyl ketone.

h) Using method of van der Baanz‘, obtained complex
mixture of products.

 

18

given acceptable yields of C-acylated product, a small
increase in yield was observed over standard reaction times.
Larger excesses of these reagents had little effect ((5%) on
the yields. Allowing the malonate ester, lithium bromide
and base to stir for longer periods of time before addition
of the acid chloride did not improve the yields. Extended
reaction times also gave no increase in yields with most
acylating agents, except for pivaloyl chloride. In the
latter case, pinacolone could be isolated in yields up to
508 but excessively long reaction times (40 hours) resulted
in lower yields.

It is interesting to note that our attempt to acylate
bis(trimethylsilyl) malonate with isobutyryl chloride using
a literature procedure that generates the enolate with g-
butyllithium was unsuccessful. Efforts to isolate the keto
acid using the literature procedure failed. Gas
chromatographic analysis for the corresponding methyl ketone
revealed a complex mixture of products. Gas
chromatographic-mass spectral analysis identified 3-methyl-
2-butanone as an insignificant component of the reaction
mixture.

The acylation of bis(trimethylsilyl) malonate using
lithium bromide with other acylating agents was next
examined (Table 5). Acyl imidazoles 11 and mixed
carboxylic acid anhydrides (acyl carbonates) 12, may be

prepared directly from the corresponding carboxylic acid

 

TIILI 5. Reaction of Bis(trimethy1silyl) Mhlonate with Acyl
Cadmmmmulamd.nfldmnfles.

1) MX, EtaN
MesSiOaCCHaCOzSiMea -———>RCOCB:COaH or RCOCH:

 

 

2) RCOX
3) 820
Entry RCOX1 Reaction Time Temp. Yieldz, x

l PhCOIm 18 hours 25° 0
2 i—PrCOIm 18 hours 25 <5
3 CvflisCOIm 18 hours 25 13*
4 PhCOIm 1 hour 0 0
5 CvnisCOIm 24 hours 0--RT 58‘
6 i-PrCOIm 24 hours 0--RT 0
7 i-PrCOIm 18 hours 0--RT 03
8 i-PrCOIm 1 hour 67 0
9 PhCOCOaEt 1 hour 0 0
10 l61815000038t 1 hour 0 35*
ll CvflisCOCOaEt 1 hour 0——RT 65*

 

2) Yields are 60 yields except where noted by "'".
Entries so marked are isolated yields.

3) MgClz used in place of LiBr.

 

20

o
M Q" fi fi
R_C-N\§' R—C-O-C-OEt
11 12

under relatively mild conditions. The acyl imidazoles,
which are much weaker acylating agents than acid chlorides,
gave acceptable yields of acylation product only after
extended reaction periods. No acylation product was
obtained using ethyl benzoyl carbonate but ethyl octanoyl
carbonate provided 3—oxodecanoic acid in relatively good
yield.

In summary, a new procedure has been developed for the
acylation of readily hydrolyzable malonic esters providing a
new route to p-keto acids and methyl ketones. The procedure
is safe, convenient and economical, particularly on a large
scale. Using these procedures, the reaction can be carried
out in a single step without complications due to reaction

of the base with the acylating agent.

 

MATERIALS

Acetonitrile, diisopropyl amine and triethylamine were
distilled from calcium hydride prior to use.
Tetrahydrofuran was distilled from sodium/benzophenone prior
to use. Diethyl ether was taken from a freshly opened can
of anhydrous ether. Methylene chloride was taken from a
freshly opened bottle of anhydrous methylene chloride.
Lithium bromide (Aldrich Chemical Company, 993) was dried in
an abderhalden flask over refluxing xylene at 0.3 torr. It
was stored and weighed under argon in a glove bog and dried
in sin: by treatment with chlorotrimethylsilane for ten
minutes followed by removal of all volatile materials under
high vacuum. All acid chlorides were obtained from Aldrich
Chemical Company and were purified by distillation. Acyl
imidazoles were prepared by the method of Staab.26
Carbonates were prepared just prior to use from ethyl
chloroformate and the corresponding carboxylic acid in
diethyl ether according to the procedure. of Tarbell and
Rice.27 The ethyl chloroformate was obtained from Aldrich
Chemical Company and was used without further purification.
The carboxylic acids were obtained from Aldrich Chemical
Company and purified by distillation except benzoic acid
which was purified by sublimation. n—Butyllithium was
obtained from Aldrich Chemical Company as a 1.6M solution in

hexane and was used directly.

21

 

22

Bisgtrimethylsilyl) malonate (28) was prepared from

malonic acid and two equivalents of chlororimethylsilane.

b.p. 50-52 (0.3 torr). 1H NMR: 6 3.26 (s, 1H), 0.1 (s, 9H).

llflflflMiOF.flouSIS

Gas chromatographic analyses were performed on a Varian
920 chromatograph equipped with a 6 ft. X 0.25 in. stainless
steel column packed with 15% SE-30 on Chromasorb-W or on a
Hewlett Packard 5880A chromatograph equipped with a 12.5
meter X 0.25 mm capillary column using crossed linked
dimethylsilicone as the liquid phase. 60 yields were
obtained using n—alkanes as internal standards. 18 NMR data
were obtained using a Finnigan 4000 EI GC/MS mass
spectrometer. Melting points were determined using a Thomas
Hoover melting point apparatus and are uncorrected.
Infrared spectra were taken on a Perkin-Elmer 599 grating

infrared spectrometer using a polystyrene reference.

General Procedure for the lation of Bis trime lsil 1
EM:

A 50 mL round bottom flask fitted with an efficient
stirrer, a septum inlet and a gas inlet tube with mercury
bubbler was flame dried under argon and charged with 5.5
mmol of anhydrous metal salt. To this was added 10 mL of
anhydrous solvent and 5.25 mmol of bis(trimethylsily1)
malonate. 5.5 mmol of triethylamine was added dropwise and a

precipitate formed immediately. After stirring ten minutes,

 

23

the flask was cooled to 0°C and 5.0 mmol of an acylating
agent was added. (Acid chlorides were added dropwise,
slowly; carbonates were added in one portion as the crude

freshly-prepared reaction mixture.)

(knenfl.Rnxn¢me fin‘the lgggggyfion of lknhzl Eehmnm.

After the reaction mixture described above was stirred
for one hour, it was quenched with 4 mL 5M HCl and then
refluxed for one hour. To this mixture was added an
appropriate n-alkane as an internal standard. The mixture
was then extracted with ether (3 X 10 mL), the organic
layers were combined and dried (MgSO¢), and this product was

analyzed by gas chromatography.

Generallkmcedmretfln‘the Egggggmion of gjknm»Acids.

After the reaction mixture described above was stirred
for one hour, it was quenched with 10 mL cold saturated
aqueous NaHCOa and stirred for ten minutes in an ice bath.
The aqueous layer was separated and acidified to pH 2-3 by
the dropwise addition of cold 4M 82804. The resulting
precipitate was extracted with ether (3 X 10 mL), the
organic layers were combined, dried (Mg504), and filtered.
The solvent was removed intmcm>(avoid excessive heat) and
the re-sulting white solid proved of sufficient purity to be

used directly.

 

24

Benzoyl Acetic Acid: m.p. 101—102 (lit. m.p. 101—102)29; 1H
NMR (CDCla): a 4.1 (s, 2H), 5.7 (enol H), 7.25—7.6 (m, 3H),

7.7—8.05 (m, 28).

2-0xodecanoic Acid: m.p. 76-77; 1H NMR (CDCla): 0.9-1.1 (m,
3H), 1.2-2.0 (br s, 12H), 3.6 (s, 1H), 10.6 (s, 1H) enol
proton not observed; m/s (m/e): 186 (M‘), 142 (-COz), 127,

85, 71, 58 (base), 43; IR (KBr): 3500-3300 (br), 2915, 2850,

1730, 1710, 1435.

 

A 50 mL flask was fitted with a septum inlet, magnetic
stirrer, a filter stick and a gas inlet tube. The flask was
flame dried under a stream of argon. The flask was charged
with lithium bromide (5 mmol, 0.44g) and diethyl ether (5
mL). Bis(trimethylsilyl) malonate (5 mmol, 1.3 mL) was
added and the solution stirred 5 minutes. Triethylamine (5

mmol, 0.7 mL) was added dropwise and a white precipitate

formed immediately. After stirring ten minutes, the mixture
was filtered. The precipitate was collected and dried in a
dessicator under vacuum (1 torr). The filtrate was

concentrated under high vacuum (0.3 torr) giving a white
residue. The dry precipitate exhibited the following: 1H
NMR (CDaCN): a 3.2 (q, 3H), 1.5 (t, 3H); m/s (m/e) 101
(EtaN), 80 (HBr79), 82 (HBrBl). The filtrate exhibited the

following: 18 NMR (CClq): 6 2.7 (s), 0.1 (s).

 

25

Characterization of 10 Ub' n—But llithium.

A 50 mL flask was fitted with a septum inlet, magnetic
stirrer and gas inlet tube. The flask was flame dried under
argon. Diethyl ether (10 mL) and bis(trimethylsilyl)
malonate (1.3 mL, 5 mmol) was introduced via syringe and the
solution was cooled to -78°C. n—Butyllithium (5 mmol, 3.125
mL 1.6M solution in hexane) was added dropwise. The
solution was stirred ten minutes and slowly brought to room
temperature. All volatile components were removed under
high vacuum to give a white solid. 1H NMR (CDCla): 6 2.8

(s), 0.1 (s).

Attggghui Agzlatian Imeiwith Duflnuagzfl Chlorflt20bigg n-
Butzllithium.

This procedure for the acylation of 9 with acid
chlorides (but not isobutyryl chloride) is described in the
literature.24 A 50 mL flask was fitted with a magnetic
stirrer, septum inlet and gas inlet tube and was flame dried
under argon. The flask was charged with bis(trimethylsilyl)
malonate (10 mmol, 2.6 mL) and diethyl ether (20 mL) and the
solution cooled to -78°C. n-Butyl-lithium (10 mmol, 6.25 mL
1.6M solution in hexane) was added dropwise and the solution
warmed to 0°C. Isobutyryl chloride (5 mmol, 0.52 mL) was
added dropwise and the reaction mixture stirred for ten
minutes. Standard aqueous work-up fails to provide any ’-
keto acid. Repeating the acylation procedure followed by
work-up previously described as the general procedure for

the preparation of methyl ketones gave a gas chromatograph

 

26
exhibiting fifteen unassignable signals and a 4% yield of 3-
methyl-Z-butanone. GC/MS did not prove useful in

determining the identity of the numerous by—products.

mu

Th Mint“ tactic. thing Trietlnl-ime
nd Mami- a- Lithi- Mich.

27

28

IIIIINBTEII
One of the most important tools available to the
synthetic organic chemist for the construction of carbon-

carbon bonds is the Wittig olefination reaction (eq. 19).

 

0
+ base + R1CR2
RaP-Cfla X‘ 4: RaP=CHa’ ----*> 81820=Cfia (19)

It provides an easy method for linking two synthons of any
size which bear various functional groups. The Wittig
reaction and its modifications provide a maJor advantage
over other olefin-forming reactions by introducing, the
double bond regiospecifically.

The Horner-Wadsworth-Emmons modification of the Wittig
reaction (eqs. 20-23) employs phosphonate esters 12 hearing
an additional-electron withdrawing group, E. One of the
most important Eorner-Wadsworth-Emmons reagents,
triethylphosphonoacetate, 19, employs an ester function as
the electron-withdrawing group. Triethylphosphonoacetate
and compounds similar to it have been used to prepare a
variety of natural products including prostaglandins30,

Juvenile hormones31, and a number of isoprenoid compounds32

29

 

0 0
fl Base, M’ -
(RO)2PCH-E : (RO)2P?-E M‘ (20)
|
R R
13
o o o
" II “\“o‘ "- ‘\\\\\“‘E
13 + Ric-32 -———> (R0)2P-C . (R0),P C (21)
\ l\fl
| R .2
-o—cum~"‘R' -o our“
R3 \n‘
14 ‘5

including p-carotene.33 The Horner-Wadsworth-Emmons reaction
has been applied intramolecularly in the synthesis of ring
compounds containing cycloalkenones3‘ and butenolide
moieties.35 The Horner-Wadsworth-Emmons reaction with
epoxides yields substituted cyclopropanes.3° The phosphonate

olefination has been employed in a number of industrial

OI
e n‘ ' ..
14 —’ Rx”: ’ (RO)2H-o (22)
17
E 1 '
15 —» R)-:<:, . 1? (23>

processes and bis-phosphonates have been examined as

polymerizing reagents.37

30

The Wittig reaction, using 19, results, overall, in the
same product obtained by an aldol condensation using an
ester enolate followed by dehydration (eq. 24).1-33 In
contrast to aldol routes, the Horner-Wadsworth-Wittig
reaction using 19 furnishes the olefinic product
regiospecifically and in one step with a high degree of
stereochemical control about the newly-formed double bond.
Acid catalyzed dehydration of p-hydroxy esters is sometimes
complicated by formation of the 3,7-unsaturated ester.3'

Base-promoted elimination of the acetate derived from the p-

0
ll
(EtO)2PCHzCOzEt

l9

hydroxy ester overcomes this problem‘0 but introduces an

extra step.

0 0‘
- II I a:
CHzC028t + ulcn2 —-—>RIR2C-cn20028t :
-HaO

 

R1830=0820038t (24)

The position of the carbon-carbon double bond formed by
the phosphonate Wittig reaction can be predicted with a high
degree of certainty, with double-bond migrations to give
p.1-unsaturated esters having been reported only in rather

unique, isolated cases.‘1

31

The stereochemistry of the newly formed double bond is
less predictable. Unstabilized Wittig reagents generally
give 2 olefins and stabilized Wittig reagents generally give
E olefins. From early studies, it was believed that
phosphonate olefinations gave exclusively E olefins.
However, the mechanism of phosphonate olefinations has since
been established (eqs. 20-23) and the olefin product
stereochemistry depends on a number of factors. The aldol
product of a p—phosphono ester (E=COzR3) with a carbonyl
substrate has never been isolated. '-
Hydroxyphosphononitriles corresponding to 14 and 15 (eq.
21, E=CN) have been isolated and separated. Subsequent
treatment with base indicates the aldol condensation is
reversible (eq. 21) and control experiments show that
collapse of 14 and 15 (eqs. 22 and 23) is highly stereo-
specific.‘3 Although postulated, there is no evidence that
14 and 15 interconvert directly. Thus, the stereochemistry
of the pro-duct olefin depends on the relative rates of
formation and decomposition of 14 and l5.‘3 In “general,
thermodynamically-controlled olefination. reactions give E
olefins while kinetically-controlled olefinations give 2
olefins. Thus, the E/z ratio of the product olefin is
dependent on temperature (eqs. 25-27) and the ability of the

counter cation‘z' to stabilize the intermediate oxyanion

32

 

 

 

o
H t—BuOE Ph
(EtO)2PCH-CN + PhCHO : “==:: + ._a—
H -73° cu Ph’ ‘cu
3
10:90 (25)
o O
t-BuOK ph
(Et0)2PCH-CN + PhCHO .: “=:f’ -+ __,,
CH 20° cu Pia/\cu
, .
40:60 (26)
0 Ph
H - grauox
(EtO)2PCH-CN + PhCHO 4.7“=<:: . __,.
(I:H 65° cu Pia/\c"
3
50:50 (27)

(eqs. 28-30). Steric effects depend on the bulk of both the
phosphonate reagent‘3"‘ and the carbonyl substrate (eqs. 31
and 32).‘35 The bulk of the phosphonate reagent is dependent
on both the size of the side chain (eqs. 33 and 34)435 and
the size of the additional electron-withdrawing group.H

Equations 35-38 illustrate that the size of the nitrile

 

 

o
H t-BuOLi Ph
_ _"’
(Et0)2PCHCN + PhCHO ;, -__—r +
| -78° ‘cu Pin/\CN
ca
3 70-30 (23)
o
t-BuONa ph
(Et0)2PCHCN + PhCHO 2;, +
| -73° ‘CN PhACN

C83

33

 

0
ll t-BuOK p
(Et0)2PCHCN + PhCHO 4 “V + _/
CN c"
C83
10:90 (30)

group is comparable to methyl, but the carbomethoxy group
has a greater effective size than methyl. The solvent“"5
has also been shown to influence the stereochemical outcome
of the reaction.

The Horner-Wadsworth-Emmons-Wittig reaction has some
strict steric-requirements. This means that the phosphonate
anions are more reactive towards aldehydes than ketones, and
certain ketones are completely unreactive towards the
phosphonate anion. In steroidal systems, for example, the
phosphonate anion 20 reacts with 3-keto steroids but does
not react with 6-keto, 7-keto, l7-keto or 20-keto
steroids.‘° Similarly, 20 is unreactive towards 20-oxo-21-
methyl steroids but does react with 20-oxo-21-hydroxy
pregnanes. Presumably, the aldol step is reversible; but
the formation of a cyclic product drives the reaction to
completion.‘7 Simple cyclohexanones are normally reactive
towards 29, but cyclohexanones bearing substituents at the 2
position are unreactive if the substituent is constrained to
an equatorial position.‘°"° Reaction of the bridged ketone
21*with 20 gives poor yields of unsaturated ester. This is

one example example where the Reformatsky reaction, followed

34

by base-promoted elimination of the acetate, was the

preferred route to the unsaturated ester.‘9 Phosphonate

0
ll -
(Et0)2P-C-002Et + caacnzcno -—-—q. + c
(in: co,£t 02“
84:16 (31)
o ..
ll - .
(EtO)2P-C-COaEt + CHaCHzCHCHa -———- +
I | 0.5:
' C83 (:30 Ofit
33:67 (32)

olefinations -with readily enolizable aldehydes or ketones,
such as acetophenone, often give poor yields of olefins.

In comparison with the normal (phosphine) Wittig
reaction, the Horner-Wadsworth-Emmons modification offers
some significant advantages. The phosphonate reagents are
easier to prepare5°, usually by the Arbuzov reaction or the
Michaelis-Becker reaction. The former, involving reaction
of a trialkyl phosphite with an alkyl halide, is the most
commonly used route. Reagent 19, for example, is prepared
readily and in high yields from triethylphosphite and ethyl
bromoacetate. The phosphonate is easily modified by
alkylation or acylation of the phosphonate anion.51
Isolation of the olefinic product is simplified because
unlike phosphine oxides, the phosphate ester by-product 17
is easily removed by aqueous extraction. The utility of the
phosphonate olefination is attested to by the extensive

reviews recently afforded it.52

35

Standard procedures for generating the phosphonate

anion 29 use strong bases such as sodium hydride, lithium

0

ll-
(EtO);P-CHCOzEt

20 21

diisopro-pylamide, or metal alkoxides which are expensive
and/or may react with sensitive functional groups in reagent
or substrate. Since compound 19 is structurally similar to
malonate esters, it appeared to be a likely candidate for
acidity enhancement through metal cation coordination,
allowing anion formation with weak bases. Clearly, a

procedure for carrying out the phosphonate olefination under

‘0
H -
(Et0)2P-08002Et + i—PrCHO -—————.. (33)

CCHEI

0
(Et0)2P-?002Et + i—PrCHO ———___.. + (34)
C H: ‘ CO,Et
. 35:65
mild conditions would be of significant value. The

stability of the metal complex 22, formed via that route
(eq. 39), should overcome the problems associated with
readily enolizable substrates in the Wittig reactions. A
weak base promoted olefination procedure would offer

significant advantages in terms of cost and convenience.

36

The diethoxyphosphinyl group is inferior to the
carboethoxy group in increasing the acidity of alpha
hydrogens. The pH of triethylphosphonoacetate is 2.5 pH
units greater than diethylmalonate (19.2 vs. 16.7 in
dimethylsulfoxide solution).53 Although the exact nature of
charge delocalization in 22 is not well defined, the. phos-
phinyl group undoubtedly makes a contribution. The
structure and reactivity of 22 appears to be dependent on a
number of factors including solvent system, the metal cation
M, and the base used to generate 22.!H The stability of 22,
and thus the acidity of 19, is greatly influenced by the
nature of the metal ion M in 22. The pH of
triethylphosphonoacetate in the presence of potassium cation
(dimethylsulfoxide solution) has been estimated at 19.2 and
in the presence of lithium cation (diglyme solution) at
12.2.53. Presumably, the difference in acidity is
attributable to tighter chelation of lithium cation

resulting in a stronger metal-oxygen bond in the resulting

 

 

enolate. Phosphonate 23, for example, which has no
0 .
fl NaH/THF P
(Et0)2PCHzCN + PhCHO ._.. + IFT\
20° C" Pb cu
85:15 (35)
0
ll NaH/THF P" P"
(“maniacs + PhCHo h. R + \u=<‘2
20°
08: C" N

60:40 (36)

37

carboethoxy group, is expected to be a poorer chelating
agent for metal cations and its acidity shows a less-
pronounced dependency on the metal cation present (4 pH
units difference for the lithium and potassium complexes).
The selection of the metal ion M in 22 is important. While
22 must be stable enough to form in useful concentrations
with weak bases, it must possess sufficient reactivity
towards carbonyl compounds. While more electropositive
metals (sodium, potassium) make 22 more nucleophilic, less
electropositive metals (lithium, magnesium) appear to result
in a more facile cycloelimination step (eqs. 20 and 21) in
Wittig olefinations.5°

The addition of metal salts has been reported to retard
the phosphine Wittig reaction by complexing with an
'intermediate in the reaction57 but there are reports that
some nucleophilic substitutions at the phosphoryl group,
similar to the one in the cycloelimination step of the
phosphonate olefination reaction, are accelerated by added
cations.58

Nearing completion of this study, an independent report
appeared in the 1iterature59 describing a procedure for
carrying out the Horner-Wadsworth-Emmons-Wittig reaction
using the relatively expensive tertiary amine bases
diisopropylethylamine (pE~10) or diaza[5.2.0]bicycloundec-7-

ene (DBU, pK~ll.6) and lithium chloride. The method was

38

0

ll
(EtO)2PCHa

23

demonstrated with three aldehydes that were readily
enolizable or base sensitive. The method proved successful

where the standard strong base conditions failed.

 

 

0
|| NaH/THF 9" ”\n/COM
(EtO):PCHzCOaMe + PhCHO : R +
2°° com.
97:3 (37)
0
ll Nail/THE P P"
(BtO):PCHCOzMe + PhCHO 4:7 + \E*<
0
('28: 20 CO2". C01".
9723 (38)

nmnnsslnmlnnmnnsnml

The ability of triethylphosphonoacetate to undergo the
Horner-Wadsworth-Emmons-Wittig reaction using triethylamine
and a number of metal salts was examined. Cyclohexanone was
stirred with triethylphosphonoacetate and triethylamine in
the presence of a number of metal halides (eq. 39) in
tetrahydrofuran solution (Table 6). After aqueous workup,
the presence of the expected product, ethyl cyclohexylidene
acetate, 24 was determined by gas chromatography. Magnesium

and lithium halides are effective in promoting the reaction.

39

I“flll&.illnthulobeehiml-nmetdthIE!bathe
lhmmmlmroflinmlllflidmhfl

I

III. + 19 + MX

 

EtaN CHCO,Et
rnr. 25°C

 

 

Entry MX (mmol) Yield, 3°
1 none 0
2 LiCl (10) 19 (50)
3 LiBr (10) 39 (85)
4 MgCl: (10) 52 (86)
5 MgBrz (5) 50 (48)
6 MgBrz (10) 62 (85)
7 MgBrz (20) 70
8 MgBrz (10) 90c
9 MgBrz (10) 0d
10 Mel (10) 0
ll ZnClz (10) 0.5
12 A101: (10) 0
13 FeCla (10) 0
14 CuClz (10) 0

 

8)

b)

e)
d)

Reaction at 25°C with 10 mmol cyclohexanone, 10 mmol EtaN,
10 mmol 19, 10 mL THE for a period of 3 h, except where

noted.

Yield of ethyl cyclohexylidene acetate determined by 18
NMR analysis. Yields in parentheses are for 24 h reaction

periods.
20 mmol EtaN was used.

No EtaN was used.

40

Control experiments demonstrate the need for both the metal
salt and the base. The magnesium halides appear to be more
effective promoters of the reaction than lithium halides

(Table 6, entries 2.3.6.7). Since lithium enolates are

 

,M\
0 ’ 0
II M... 9 R3 N Et0\ %
(EtO):PCHzCOzEt : 1' 39
E OEt ( )
19 22
generally more reactive than magnesium enolates. this

difference may be due to the magnesium enolate (22. M=Mg)
forming faster than the lithium enolate (22, M=Li). The
reaction appears to require stoichiometric amounts of metal
salt. Excellent yields of ethyl cyclohexylidene acetate 24
are obtained over relatively short reaction times with an
excess of triethylamine. However, excellent yields may also
be obtained using stoichiometric amounts of base over longer
reaction times (24 hours).

The reaction of triethylphosphonoacetate with
cyclohexanone of benzaldehyde (eq. 40) was examined (Table

7) using a variety of solvents (acetonitrile. diethyl ether.

'0
EtaN
19 + RCR1 + Mx =5 RRIC=08002Et (40)
solvent, 25°C

 

41

'flflfll'fi. 8mmfiian¢flf0mflxmul(kmpommhlwifli192n|a
lmdm¢y¢MFSdhnnhmfi

 

 

 

fl EtaN
RICH2 + MX + 19 :: R1R20=CHCOzEt
solvent, 25°C
3 hours
0
Solvent "
Yield, 2° Rlch MX Yield".c
Acetonitrile Benzaldehyde LiCl 77
LiBr 93
MgClz 15
MgBrz 71
Diethyl ether LiCl 77
LiBr 71
MgBra 80
Tetrahydrofuran LiCl 86
LiBr 96
MgBr 81
Methylene chloride LiCl 56
LiBr 70
MgBrz ' i 47
Dimethylformamide LiBr 25
MgBrz 10

Benzene LiBr 93

42

{hideih cuMfinmui

 

 

O
Solvent "
Yield, 3° R1CR2 MX Yield°.c
Benzene Cyclohexanone MgBrz 82
Ethanol MgBrz 30
Water MgBrz 0

 

a) Reaction at 25° with 10 mmol carbonyl substrate, 10 mmol

19. 10 mmol EtaN.

10 mL solvent for a period of 3 hours.

b) GLC yields of ethyl cinnamate or ethyl cyclohexylidene

acetate.

c) Only trans ethyl cinnamate detected.

43

tetrahydrofuran. methylene chloride. dimethylformamide,
benzene. glyme, ethanol). Only lithium chloride was
completely soluble in any of the solvents. However,

addition of triethylphosphonoacetate invariably resulted in
complete dissolution of the salts and homogeneous solutions.
Addition of triethylamine to solutions containing magnesium
halides resulted in instantaneous formation of a
precipitate. Addition of triethylamine to solutions
containing lithium halides resulted in no observable change
in the reaction mixture. Addition of the carbonyl substrate
to a solution of 19, lithium halide and triethylamine re-
sulted in the rapid formation of the same precipitate.

Most common solvents give satisfactory results. It is
interesting that while yields are low. the reaction occurs
in some protic solvents (ethanol). It is noteworthy that
the reaction can be carried out in methylene chloride, a
solvent incompatible with the strong bases usually employed
in the Horner-Wadsworth-Wittig reaction. The low
conversions realized in dimethylformamide solution probably
reflect the high coordination power of the solvent for metal
ions, disfavoring formation of the internally coordinated
enolate 22.

The reaction of a variety of aldehydes and ketones with
triethylphosphono-acetate in the presence of either
magnesium bromide or lithium bromide was conducted on a
preparative scale with the results shown in Table 8.

Excellent yields are obtained with aldehydes or the reactive

44

nuu31a. Bunniescdfa‘fludetycfl?0udmmwlIkmnommhtwfluil9:u:
thelhemmmx:ofihdeflhdamhmsandllnmlliflidmh!

 

 

 

fl EtaN
R1082 + MX + >VR1RZC=08002Et
solvent, 25°C
3 hours
Carbonyl Metal
Compound Halides (solvent) Product Yield, 3°
CsHsCHO LiBr (CHaCN) CsHsCH=08002Et 84
MgBrz (THF) 85
(caa)2cncao .LiBr (CHaCN) (C83)2CHCH=08002Et 80
MgBrz (THF) 40
ngeHiacHO LiBr (CHaCN) ngsfiiacH=08002Et « 75
agar. (THE) 100
cyclohexanone LiBr (CHaCN) (CHz)sC=CHCOzEt 85
cyclopentanone LiBr (CHaCN) (CHz)4C=08002Et l5
CeHsCH=CHCHO LiBr (CH:CN) CsHsCH=CHCH=08002E 65
08300083 LiBr (CHaCN) 0
MgBrz (THF) 0
CsHsCOCHa LiBr (CHaCN) 0
MgBra (THF) 0

 

a) Reaction at

25°C for 12 h. 25 mmol scale (carbonyl
compound:l9:EtaN:metal halide-l:l:l.l:l.2).

b) Isolated yield, based on weight of distilled product.

45

ketone cyclohexanone. However, simple methyl ketones such
as acetophenone or acetone fail to react.

In all cases. the unsaturated ester product was the
conjugated E isomer. No p.7-unsaturated ester was detected
and none of the corresponding 2 isomer was detected by CC or
18 NMR analysis under conditions judged sufficient to detect
0.53 of the other isomers. Similar high E/Z ratios have
been reported for the Horner-Wadsworth-Emmons Wittig
reaction.5° The procedure appears to be somewhat tolerant of
less than rigorous exclusion of moisture from the reaction
flask. In a number of experiments, anhydrous metal salts
were weighed in the atmosphere and solvents and
triethylamine could be used directly from a freshly opened
bottle without significant losses (<58) in yields' of
olefination product.

We attempted to verify that the reaction was proceeding
via production of the phosphonate anion. When magnesium or
lithium bromide is added to triethylphosphonoacetate, its 18
NMR spectrum exhibits a small downfield shift (0.16 ppm) of
the alpha hydrogens. When triethylamine is added to a
solution of triethylphosphonoacetate. the alpha hydrogen
doublet (JP-H=24Hz) coalesces into a broad singlet. A
mixture of triethylphosphonoacetate. lithium bromide and
triethylamine exhibits a pair of quartets around 3.2ppm and
2.4ppm. The alpha hydrogens are no longer observable.) This
is indicative of an equilibrium with a rapid proton exchange
between the enolate and triethylamine. 18 NMR analysis of a

heterogeneous mixture of triethylphosphonoacetate, magnesium

46

bromide and triethylamine exhibits one quartet at 5 3.2
indicative of quantitative ammonium salt formation. No
signal is observed for the alpha protons of
triethylphosphono-acetate. Filtration of a mixture of
triethylphosphonoacetate, triethylamine and magnesium
bromide gave. after evaporation of all volatile components.
a precipitate identified as triethylamine hydrobromide and a
white solid exhibiting the following 18 NMR: 5 4.4 (m). 3.3
(d. J=20 Hz). 1.4 (m). Comparison of these spectra with the
18 NMR spectrum reported for the calcium enolate°4° of 19
and examination of solvent effects on the chemical shifts in
the lithium enolate“c of 19 led to the conclusion that the
white solid was the magnesium enolate of 19 (22, M=Mg).

Opposed to this is a report59 that 19. in the presence
of lithium chloride (LiCl) and diazabycyclo[5.4.0]undec-7-
ene (080) in acetonitrile-63 solution, gives a 31P NMR
spectrum that exhibits two a1P signals. The authors’
interpretation is that two LiCl/l9/DBU complexes exist and
that one of them is in-terconverting to 22 (M=Li) at a rate
comparable to the NMR time scale.

Procedures for the Horner-Wadsworth-Emmons Wittig
reaction with aldehydes using lithium bromide or magnesium
bromide and triethylamine give yields of unsaturated esters
comparable to those procedures using other. stronger
bases.52°'5° The negative results obtained with ketones
using this procedure. while disappointing. are not entirely
unsatisfactory. With a few exceptions, especially

cyclohexanones. the phosphonate Wittig reactions using 19

47
with ketones are difficult. There are reports in the
literature that metal salts inhibit the normal Wittig
reaction, presumably by forming unreactive complexes with an
intermediate in the reaction. However. our results with
lithium and magnesium salts have been entirely satisfactory.
This new procedure for the phosphonate Wittig reaction seems
especially useful for large scale preparations where
triethylamine possesses significant handling and cost

advantages over other bases.

MNflflUJLS

Tetrahydrofuran was distilled from sodium/benzophenone
just prior to use. Acetonitrile and triethylamine were was
distilled from calcium hydride. Dimethyl formamide was
distilled from phosphorous pentoxide. Diethyl ether was
taken directly from a freshly opened can of anhydrous ether.
Methylene chloride was taken from a freshly opened bottle of
anhydrous methylene chloride. Lithium bromide (Aldrich
Chemical Company, 99+X) was dried in an abderhalden flask
over refluxing xylene at 0.3 torr. Magnesium bromide was
prepared from dibromoethane and magnesium metal and dried
under vacuum at 15000.30 Zinc chloride (Aldrich Chemical
Company, 98%) was dried with thionyl chloride followed by
removal of excess thionyl chloride under high vacuum.61
Lithium chloride (Aldrich Chemical Company, 99%) was dried
in an abderhalden flask over refluxing xylene at 0.3 torr.
The remaining metal salts were obtained as anhydrous
materials from commercial sources. All metal salts were
stored in a dessicator and transferred under argon in a
glove bag. Triethylphosphonoacetate was prepared from

ethylbromoacetate and triethylphosphite.5°

MEEIDStNPAMNMmflS
18 NMR data were obtained on a Varian T-60 spectrometer

at 60MHz. Chemical shifts are reported on the delta scale

48

 

49

relative to an internal tetramethylsilane standard. Gas
chromatographic analyses were performed on a varian 920
chromatograph equipped with a 6 ft. X 0.25 in. stainless
steel column packed with 153 SE-30 on Chromasorb W. 00
yields were obtained using n-alkanes as internal standards.
NMR yields were obtained using acetophenone as internal

standard.

Huxmdmnafor te(fl. an ion.

The following procedure. with modification of scale. is
representative of the procedure used to obtain the results
in Tables 5-7. A 50 mL flask with a septum inlet and
magnetic stirrer was flame dried under argon. Anhydrous
metal salt (30 mmol) was weighed in a glove bag and
transferred under a stream of argon to the flask. Solvent
(25 mL) and triethylphosphonoacetate (25 mmol. 5.54g) were
added and the mixture stirred 5 minutes. Triethylamine (28
mmol, 3.9 mL) was added and the mixture stirred an
additional 10 minutes. the carbonyl compound was then added.

and the reaction mixture stirred overnight. After quenchins

 

with dilute aqueous HCl, the reaction mixture was extracted
with ether (3 x 25 mL). The organic extracts were combinerd

and dried over magnesium sulfate.

MW

To the combined extracts described above was snided a

known quantity of acetophenone as internal 18 NMR standard'

 

 

50

The solvent is removed in vacuo. Yields are based on the

relative integration of product olefin to acetophenone.

(knead.anmdmnaumaitotfigggztkflvam:Signal.

To the combined extracts described above was added a
known quantity of n-alkane as internal CC standard. The

solution was then analyzed by CC for product olefin.

Gammmllfincahmetfiu'theImohndnn«MfUhNMHumtuiEshum.

The combined extracts described above were filtered and
the solvent removed in vacuo. The crude product was puri-
fied by short-path distillation.

Ethyl cinnamate'!‘2 was prepared from 19 and
benzaldehyde: b.p. 75°C (0.2 torr); 1H NMR (CDCla): a 1.3
(t. 3 H), 4.18 (q, 2 H), 6.33 (d, l H), 6.7—7.7 (m, 6 H).

Ethyl-4-gethyl:2-pentenogtg°2 was prepared from 19 and
isobutyraldehyde: b.p. 60°C (30 torr); 1H NMR (CDCla): 5
0.96-2.48 (m. 9 H), 2.4 (septet, 1 H). 4.16 (q. 2 H).

Ethyl-g-noneggtg93 was prepared from 19 and
heptaldehyde: b.p. 72°C (2 torr); 1H NMR (CDCla): 6 0.8-1.1
(m, 3 H), 1.1-1.7 (m, l H). 1.9-2.5 (m. 2 H). 4.2 (q, 2 H),
5.75 (d. 1 H), 6.95 (m, l H).

Ethyl cyclohexylidene acetate54 was prepared from 19
and cyclohexanone: b.p. 50°C (0.2 torr); 18 NMR (CDCle): 6
1.23 (t, 3 R), 1.4-1.8 (m. 6 R), 1.9-2.5 (m. 2 H), 5.5 (s, 1

H).

51

Ethyl cyclopentylidene acetate°5 was prepared from 19

 

and cyclopentanone: b.p. 85°C (10 torr); 18 NMR (CDCla): a
1.25 (t, 3 H). 1.8 (m, 6 H), 2.5 (m. 2 H). 4.2 (q. 2 H), 5.8
(m. 1 H).

Ethyl-S-phenyl-g.4-pentadienoatg°° was prepared from 19
and cinnamaldehyde: b.p. 90°C (0.2 torr); 1H NMR (CDCla): 6
1.33 (t, 3 H), 4.2 (q, 2 H), 5.95 (d, l H), 6.7-7.6 (m, 8

H).

1 H Ill Stuw - of Trietliylphosphmoacetate/Lithit-
lhmmifiwflrhmnwimmhn:Mhdmme.

Triethyl phosphonoacetate (1.0 M. CD3CN solution)
exhibits the following 1H NMR spectrum: 5 4.1 (m). 3.10 (d.
J=22 Hz). 1.3 (t), 1.2 (t).

Triethylphosphonoacetate (0.5 mmol, 0.09 mL) was added
to CDaCN (0.5 mL) and lithium bromide (0.5 mmol. 0.044g)
stirring under argon. The homogeneous solution was
transferred to an NMR tube that had been flushed with argon.
The solution exhibits the following 18 NMR spectrum: a 4.33-
3.83 (m), 3.10 (d, J=2282). 1.30 (t). 1.20 (t).

Triethylphosphonoacetate (0.5 mmol, 0.09 mL) was added
to an NMR tube containing triethylamine (0.5 mmol. 0.07 mL)
in CDaCN solution (0.5 mL). The solution exhibited _the
following 1H NMR spectrum: 5 4.1 (m), 2.93 (br s), 2.4 (q).
1.4-0.9 (m).

Triethylamine (0.5 mmol) was added to an NMR tube

containing a solution of a solution of lithium bromide (0.5

52
mmol) and triethylphosphonoacetate (0.5 mmol) in CDaCN (0.5
mL). The sample exhibited the following 1H NMR spectrum: 6

4.45-3.97 (m), 3.2 (q), 2.4 (q), 1.4-0.83 (m).

1 H n Study of Triethylphosphonoacetate/Triethylnine/
Hmullflulnnmfldeltbdmre.

Triethylamine (0.07 mL, 0.5 mmol) was added to an NMR
tube containing a solution of triethylphosphonoacetate (0.5
mmol. 0.09 mL) and magnesium bromide (0.5 mmol, 0.052g) in
CDaCN (0.5 mL). A white precipitate formed instantly. The
sample exhibited the following 18 NMR spectrum: 6 4.45-3.9

(m), 3.2 (q). 1.4-0.83 (m).

Isolation of the Ihgnesit- Enolate of Triethyl-
phaummnuuxmaum

A 50 mL flask fitted with a septum inlet. gas inlet
tube with bubbler, magnetic stirrer and filter stick was
flame dried under argon. The flask was charged with 5 mmol
(0.92g) of anhydrous magnesium bromide 'which had been
weighed and transferred under argon. Anhydrous diethyl
ether (5 mL) and triethylphosphonoacetate (5 mmol. 0.97 mL)
was added via syringe and the mixture stirred 5 minutes.
Triethylamine (5 mmol. 0.70 mL) was added dropwise and a
white precipitate formed instantaneously. After stirring
ten minutes, the mixture was filtered. The filtrate was
concentrated under high vacuum overnight and gave a white
solid. The precipitate was dried in a vacuum dessicator

overnight.

53
Filtrate residue: 1H NMR (CDCla): a 4.5-4.2 (m). 3.3
(d, J=ZOHz), 1.5-1.1 (m).
Precipitate: 18 NMR (CDCla): 5 3.2 (q), 1.45 (t); m/s

(m/e): 101 (-HBr). 86 (base). 82 (HBrBl), 80 (HBr79).

 

BIBLIM

10.

11.

12.

strum

For the use of lithium dialkylamides in organic
synthesis, see: (a) House, 8. ”Modern Synthetic
Reaction”. W. A. Benjamin, Inc., Menlo Park, 293
ed., 1912. (b) Chevrot. 0.; Perichon, J. Bull. Soc.
Chim. D‘. 1977, 5-6 pt. 1, 421.

(a) Roughton. R. P. ”Metal Complexes in Organic
Chemistry”. Cambridge University Press, Cambridge
1979. (b) Mehrorta. R.; Gear, 1). P. "Metal '-
Diketonates and Allied Derivatives”, Academic press,
San Francisco, 1978.

Rathke, M. W.; Deitch, J. htmbetb'an Lett. 1971, 12,
2953.

Reutov, O. A.; Beletskaya, I. P.; Butin, E. P. "CH-
Acids”, Pergamon Press, Oxford. 1918.

(a) Collman. J. R.; et. a1. Chem. Mind. I”, 1213.
(b) Collman. J. P.; et. al. J. Org. 0b.. 1m, 28,
1449. (c) Collman. J. P. Angew. Chem. Int. Ed. 1955.

4, 132.

Augustine, R. L. "Carbon-Carbon Bond Formation”.
Marcel Dekker. Inc., New York, 1919, Vol. 1, p.250-
263.

COwan, P. J. ”The Acylation of Trimethylsilyl Acetate",
Ph.D. Dissertation (1983). Michigan State University.

Wissner, A. J. Org. Chem. 1979, 44, 4617.

Euo, Y. N.; Yahner, J. A.; Ainsworth, C. J. Am.
Chem. Soc. 1911, 3?, 6321.

Crombie, L.; Hemesley, P.; Pattenden. G. htrabedron
Lott. 1m, 3021. ‘

Herbert. R. 8.; Jackson, E. B.; Nicholson. I. T. J.
Chem. Soc. Chem. 62-. 1976, 450.

Pollack, R. M. in Gandor R. D.; Schowen. R. I.

"Transition States of Biochemical Processes". Plenum
Press, New York, 1918, p.46.

54

13.

14.

15.
16.

17.
18.
19.

20.

21.

22.

23.

24.

25.
26.
27.

28.

29.

55

Logue, M. W.; Pollack. R. M.; Vitullo. V. P. J. Am.
Chem. Soc. 150, 72, 1231.

(a) Oshry, L.; Rosenfeld. S. M. Org. Prep. and
Proceed. 1-, 251. (b) Brown. J. M. ”Oxo Carboxylic
Acids”. Sec. 9.5 in Barton, 0.; Ollis, W. D.
”Comprehensive Organic Chemistry". Pergamon Press,
Oxford. 1979.

Stenhagen, E. Ark. 5m. 1951, 3, 381.

Mitz. M. A.; Axelrod, A. E.; Hoffman, K. J. Am. Chem.
Soc. w. 72, 1231.

Logue, M. W. J. Org. Chem. 1914, 39, 3455.
Stiles. M. J. Am. Chem. Soc. 159, 81. 2598.

Bottachio, 0.; Chiusoli, G. P. Chen. Calm)”.
6180 '

Bottachio, 0.; Chiusoli, G. P.; Felicioli. M. G.
Gazz. Chin. Ital. 1913, 103, 105.

Tirpak, R. ”Acylation of Ketone Enolates and Their
Trimethylsilyl Derivatives”, Ph.D. Dissertation, 1982.
Michigan State University.

Erapcho. A. P.; Eashdan, D. 8.; Jahugen, E. G. E..
Jr. J. Org. Chem. 1917. 42, 1189.

(a) Cowan, P. J.; Rathke. M. W. Syntb. Om. 1m,
13, 183. (b) Moreau. J. L.; Couffignol. R.; Arous-
Chtara, R. mtrabea‘an 1931, 307.

Barnick, J. F. W. R.; van der Dean, J. L.;
Bickelhaupt. J. Synthesis 1979, 10. 787.

See Reference la. p.763.
Staab, R. A. Chem. Ber. 157, 90, 1321.

Tarbell. D. 8.; Rice. J. A. J. Org. Chem. 157. 22.
245.

Schmidt, 0.; Schwochau, M. Mastscb. Chem. 1”, w,
1492.

Miyano, M.; Dora, 0. R. J. Org. Chem. 1912. 37, 1818.

30.

31.

32.
33.

34.

35.

36.

37.

38.
39.
40.

41.

42.

56

(a) Plantema, O. 0.; deKonig, R.; Huisman. H. O. J.
Chem. Soc. Perk. 11918, 304. (b) Taub, D.; et. a1.
Chem. Cm. 1970, 1258. (c) Corey, E. J.; et. al.

J. Org. Chem. 1912, 37, 3043.

(a) Dahm. E. R.; Trost, B. M.; Roller. R. J. J. Am.
Chem. Soc. 187, 39. 5292. (b) Schultz. R.; Sprung, I.
Angew. Chem. Intern. Ed. Eng]. 1”, 8. 2 71 . (c)
Braun, B. R.; et. al. J. Econ. Kata-oi. 1m, 6']. 866.

(d) Cavill, 0. W. R.; Laing. D. 0.; Williams, P. J.
Aust. J. 62.. 19a, 22, 2145. (e) Cochrane, J. 8.;
Hanson, J. R. J. Chem. Soc. Pent. [1912, 361.

Vig, O. R.; Anand, R. 0. Indian J. 0b.. 1m, 1111.

Surmatis, J. D.; Thommen. R. J. Org. Ohm. 1m. 34.
559.

(a) Clark, R. D.; Kozar. L. 0.; Heathcock, L. H.
Sntn. (lo-In. 1915. 5, l. (b) 0rieco, P. A.;
Pogonowski, C. S. Syntbais I913, 425. (c) Benrick. 0.
A.; et. al. J. Am. Chem. Soc. 1m. 90, 5926.

Lehman, H. 0.; Wiechert, R. Angew. Chem. Int. ht.
Ed. I”. 7, 300.

Sturtz, 0. Bull. Soc. Chim. h. I”, 2349.

(a) BASE A0 0erm. Pat. 1,108,219 (1959) and 1,112,072.
(b) Drefahl, 0.; Roerhold. R. R.; Wildner, H. 8.:
0erm. Pat. 51,436 (Chem. Abstr. 66:105346h).

Rathke, M. W. Organic Reactions 1975, 22, 423.
Shriner, R. L. Organic Reactions 1942, I. 1.

Fung. K. R.; Schmalzl. K. J.; Mirrington. R. N.
mtrabedran Lett. 1914, 1733. .

(a) Quin. 0. D.; et. al, J. Org. 05.. 1911, .96. 1495.
(b) Wadsworth. W. 8.. Jr.; Emmons, W. D. Organic
Synthesis 1”, 45. 44. (c) Sunberg. R. J.; Bolcombe,
I. 0. J. Org. Obs. m. 34. 3273.

(a) Redjal. A.; Seyden-Penne, J. ktmbedran Lett. 1914,
1733. (b) Lefebvre, 0.; Seyden-Penne J. Chem. Soc.
flea. O... 1970. 1308. (c) Lavielle, 0.;
Carpentier, M.; Savignac. P. 179th Lett. 1973.

173. (d) Deschamps. 8.; Lefebvre, 0.; Seyden-Penne, J.

htrabedran 1912, 28, 4209. (e) Danion. D.; Carrie. R.
htrabedroo 1912. 28, 4223.

 

43.

44.
45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

57

(a) Jones, 0.; Maisay. R. F. J. aha. Soc. One.
Oman. 1”, 543. (b) Einstle. J. R.; Mandanas, B.
Y. Chem. Oo-un. m. 1699. (c) Sasaki, K. Bull.
Chem. Soc. Jap. m, 39, 2703.

Bottkin-Strzalko. T. ktrabedran, 1913. 29, 4199.

(a) Normont. R.; Sturtz, 0. Cent. Hand. 6'. m. 256,
1800. (b) Gallagher. 0.; Webb. R. L. Synthesis 1914.
122. (c) Deschamps. B.; Lefebvre, 0.; Redjel, A.;
Seyden-Penne, J. ktrabecb'an 1913, 29, 2437. (d)
Bottkin-Strzalko. T.; Seyden-Penne. J. htrabedron
Lett. 1912, 1945. (e) Danion, D.; Carrie. R.
Tetrahedron 1972, 28, 4223.

Bose. A. R.; Dahill. R. T., Jr. J. Org. Chem. 1”. 30.
505.

Fritsch. W.; Stache, V.; Ruschig. 8. Ann. I”, M,
1950 .

Harding. E. R.; Tseng. C. Y. J. Org. mam. 1915. 40.
929. -

Merrington. R. N.; Scmalzl, R. J. J. Org. (mam.
1912, 37. 2871.

Kosolapoff, 0. M. ”Organophosphorous Compounds".
Wiley. New York, 1950, Chap. 7.

(a) Clark. R. D.; Kozar. L. 0.; Heathcock. C. H.
Synthesis 1975, 635. (b) Grieco, P. A.; Pogonowski, C.
S. J. Am. Chem. Soc. 1913, 85, 3071.

(a) Wadsworth, W. 8., Jr. Organic Reactions 1919. 25,
73. (b) Walker, B. J. ”PO-Activated Olefinations”,
Chap. 3 in Cadogan, J. I. 0. "Organophorous Reagents
in Organic Synthesis”, Academic Press. London, 1979.

(a) Petrov, E. S.; et. a1. Zb. wean. nimii 1915, 45,
1529. (b) Petrov, E. 8.; et. a1. Izv. Akad. Abuk.
SSSR, Ser. nim. 1916, 534.

(a) Kirilov, M.; Petrov, 0. Naoatscb. Chem. 1912. 103,
1651. (b) Bottin-Strzalko, T.; Seyden-Penne. J.;
Simmonin, M. P. J. Chem. Soc. Chem. (Ia-n. 1916.-
905. ’

(a) Mesyats. S. P.; et. al. In. Akad. Abuk. SSSR,
Ser. Dim. 1914, 2497. (b) Mesyats. S. P.; et. al.
Izv. Akad. Mink. SSSII, Ser. nim. 1914. 2489.

56.

57.

58.

59.

61.

62.

63.

64.

65.

Redjel,

16; 1

Schlosser,

650.

(a) Wadsworth, W.

733.

A.;

Seyden-Penne, J.

58

Anna» Chan. Jbt.

Jr.;

Jr. J. 011. Chen. 1973, 38, 2921.

Ibhnwmdnuvzcmh IQWL

Ed. Eng. 1963, 7,

Larsen,
(b) Wadsworth, H. 8.,

el.

8.; Horten, R.

Jbtndmdhxllehh

J. ammuniet. aha»

"Inorganic Synthesis”, Riley, New York,

D O J. Ale Chas

BuLL 8mm can»

BuLL chum 8mm

Mesa-une, 8.; Roush, W. R.; et.
13:, 25, 2183.

Ashby, B. 0.; Arnott, R. C.
1988, Lfl 1”

Pray, A. R.

1973, 0011. Vol. 7, 9.1536.
Wadsworth; W. 8., Jr.; Ennons, W.
Soc. 1931, w, 1733.

Rory, 8.; Lin, D. J.; Fetizon, M.
fr. M1, 1298.

Tekenishi, R.; Fijiwnre, K.; Ohte,
Jap. 1‘, 35, 1498.

Richerdt, 0.; Renee, P.; Bichler,

10%

1144.

(men. Bun.1987,

L.

169 0559

1293 03

3