MET ALATED ORGAHOBORANES

Thesis for the Degree of Ph. D.
iiCHéGAN STATE UNSVERSITY
RONALE KM
1975

‘1
3
\f

LIBRARr

{410114an State
0:16???th

     
       

This is to certify that the

thesis entitled

Metalated Organoboranes

presented by

Ronald Kow

has been accepted towards fulfillment
of the requirements for

 

Ph.d. degree in Chemistry
Major professor
Date I 7 S

0-7639

n\r;r‘\ -2“:

   

F BINDING BY ‘
HUAG & SUNS'
800K BINDERY INC.
LB! may BINDERS
summer rumm- HI

 
      

ABSTRACT

METALATED ORGANOBORANES

By

Ronald Kow

The formation of metalated organoboranes by base-promoted a-proton
removal from organoboranes has never been previously reported. In
large part, this must be due to the tendency of bases to coordinate to
the boron atom. The use of a sterically-congested base, lithium 2,2,6,6-
tetramethylpiperidide, overcomes this difficulty, and with the use of
this base metalated organoboranes can be prepared from trialkylboranes.
The metalated derivatives react with alkyl halides to give alcohols
after oxidation, and with cyclohexanone to give alkenes after elimination.

The yields of metalation with the trialkylboranes seldom exceeded
50 percent and a search for more acidic boron compounds was undertaken.
Vinyldialkylboranes (from hydroboration of alkynes) were found to
react with lithium 2,2,6,6-tetramethylpiperidide to furnish the
metalated derivatives in good yields. The reactions of these
metalated derivatives with a variety of electrophilic reagents were
investigated.

Attempts were made to metalate other boron compounds. The results
of the studies with diphenylmethylborane, geminal diboraalkanes, and
o-phenylene methylboronate are presented.

The reaction of trimethyl borate with lithium t-butylacetate gives

Ronald Kow
the salt XXXIX. The reactions and properties of this salt were

investigated.

Li

(CH3O)3B-CH2C02——+—-

XXXIX

METALATED ORGANOBORANES

By

Ronald Row

A THESIS

Submitted to

Michigan State University

in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

1975

To

Helen, With Love

ii

ACKNOWLEDGEMENT

The author wishes to extend his appreciation to Dr. Michael
W. Rathke for his ideas, assistance and time throughout this project.
Thanks are also given to Dr. William H. Reusch for serving as Second

Reader and for his many helpful comments.

The financial support of Michigan State University and the

National Science Foundation is gratefully acknowledged.

The Yellow Menace would also like to thank all the ”haoles" in
Dr. Rathke's group and on the fifth floor for making life more

enjoyable and his stay at M.S.U. longer.

And finally the author wishes to thank Helen for the typing
of this thesis, for her love and encouragement, and.....for being

HELEN !!

iii

TABLE OF CONTENTS

CHAPTER I

THE PREPARATION AND REACTIONS OF METALATED TRIALKYLBORANES

Page
INTRODUCTION . . . . . . . . . . . . . . 2
Literature Review . . . . . . . . . . 3
Discussion of the Problem and an
Outline of the Research . . . . . . . . 23
RESULTS . . . . . . . . . . . . . . . 15

Initial Studies Using PMR . . . . . . . . 15
Use of Benzyl Chloride as 3 Probe . . . . . 16
Use of Deuterium Oxide as a Probe . . . . . 21
Reactions of Metalated B-methyl-BBN . . . . 22

Formation of Metalated Derivatives
from Other Organoboranes . . . . . . . . 27

DISCUSSION . . . . . . . . . . . . . . 28

Use of Sterically Hindered Reagents

in Reactions of Organoboranes . . . . . . 28
Steric Effect of the Bases . . . . . . . 31
Solvent Effects . . . . . . . . . . . 33
Reaction of Metalated Organoboranes . . . . 34
EXPERIMENTAL . . . . . . . . . . . . . . 38
I. General . . . . . . . . . . . . . 38
Spectra . . . . . . . . . . . . . 38

iv

TABLE OF CONTENTS - Continued

Page
Gas Chromatography . . . . . . . . . . 38
II. Materials . . . . . . . . . . . . 38
Handling of Materials . . . . . . . . . 38
Amines . . . . . . . . . . . . . . 39
Preparation of 3-chloro—3-methyl-l-butyne . . 39
Preparation of 2-t-butylamino-2-methylbutane . 39
Preparation of 9-borabicyclot3.3.l]nonane , , 40
Preparation of B-methyl-9-
borabicyclo[3.3.1]nonane . , , . , , , , 40
Preparation of B-cyclOpentyl-9-
borabicyclo [3 . 3 . l] nonane . . . . . . . . 41
Triethylborane . . . . . . . . . . . 41
Cyclohexanone . . . . . . . . . . . 41
Alkyl Halides . . . . . . . . . . . 41
III. Bases . . . . . . . . . . . . . 41
Preparation of lithium 2,2,6,6-
tetramethylpiperidide . . . . . . . , . 42
IV. Studies with Metalated Organoboranes . . . 42
PMR Studies of the Reaction Mixtures . . . 42
Deuterium Oxide Quenching of the Reaction
Mixture of B-methyl-BBN . . . . . . . . 43
Deuterium Oxide Quenching of the Reaction
Mixtures of B-cyc10pentyl-BBN . . . . . . 43
Stability of B-methyl-BBN to Reaction
Conditions and water . . . . . . . . . 44
Removal of Amine by Distillation
from Reaction Mixture . . . . . . . . . 44
V. Preparation and Reactions of
Metalated Organoboranes . . . . . . . . 45

TABLE OF CONTENTS - Continued

Page
Preparation of metalated B-methyl-BBN . 45
Reaction of Metalated Organoboranes
with Alkyl Halides 45
Reaction of Cyclohexanone ‘with Metalated
B-methyl-BBN 46

CHAPTER II

THE PREPARATION AND REACTIONS OF METALATED VINYLBORANES

INTRODUCTION . . . . . . .
RESULTS 0 O O O O O O O O C O O O 0 O I 0
Synthesis of Vinylboranes

Studies with 2-1ithio-2-phenyl-4,
4-dimethylpentane. . . . . .

Initial Studies of Vinylboranes with LiTMP . . .
Quenching of the Reaction Mixture with Water.
Alkylation of Metalated Vinylboranes . . . . .

Reaction of Metalated Vinylboranes
with Trimethylchlorosilane.

Reaction of Carbonyl Compounds with
Metalated Vinylboranes . . . . . . .

Miscellaneous Studies . . . . . . . . . .
DISCUSSION.

Effect of Different Bases . . . . . . .

Sites of Reactions of Metalated Vinylboranes.

Preparation of Allyldialkylboranes .

Comparison of Metalated Vinylboranes to
Metalated Trialkylboranes .

vi

48

SO

50

51

51

54

56

59

59

63

66

66

67

68

7O

TABLE OF CONTENTS - Continued

Page
EXPERIMENTAL . . . . . . . . . . . . . . . 71
I. Materials . . . . . . . . . . . . . . 71
Alkyl Halides . . . . . . . . . . . . . 71
Alkenes . . . . . . . . . . . . . . . 71
Alkynes . . . . . . . . . . . . . . . 71
Carbonyls . . . . . . . . . . . . . . 71
Trimethylchlorosilane . . . . . . . . . . 71
II. Vinyldialkylboranes . . . . . . . . . . 72
Preparation of Dicyclohexylborane . . . . . . 72
Preparation of Disiamylborane . . . . . . . 72
Preparation of Pentenyldicyclohexylborane. . . . 72
III. Studies with 2-lithio-2-phenyl-4,
4-dimethylpentane . . . . . . . . . . 73

Preparation of 2-1ithio-2-phenyl-4,
4-dimethylpentane . . . . . . . . . . . 73

Attempted metalation of hexenyldicyclohexyl-

borane with the alkyllithium reagent . . . . . 73
IV. Preparation of Metalated Vinylboranes . . . . 73
Preparation of Lithium Pentenyldicyclohexylborane . 74
V. Reactions of Matalated Vinylboranes . . . . . 74
Product Analysis . . . . . . . . . . . . 74
Reaction of Water with Metalated Vinylbornaes . . 74
2-pentene-l-ol . . . . . . . . . . . . 75

Alkylation of Lithium Vinyldialkylboranes. . . . 75
Mixture of Decenes . . . . . . . . . . . 76

Reaction of Trimethylchlorosilane with
Metalated Vinylboranes . . . . . . . . . . 76

3-(trimethylsilyl)-hexana1. . . . . . . . . 76

vii

TABLE OF CONTENTS - Continued

Page
5-(trimethylsilyl)-3-hexanone . . . . . . . 77
3-(trimethylsilyl)-pentanal . . . . . . . 77
Reaction of Carbonyl Compounds
with Metalated Vinylboranes . . . . . . . 77
Attempted Cyclization of XXVI . . . . . . 78

CHAPTER III

METALATED GEMINAL DIBORAALKANES AND MISCELLANEOUS STUDIES

INTRODUCTION . . . . . . . . . . . . . . 80
Geminal Diboraalkanes . . . . . . . . . 8O
Diphenylalkylboranes . . . . . . . . . 81
a-Borylesters . . . . . . . . . . . . 82

RESULTS . . . . . . . . . . . . . . . . 84
Preparation of Geminal Diboraalkanes . . . . 84

Studies with the Dihydroborated Compounds . . . 84
Preparation of Diphenylmethylborane . . . . . 89

Studies with the Pyridine Complex of
Diphenylmethylborane . . . . . . . . . . 89

Preparation of o-phenylene methylboronate . . . 90
Attempted Metalation of o-phenylene Methylboronate 91

Reaction of Trimethyl borate with

Lithium t-butylacetate . . . . . . . . . 92
DISCUSSION . . . . . . . . . . . . . . . 95
O-Phenylene methylboronate . . . . . . . . 97

Reaction of Trimethylborate with
Lithium t-butylacetate . . . . . . . . . 97

viii

TABLE OF CONTENTS - Continued

Page
EXPERIMENTAL . . . . . . . . . . . . . . 100
I. Materials . . . . . . . . . . . . . 100

Preparation of Bis-l,l—(dicyclohexylboryl)pentane 100

Preparation of Bis-1,1-(borabicyclononyl)pentane 100
Preparation of Trimethylboroxine. . . . . . 101
Preparation of 0-pheny1ene methylboronate. . . 101
Preparation of Aminoethyl Diphenylborinate . . 101
Preparation of Pyridine-Diphenylmethylborane. . 102
II. Studies with Geminal Diboraalkanes . . . . 102

Reaction of Methyl Iodide with Metalated

Bis-l,l~(borabicyclononyl)-pentane . . . . . 102

Reaction of Benzaldehyde with Metalated

Bis-1,1-(borabicyclononyl)pentane . . . . . 103
III. Miscellaneous Reactions. . . . . . . . 104

Attempted Metalation of o-Phenylene

Methylboronate . . . . . . . . . . . 104

Attempted Metalation of Pyridine-

Diphenylmethylborane. . . . . . . . . . 104

Reaction of Trimethyl borate with

Lithium t-butylacetate . . . . . . . . . 104

Reaction of Cyclohexanone with XXXIX . . . . 105
BIBLIOGRAPHY . . . . . . . . . . . . . . 106

ix

10.

11.

12.

LIST OF TABLES

Effect of Solvent on the Formation
of Metalated B-methyl-BBN .

Effect of Different Bases on the
Formation of Metalated B-methyl-BBN.

Amount of Deuterium Incorporation
in B-methyl-BBN with Different Bases

Amount of Deuterium Incorporation
in B-methyl-BBN with Different Solvents . .

Reaction of B-methyl-BBN with Varying Amounts
of Lithium 2,2,6,6-tetramethylpiperidide in
Benzene . . . . . . . . . . .

Alkylation of Metalated B-methyl-BBN

Effect of Different Bases on the MEtalation
of hexenyldicyclohexylborane . . . . .

Effect of Time and Temperature on the
Formation of Metalated Hexenyldicyclohexylborane

\Quenching of the Reaction Mixtures with Water

Conversion of Disiamylvinylboranes to
B-trimethylsilyl carbonyl Compounds.

Reaction of Methyl Iodide with Metalated
Geminal Diboraalkanes . . . . . . . .

Reaction of Benzaldehyde with Metalated
Geminal Diboraalkanes

18

20

23

24

24

26

53

55

57

60

86

88

LIST OF FIGURES

Enolate Anion

Organoborane Anion

Sterically Hindered Organoboranes
Poor Resonance Contributor
Allylic Anion

Metalated Vinylborane

xi

14

66

66

CHAPTER I

THE PREPARATION AND REACTIONS OF METALATED TRIALKYLBORANES

INTRODUCTION

The electronic configuration of trivalent boron is similar to
that of carbon in a carbonyl grouping. Because of this similarity
many of the reactions of aldehydes and ketones have counterparts in the
1
reactions of trialkylboranes. For example, the Baeyer-Villiger and
2
Beckmann rearrangements (eq. 1) of ketones are analogous to the

oxidation3 and amination reactions (eq. 2) of organoboranes.

C R B R
/
/ / \O/ \O/
.0, [0] <1) I [0] (2)
c . B
/ \R / \R
k 9 NHZX g
c R B R
/’ \xNH/ /’ \xNH/

The addition of Grignard reagents to aldehydes and ketones (eq. 3) is

5
also duplicated with boron compounds (eq. 4).
- +
g RMgX o ng | RMgX |_ +
C ---—€> --—€> B M.X
/ \ /g\ (3) /B\ /1.{\ g (47

Enolate anions, anions stabilized by an adjacent carbonyl group,
(Figure 1), are well-known and provide the organic chemist with many

2

O
" <—-> é
/ \- / §C/
I

Figure l. Enolate Anion

useful and versatile synthetic procedures. Carbanions stabilized by an
adjacent boron atom, the organoborane counterpart of enolate anions

(Figure 2), are less well-characterized and have not been as widely

\ -
B—C — e) B:
/’ | /’

Figure 2. Organoborane Anion

used as the enolate anions. Since the boron atom is one of the easi-

est of all functional groups to introduce within a molecule (through
. . . 6 . .

the hydroboration reaction of olefins), carbanions derived from tri-

alkylboranes have the potential of being useful synthetic intermediates.

Literature Review.

In 1956 and 1957 H. C. Brown and B. C. Subba Rao reported that
olefins can be readily and quantitatively converted into organoboranes
under mild conditions (eq. 5), thus providing a new route to organo-

boranes: ’

the hydroboration reaction.

ORCHZCHZ + BZHO _—> 2(RCH2CH2)3B (5)

Further study of the hydroboration reaction revealed that the addition

is controlled in part by steric factors. Simple l-alkenes, such as
l-pentene and l-hexene, react to place 94 percent of the boron on the
terminal carbon, the remaining 6 percent is placed on carbon 2. The
use of sterically hindered dialkyboranes [ERZBH);] as hydroborating
reagents with 1-alkenes permits even greater regiospecificity, giving
hydroborated products with more than 99 percent of the boron attached
to the terminal carbon. A few of the sterically hindered organoborane

reagents currently used are shown in Figure 3.

 

 

r— --1
h-BH Z- (Cy)2BH Dicyclohexylborane
L I 2
H3C CH3 _
I I :' (Sia)2BH Disiamylborane
CHBCHC BH

H
2
BH
\‘ ©BH 9-Borabicyclo [3. 3 . 1] nonane

Figure 3. Sterically Hindered Organoboranes

The addition of diborane to terminal alkynes produces mixtures of

addition products, both 1,1- and 1,2- addition (eq. 6).

\B/ \B/
RCECH + 32116 ——> RCHz-C-H + R-CH-CHZ (6)

/ x /B\

The use of two equivalents of a sterically hindered organoborane as the
hydroborating reagent with an alkyne will give an almost quantitative
yield of only the 1,1- addition product,11 while the addition of only
one equivalent of the same reagent to an alkyne will give a Vinylborane

1
derivative (eq. 7). 2

R H
/
RCE-CH + R2311 ——> ‘c=c (7)
H/ \I
I

Oxidation of trialkylboranes with alkaline hydrogen peroxide gives

alcohols (eq. 8) in essentially quantitative yields.3

R3B + 3H202 + NaOH -——e>- 3ROH + NaB(0H)4 (8)

The expected product of oxidation with alkaline hydrogen peroxide of
the dihydroborated derivative with both borons on the same carbon, was
the carbonyl derivative (eq. 9). Surprisingly, oxidation of the di-
hydroborated 1- and 3-hexynes did not give any carbonyl compounds, but

yielded 1- and 3-hexanol as a major product in each case (eq. 10).

 

[d] OH 0
k g) R-¢-H <—> R-fl-H (9)
OH
\§/'
R-c-H
/B\-
[0]
;; RCHZOH (10)

 

Brown suggested that the geminal diboroalkane was unstable to sodium

13
hydroxide and proposed the following mechanism (eq. 11):

u 9H
- .- I

/B\ HO v/B\ - / H20 /
RCH ———> RCH ——> RCHB ———> RCH B\ (11 )

\B/ \B/ \ 2

‘ ' ¢ ¢ H202
RCH:B< RCHZOH
I

The carbanion, which is formed by attack of the base on one of the

electrophilic boron atoms, is stabilized through resonance by the vacant

orbital of the second boron atom. Thus, in 1961, Brown was the first

to prepose the possibility of an organoboron carbanion intermediate (I).
Although Brown was the first to suggest the existence of an

organoboron carbanion intermediate, other researchers began to eXplore

the possibility of using these anions as useful intermediates in

synthetic reactions. One of the first reports to appear (submitted

by Cainelli and co-workers in 1965-1966)]A’15 described the reaction

of metalated organoboranes with carbonyl compounds (eq. 12) to give

olefin derivatives.14 Generation of the metalated derivative was

accomplished by reaction of the base, n-butyllithium, with a geminal

diboron compound:

\

B/ nBuLi _
RCECH + 2(Cy)zBH -—) RCHz-C-H —-—> RCHZCHB:

B

/\ (12)

- H 1}
RCHZCHB:+ R'CHO ——> RCH2-<|E-<|:-R' -——> RCHZCH:CHR'

I
B‘OLi

/ \

II

After formation of the metalated derivative, the reaction involves a
nucleophilic attack of the anion at the carbonyl grouping, leading to
the intermediate II. II then undergoes a 1,2-elimination to give the
olefin. 1,2-e1iminations are fairly common for B-halo, B-hydroxy, and
B-acetoxyalkylboranes.16 One of the examples given by House is shown

below (eq. 13):

OCH3
OCH
3H3 3 BHZ
+
THF
. 13
0 3H2 ( )
F" .__

‘\ l

 

 

The yields of olefins from the metalated organoboranes ranged from 20
to 50 percent for various carbonyl compounds (i.e. benzaldehyde, cyclo-
hexanone, cholestanone).

During the same period, George Zweifel and H. Arzoumanian investi—
gated the reaction of metalated organoboranes with ethyl bromide.17
Zweifel used essentially the same method as Cainelli to generate the
metalated organoboranes for his study. The base induced cleavage of
geminal dihydroborated intermediates was accomplished with a variety
of bases including sodium methoxide, lithium methoxide, n-butyllithium,
methyllithium . After reaction of the anion with ethyl bromide and
oxidation with alkaline hydrogen peroxide, Zweifel obtained the alcohol

in yields of 35 to 90 percent, depending on the reaction conditions

(eq. 14).

[0]

RCHz-C-H + CZHSBr ——> RCHz-CH-CHZCH3 ——-> RCHZ—CH-CHZCH3
/B\ /B\ OH (14)

The best results were obtained with methyllithium as the base.
In 1969 Brown used metalated organoborane intermediates to synthe-

size cyclopropane (eq. 15) and cyclobutane derivatives (eq. 16).18

6

B
HCECCHzBr + BH —-> H-C-CHZCHzBr + MeLi -—->

9
- [o]
u-c-cnzcnznr -———> 13$ -———> on (15)

Q 657.
63

B
Hcsccuzcnzors + @341 -—-> H-{z-cuzcazcuzors + MeLi -—>
B

Q

- O
B \
g :2 65%

Generation of metalated organoborane intermediates was again accom-
plished by the addition of a base, methyllithium, to geminal dihydro-
borated derivatives. The intermediates then cyclized through an

intramolecular nucleophilic substitution.

D. S. Matteson, working along the same lines but using a different

set of organoboron substrates, also studied the formation and reactions
of metalated organoboranes.19'29 The substrates Matteson chose to work
with were the methanetetraboronic esters, C[B(OR)2 4' These boronic

esters were synthesized by Matteson's group by a difficult process

(eq. l7).19’23

B<OM€)2
THF
cc14 + 8Li + 4(Me0)2BC1 -———e> (MeO)2B-C-B(0Me)2 + 8LiCl (17)

B(OMe)2

Tetrakis(dimethoxyboryl)methane III can be converted into other boronic

esters by a transesterification reaction (eq. l8).25’26

l0-C(CH3)2
CE(OMe)EI + 4(CH3)2C-C(CH3)2 ——> c B l + 8MeOH
4 HO OH O-C(CH3)2
4 (18)
111 IV

The boronic esters react readily with alkyllithium to form tri-

borylmethide anions. For example, treatment of the pinacol boronic es-

ter IV with methyllithium yields the corresponding triborylmethide an-

ion V.26 The anion reacts with bromine to give the bromomethanetribo-

ronic ester VI (eq. 19).

O-C(CH3)2 O-C(CH3)2 Br lO-C(CH )2
c B: | + meLr -+>»’c B: I -—3¥> BrC B\ l 3
O-C(CH3)2 O-C(CH3)2 O-C(CH3)2
4 3 3
IV V VI (19)

Matteson also discovered that the bromomethanetriboronic ester VI can

10

react with more alkyllithium to form the carbanion, -CBr(B02C2Me4)2.
This carbanion reacted with bromine to give the dibromomethanediboronic
ester, BrzC(BOzC2Me4)2.

Triborylmethide ions, RRO)2B]3C-, condense readily with ketones
to form alkene-1,l-diboronic esters, RZ'C=C[B(OR)2]2.27 This reaction
is general and functional groups, such as a-chloro, carbethoxy, and
tertiary amino substituents, on the ketone do not affect it adversely.
Matteson also found that alkene-l,l-diboronic esters are potentially
useful synthetic intermediates. He was able to convert the alkenes to
carboxylic acids (eq. 20), a-bromoalkeneboronic esters (eq. 22), and

alkenyl-l,l-dimercuric chlorides (eq. 21).

B(OH)2 H202
C[B(OMe)2]3 + 0:0 —> Os.“ O=C02H
‘B(on)2

(20)
HgClZ CH3 ngCl
,‘C- (21)
NaOAc CH3 \HgCI
C ‘ ’0
( H3)2C-C B\
0
2
BIZ CH3 [Br
,c:c\ o (22)

Matteson has also reported on the reaction of the triborylmethide

ions with alkyl halides (eq. 23),22 and the metalation of the anions

with group IV triarylmetal halides (eq. 24)?!"25

11

R'X R'c[B(0R)2]3 (23)

-C[B(OR)2]3

Ar3MX Ar3MC[B(OR)2]3 (24)

One of the more interesting aspects of Matteson's work is the
fact that he was able to isolate a boron-substituted carbanion.28 The
lithium salt of tris(trimethylenedioxyboryl)methide ion VII was isolated
in yields of 75 to 100 percent from the reaction of butyllithium and
tetrakis(trimethylenedioxyboryl)methane in tetrahydrofuran at -70°C

(eq. 25).

VII (25)

Since VII chars in air and reacts very rapidly with water, it must be
handled under argon. This salt is very reactive and undergoes all the

reactions of triborylmethide ions discussed earlier.
Discussion of the Problem and an Outline of the Research.

In all the cases of metalated organoboranes discussed above the
metalated derivative was generated by the base-promoted cleavage of a
carbon compound containing two or more boron atoms on the same carbon.

This process is analogous to the cleavage of B-diketones by base

(eq. 26).30

12

9 +
I \ 158 0
R-CH + NaOCH ——-—> R-CH + CH 0-3, (26)
\C/ 3 \C/ 3 \
b' 5

The disadvantage of base-cleavage for the formation of metalated or-
ganoboranes is that one is limited to alkynes for the formation of
dihydroborated derivatives, or to a difficult and expensive synthesis
to obtain the tetraborylmethane or triborylmethane compounds.23
In most synthetically useful examples of enolate anion formation,
the anions are generated by the removal of an a-hydrogen by a base
(eq. 27). It was thought that a similar reaction might be possible
with organoboranes, and a study was undertaken on the feasibility of

a-proton removal from organoboranes by the action of strong bases

(eq. 28).

9 base i. g-
/C\c’H > /C\c" <—> , \\c - (27)

/ \ / \ ‘

n b I I-
,B\ ,H ase /B\_ e 3 /B\ ’ (28)

C C" C

A major obstacle to this base-promoted deprotonation of organo-
boranes is their strong Lewis acidity. Normally, addition of a base
to an organoborane results in coordination of the base to the boron
atom (eq. 29).5 In some cases organoboranes can even be titrated

quantitatively with alkoxide bases.31

13

. C-H
\B-C-H + Base ———)\B/' (29)
/' /’ \

I Base
This acid-base reaction (eq. 29) ties up the base and is undoubtedly
the major reason why alpha metalation of organoboranes by base has

not previously been observed (eq. 28).

Our study began with B-methyl-9-borabicycloC3.3.I]nonane (B-methyl-

BBN) VIII as the organoboron substrate. B-methyl-BBN was chosen be-

B-CH

\

3

VIII

cause it is easily synthesized (eq. 30)32 and because of the expected

differences in the rate of a-proton abstraction between the methyl

and tertiary protons.

-,n m
B-H + MeLi -———€’ 3‘ —-"-€} B-Me + H2 (30)
+ M8
L1

We assumed the methyl protons to be more acidic than the tertiary-bridge-

head protons because primary carbanions are more stable than tertiary

33

carbanions, and because resonance stabilization by the boron atom

of the tertiary carbanion would involve a double bond to a bridgehead
carbon (Figure 4), which is expected to be much less stable than the

corresponding contributor for the methyl anion.

l4

B-CH3 'B-CH

_\ <——-> \t

Figure 4. Poor resonance contributor

Once VIII was prepared we began a search for a base so sterically
hindered that coordination to the boron atom of VIII would be difficult
or impossible, thereby preventing the unwanted neutralization reaction
of eq. 29. Once such a base was found, studies were undertaken to prove
that a metalated organoborane was formed, and to find the optimum con-
ditions necessary for complete metalation. Another aspect of this
research was to search for useful synthetic reactions of the metalated
organoboranes. Lastly, studies were conducted to learn if other organo-
boranes, besides B-methyl-BBN, form metalated derivatives. The results

of these studies are presented in the following section.

RESULTS

Matalation of organoboranes was studied with B-methyl-BBN and a
variety of lithium dialkyl amides. The lithium dialkyl amides were
prepared by the reaction of the corresponding secondary amine at 00

with n-butyllithium in hexane (eq. 31).

R 0° R
:NH + n-BuLi ‘---€> \NLi + n-butane (31)
R hexane R

Initial Studies UsingAPMR.

Early studies consisted of proton nmr examination of the reaction
mixtures of VIII and the lithium amides in benzene. Benzene was chosen
as the initial solvent because the resonance signals from the benzene
protons would not interfere with the signals from B-methyl-BBN, and
because benzene is relatively inert to the amide bases. The first
three bases tried, lithium diisopropyl amide, lithium diethyl amide,
and lithium isopropylcyclohexyl amide, appeared to coordinate to the
boron atom of VIII. In the pmr spectra, the methyl group of B-methyl-
BBN was shifted from a 5 value of 0.90 (TMS internal reference) to a
6 value of 0.2-0.4 in the presence of the base. Refluxing the reaction
mixture for a few hours produced no change in the pmr spectra. Addition
of Banethyl-BBN to the hindered base IX, lithium 2,2,6,6-tetramethy1pi-

peridide (LiTMP), in benzene resulted in a pmr spectrum which

15

16

H3C CH3
H3C 1? CH3
Li
IX

correSponded closely to a mixture of the two components. Apparently
this base is too hindered for coordination with the boron atom of

VIII to take place. After refluxing, this reaction mixture exhibited
a significant change in its pmr Spectrum. The methyl signal of VIII
disappeared and new signals appeared at fields higher than TMS.34 From

these results, we assumed that LiTMP was the sterically crowded base

that we were searching for.

Use of Benzyl Chloride as a Probe.

 

Since the pmr Spectra did not provide unambiguous evidence for
the formation of a metalated organoborane, more conclusive proof was
needed. We decided to quench the reaction mixture with an alkyl hal-
iide and then oxidize the resulting organoborane with alkaline hydro-
gen peroxide. If a metalated organoborane were formed, this sequence
of reactions, with benzyl chloride as the alkylating agent, would yield

B-phenylethanol (eq. 32).

17

@3415 + LiTMP -——> ©B-CH2L1 + TMP

VIII IX
[0]
B-CHZLi + C6H5CH2C1 -——-> B-CH2CH2C6H5 '—-_>
(37-)
HO
C6H5CH2CH20H + .

OH

In the initial experiment, VIII was refluxed with one equivalent
of IX for 24 hours, following which the reaction mixture was quenched
with benzyl chloride at room temperature. After oxidation, B-phenyl-
ethanol was detected in 7 percent yield. In an attempt to improve the
yield of B-phenylethanol, various reaction conditions were tried. The
effect of solvents, temperature of reaction mixture, time, and quantity
of reagents used were studied. The optimum conditions, using one equi-
valent of base, occurred when the reaction mixture of the organoborane
and the base was allowed to stir for one hour at room temperature.

The yield of B-phenylethanol obtained in this case, as determined by
glpc, was 21.4 percent. An excess of base increased the yield of
alcohol to 46 percent; however, the use of more polar solvents did not
improve the yield. Table 1 summarizes a few of the reactions studied.

The effect of different size bases was also studied with the

18

TABLE 1

Effect of solvent on the formation of metalated B-methyl-BBN.

 

 

Ratio of Reaction Temperature of Solvent Z Yield8 of
LiTMP/VIII Time Reaction Mixture B-phenylethanol

1/1 24 hrs. Reflux Toluene 6

1/1 24 hrs. Reflux Benzene 7

1/1 24 hrs. Reflux HMPA 7

1/1 24 hrs. Reflux Ethyl 9

Ether

1/1 15 min. Reflux THF 20

1/1 1 hr. R. T Benzene 21

2/1 1 hr. R. T Benzene 46

 

aYield determined by glpc with an internal standard.

l9

benzyl chloride alkylation reaction. As expected, the less hindered
bases gave little or no B-phenylethanol after quenching with the hal-
ide. 3-t-buty1amino-3-methyl-1-butene (X) appears to be as steri-
cally hindered as IX and gave comparable results. X was obtained

through a substitution reaction (eq. 33).35

CH3 CH3 CH3 CH3
l I_ I I “2
CH3-c-NH2 + CH3-C-C1CH. -———-—e> CH3-C-N-C-CECH -—————e>
I I I H I Ni
CH3 Cl CH3 CH3 (33)
EH3 CH3
CH3-C-N-C-CH:CH2
I H I
CH3 CH3
x

LiTMP was used in most of these experiments because the amine is
commercially available and relatively inexpensive. The results of
using different bases to accomplish metalation are presented in Table
2.

Several reactions were tried in which the amines were removed by
distillation. If the formation of the metalated organoborane is an

equilibrium reaction (eq. 34), the removal of the amine should force

the equilibrium to the right.

. <-——— .
R2NL1 + ©B-CH3 3 RZNH + ©BCH2L1 (34)

20

TABLE 2

Effect of different bases on the formation of metalated B-methyl-BBN.a

 

Base Z Yield of
B-phenylethanol

 

CH3(CH2)2CH2-N-CH2(CH2)2CH3
Li 0

 

(CH3)2CHCH2-E:CH2CH(CH3)2
1

 

 

 

 

 

0
Li 3
H3O CH3 18
H C N
3 ti CH3
?H3 CH3
CH3-C-g-C-CHZCH2 20
I Lil
CH3 TH3
I
CH3-Si-N-Si-CH3
| Lil 5
CH3 CH3
CH3
I
CH3-?-L1 0
CH3

 

aB-methyl-BBN was added to the base in benzene and the mixture refluxed

for 15 min. before the addition of benzyl chloride.

21

The results from these experiments were usually very poor, giving

little or no B-phenylethanol after quenching.
Use of Deuterium Oxide as a Probe.

Since the yields of the alkylation reactions using benzyl chloride
were not large, the possibility existed that the alkylation reaction
was not going to completion and therefore, gave an incomplete picture
of the extent of metalation. Clearly, another method for determining
the amount of metalation was needed. A few experiments showed that
B-methyl-BBN was fairly stable to water for short periods. Consequent-
ly, the decision was made to use deuterium oxide as a quenching reagent
to establish the proportion of starting organoborane that was being
converted to the metalated derivative. The final product of this

quenching experiment would be the deuterated B-methyl-BBN (XI). The

XI

resulting mixture was analyzed by mass Spectrometry.

Since the best results with the benzyl chloride alkylations were
obtained when the base and the boron compound were reacted at room
temperature, all the reactions with deuterium oxide were done at this
temperature. In each reaction, the boron compound was added to one
equivalent of the base suSpended in benzene, quenched with deuterium
oxide after an appropriate reaction period, and isolated by glpc for

mass Spectral analysis. The results of mass spectral analysis confirmed

22

the findings of the benzyl chloride alkylations, that the less hindered
bases do not effect metalation. The results from the reactions of
three different bases are given in Table 3.

Since the use of different solvents in the alkylation reactions
did not improve the yields of B-phenylethanol, only a brief study of
solvent effects was undertaken for the deuterium oxide reaction. The
results were quite similar to those from the benzyl chloride reaction,
and the data are summarized in Table 4.

The effect of different quantities of base on the reaction was

also studied. The results are given in Table 5.
Reactions of Metalated B-methyl-BBN.

The results of quenching with deuterium oxide seemed to indicate
that the metalation of B-methyl-BBN was not quantitative. For this
reason only a few reactions of the metalated derivative were studied.

Reaction of the metalated derivative of B-methyl-BBN with cyclo-
hexanone gave a 55 percent yield of methylenecyclohexane (XII), based

on the quantity of B-methyl-BBN used (eq. 35).

0
L10

(35)
XII
With a 50 percent excess of base, the yield of XII was increased to
66 percent. This reaction is similar to the reaction of organoboron

carbanions and carbonyl compounds reported by Cainelli14 and Matteson.

23

TABLE 3

Amount of deuterium incorporation in B-methyl-BBN with different bases.a

 

 

Lithium amide Reaction time Deuterium b
(hours) Incorporation, %
/FH(CH3)2
LiN 12 0
H3O CH3 1 40
12 50
24 60
. 48 65
LlN 7 days 62
H3C CH3
C(CH3)3
12 40
LiN
4 days 50
C (CH3) 2CH2CH3

 

aReaction mixtures are 1M in boron compound. Equivalent amounts of
base and boron compound were utilized.

From mass spectral analysis of recovered boron compound.

24

TABLE 4

Amount of deuterium incorporation in B-methyl-BBN with different
solvents.a

 

 

Solvent Z Deuterium Incorporation
Benzene 60
THF 32
4Z HMPA/Benzene 58
20Z HMPA/Benzene 26

 

8Reaction conditions: B-methyl-BBN was stirred with one equivalent of
LiTMP at room temperature for 24 hours before quenching.

TABLE 5

Reaction of B-methyl-BBN with varying amounts of lithium 2,2,6,6-tetra-
methylpiperidide in benzene.b

 

 

Base/Boron Compound Z Deuterium Incorporation
0.5 36
0.75 49
1.0 60
2.0 75
4.0 73

 

bReaction times are 24 hours.

25

A reaction of metalated B-methyl-BBN with an aldehyde, heptanal,

was also attempted (eq. 36).

©BCH2Li + CH3(CH2)SCHO ——> CHfCH(CH2)5CH3 (36)

The yield of l-octene was very low. Aldehydes have a tendency to under-
go condensation reactions in the presence of strong bases. Since the
metalated B-methyl-BBN and the unreacted LiTMP are strong bases, self-

condesation of the aldehyde must be a competing reaction in this case.

Cainelli14

also commented on the fact that competing condensation
reactions were probably a factor in the low yields of some of the
olefins formed in his studies.

Other alkyl halides (butyl bromide, pentyl chloride, and octyl
chloride) were also studied as alkylating agents (Table 6). In the

case of butyl bromide, a 42 percent yield of n-pentanol was obtained

(eq. 37).

[0]
©B-CH2Li + BuBr ——> ©B-CH2Bu —->HO-CHZBu (37)

Cyclohexane was also tested as a solvent with butyl bromide as the
alkylating agent, the results were the same as when benzene was the

solvent (38 percent of n-pentanol was detected).

26

TABLE 6

Alkylation of metalated B-methyl-BBN.a

 

 

Alkyl halide Alcohol obtained Z Yield of Alcoholb
Butyl bromide Pentanol 42C

Pentyl chloride Hexanol 37

Octyl chloride Nonanol 36

Benzyl chloride 2-phenylethanol 21

 

3Reaction conditions for formation of metalated B-methyl-BBN: One
equivalent of LiTMP in benzene at room temperature for one hour.

inelds determined by glpc analysis.

CB-methyl-BBN and LiTMP were stirred for 12 hours at room temperature
before addition of butyl bromide.

27

Formatiogfof Metalated Derivgtives from Other Organoboranes.

The success of B-methyl-BBN in forming a metalated compound
could not be duplicated with other alkylboranes. Triethyl boron,
on reaction with LiTMP, followed by alkylation with butyl bromide
and oxidation by alkaline hydrogen peroxide, gave very low yields

of 2-hexanol (eq. 38).

LiTMP BuBr [0] OH
Etchnzcu3-————e> EtZBCHCHB —-———€>- EtZBCHCH3 ————— >‘BuCHCH3
Bu (38)
5-101

Another trialkyborane, B-cyclopentyl-9-borabicycloI3.3.1]nonane
(XIII) also gave poor conversions to the corresponding metalated deri-
vative. The amount of metalation was determined by deuterium oxide

quenching (eq. 39).

B -—-—E> B

/’ // [012. +:IIIOIH
XIII (39)

After oxidation of the organoborane the cyclopentanol was collected

by glpc and analyzed by pmr. The amount of deuterium incorporation

in the product ranged from 10-15 percent.

DISCUSSION

The results presented in the previous section represent the first
example of base-promoted G-proton removal from organoboranes to produce
metalated organoboranes. The removal of a Y-proton from l-phenyl-l,4-
dihydroborabenzene by a base to furnish a metalated organoborane has

been reported recently (eq. 40).36

/
I I + t-C4H9Li ‘—-———€>- [::j\::] (40)
‘\.B-

B 0
C6115 c611S

XIV

The metalated derivative XIV is probably stabilized by aromatic resonance
and therefore represents a special case of metalated organoboranes.
This section will include a discussion of some of the ideas behind this

research and on the results of the study.
Use of Sterically Hindered Reagents in Reactions of Organoboranes.

Because the boron atom of trialkylboranes has only six electrons
in its outer shell and can accomodate eight, trialkylboranes are
classified as Lewis acids. Many of the reactions of trialkylboranes
are assisted by this empty orbital of boron. For example, oxidation
of organoboranes by alkaline hydrogen peroxide probably involves an

28

29

3
initial coordination of a hydroperoxy anion to the boron atom (eq. 41).

H202 + -0H --€> HOO

R
R3B + ‘OZH ——> R-B-OOH
R
13 _
R-B-oon -——-> R-B-OR + OH (41)
R R

The initial coordination of a nucleOphilic reagent with organoboranes
may be of assistance in some of their reactions, but can be a major
obstacle to a base-promoted metalation of trialkylboranes. This
undesired coordination of a nucleOphile with boron can usually be pre-
vented by using a reagent that is so sterically crowded that coordination
becomes impossible.

The use of a sterically hindered reagent in reactions with organo-
boranes to prevent coordination to the boron atom is not a new concept.
For example, the alkylation reaction of G-haloalkanoic esters with

organoboranes utilizes a sterically hindered base (eq. 42).37

Base
I
R3'B + RCHBrC02C2H5 -—--€> RR CHCOZCZH5 (42)

The mechanism proposed for the above reaction is given in eq. 43.

30

+
K
Base:K + RCHB 00011 ———> ' .
r 2 2 5 RCBrC02C2H5 + BASE.H
+ -
I: + 8'13
RCBrCOZCZHS + R'3B —> K R'-B-C-C02C2H5
'Br
. 8'13 13'
K R'-B-C-C02C2H5 ——> R'-B-c.-c:02cza5 + KBr
R'Br R'R
RI
R'-1.3-c::-cozcznS + R"OH ——> RR'CHC02C2H5 + R"O-B-R' (43)
' O

R!

Optimum results were obtained by using the base, potassium 2,6-di-t-

butylphenoxide (XV). According to Brown the reason why this phenoxide

O K

XV

is so effective is because "the large steric requirements of the base
prevents coordination to the trialkylboranes. Coordination would be
unfavorable for the rapid capture of the a-halocarbanions"37 by the
organoboranes. In our own case coordination of the base to the organo-
boranes would tie up the base and prevent it from removing a proton

from the organoborane.

31

Steric Effect of the Bases.

For our studies on the ionization reaction we needed a variety
of bases having different steric requirements. We chose the lithium
dialkyl amides as bases because they are easy to prepare and because
they are very strong bases, pKa of these amines is approximately 36 pKa
units.38 Furthermore a variety of secondary amines of different sizes
are readily available. We tested a number of amides of increasing size
until we found one that gave satisfactory results with B-methy1-9-
borabicyclononane.

The lithium amides of di-n-butyl, diisobutyl, and diethyl amine,
the bases with the least steric requirements in the series tested, do
not give a metalated derivative with B-methyl-BBN. Another base, t-
butyllithium, is also unsuccessful in the base-promoted ionization
reaction of trialkylboranes. These bases are undoubtedly too small

and are coordinating to the boron atom (eq. 44).
[CH3
B-CH3 + Base” ———> B: (44)
Base

Lithium diisopropylamide and lithium isopropylcyclohexylamide are
bulkier bases, but still do not form a metalated compound on reaction
with organoboranes.

The next base that we wanted to try was lithium di-t-butylamide,
which is more sterically hindered than lithium diisopropylamide, but

the amine was not readily available.

32

pu3 pn3 pH, puz
HC—£ c~cH IiC-C c-cn
3 \/ 3 3 \/
INI /N 3
H3C H CH3 H3C H CH3
xvr xvrr

The only difference between the two amines XVI and XVII is an extra
methylene group in the cyclic compound joining the two t-butyl groups
together. Since XVII is probably as sterically hindered as XVI and is
commercially available, we chose the piperidide as the next amide to
study. The piperidide gives satisfactory results for metalation and
therefore must be hindered enough to prevent coordination to organo-
boranes.

Another amine, which is approximately the same size as the tetra-
methylpiperidine, is 2-t-butylamino-2-methy1butane (XVIII). The lithium

amide of XVIII gave satisfactory metalation with B-methyl-BBN.

Surprisingly, lithium bis(trimethylsilyl)amide (XIX), which appears
to be as bulky as the tetramethylpiperidide, gave very poor yields of
metalated organoboranes. No data on the pKa of the bis(trimethylsilyl)-

amine could be found in the literature, but we assume the pKa to be

33

H3? CH3
CH -Si Si-CH
3 \\ ,/ 3
I N \
H3C L1 CH3
XIX

lower than that of the dialkyl amine, since the silicon can provide
charge delocalization with its empty 3d orbital.39 The lower basicity
of the silyl amide could be a factor in the poor yields of organoborane
metalation. It is also possible that the Si-N bond length is longer
than the C-N length. This would make XIX actually less hindered than it

appears.
Solvent Effects.

Solvents can play a large part in determining the rate of carbanion
formation. Thus, an increase in the rate of carbanion generation is
usually observed when more polar solvents are used. For example, the
base-catalyzed racemization of optically active 2-methyl-3-phenylpro-
pionitrile (XX) is much faster in DMSO than in alcohol solvents.40

rs
C6H5CH2 -C -H

EN

Because the lithium amides and the metalated organoboranes are

extremely basic reagents, we were limited to non-proton-donating

34

solvents in our own studies. Although lithium 2,2,6,6-tetramethylpi—
peridide is much more soluble in more polar solvents (tetrahydrofuran,
ethyl ether, and HMPA) than in benzene or cyclohexane, the rate of
organoborane metalation in these solvents is usually the same or lower
in comparison to benzene. Possibly, the more polar solvents coordinate
to the organoborane and in this manner, as Shown for tetrahydrofuran

(eq. 45), hinder the desired ionization reaction.

.0
©B-CH3 + —+ ©B£ (45)
O

\
CH3
Matteson has indicated the possibility of metalated organoboranes,
-CCB(OR)2]3 , abstracting protons from supposedly "aprotic" solvents,
6,27
such as tetrahydrofuran and dimethyl sulfoxide. We do not have

any evidence for proton abstraction from tetrahydrofuran in our own

investigations, but this possibility cannot be ruled out.
Reactions of Metalated Organoboranes.

The yields of the alkylation, deuteration, and condensation
reactions with the metalated derivatives of B-methyl-9-borabicyclononane,
obtained with one equivalent of base, seldom exceeded 50 percent, show-
ing incomplete formation of the metalated derivative. One possible
reason for the incomplete formation of the metalated organoborane
may be the formation of a complex. The newly formed metalated organo-

borane XXI can attack the boron atom of unreacted starting material to

35

form an addition complex (eq. 46).

B- CH2 Li +<3:>\B— ~CH3 '—-—-€>' B- ~CH2©
CH3 (46)

VIII

If formed, the complex XXII would prevent at least half of the starting
organoborane from reacting with the amide. Matteson reported a similar

problem in his own studies (eq. 47).23

OMe
[(MeO)2B] 3c + [(Me0)2B] 4C ——> I:(Me0)zB]3C‘g;1: @(0M9)213

 

 

XXIII
F. _
4%» [(MeO)2B]3-C—B-C[B(0Me)2]3 —> —, -B-C- (47)
0M?— I I
MeO B(OMe)
L . L n

In Matteson's case, the resulting complex XXIII can lose a methoxide
group, add another carbanion, and continue the process to form a
polymer. There are no good leaving groups in XXII and consequently,
formation of a polymer is prevented. Formation of the complex XXII,
if it occurs, is probably an equilibrium reaction, as excess base
usually increases the amount of metalation to greater than 50 percent,
and because the metalated organoborane can be alkylated.

Alkylation of metalated organoboranes takes place readily with alkyl
halides. We found that benzyl chloride gave lower yields of substitution

product compared with the primary alkyl halides and the deuteration

36

studies. Matteson also found the benzyl halides to be poor alkylating
agents with his triborylmethide anions.20’23 The difficulty with the
benzyl halides could be due to a competing metal-halogen exchange be-
tween the halide and the metalated organoborane (eq. 48).

1,1 x

‘\ ‘\ - .
"B-?- + C6H5CH2X ‘--¥):/B-C- + C6H5CH2L1 (48)

Matteson examined one of the reaction mixtures of benzyl bromide and
a triborylmethide anion by treatment with carbon dioxide, and verified

the metal-halogen side reaction (eq. 49).

. +' .
(49)
1. C02

+> PhCH2C02H
2. H3O

 

58%
Limitations of Metalated Organoboranes.

Although B-methyl-9-borabicyclononane gives reasonable yields of
metalated organoborane, other trialkylboranes do not form metalated
derivatives readily. Part of the difficulty could rest in the difference
in the acidity of the n-protons present in the different organoboranes.
The pKa's of the O-protons of organoboranes are not known, but if we
assume that the effect of the boron atom is equal in all organoboranes
and that the pKa's for methane, ethane, and cyclopentane are 40,

42, 44 respectively,41 then the methyl protons of B-methyl-BBN should

be more acidic than the u-protons of triethylboron, which should be more

37

acidic than the a-proton in B-cyclopentyl-BBN. Triethylboron and B-
cyclopentyl-BBN could be such weak acids that lithium 2,2,6,6-tetra-
methylpiperidide is not sufficiently strong as a base to give good
yields of metalation.

Another factor that may account for the lower yields of metalated
derivatives with other organoboranes could be that the anion of B-methyl-
BBN forms a complex (XXII) with unreacted starting material, and that
such complexation helps to stabilize the metalated compound. The
metalated derivatives of the other trialkylboranes might be too hindered
to form such complexes.

Since the formation of metalated derivatives from organoboranes is
not a quantitative reaction, the use of the metalated organoboranes as
a synthetic intermediate will be limited. Unreacted base in the reac-
tion mixture will react with other reagents and contaminate the desired
reaction. For example, the addition of enolizable ketones and alde-
hydes to the metalated organoborane mixture will undergo self-conden-
sation because of the unreacted amide. Because the formation of meta-
lated organoboranes is not quantitative, much of the starting material

is also lost in a synthetic process.

EXPERIMENTAL SECTION

I. General

Spectra.

Proton magnetic resonance spectra were measured using a Varian
T-60 Spectrometer. The mass spectra were determined by Mrs. Lorraine
Guile of this department with a Hitachi-Perkin-Elmer model RMU-6

spectrometer.

Gas Chromatography.

Vapor phase chromatographic analysis and preparative work were
carried out on a Varian Aerograph A-9O thermal conductivity chromato-
graph. Relative peak areas were determined by a Varian disc chart

integrator-model 255.

II. Materials

Handling of Materials.

All reactions were carried out under a nitrogen atmosphere and all
liquids were transferred in glass syringes under nitrogen. All solvents
were carefully dried and stored under a nitrogen atmOSphere over mole-
cular sieves. Extra precautions must be taken with the organoboranes,

as they tend to oxidize quite rapidly in the presence of oxygen.

38

39
Amines.

All amines, except for 2-t-butylamino-2-methylbutane, were commer-

cially available and were distilled before use.
Preparation of 3-chloro-3-methy1-l-butyne.

84 gms. of dimethylethynylcarbinol (1 mole) was stirred with 420 ml.
of concentrated hydrochloric acid (5 moles) containing 1 gm. of hydro-
quinone while the reaction flask was cooled with a water bath. After 1
hour of stirring the two layers were separated and the organic layer
dried over anhydrous potassium carbonate. The organic layer was dis-
tilled and the fraction boiling at 73-8000 was collected, this fraction
was redistilled using a Vigreaux column and the product collected at

74-75°c.
Preparation of 2-t-buty1amino-2-methylbutane.

A solution of 33.6 gms. (.6 moles) of potassium hydroxide, 30
milligrams of powdered copper, and 126 ml. (1.2 moles) of t-butyl amine
in 50 ml. of water was placed in a 500 ml. three-necked flask equipped
with a mechanical stirrer and addition funnel. 45.3 ml. (.4 moles)
of 3-chloro-3-methy1~1-butyne was then added through the addition funnel
over a 2 hour period with vigorous stirring. After stirring for an
additional two hours the mixture was allowed to stand overnight. The
organic layer was then separated and the aqueous portion extracted with
pentane, which was added to the organic layer. The combined organic

extracts were washed four times with 100 m1. of water and then dried

40

over potassium carbonate. After removal of the solvent by a rotary
evaporator , the remaining organic mixture was added to a hydrogena-
tion flask with 30 ml. of ethanol and 1 gm. of RaNi. The mixture
was hydrogenated for 24 hours at room temperature. After the hy-
drogenated mixture was filtered distillation gave 16 gms. of pro-

duct (28%), b.p. 140°C.
Preparation of 9—Borabicyc10C3.3.Ilnonane.

A 500 ml. flask equipped with a reflux condenser, an addition fun-
nel, a magnetic stirring bar, and a side-arm fitted with a rubber serum
stopper was flushed with nitrogen and maintained under a static nitrogen
atmosphere. In the flask was placed 163 m1. of .61M borane (.1 mole)
in tetrahydrofuran and the flask was immersed in an ice bath. In the
drOpping funnel was placed 12.2 ml. (.1 mole) of 1,5-cyclooctadiene.

The diene was added drOpwise to the BH3~THF solution with vigorous
stirring. The mixture was then heated under reflux for one hour. The
9-borabicyclo[3.3.fl nonane was left in the flask in the THF solution

for subsequent steps.
Preparation of B-methyl-9-borabicycloC3.3.ILnonane.

The same set-up as that used for the preparation of 9-BBN was
used for the preparation of the methyl derivative. 100 ml. of a .1M
solution of methyllithium (.1 mole) in ether was placed in the cleaned
dropping funnel and added dropwise to the solution of 9-BBN at 0°C over
a period of one hour. This was immediately followed by drOpwise addi-

tion of 6.5 ml. of methanesulfonic acid (.1 mole) and approximately .1

41

mole of hydrogen was rapidly evolved. The salt was allowed to settle,
and the clear solution was transferred under nitrogen to a distilling
flask and the product distilled. A 70 percent yield of B-methyl-9-BBN,

b.p. 67-6800/14 mm., was obtained.

Preparation of B-cyclopentyl-9-borabicyc10[3.3.I]nonane.

 

9.8 ml. of cyclopentane (.1 mole) was added to a solution of
9-BBN (.1 mole) dropwise at 00C. The mixture was allowed to stir for
2 hours at room temperature. After removal of the solvent by a vacuum

the product was distilled, 110-11400/7mm.(78%).
Triethylborane.

Triethylborane was commercially available and was used without

further purification.
Cyclohexanone.

Cyclohexanone was commercially available and was distilled before

use.
Alkyl halides.
All the alkyl halides were commercially available and were used
without further purification.
III. Bases

All the lithium amides were prepared in the manner to that illus-

trated below for lithium 2,2,6,6-tetramethy1piperidide. t-butyllithium

42
was commercially available as a solution in pentane.
Preparation of lithium 2,2,6,6-tetramethylpiperidide.

A 50 m1. round-bottomed flask equipped with a side-arm fitted with

a rubber serum stopper was attached to a gas connection tube and mercury
bubbler. The apparatus was flame dried under nitrogen and allowed to
cool to room temperature. To this flask was added 2.07 ml. (5 mmoles)
of 2.4M BuLi in hexane. At 00, 0.85 ml. of 2,2,6,6-tetramethylpiperi-
dine was added dropwise over a period of a few minutes to the butyl-
lithium. As the reaction proceeded, butane was evolved and the tempera-
ture of the reaction mixture increased. After the addition of the amine
the mixture was stirred for a few minutes and the solvent removed with

a water aspirator to give a whitish-yellow solid of the lithium amide.

The base was left in the flask for subsequent reactions.

IV. Studies with Metalated Organoboranes
PMR Studies of the Reaction Mixtures.

In the pmr studies of the reaction mixtures 5 m1. of benzene or
tetrahydrofuran was added to the solid lithium amide (5 mmoles) in the
50-ml. round-bottomed flask equipped with a reflux condenser under a
nitrogen atmosphere. 0.75 ml. of B-methyl-BBN (5 mmoles) was then added
to the mixture and the solution refluxed. Although the amides are not
soluble in the benzene the amides did disolve after the addition of the
organoborane and the application of heat. 0.4 ml. aliquots of the re-
action mixtures are removed at intervals and transferred to a pmr

tube under nitrogen, at which point pmr spectra are taken.

43

Deuterium Oxide Quenching of the Reaction Mixture of B-methyl-BBN.

5 m1. of benzene, or whatever solvent was being tested, was added
to the lithium amide (5 mmoles). 0.75 ml. of B-methyl-BBN (5 mmoles)
was then added to the reaction mixture and stirred for the apprOpriate
time interval at room temperature. The mixture was quenched with 1 ml.
of deuterium oxide, followed by the addition of 1.0 ml. 3N hydrochloric
acid a few minutes later. The water layer was removed with a syringe
and the remaining organic layer was saturated with anhydrous potassium
carbonate. The B-methyl-BBN was collected by preparative glpc using
a % inch by 6 ft. SE-30 column. The collected organoborane was then

analyzed by mass spectrometer.
Deuterium Oxide Quenching of the Reaction Mixtures of B-cyclOpentyl-BBN.

5 ml. of benzene was added to the lithium amide (5 mmoles), and
was followed by the addition of 0.99 ml. of B-cyclopentyl-BBN (5 mmoles).
After stirring for 1 hour at room temperature, 1 ml. of deuterium oxide
was added to the reaction mixture and then stirred for a few minutes.

2 ml. of ethanol was then added to the mixture, and was followed by 1.9
ml. of 3 N sodium hydroxide. After cooling the reaction flask to 0°C
with an ice bath, 1.7 ml.of a 30% solution of hydrogen peroxide was
added to the solution dropwise. After 10 minutes the organic layer

was washed successively with 3 N hydrochloric acid, water, and saturated
sodium chloride solution. After drying the organic layer with anhydrous
magnesium sulfate, cyclopentanol was collected by preparative glpc using
a % inch by 6 ft. carbowax column. The collected cyclopentanol was

then analyzed by pmr.

44

Stability of B-methyl-BBN to Reaction Conditions and Water.

The stability of B-methyl-BBN to the reaction conditions was tested
by stirring the B-methyl-BBN (5 mmoles) at room temperature in benzene
containing the lithium amide (5 mmoles) and 5 mmoles of undecane (for
glpc analysis). Aliquots were removed at appropriate intervals and
added to water. The water was removed and the organic layer dried with
anhydrous potassium carbonate. The mixture was then analyzed by glpc.
Aliquots at the end of 4 days showed the B-methyl-BBN present in 90-

100 percent quantities.

Refluxing of the reaction mixture and testing of the aliquots for
B—methyl-BBN showed the organoborane stable for short periods of reflux,
but only 30 percent of the organoborane could be recovered after 22

hours of reflux.

Removal of Amine by Distillation from Reaction Mixture.

 

Sixty ml. of xylene was added to the lithium amide IX (5 mmoles)
in a 100 ml. round-bottomed flask equipped with a distilling head and
a side arm fitted with a rubber serum cap. B-methyl-BBN (5 mmoles)
was added to the mixture and the solvent and amine were distilled at
40°C/22 mm. Fresh xylene was added periodically to the flask as the
distillation proceeded. A total of 180 ml. of xylene was distilled and
the recovered xylene was analyzed by glpc for 2,2,6,6-tetramethylpi-
peridine (4.3 mmoles recovered). The reaction flask, containing approx-
mately 10 ml. of solution, was cooled to room temperature and 0.58 ml.
of benzyl chloride (5 mmoles) was added to the reaction flask and

stirred for 15 minutes. The reaction flask was then cooled to 00 with

45

an ice bath and 3 ml. of ethanol was added, followed by 1.9 ml. of 3N
sodium hydroxide. 1.7 ml. of 30% hydrogen peroxide was added dropwise
to complete the oxidation of the organoborane. The mixture was saturat-
ed with potassium carbonate and analyzed by glpc for B-phenylethanol

(6 percent).

V. Preparation and Reactions of Metalated Organoboranes.

The preparation given below is representative of the many re-

action procedures used.

Preparation of Metalated B-methyl-BBN.

5 ml. of benzene was added to the lithium amide (5 mmoles) in the
50 ml. round-bottomed flask. The B-methyl-BBN (5 mmoles) was added
to the undissolved amide at room temperature. The amide gradually
dissolves over a period of an hour to give a clear yellow solution.
The mixture was stirred for 12 hours, during which time a yellow pre-
cipitate forms. The metalated derivative was left in the flask for

subsequent reactions.
Reaction of Metalated Organoboranes with Alkyl Halides.

The procedure given below is representative.

1.1 m1. of n-butyl bromide (10 mmoles) was added to the solution
of metalated B-methyl-BBN (5 mmoles) and the mixture stirred for one
hour at room temperature. The reaction flask was cooled to 0°C by an
ice bath and 2 ml. of ethanol and 1.9 ml. of 3N sodium hydroxide was

added. 1.7 ml. of 30% hydrogen peroxide was added dropwise and the

46

mixture stirred an additional 15 minutes at room temperature. The
solution was saturated with potassium carbonate and the organic layer
removed. 0.3 ml. of ethylbenzene was added to the organic layer as an
internal standard for glpc analysis. The mixture was analyzed by
glpc on a carbowax column; yield of n-pentanol was 42 percent.

All alcohols from the alkylation reactions had identical nmr

spectra compared with the commercially available product.

Reaction of Cyclohexanone with Metalated B—methyl-BBN.

0.75 ml. of cyclohexanone (7.5 mmoles) was added to metalated B-
methyl-BBN (5 mmoles) dropwise at room temperature and the mixture
stirred for an hour. 2 ml. of ethanol was added to the reaction mixture
and the flask was cooled to 00 with an ice bath. 1.9 ml. of 3N sodium
hydroxide was added to this solution, followed by 1.7 ml. of 30%‘hydro-
gen peroxide added dropwise. The mixture was stirred for an additional
fifteen minutes at room temperature. 0.35 ml. of cumene was added as
an internal standard for glpc analysis, and this organic layer was
washed successively with 3N hydrochloric acid, water, and saturated
sodium chloride. After drying over anhydrous magnesium sulfate the
mixture was analyzed by glpc using an SE-30 column. 53 percent

methylenecyclohexane was detected.

CHAPTER II

THE PREPARATION AND REACTIONS OF METALATED VINYLBORANES

47

INTRODUCTION

The formation of metalated derivatives of trialkylboranes using
lithium amide bases is not a quantitative reaction. This limits its
usefulness in organic synthesis. One of the reasons suggested for the
poor yields of metalation is the weak acidity of the alpha protons in
the trialkylboranes. There are two possible solutions to this problem:
1) a stronger base (for example, an alkyllithium compound) could be
used for the proton abstraction, or 2) the metalated organoboranes could
be made from more acidic boron compounds (for example, the Vinylboranes).

Initially, we worked on the first solution, but the synthesis of
alkyllithium compounds of large steric requirements gave poor yields

of alkyllithium reagents (for example, eq. 50),42 and after a few

CH3 CH3
I TMEDA
C6H5-C3CH2 + (CH3)3CL1 ---€>- C6H5-E;CH2C(CH3)3 (50)

65%

experiments we decided to explore the possibility of using the more
acidic Vinylboranes to form metalated organoboranes. The Vinylboranes
are eXpected to be more acidic than trialkylboranes because of increased

delocalization of charge in the anion (eq. 51).!“3

48

49

The Vinylboranes are easy to prepare and react quite readily with
lithium 2,2,6,6-tetramethylpiperidide to form metalated Vinylboranes in
good yields. The metalated Vinylboranes were reacted with a number of
electrophilic reagents to give a variety of products. The results of

these studies are presented in the following section.

RESULTS

Synthesis of Vinylboranes.

Vinylboranes can be prepared by the addition of dialkylboranes to

alkynes.12

For example, the hydroboration of 3-hexyne with disiamyl-
borane or dicyclohexylborane goes quantitatively to the Vinyldialkyl-
borane (eq. 52).

RZBH H5C2\ [CZHS

H BR2

(R = cyclohexyl or siamyl)

This reaction proceeds by a cis addition to the triple bond. Terminal
alkynes react with these dialkylborane reagents to give a vinyl deriva-

tive with the boron on the terminal carbon (eq. 53).

R'CECH + RZBH ‘———1>' R'CH=CHBR2 (53)

(R = cyclohexyl or siamyl)

The preparation of these vinylboron compounds was carried out in tetra-
hydrofuran and the compounds were not isolated.12 In our studies the

solutions of the vinylboron compounds were added directly to the base.

50

51

Studies with 2-lithio-2-phenyl-4,4-dimethylpentane.

 

The addition of one equivalent of t-butyllithium in pentane to
one equivalent of a-methylstyrene in the presence of TMEDA gave a 65
percent yield of 2-1ithio-2-phenyl-4,4-dimethylpentane (eq. 50). The
yield of alkyllithium reagent was determined by quenching the reaction
mixture with water and analyzing for 4-phenyl-2,2-dimethylpentane by
glpc. A solution of hexenyldicyclohexylborane in tetrahydrofuran was
added to this alkyllithium reagent and the mixture was stirred for an
hour. Analysis of the mixture by glpc, after quenching with n-octyl
bromide and oxidation with alkaline hydrogen peroxide, did not reveal
any products from the reaction of n-octyl bromide with metalated
hexenyldicyclohexylborane. Possibly, the alkyllithium reagent is not
bulky enough to prevent coordination to the boron atom in the vinyl-

borane.

Initial Studies of Vinylboranes with LiTMP.

 

The addition of hexenyldicyclohexylborane in tetrahydrofuran to
lithium 2,2,6,6-tetramethylpiperidide at room temperature produced
a light red solution. After stirring for 15 minutes the mixture
was quenched with benzyl chloride. Oxidation of the quenched reaction
mixture with sodium hydroxide and hydrogen peroxide gave a 50 percent
yield of two C13H18 alkenes. The two alkenes, XXIV and XXV were assumed

to be formed through the following mechanism (eq. 54).

52

LiTMP
(Cy)2BCH=CHCH2C3H7 —T—};> (Cy)2BCH=CHCHC3H7 <—> (Cy)2BCHCH=CHC3H7
C6HSCH2c1 NaOH
———> (Cy)2B-CH-CH=CHC3H7 -———> C6H5CHZCH2CH=CHC3H7 +
CH2C6H5

XXIII XXIV
C6H5CH2CH=CHCH2C3H7

XXV (54)

The formation of the alkenes was surprising, as we had expected
the allyl portion of XXIII to be oxidized to the corresponding allyl
alcohol. A search of the literature revealed that allyldialkylboranes

44 and

(e.g. XXIII) are very reactive and hydrolyze rapidly in water
therefore, XXIII must be undergoing hydrolysis faster than oxidation.
Since benzyl chloride did not give very good results in our earlier
studies we decided to use other alkyl halides to quench the reaction
mixture. Under the same conditions as those used for benzyl chloride,
n—octyl bromide gave a 70 percent yield of C14H28 alkenes.
Hexenyldicyclohexylborane was reacted with a variety of bases to
determine their relative effectiveness in forming metalated derivatives.
No bases, other than LiTMP, produced a red solution when added to the
Vinylboranes and only the reaction mixtures with LiTMP gave alkenes on
quenching with an alkyl halide (Table 7).
Reaction conditions were varied in an attempt to increase the
yield of alkenes. Formation of the metalated compounds was essentially

complete in 15 minutes at room temperature. Reactions ran at lower

temperatures (e.g. 0°C) usually gave poorer yields of alkenes

53

TABLE 7

Effect of different bases on the metalation of hexenyldicyclohexylborane.

 

Base Alkyl Halide % Yield Conditionsb

of Alkenesa

 

[:::l\ EL~§ C6H5CH2C1 O 15 minutes/r.t.

C6H5CH201 50 15 minutes/r.t.

(CH3)BSi\N,Si(CH3)3

n-C8Hl7Br O 20 hours/r.t.
L1

n-08H17Br 70 15 minutes/r.t.

(CH3)3CLi n-08H17Br O 15 minutes/r.t.

 

aYields determined by glpc.

b
Conditions listed are the reaction time of the base and hexenyldicyclo-
hexylborane.

54

(Table 8). No solvents other than tetrahydrofuran were tested.

Quenching of the Reaction Mixture with Water.

Once the optimum conditions for the formation of the metalated
organoboranes were established, we studied the reactions of these
derivatives with different electrophilic reagents. Because they are

resonance hybrids, metalated Vinylboranes have two sites of reactivity

 

 

(eq. 55).
gamma E
> B-c:c-¢-
I I I
attack
(-)
31356-151251 + a (55)
alpha
>:B-C-C:C-
o . '
attack E

Attack of a proton at the alpha carbon, followed by hydrolysis, would
give an alkene (eq. 57), whereas attack at the gamma carbon, followed

by oxidation would give an aldehyde (eq. 56).

H 00 O H
\ ' I. l

,B-CZC-C- -—-> C-CH-C-
I t I I I I

H
(-> (56)
\ ......
H20

(57)

55

TABLE 8

Effect of time and temperature on the formation of metalated hexenyl-
dicyclohexylborane.

 

 

Conditionsa Alkyl Halide Z Yield f
alkenes
o°/1 hour C8H17C1 62
00/4 hours C8H17C1 67
r.t./15 min. C8H17Cl 66
r.t./15 min. C8H17Br 70
r.t./l hour C8H17Cl 72

 

aConditions listed are for the reaction of LiTMP with hexenyldicyclo-
hexylborane.

inelds determined by glpc.

56

In our first experiment with water as the quenching reagent, the
metalated compound from octenyldicyclohexylborane was quenched with 3N
sodium hydroxide, followed by hydrogen peroxide oxidation. After
work-up, only a 22 percent yield of the alkene and approximately 6
percent of the aldehyde was obtained. Further analysis of the reaction
mixture showed a 72 percent yield of 2-octene-1-ol. This was unexpected,
since our initial studies and other reports44 indicate that allyldi-
alkylboranes are unstable in the oxidation mixture and are hydrolyzed
before oxidation can take place. A review of the laboratory procedures
for our earlier oxidations of the allyldialkylboranes showed that the
sodium hydroxide solution was added at room temperature and in the later
oxidations the addition of the sodium hydroxide was done at 0°C. The
allyldialkylboranes must be less reactive to water at the lower temper-
atures and therefore can be oxidized to the allyl alcohols.

A number of experiments were conducted to determine the Optimum
conditions for the formation of the allyl alcohols. Refluxing of the
quenched reaction mixture with 3N sodium hydroxide hydrolyzed the
allylborane completely and produced a 95 percent yield of alkenes
(no allyl alcohol was detected). Slow quenching of the metalated
Vinylboranes at low temperatures (0°C), followed by a slow oxida-
tion (addition of the hydrogen peroxide at a slow rate) gave good

yields of the allyl alcohol (Table 9).

Alkylation of Metalated Vinylboranes.

 

After we discovered that the oxidation of allyldialkylboranes can
be controlled to give allyl alcohols, we studied the reactions of

alkyl halides with the metalated Vinylboranes using the new oxidation

57

.pocflsuoump uoc mums umfifi ago so mocmuoaamca> osu ummH man How mmvzsopflm mam mmcmxfim man no mpamflwn

C03“ 0Hw3 mwhduxflz

.oan up umumamcm mam 000 um mpwxouoa cowouvhn Sufi: wmmwpfixo
.mocmuonamcw> woumamuoa mam mo maggocodv osu pom pom: mmonu mum poumHH mdoaufipcoom

 

 

Amov Ho-H-odoudoo-~ omn-\mouz 2m Nmnemoumommiov
«Asovmx x:
Ammo Ho-N-ouuxue-m we-\mouz 2m one
0

mmmox rmmmo
oaouooum

Aowv Ho-H-ooouuo-N Ame Hadduoo Ame + owa-\mooz Zm mamoomoumomNAaov
mcmuoona
ocmuoonm

on Ho-H-udouoo-N Ame Huduuoo Amov + .dea om xaamou\m0dz 2m mmmoomoumommhxov
mamuoOIH
ocouoonm

Ammo Ho-H-udouoo-N Aev Houuuuo Ammo + oo\mouz 2m mameomoomommxaov
mcouoOuH
Assess NV Aofioaa NV eaeeues NV

Hosooa< mpzsmpa< mcoxH< mcofiufiwcoo mcmuon0dmwuo

m

 

m mqm<8

.umumB saws mmusuxfla cowuommu

ozu mo magnoamso

58

procedure. Thus, the reaction of n-octyl bromide with the metalated
derivative of hexenyldicyclohexylborane was repeated. When the mixture
was refluxed for 15 minutes with 3N sodium hydroxide following the
addition of the alkyl halide, the yield of alkenes increased to 80
percent (eq. 58). Reaction of the metalated derivative of octenyldi-
siamylborane with methyl iodide, followed by careful oxidation at 00C
yielded 70 percent of 3-nonene-2-ol, and approximately 10 percent of

isomeric nonenes from the hydrolysis reaction (eq. 59).

(-)

---------

RZB-CH-CH-CHCSHll

 

 

C8H17Br CH31
RzB-CH-CH=CHC5H11 RzB-CH-CH2CHC5H11
C8H17 CH3
0
NaOH H20, [0] 0
reflux
\/ \/
C8H17CH2CH=CHC5H11 HO-gg-CfliCHCSHll
+isomer 3
(R = cyclohexyl) (R = siamyl)

(58) (59)

59

Reaction of Metalated Vinylboranes with Trimethylchlorosilane.

Addition of trimethylchlorosilane to the metalated derivative of
octenyldisiamylborane gave an immediate precipitate of lithium chloride,
and subsequent hydrogen peroxide oxidation furnished 3-(trimethylsily1)-

octanal in 75 percent overall yields (from l-octyne):

 

Si(CH ) Si(CH )3
Li (CH3)3SiCl | 3 3 flfl l 3
(Sia)zBCH:CHCHC5H11 —e> SIaZCH=CHCHC5H11-——e> CSHIICHCHZCHO +
SiaOH
(60)

Other Vinylboranes gave similar results and the yields are given in
Table 10. In reactions with vinyldicyclohexylboranes the alkenes
resulting from attack of trimethylchlorosilane at the gamma carbon,
followed by hydrolysis, were identified, but the yields were not de-

termined.

Reaction of Carbonyl Compounds with Metalated Vinylboranes.

The addition of benzaldehyde to hexenyldicyclohexylborane, followed
by oxidation, gave a very low yield (less than 10 percent) of l-phenyl-
1,3-heptadiene. This product was probably formed by reaction of

benzaldehyde at the alpha carbon of the metalated derivative (eq. 61).

c H CHO
- 6 5
BCHCH=CHC H -—-—-€>"B-CH-CH=CHC H ————e» c H CH:CHCH=CHC H
3 7 .q\‘| 3 7 6 5 3 7
LiO-C-C H
' o 5

H (61)

60

TABLE 10

Conversion of disiamylvinylboranes to B-trimethylsilyl carbonyl
compounds.

 

 

a
Alkyne Silylated Product Yield , %
l- t CH CH CHS' CH CH CHO 72

pen yne 3 2 1( 3)3 2
l-hexyne CH3(CH2)2CHSi(CH3)3CH2CHO 66
l-octyne CH3(CH2)4CHSI(CH3)3CHZCHO 75
3-hexyne CH3CH8i(CH3)3CH2COCHZCH3 90

 

aGlpc, based on starting alkyne.

61

No identifiable product from reaction of benzaldehyde at the gamma
carbon could be detected by gas chromatography.

The reaction of several ketones with metalated Vinylboranes gave
no dienes (attack at alpha carbon), although very low yields of alkenols
(XXVII) were isolated from some of the reaction mixtures. These com-
pounds were probably formed by hydrolysis of the intermediate XXVI
from the reaction of the ketone at the gamma carbon of the metalated

derivative (eq. 62).

R' H20
- \. '\
‘B-CH:CHCHR + ,c:o -—>/ BCH=CHCHR ——> HZCICHCHR (62)
” R" LiO-C-R' HO-C-R'
é" fin
XXVI XXVII

The expected product from the reaction of a carbonyl compound
with the gamma carbon of a metalated Vinylborane, followed by peroxide

oxidation, would be a Y-hydroxyaldehyde (XXVIII), which might cyclize

o
CHCH CHR
2|
Ho-c-R'
[in

XXVIII

to a hemi-acetal.45 No Y-hydroxyaldehydes or hemi-acetals could be
detected in any of the reaction mixtures from carbonyl compounds and
metalated Vinylboranes. Possibly, XXVIII condensed in the basic
solution used for the oxidations and the condensation products were

not detected, or the products were lost in the work-up procedure.

 

JEJr‘III‘lllal-Il. l].

62

Whatever the case, a new work-up procedure was needed.
Vinylboranes readily undergo protonolysis with carboxylic acids at

00C to give olefins (eq. 63).12

R ,R CH3C02H R R

\ \ I

’CzC ______4>. /C:C\ (63)

H B‘. 0° H H

If XXVI is an intermediate in the reaction mixture, treatment with a
carboxylic acid should give an alkenol, XXVII. To test this possibility,
acetone was added to the metalated octenyldisiamylborane at room
temperature. After 15 minutes prOpionic acid was added and the mixture
was refluxed for 45 minutes, and then oxidized with alkaline hydrogen

peroxide. The yield of alkenol, as determined by glpc analysis, was

72 percent (eq. 64).

 

_ CH3ECH3 CH3CH2C02H
(Sia)zBCH:CHCHC5H11 -——-> (3 1a) 2BCH=CH('3HC 5H11 >
CH3-c|:-CH3
OLi

(64)

CH2z CHCHC 5H1 1
'C'CH3

OH

CH3

A similar reaction with acetophenone and metalated pentenyldisiamyl-

borane produced a 75 percent yield of XXIX.

63

CH2=CHCHC2H5

I
OH

XXIX

Miscellaneous Studies.

The reaction of iodine with metalated octenyldicyclohexylborane
was attempted. Possibly, iodine would react at the alpha carbon of the
metalated derivative to give XXX, which could then rearrange by trans-

ferring a cyclohexyl group to the octene chain (eq. 65).46

O... 0
B-CHCHZCHCSH11 -——€> lB-CHCHICHCSHll (65)
I

XXX

To this end, one equivalent of iodine was added to the metalated inter-

mediate and the mixture stirred for one hour. After hydrogen peroxide
oxidation at 00C and work-up the mixture was analyzed. No products
could be detected on the gas chromatograph (except for cyclohexanol
and solvents). Other reaction conditions and work-up procedures were
tried, but all gave the same negative results.

The intermediate XXVI, from the reaction of carbonyl compounds

with metalated Vinylboranes, is an interesting compound, which could

conceivably undergo a 1,4-elimination to give a cyclobutene derivative

(eq. 66).

64

)B-CH=CH
\\ l HC=CH
LiO-C-CHR ———-> :B-OLi + | | (66)
/ \ R'-C-CHR
R" RI 1""
XXVI

Therefore, to test this possibility, acetophenone was added to the
metalated derivative of pentenyldicyclohexylborane and the mixture
refluxed for one hour. After alkaline hydrogen peroxide oxidation a

70 percent yield of XXIX was obtained.

H2C=CHCHCH2CH3

I
c H -C-CH

6 5 OH 3

XXIX
This product is the same as that obtained from the protonolysis reaction

discussed earlier. Possibly the intermediate XXVI abstracts a proton

from one of the reagents or solvent (eq. 67).

proton source?

/

:B-CHr-CH 06H5 H2C =CHCHR
T | | H20
-0;;C-CHR -——e> ::B-O-CCHRCH=CH2 ——————e> C6H5-C-CH3 (67)
l
C6H5 CH3 CH3 0H
XXVI

In the same manner, the addition of benZOphenone to metalated

pentenyldicyclohexylborane, followed by two hours of reflux and hydrogen

65

peroxide oxidation, gave a 90 percent yield of XXXI. The initial
product is probably the alkenol, which loses water readily to give

the diene (eq. 68).

R C20
- 2
(Cy)2BCH=CH-CHC2H5 -——-——e> (Cy)zBCH:CHCHCZH5 -———-€>

R-C-R
I
OLi
R R R CH=CH2
| HQ | -H20 \/
(Cy) B-O-C-CHCH=CH -————e> H0-C-CHCHzCH -——-—e> 0:0
2 ll 2 ll 2 /\
R CZHS R CZHS R CZHS
XXXI ((68)

«o

DISCUSSION

The results presented in the previous section show that vinylboranes
can be metalated by lithium 2,2,6,6-tetramethylpiperidide in high yields.
The yields are much higher than those obtained with the saturated
organoboranes discussed in the first chapter. This change in the extent
of metalation between the two groups is probably caused by the differ-
ences in the acidity of the allyl protons in vinylboranes and the
alpha protons in saturated organoboranes. Allylic protons (pKa 36)3
are more acidic than the protons on saturated hydrocarbons (pKa 40)
because of the delocalization of charge in the allylic anion (Figure

43

5). The increased acidity of the allyl protons in the vinylboranes

-<,':-c:<,:- 4., -g=g—é— :2. .6616:
I I

Figure 5. Allylic Anion

is probably due to a similar delocalization of charge in the metalated
Vinylborane (Figure 6).

B-C-C=C- <—-—> :3-020-6- <—-—> \fszc-czc- E)B-c-c-c-
OUI I I ll

\
/

Figure 6. Metalated Vinylborane
Effect of Different Bases.

Although the protons in the vinylboranes are more acidic than the

saturated organoboranes, a bulky base is still needed to effect

66

67

deprotonation. Coordination of less hindered bases with boron probably
takes place with the vinylboranes, since lithium diiSOpropylamide is

still ineffective in the metalation reaction.

Sites of Reactions of Metalated Vinylboranes.

Metalated vinylboranes can conceivably react at either the alpha
or gamma carbon. Our results show that metalated vinylboranes react
with ketones and trimethylchlorosilane predominantly at the gamma
carbon (eq. 69), and with alkyl halides and water predominantly at

the alpha carbon (eq. 70).

 

 

E = ketone or E

3: :B-CzC-C- (69)
TMCS
(-)
/B-C-C-C-

E = RX or E

> :B-C-CIC- (70)
H O
2

To some extent, these results may be due to steric factors. The larger
reagents, because of steric repulsion by the dialkylboron group,
attacks the position remote from the boron. This selectivity of
reaction sites with metalated vinylboranes is different from those
obtained with the enolate anions of 0,8-unsaturated ketones or esters.
Enolate anions derived from a,B-unsaturated ketones or esters undergo

reaction almost exclusively at the carbon alpha to the carbonyl

(eq. 71).47

68

-C=C- (71)

The size of the dialkylboron group may also be a factor in
controlling the site of reactivity. In the reaction with trimethyl-
chlorosilane, the metalated vinyldicyclohexylboranes gave a mixture of
of products (products from attack at the alpha carbon could be iso-
lated), whereas the metalated vinyldisiamylboranes gave mostly one
product (from attack at the gamma carbon). Since the disiamylboron
group is probably larger than the dicyclohexylgroup, this difference
is reasonable.

Because of the Lewis acidity of boron, the metalated vinylboranes

can exist as dimers (for example, XXXII).

,B-CZC-C— 'j——%r

The effect of such dimers on the preparation and reactions of metalated
vinylboranes is not known, but if these dimers do exist, they certainly

do not seem to stop either process.
Preparation of Allyldialkylboranes.

From this short study of metalated vinylboranes, these derivatives
seem likely to have many synthetic applications. One of these is the
synthesis of allyldialkylboranes. Allyldialkylborane compounds have
previously been prepared by the addition of tetraalkyldiborane to

48
allenic hydrocarbons (eq. 72).

69

CH CH

3 3

\ \

/C:C:CH2 + RABZHZ -——> C=CHCH2BR2
CH3 CH3

(72)

and from the addition of Grignard reagents to boron trifluoride (eq.

73).49

(73)

Metalated vinylboranes offer a new route to these reactive intermediates;

allyldialkylboranes can be prepared from metalated vinylboranes by the

addition of water (eq. 74) or by reaction with an alkyl halide (eq. 75).

H20

R-X

(74)

(75)

The allyldialkylboranes are very reactive intermediates and can undergo

several transformations. They can be hydrolyzed to an alkene (eq. 76),

or oxidized to an allyl alcohol (eq. 77).

H20

 

V/
I
-0-
('3
V

I
\B-C-C=C-
’ I n I [O]

00

 

I
>> HO-C-CtC-

(76)

(77)

Most trialkylboranes are inert to carbonyl compounds, but allydialkyl-

4
boranes react readily with carbonyl compounds (eq. 78). 8

70

O R
x - .. -Io— ' \ - -.- -
,B CHZCH_CH2 + R C R --9> ’,B O g’CHZCH-CH2 (78)

Comparison of Metalated Vinylboranes to Metalated Trialkylboranes.

The metalation of vinylboranes not only occurs in higher yields
than the metalation of trialkylboranes, but the procedure for the
formation of metalated vinylboranes is much simpler. The Vinyldialkyl-
boranes are made in tetrahydrofuran and do not need to be isolated;
the formation of the metalated vinylboranes occurs in the same solution,
as do subsequent reactions. Also, the metalation takes only 15 minutes
at room temperature, compared to several hours for the trialkylboranes.
The synthesis of vinylboranes is one possible limitation in its use-
fulness in organic synthesis. Vinylboranes are prepared from alkynes

and therefore, one is limited to alkynes for starting compounds.

EXPERIMENTAL SECTION

I. Materials

Alkyl halides.

 

All alkyl halides were commercially available and were used without

further purification.
Alkenes.

Commercial cyclohexene and 2-methyl-2-butene were used. Both were
0
distilled before use, cyclohexene at 82-83 C and 2-methy1-2-butene at

38-4OOC.

Alkynes.

All alkynes were purchased from Chemical Samples and were used

without further purification.

Carbonyls.

 

All carbonyl compounds were commercially available and were used

without futher purification.

Trimethylchlorosilane.

 

Trimethylchlorosilane was commercially available and was distilled

before use.

71

72

II. Vinyldialkylboranes

All preparations and reactions were conducted in flame-dried flasks

and under a nitrogen atmOSphere.
Preparation of Dicyclohexylborane.

In a 50 ml. round-bottomed flask equipped with a side-arm fitted
with a rubber serum stopper was placed 2.03 ml. ( 20 mmoles) of cyclo-
hexene and 10 ml. of tetrahydrofuran. The flask was cooled to 00, then
4.15 ml. of a 2.48M solution of borane (10 mmoles) in tetrahydrofuran
was added slowly. During the borane addition, the dicyclohexylborane
precipitated from solution. The mixture was stirred for an additional

0
hour at O C prior to its use.
Preparation of Disiamylborane.

Essentially the same procedure was followed for the preparation
of disiamylborane as that used for the preparation of dicyclohexyl-
borane. After addition of the borane the mixture was stirred for two
hours at 00C before use. The disiamylborane is soluble in tetrahydro-

furan and does not form a precipitate.
Preparation of Pentenyldicyclohexylborane.

The procedure given below for the preparation of pentenyldicyclo-
hexylborane is representative of all the Vinyldialkylboranes prepared.

To the dicyclohexylborane (10 mmoles) was added at 00C 0.98 ml.
of pentyne (lO mmoles). The mixture was allowed to warm to room

temperature and stirred an addition hour.

73
III. Studies with 2-lithio-2-phenyl-4,4-dimethylpentane
Preparation of 2-lithio-2-phenyl-4,4-dimethylpentane.

A 50 ml. round-bottomed flask equipped with a side arm fitted
with a rubber serum cap was charged with 8.05 ml. of a 1.24 M solution
of t-butyllithium (10 mmoles) in pentane. To this solution at 00C
was added 1.28 ml. of TMEDA (10 mmoles). 1.29 ml. of a-methylstyrene
(lO mmoles) was added dropwise to the slurry of t-butyllithium-TMEDA
at 0°C. A deep red solution was obtained.

Analysis of the reaction mixture by quenching with water showed a

65 percent yield of 2,2-dimethyl-4-phenylpentane.

Attempted metalation of hexenyldicyclohexylborane with the alkyllithium
reagent.

A 10 mmole solution of hexenyldicyclohexylborane was added to the
solution of 2-lithio-2-phenyl-4,4-dimethylpentane and the mixture was
stirred 20 minutes at room temperature. 10 mmoles of n-octyl bromide
was then added to the solution. After 20 minutes the organoborane
was oxidized with 4 ml. 3N sodium hydroxide and 3.4 ml. 30% hydrogen
peroxide. Glpc analysis revealed no products from reaction of n-octyl

bromide with metalated hexenyldicyclohexylborane.

IV. Preparation of Metalated Vinylboranes

 

The metalated vinylboranes were prepared in a manner similar to

that illustrated for pentenyldicyclohexylborane.

74
Preparation of Lithium Pentenyldicyclohexylborane.

The solution of pentenyldicyclohexylborane (10 mmoles) was added
at room temperature to the solid lithium 2,2,6,6-tetramethylpiperidide
(10 mmoles) in a 50 ml. round-bottomed flask equipped with a side-arm
fitted with a rubber serum cap. The solution became slightly warm
and turned light red. The mixture was stirred for an additional 15 to

20 minutes.
V. Reactions of Metalated Vinylboranes
Product Analysis.

Most structures of products obtained were determined by PMR
analysis. IR and mass spectra (molecular ion peak) data are also

included for some compounds.
Reaction of Water with Metalated Vinylboranes.

Two procedures were used in the reaction of water with a metalated
Vinylborane. The procedures given below are representative. Internal
standards are used for all glpc analysis.

(A) 4 ml. of 3N sodium hydroxide was added to 10 mmoles of lithium
octenyldicyclohexylborane in THF and the mixture refluxed for 15 minutes.
At 00C 4 ml. 30% hydrogen peroxide was added dropwise. After 10 minutes
of stirring 10 ml. of pentane was added and the two layers separated.
The organic layer was washed successively with 10 ml. water, 10 m1. 3N
hydrochloric acid, and 10 ml. saturated sodium chloride solution.

The organic layer was dried over anhydrous magnesium sulfate and the

75

mixture analyzed by glpc using and SE-30 column. A 95 percent yield of
2- and 3-octene was obtained.

(B) The flask containing the solution of lithium octenyldicyclo-
hexylborane was cooled to 0°C. 2 ml. H20 was added to the solution
dropwise, followed by .2 ml. 3N sodium hydroxide. After 10 minutes
of stirring the two layers were separated and the organic layer washed
with saturated sodium chloride solution. The organic layer was dried
with anhydrous magnesium sulfate. Analysis of the reaction mixture by
glpc with an SE-3O column showed a 72 percent yield of 2-octene-l-ol.

PMR(CC14): 5.66(m,2H), 4.05(m,2H), l.36(m,7H), l.O86(t,3H).
2-pentene-l-ol.

PMR(CC14): 5.66(m,2H), 4.06(m,2H), 2.16(m,3H), l.06(t,3H).

1

Mass Spect.: 86. IR(neat): 3350 cm- , 1680 cm-1.

Alkylation of Lithium Vinyldialkylboranes.

Two methods were used to convert the product from the alkylation
reaction into easily analyzable products. The first method (C) was
to hydrolyze the allyldialkylboranes to alkenes, and the second method
was to oxidize the allyldialkylboranes to allyl alcohols. The
following procedures are representative.

(C) 1.0 ml. of n-butyl bromide (10 mmoles) was added to 10 mmoles
of lithium hexenyldisiamylborane at room temperature. The mixture was
then stirred for an additional 15 minutes. Procedure (A) was then
followed and the mixture analyzed by glpc using an SE-30 column.

An 80 percent yield of 4— and 5-decene was obtained.

76
Mixture of Decenes.

PMR(CC14): 5.46(m,2H), 2.06(m,4H), 1.06 and l.46(m,l4H).
Mass Spectrum: m/e 140

(D) 0.62 ml. of methyl iodide (11 mmoles) was added to 10 mmoles
of lithium octenyldisiamylborane at room temperature. The mixture was
stirred for an additional 15 minutes. The flask was cooled to 00C
and 4 ml. of 3N sodium hydroxide was added slowly. 4 m1. of 30%
hydrogen peroxide was then added drOpwise. After 10 minutes of stirring
10 ml. of pentane was added and the two layers separated. The organic
layer was washed successively with 10 ml. water, 10 ml. 3N hydrochloric
acid, and 10 ml. saturated sodium chloride solution. The organic
layer was dried over anhydrous magnesium sulfate and the mixture
analyzed by glpc. A 70% yield of 3-nonene-2-ol was obtained.
PMR(CC14): 5.56(m,2H), 4.16(m,lH), 2.06 to l.06(m,15H). IR(neat):

-l -l
3350 cm , 1680 cm
Reaction of Trimethylchlorosilane with Metalated Vinylboranes.

1.3 ml. of trimethylchlorosilane (10 mmoles) was added to 5
mmoles of lithium octenyldisiamylborane at room temperature and the
mixture stirred 30 minutes. The mixture was then oxidized using the
same procedure outlined in (D). Glpc analysis with an internal standard
showed a 73 percent yield of 3-(trimethylsilyl)-octanal. PMR(CC14):

9.9WMt,lH), 2.36(m,2H), 1.25(m,9H), .96(t,3H), 0.06(s,9H).
3-(trimethylsilyl)-hexanal.

PMR(CC14): 9.96(t,lH), 2.36(m,2H), 1.36 and .96(m,8H), 0.06(s,9H).

77
Mass Spectrum: m/e 172. IR(neat): 1725 cm.-1
5-(trimeghylsilyl)-3-hexanone.

PMR(CC14): 2.36(m,4H), 1.06(m,7H), 0.06(s,9H).

-1
Mass Spectrum: m/e 172. IR(neat): 1710 cm

3-(trimethylsilyl)-pentanal.

 

PMR(CC14): 9.96(t,1H), 2.36(m,2H), 1.36 and 0.96(m,6H), 0.06(s,9H).

IR(neat): 1725 cm-1.

Reaction of Carbonyl Compounds with Metalated Vinylboranes.

0.76 ml. of acetone (10 mmoles) was added to 5 mmoles of lithium
octenyldisiamylborane at room temperature and the mixture was stirred
15 minutes. 1.5 m1. of prOpionic acid (20 mmoles) was then added and
the mixture oxidized using the same procedure outlined in (D). Glpc
analysis with an internal standard showed a 72 percent yield of 2-methyl-
3-pentene—2-ol. PMR(CC14): 5.46(m,3H), 1.46(m,l9H). IR(neat): 3400 cm-1,
1670 em‘l.

A similar reaction was done using acetophenone and lithium pentenyl-
dicyclohexylborane. A 75% yield of 2-phenyl-3-ethyl-4—pentene-2-ol
was obtained. PMR(CC14): 7.36(m,5H), 5.26(m,3H), 2.2 to 0.96(m, 10H).

IR(neat): 3450 cm'l, 1660 cm'l, 1600 cm'l.

78

Attempted Cyclization of XXVI.

1.3 gms. of benzophenone (7.5 mmoles) was added to lithium pentenyl-
dicyclohexylborane and the mixture refluxed for two hours. The mixture
was cooled to 0°C and the solution oxidized according to procedure (D).

A 90 percent yield, by glpc analysis, of l,1-diphenyl-2—ethy1-1,3-
butadiene was obtained. Mass Spectrum: m/e 234. PMR(CC14): 7.36(m,10H),

6.5 and 5.956(m,3H), 2.16(m,2H), 1.06(t,3H).

CHAPTER III

METALATED GEMINAL DIBORAALKANES AND MISCELLANEOUS STUDIES

79

INTRODUCTION

The results presented in the previous chapter on metalated
vinylboranes encouraged us to continue our search for more acidic
organoboranes. To this end, our studies of metalated organoboranes
were continued with substrates such as the geminal diboraalkanes,
diphenylalkylboranes, and a-borylesters. These classes of compounds
are expected to be more acidic than trialkylboranes because of in-
creased delocalization of charge in the anion or because of electron-
withdrawing effects of the substituents. However, several difficulties

were anticipated in the studies with these organoboron compounds.

Geminal Diboraalkanes.

B-Dicarbonyl compounds are more acidic than the corresponding

monocarbonyl compounds because of increased delocalization of charge

in the anion (eq. 79).43
9 9 9 9-
/C\ )1 /C\_ / \ /C§
C + Base —-€’ C- <E€> /C- <E> C- (79)
\ / \ \ / \ / \ /
II 9 CI 9
O O O- O

Geminal diboraalkanes are expected to be more acidic than trialkyl-

boranes for a similar reason (eq. 80).

80

81

I
/B\ [H /B ,,B§ ,B\
C + Base ——-> C— <9 'C— <-> C— (80)
\B/ \ / /
I l I I

We expected the formation of metalated geminal diboraalkanes

to be hindered by a competing base-cleavage reaction, which would

give a metalated alkylborane (eq. 81).13

H H
/ \ / — — /

c + Base ——‘> \B-C- <——> \8=C (81)
\ / \ / 1'1 / \

“W

This cleavage of the geminal diboraalkanes would also hinder analysis
of the reaction mixture, since the metalated alkylborane, if formed,

would react with the reagents used to quench the reaction mixture.
Diphenylalkylboranes.

Diphenylalkylboranes are expected to be more acidic than tri-

alkylboranes because phenyl substituents are electron-withdrawing50

and are expected to stabilize the anion by an inductive effect51

(eq. 82). For example, triphenylborane is a stronger Lewis acid

than trialkylboranes.52

.B-C- + Base -——> /\ B-C- <——> .\B'm: (82)
O ‘

82

Although mixed trialkylboranes can be prepared, isolation of pure
compounds is usually difficult because of the tendency of the alkyl

groups of trialkylboranes to redistribute upon heating53 (eq. 83).
RR'R" '————€>’ R3B + R'BB + R"3B + isomers (83)

A similar problem was anticipated for the preparation of diphenylalkyl-

boranes (eq. 84).

thRB -—-——€> Ph3B + PhR2B + isomers (84)

Mixed trialkylboranes can be isolated as stable pyridine complexes.
a-Borylesters.

a-Borylesters (i.e. XXXIII) are expected to be more acidic than
trialkylboranes because of increased delocalization of charge afforded

by the carbonyl group (eq. 85).

o 0‘ o
)B-CHZEOR + Base ——‘> >B-CHC/zé—->>B-Czc: <—-9 )8:g-c’\’
OR H OR H OR
XXXIII
(85)

The synthesis of these a-boryl compounds posed a major problem. Al-
though these compounds have been postulated as intermediates in re-
actions, they have not been isolated. Attempts to isolate a carbon
borylated derivative resulted in the isolation of an oxygen borylated

compound (eq. 86).54

83

9 -N2 /OC2H5
(CH3CH2CH2)3B + NZCHCCZH5 -———€> (C3H7)CH=C (86)

OB(C3H7)2
We attempted the synthesis of an a-borylester by the addition of tri-

methyl borate to lithium t-butylacetate.55 The results of this reaction,

along with the results of studies on geminal diboraalkanes, diphenyl-

methylborane and o-phenylene methylboronate are presented in the

following section.

RESULTS

Preparation of Geminal Diboraalkanes.

 

The geminal diboraalkanes used in our experiments were prepared

by the addition of 2 equivalents of a sterically hindered organoborane

12
to an alkyne in tetrahydrofuran solution (eq. 87).

\ /

B
RCECH + 2:8-H ———> RCHZ-C-H (87)
B

/ \
These dihydroborated compounds were not isolated.

Studies with the Dihydroborated Compounds.

The dihydroborated derivative XXXIV was treated with one

6)

B

Q

XXXIV

equivalent of LiTMP in tetrahydrofuran for one hour at room temperature,

84

85

at which point the mixture was quenched with methyl iodide. Deuterium
oxide and butyllithium were then added to hydrolyze the product of

the alkylation reaction (eq. 88).

\/ \

B LiTMP 8’ CH3I \8’
CH3(CH2)4-C- -H ——> CH3(CH2)4-C —> 0113(CH2)4-(:I--CH3

B 1 hr. B 1 hr. B

/ \ / \ / \

\ /

1. D20 B [0] OH
———-> CH 3(CH2)4-C- ~3CH ———> CH3(CH2)4-C- CH3 (88)
2. BuLi D

After alkaline hydrogen peroxide oxidation, a 21 percent yield (glpc
determination) of 2-D-2-heptanol was obtained. Since no 2-heptanol

was detected in the pmr spectrum of 2-D-2-heptanol, base-cleavage of
the geminal diboraalkane probably does not occur with the sterically

hindered LiTMP (eq. 89).

‘8’ LiTMP 1. CH I

CH3 (CH 2)4-C- -H ——-——> CH3(CH2 )h-C- -H ——-> CH3(CH2)4- -CH- -CH3
B 2. [01011
I \ l \

0% (89)

Various reaction conditions (with the deuterium oxide and butyllithium
step left out) were studied, but the yield of 2-heptanol never exceeded
32 percent (Table 11).

Benzaldehyde was added to the reaction mixture of LiTMP and XXXIV.

After oxidation, l-phenyl-Z-heptanone was isolated by glpc from the

reaction mixture (eq. 90).

86

TABLE 11

Reaction of methyl iodide with metalated geminal diboraalkanes.

 

 

Reaction Conditionsa % Yield of 2-heptanolb
1 hr. r.t. 21
15 min. reflux 32
1 hr. reflux 30

 

8Reaction conditions listed for reaction of LiTMP with XXXIV in
tetrahydrofuran.

ineld determined by glpc based on starting alkyne.

87

\
I] C6H5CH0 \I/fi
CH3(CH2)4-C’ ——> CH3(CH2)4-C-C-C6H5 ————>
B B OLi
I \ I \
‘9’ [0] 9
CH3(CH2)4-C:CHC6H5 --——a> CH3(CH2)4CCH2C6H5 (90)

The yield of l-phenyl-Z-heptanone ranged from 30 to 39 percent for
various reaction conditions with LiTMP. In some reactions a small
yield (about 5 percent) of l-phenyl-l-heptene was isolated; we assume
the alkene was formed by reaction of benzaldehyde with the metalated

derivative from base-cleavage of the diboraalkane (eq. 91).

_ C6H5CHO H
CH3(CH2)4-C- ——————e> CH3(CH2)4-C-CH-C6H5 ———€>’ CH3(CH2)4CH=CHC6HS
’B\ /B\OLi

(91)

Other bases, such as lithium diisoprOpylamide and lithium bis(tri-
methylsilyl)amide, gave poor yields of the 2-heptanone derivative

(Table 12). The dihydroborated derivative from dicyclohexylborane
and pentyne XXXV, on reaction with LiTMP and benzaldehyde under

various reaction conditions, gave no ketone after work-up.

C(Cy)2
CH3(CH2)3-C-H
B(Cy)2
XXXV

The reaction of benzoyl chloride with the metalated derivative of

XXXIV was also examined (eq. 92).

88

TABLE 12

Reaction of benzaldehyde with metalated geminal diboraalkanes.

 

Base Reaction Conditionsa % Yield of l-phenyl-
2-heptanone

 

 

2 hr. r.t. 30
15 min. reflux 39
20 min. reflux 30
N
Li
(CH3)BSI\N/SI(CH3)3 30 min. reflux 5
Li
(CH3)2CH\N’CH(CH3)2 30 min. r.t. 0
Li
a
Reaction conditions listed for reaction of LiTMP with XXXIV in
tetrahydrofuran.
b

Yield determined by glpc based on starting alkyne.

89

o
‘8’ C6H5CC1 ‘8’9 1. NaoH 9
CH3(CH2)4-C -—-———>» CH3(CH2)4-C-C-C6H5 -—————e> CH3(CH2)5CC6H5
/ \ , \ 2. H202

(92)
XXXVI
Since the initial product XXXVI is expected to be unstable in water,
it should hydrolyze rapidly in the alkaline oxidation conditions to
give a ketone. A very low yield of l-phenyl-l-heptanone (5 percent)

was obtained (glpc analysis).

Preparation of Diphenylmethylborane.

 

Diphenylmethylborane was prepared by the addition of methyl-
magnesium iodide to aminoethyl diphenylborinate, and was isolated as

a pyridine complex (eq. 93).

 

1. 8(oBu)3 1. 2MeMgI /CH3
2. NHZCHZCHZOH, 1 | 2. C6H5N +N
+ NHZ-CHZ l
H XXXVII
(93)

The pyridine complex XXXVII was obtained as a white crystalline solid
(m.p. 145-1500C) which was fairly stable to air. The proton nmr spec-
trum of a sample eXposed to air on a watch glass for two hours was

identical to that of freshly prepared complex.
Studies with the Pyridine Complex of Diphepylmethylborane.

Pyridine-diphenylmethylborane is fairly stable to acids. XXXVII
was treated with 3N hydrochloric acid, concentrated hydrochloric acid,

and methanesulfonic acid under various conditions in an effort to

remove the pyridine, no free diphenylmethylborane was formed.

treatment with hydrochloric acid did not cause dissociation of XXXVII,

and the methanesulfonic acid seemed to react with the diphenylmethyl-

borane to give new compounds (the methyl peak of starting boron com-

pound disappeared from the pmr Spectrum and was replaced by four

new singlets).

Since the pyridine complex could not be dissociated to give pure

diphenylmethylborane, several attempts were made to metalate XXXVII

directly with LiTMP.

tried, but none of the products expected from the reaction of a

Various solvents and reaction conditions were

metalated derivative with the quenching reagent were isolated (eq. 94

and 95).

2

Li

CH I

 

€> (C6H5)2B-CH2CH3

0%

 

>0

0%

(94)

(95)

Presumably, the pyridine, by coordinating to the boron atom, hinders

the formation of the metalated derivative.

Preparation of o-phenylene methylboronate.

O-phenylene methylboronate has previously been prepared by the

addition of catechol to trimethylborane (eq. 96).5

6

91

OH 290° 0‘
+ B(CH3)3 ——>O ’B-CHB + 20H4 (96)
0H 0

We were able to prepare the boronate in good yields by the addition

of trimethylboroxine (from carbon monoxide and diborane)57 to catechol
(eq. 97).
T“3
B
/’ \\
O O OH 0‘
C0 + B113 —-—§ I I + O -> O IB-CH3 + H20
B B OH O
\
H3C/ \0/ CH3 (97)

802
The o-phenylene methylboronate was fairly stable to air, a small sample
left on a watch glass for three hours gave a pmr Spectrum identical to

that of a freshly distilled sample.

Attempted Metalation of o-phenylene Methylboronate.

The reaction of o-phenylene methylboronate with lithium 2,2,6,6-
tetramethylpiperidide was studied using different solvents and various
reaction conditions. The reaction mixtures were analyzed for possible
metalation by quenching with either cyclohexanone or an alkyl halide.
None of the reaction mixtures yielded any of the products expected
from the reaction of a metalated o-phenylene methylboronate with the
quenching reagent.

Metalation of o—phenylene methylboronate was also attempted with
sodium hydride. If metalation were to occur with this reagent, hydrogen

gas would be produced (eq. 98).

92

O\ \ 0\
B-CH3 + NaH -——-—€> ' B-CH Na + H (98)
O/ ,/ O/ 2 2

Therefore, to detect and measure the amount of gas produced the reaction
flask was connected to a gas buret, but after 3 hours of reflux no

gas collected in the buret, nor did the sodium hydride dissolve.
Reaction of Trimethyl borate with Lithium t-butylacetate.

Trimethyl borate was added to lithium t-butylacetate in an

attempt to synthesize dimethoxyboryl-t-butylacetate (XXXVIII) (eq. 99).

 

MeO
- I w +-
3(ocu3)3 + L1CH2C02 l .rMéo/B-CHZCOZ + L18(0CH3)4 (99)

XXXVIII

In the initial reaction, the lithium t-butylacetate, from the addition
of t-butylacetate to lithium diisopropylamide in tetrahydrofuran,
was not isolated (eq. 100); and trimethyl borate was added directly

to the metalated ester.

THF
CH CH CH CH + CH C0 ————e> L’CH C0 —+—— 100
( 3)2 \N’ ( 3)2 3 2 1 2 2 ( )

Li

Removal of the solvent after 45 minutes left a white precipitate.
Since the precipitate produced only a red flame on ignition, we
assumed the solid to be largely unreacted lithium t-butylacetate.

In a second attempt to synthesize XXXVIII, lithium t-butylacetate

93

was prepared and isolated as an amine free solid55 before the addition
of trimethyl borate. The addition of trimethyl borate to lithium t-
butylacetate in pentane yielded a white precipitate suspended in
solution. After evaporation of the pentane, a white solid, which

gave a red (lithium) and green (boron) flame on ignition, was obtained.
The pmr Spectrum of this solid in carbon tetrachloride solution ex-
hibited 3 singlets: 1.46(9H), l.85(2H), 3.16(9H); the ir spectrum
displayed a strong band at 1660 cm"1 (the carbonyl stretch of t-
butylacetate appears at 1740 cm-l). From these data we assumed the

compound has the structure XXXIX.

Li

(CH3O)3B-CH2COZ-—+—

XXXIX

XXXIX was added to water and the resulting solution was analyzed
by glpc. T-Butylacetate was detected in quantitative yield (eq. 101).
The addition of cyclohexanone to XXXIX gave a 73 percent yield of the

B—hydroxyester XL (eq. 102).

H20
CH3C02-—+- (101)

Li
(CH3O)3B-CH2C02 +

H0 CHZCOZ +
0 >4
' [\v/J (102)

XL

94

A reaction of XXXIX with benzyl chloride was also attempted. After
one hour at reflux, the reaction mixture of XXXIX and benzyl chloride
was added to water and analyzed by glpc. Only two compounds were
detected, t-butylacetate and benzyl chloride.

The reaction of XXXIX with lithium diisopropylamide was examined.
XXXIX was added to lithium diiSOprOpylamide in tetrahydrofuran at
-78OC and the mixture was stirred for 1 hour. After the addition of
cyclohexanone a very low yield (about 5 percent) of the unsaturated

ester was detected (eq. 103).

 

Li 1 Li ’C02+-
(CH3O) BBCHZCOZ—l O) O C (103)

DISCUSSION

Two carbonyl groups should exert a greater stabilizing influence
on an u-carbanion than a single carbonyl group (pKa of XLI is 20 vs.

11 for XLII).38 We had hoped that two boron atoms would exert a

O 0
II II II
CHBCCH3 CH3CCH2CCH3
XLI XLII

similar effect and stabilize an a-carbanion so that metalation of
geminal diboraalkanes would occur much more readily than the metalation
of trialkylboranes. The results of the previous section Show that
metalation of geminal diboraalkanes occurs to a lesser extent than
metalation of B-methyl-BBN, but to a greater extent than the metalation
of other trialkylboranes.

The poor yields of metalation with the diboraalkanes might be
due to the low acidity of the a-hydrogens. Conceivably, the a-protons
of geminal diboraalkanes might not be much more acidic than those of
trialkylboranes.

A more likely explanation for the low yields (20 to 40 percent)
of metalation could be the long sequence of reactions used, from the
synthesis of the geminal diboraalkanes to the final oxidized products

obtained at the end.

95

96

8H3 R-CECH ‘8’ LiTMP ‘8’
———> @B-H ———> R-C-H ———> R-C'
/ 8 8
/ \ / \

MeI ‘8’ NaOH 3 H202 H

———-> R-C-CH3 -——> R-.-CH3 ——> R-C-CH3 (104)
0H
/ \ / \
(R = pentyl)

This sequence of reactions, illustrated here for the preparation of
2-heptanol, involves six steps (eq. 104). If each step is not quanti-
tative, the yield of 2-heptanol would be lowered accordingly.

The poor results observed with the dihydroborated derivative
obtained from dicyclohexylborane XXXV, in comparison to the results
from.XXXIV, were somewhat surprising. The difference in the degree
of metalation could possibly be attributed to the size of the alkyl
groups of the dialkylboranes. In the dicyclohexylboryl derivative
XXXV the cyclohexyl groups are freer to move about, whereas the alkyl
groups of XXXIV are tied back in the bicyclic system. Thus, the

a-protons of XXXIV are probably more accessible to the bulky LiTMP.

\B/ \g/
~C-H -C-H
B B

XXXIV XXXV

97
O-Phenylene methylboronate.

One disadvantage of working with organoboranes is their rapid
and exothermic reaction with oxygen. The study of o-phenylene
methylboronate was undertaken because of the low reactivity of
o-phenylene alkylboronates to oxygen and because a variety of these

compounds are readily available through a Simple reaction (eq. 105).

O O
\ \
(lo/B-.. . Ram, _._> ‘Olmzcnp do.)

Unfortunately, o-phenylene methylboronate failed to undergo
metalation. The boron atom of o-phenylene methylboronate is not as
effective in the stabilization of the carbanion formed from a-proton
removal because of back-coordination from oxygen to boron (-O?B:).6O
Thus, the o-phenylene alkylboronates are probably weaker acids than
the trialkylboranes. Another reason for the failure of o-phenylene
methylboronate to undergo metalation may be that coordination of the
base with the boron atom is permitted by the decreased steric hindrance

in this system. Coordination of the base could even cause displacement

of one of the oxygen atoms.
Reaction of Trimethylborate with Lithium t-butylacetate.

The addition of trimethyl borate to lithium t-butylacetate could
have occurred at either of two sites to give the carbon XXXIX or

oxygen borylated products XLIII. Compounds similar to XLIII have

98

u 9 0?-
~B-CH2-C-O-—+—— IC—C

OR

XXXIX XLIII

been prepared, for example, XLIV was formed in the reaction of tri-

54
n-propylborane with ethyl diazoacetate (eq. 106).

OC2H5
2

-——-—€> (C3H7)CH:C (106)

O -N

CHCOC H

(CH3CH CH2)3B + N2 2 5

2
OB(C3H7)2

XLIV

XLIV displayed no absorption in the carbonyl region of its ir spectrum.
The vinyloxyboranes react with water (eq. 107) and carbonyl compounds

6
(eq. 108) 1 in a manner similar to that of XXXIX.

 

 

H20 3 9
>: -C-C-0R (107)
\ /OR
/C:C\
OB
1. R'CHO H . 0
;. R'-C-C-C-OR (108)
2. H20 0H

Compound XXXIX seems to be less reactive than the corresponding
lithium t-butylacetate. Thus, lithium t-butylacetate reacts with

cyclohexanone in 5 minutes to give a 95 percent yield of the

99
B-hydroxyester (eq. 109),57 whereas XXXIX, on addition of cyclohexanone,

HO CH2C02

O
O
mama. . 0 __. ‘ do.)

gives only a 48 percent yield of the B-hydroxyester after 10 minutes,
which then increases to 73 percent after two hours. Although lithium
t-butylacetate is alkylated by benzyl bromide58 to give a 96% yield of

the ester (eq. 110), XXXIX does not react with benzyl chloride at

 

C6H5CH2 Br‘ 0
L1CH2C-O .7, C6H5CH2CH2C-O (110)

reflux temperature. It would be interesting to study the different
reactions of XXXIX and determine the usefulness of this compound for

further synthetic sequences.

EXPERIMENTAL SECTION

I. Materials

Preparation of Bis-1,l-(digyclohexylboryl)pentane.

 

In a 50 m1. round~bottomed flask equipped with a side-arm fitted
with a rubber serum stopper under nitrogen was placed 2.03 ml. of
cyclohexene (20 mmoles) in 5 m1. of tetrahydrofuran. The flask was
cooled to 0°C, then 4.7 ml. of a 2.13 M solution of borane (lO mmoles)
in tetrahydrofuran was added slowly. The mixture was stirred for an
additional hour before its use.

To the dicyclohexylborane was added at room temperature .5 ml.
of l-pentyne (5 mmoles). The mixture was stirred for an additional

6 hours prior to its use.
Preparation of Bis-1,1;(borabicyclononyl)pentane.

In a 50 ml. round-bottomed flask equipped with a side-arm fitted
with a rubber serum stopper and a reflux condenser under nitrogen was
placed 4.7 ml. of 2.13 M borane (10 mmoles) in tetrahydrofuran. The
flask was cooled to 00C, then 1.22 ml. of cyclooctadiene (10 mmoles) was
added drOpwise through the rubber septum with a syringe. The mixture
was then refluxed for an additional hour prior to its use.

To the borabicyclononane formed was added at room temperature
.5 m1. of l-pentyne (5 mmoles). The mixture was stirred for an addition-

a1 4 hours before its use.

100

 

101

Preparation of Trimethylboroxine.

.5 mole of borane in 200 ml. of tetrahydrofuran and .2 gms. of
sodium borohydride (.005 mole) were placed in a 500 ml. round-bottomed
flask. The flask was flushed with carbon monoxide and connected to a
gas buret filled with the same gas. The reaction was initiated by
magnetically stirring the contents of the reaction flask. Approximately
12 liters of carbon monoxide were delivered to the flask over a period
of 12 hours. The trimethylboroxine was not isolated, but left in

solution for the following reaction.

Preparation of o-phenylene methylboronate.

22 gms of catechol (200 mmoles) was dissolved in 40 m1. of tetra-
hydrofuran in a 500 ml. round-bottomed flask. 100 ml. of the trimethyl-
boroxine solution (250 mmoles) was added to the catechol at room
temperature and the reaction mixture stirred for 15 minutes. 20 gms.
of sodium sulfate was then added to remove the water formed. After
15 minutes the solution was filtered and the solvent evaporated.
Distillation of the remaining mixture (160-1650C) gave an 80 percent
yield of o-phenylene methylboronate.

NMR (CClA): 7.1 (m,4H); 0.8 (s,3H)

Preparation of Aminoethyl Diphenylborinate.

 

52 ml. of a 1.92 M solution of phenylmagnesium bromide (100 mmoles)
in ether was added dropwise over a period of 1.5 hours to a well-stirred
solution of 14.5 ml. butyl borate (50 mmoles) in ether at dry ice

temperature. The mixture was allowed to warm to room temperature and

102

stirred overnight, then hydrolyzed with 20 ml. of 3N hydrochloric acid.
The ether layer was mixed with 25 ml. of 75% ethanol and 6 ml. of
ethanolamine. After this solution had been warmed on the steam-bath
for a short time, the ester precipitated. Recrystallization from
aqueous alcohol yielded the purified ester as fine white needles,

m.p. 190-192°C (50% yield).
Preparation of Pyridine-Diphenylmethylborane.

28.4 gms. of aminoethyl diphenylborinate (125 mmoles) was dis-
solved in 200 ml. of tetrahydrofuran in a 500 ml. round-bottomed flask
equipped with an addition funnel. The flask was cooled with an ice-
bath and 100 ml. of a 2.5 M solution of methylmagnesium iodide (250
mmoles) in ether was added dropwise. The mixture was stirred overnight
at room temperature, then 20 m1. of pyridine was added to form the
complex. 3N hydrochloric acid was then added to dissolve the magnesium
salts. The layers were separated and the organic layer washed with
water. Evaporation of the solvent gave white powdery crystals, which
were recrystallized in carbon tetrachloride, m.p. 145-1500C (50%
yield).

NMR(CC14): 7.6 and 8.66(m,5H); 7.16(s,lOH); 0.66(s,3H).
II. Studies with Geminal Diboraalkanes.

Reaction of Methyl Iodide with Metalated Bis-l,-l(borabicyclononyl)-
pentane.

A solution of bis-1,1-(borabicyclononyl)pentane ( 10 mmoles)
was added to the solid LiTMP (10 mmoles) and the mixture stirred for

one hour at room temperature. 1.3 ml. of methyl iodide (20 mmoles)

103

was added to the dark red solution and the mixture stirred an additional

hour. 3 ml. deuterium oxide was added to this mixture, followed by 10
mmoles of butyllithium. The reaction flask was cooled to 0°C and the
organoborane oxidized by the addition of 6 m1. 3N sodium hydroxide and
the dr0pwise addition of 8 ml. 30% hydrogen peroxide. 20 ml. of pen-
tane was added and the two layers separated, the organic layer was
washed successively with 3N hydrochloric acid, water, and saturated
sodium chloride solutions. After drying over anhydrous potassium car-
bonate the mixture was analyzed by glpc. A 21 percent yield of 2-D-2-

heptanol was detected.
Reaction of Benzaldehyde with Metalated Bis-1,1-(borabicyclononyl)pentane.

A solution of the diborapentane (5 mmoles) was added to the solid
LiTMP (5 mmoles) and the mixture stirred for one hour at room tempera-
ture. 0.6 ml. of benzaldehyde (6 mmoles) was added to the dark-red
solution and the mixture stirred for 3 hours at room temperature. The
reaction flask was cooled to 0°C and the organoborane oxidized by the
addition of 4 ml. 3N sodium hydroxide and the dropwise addition of 4 ml.
30% hydrogen peroxide. The same work-up procedure as that used above
was then carried out. A 35 percent yield of l-phenyl-Z-heptanone was
detected. NMR(CC14): 7.26(s,5H), 3.66(s,2H), 2.36(m,2H), 1.3 and

1.05(m,9H).

104

III. Miscellaneous Reactions

Attempted Metalation of o-Phenylene Methylboronate.

.6 ml. o-phenylene methylboronate (5 mmoles) was added to LiTMP
(5 mmoles) in tetrahydrofuran and the mixture was stirred for one hour
at room temperature. .5 m1. cyclohexanone (5 mmoles) was then added and
the mixture stirred an additional 15 minutes. 2 ml. 3N sodium hydroxide
was added and after 10 minutes the solution was saturated with anhydrous
potassium carbonate. Analysis of the organic layer by glpc showed no

methylenecyclohexane.

Attempted Metalation of Pyridine-Diphenylmethylborane.

1.3 gms. of pyridine-diphenylmethylborane (5 mmoles) was added to
LiTMP (5 mmoles) in benzene and the mixture was stirred for one hour at
room temperature. .5 m1. cyclohexanone (5 mmoles) was then added and
the mixture stirred 30 minutes. The organoborane was oxidized at 0°C
by the addition of 1 m1. 6N sodium hydroxide and 2 ml. 30% hydrogen
peroxide. The resulting mixture was saturated with anhydrous potassium
carbonate and the organic layer analyzed by glpc. No methylenecyclo-

hexane was detected.

Reaction of Trimethyl borate with Lithium t-butylacetate.

0.47 gms. of lithium t—butylacetate (3.9 mmoles) was added to 5
ml. of pentane in a 50 ml. round-bottomed flask under nitrogen. To
this suspension was added 0.45 ml. (3.9 mmoles) of trimethyl borate.

The reaction flask warmed up Slightly, but no change in the susPension

105

could be detected. After 30 minutes the pentane was evaporated to give
0.852 gms. of a white glassy solid. 3.9 mmoles of XXXIX is expected

to weigh 0.884 gms. NMR and ir data are given in the results section.

Reaction of Cyclohexanone with XXXIX.

 

0.5 ml. of cyclohexanone (5 mmoles) was added to 5 mmoles of XXXIX
in pentane. After stirring for two hours, the reaction mixture was
quenched by the addition of 3N sodium hydroxide. A 73% yield of
t-butyl(l-hydroxycyclohexyl)acetate was detected by glpc. The B-hydroxy-
ester was identified by comparison of glpc retention times with authentic

material.

BIBLIOGRAPHY

10.

ll.

12.

13.

14.

15.

16.

17.

l8.

l9.

BIBLIOGRAPHY

C. H. Hassall, Org. Reactions, 8, 73 (1957).

 

G. Donaruma and W. Heldt, Org. Reactions, ll, 1 (1960).

 

H. C. Brown, "Hydroboration", W. A. Benjamin, Inc., New York
(1962), p. 69.

M. W. Rathke, N. Inoue, K. Varma, and H. C. Brown, J. Amer.
Chem. Soc., 88, 2870 (1966).

Brown, 92' Cit.3, p. 56.

H. C. Brown, "Boranes in Organic Chemistry", Cornell Univ. Press,
New York (1972).

 

H. C. Brown and B. C. Subba Rao, J. Org. Chem., 88, 1136 (1957).

 

H. C. Brown and B. C. Subba Rao, J. Org. Chem., 88, 5694 (1956).

Brown, Qp. Cit.3, p. 114.

6
Brown, . Cit. , p. 281.

 

G. Zweifel and H. Arzoumanian, J. Amer. Chem. Soc., 82) 291 (1967)

 

H. C. Brown and G. Zweifel, J. Amer. Chem. Soc., 88, 3834 (1961).

 

Ibid.

G. Cainelli, G. Bello, and G. Zubiani, Tetrahedron Lett., 4315
(1966).

 

G. Cainelli, G. Bello, G. Zubiani, Tetrahedron Lett., 3429 (1965).

 

H. 0. House, "Modern Synthetic Reactions", 2nd edition, W. A.
Benjamin, Inc., Menlo Park (1972) p. 116.

G. Zweifel and H. Arzoumanian, Tetrahedron Lett., 2535 (1966).

 

H. C. Brown and S. Rhodes, J. Amer. Chem. Soc., 21, 4306 (1969).

 

R. Castle and D. Matteson, J. Amer. Chem. Soc., 88, 2194 (1968).

 

106

2().

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

107
D. Matteson and C. Larson, J. Amer. Chem. Soc., 21, 6541 (1969).

D. Matteson and P. Tripathy, J. Organomet. Chem., 22, P6 (1970).

 

D. Matteson and J. Thomas, J. Ogggnomet. Chem., 82, 263 (1970).

 

D. Matteson and R. Castle, J. Organomet. Chem., 28, 19 (1969).

 

D. Matteson and G. Larson, J. Organomet. Chem., 81, 225 (1973).

 

D. Matteson and R. Wilcseh, J. Organomet. Chem., 21, 231 (1973).

 

D. Matteson, R. Davis, and L. Hagelee, J. Organomet. Chem., 82,
45 (1974).

 

D. Matteson and P. Tripathy, J. Organomet. Chem., 82, 53 (1974).

D. Matteson, L. Hagelee, and R. Wilcseh, J. Amer. Chem. Soc.,
95, 5096 (1973).

 

D. Matteson and M. Furve, J. Organomet. Chem., 82, 63 (1974).

 

H. Stetter and W. Dierichs, Chem. Ber., 88, 1061 (1952).

H. C. Brown and V. Dodson, J. Amer. Chem. Soc., 12 2302 (1957).

 

H. C. Brown and M. Rogic, J. Amer Chem. Soc., 21, 4304 (1969).

 

D. Cram, "Fundamentals of Carbanion Chemistry" , Academic Press,
New York (1965), p. 21.

G. Fraenkel, D. Adams, and J. Williams, Tetrahedron Lett., 767
(1963).

G. Hennion and C. Digiovanna, J. Org, Chem., 82, 2645 (1965).

 

A. Ashe and P. Shu, J. Amer. Chem. Soc., 28, 1804 (1971).

 

H. C. Brown, M. Rogic, M. Rathke, and G. Kabalka, J. Amer. Chem.

 

Soc., _9_g, 818 (1968).

House, 9p. Cit.16, p. 494.

Gram, 9p. Cit.33, p. 71.

D. Cram and M. Sahyun, J. Amer. Chem. Soc., 22: 1734 (1962).

33
Gram, . Cit. , p. 19.

 

G. Fraenkel, J. Russell and Y. Chen, J. Amer. Chem. Soc., 22,
3208 (1973).

 

Cram, Qp. Cit.33, p. 52.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

108

B. M. Mikhailov, Organometal. Chem. Rev. A, 8, l (1972).

 

J. March, "Advanced Organic Chemistry: Reactions, Mechanisms and
Structure," McGraw-Hill, Inc., New York (1968), p. 661.

H. C. Brown and Y. Yamamoto, J. Amer. Chem. Soc., 22, 2796 (1971).

 

M. W. Rathke and D. Sullivan, Tetrahedron Lett., 4249 (1971).

 

B. M. Mikhailov, V. N. Smirnov, and O. D. Ryazonova, Dokl. Akad.
Nauk SSSR, 204, 447 (1972).

 

 

B. M. Mikhailov and F. B. Tutorskaya, Dokl. Akad. Nauk SSSR, 123,
479 (1958).

 

March, 9p. Cit.45, p. 21.
Cram, Op. Cit.33, p. 55.

G. E. Coates, "Organo-Metallic Compounds," John Wiley & Sons, Inc.
New York (1960), p.112.

W. Gerrard, "The Organic Chemistry of Boron," Academic Press, New
York (1961) p. 102.

D. J. Pasto and P. Wojtkowski, Tetrahedron Lett., 215 (1970).
M. W. Rathke and D. Sullivan, J. Amer. Chem. Soc., 22, 3050 (1973).

D. Ulmschneider and J. Goubeau, Chem. Ber., 22, 2733 (1957).

 

M. W. Rathke and H. C. Brown, J. Amer. Chem. Soc., 82, 2737 (1967).

M. W. Rathke and A. Lindert, J. Amer. Chem. Soc., 93, 2318 (1971).

 

H. C. Brown and S. K. Gupta, J. Amer. Chem. Soc., 22, 1816 (1971).

 

Gerrard, Qp. Cit.53, p. 12.

T. Mukaiyama, K. Inomata, M. Murake, J. Amer. Chem. Soc., 22, 967

ilfil l.
I .. I... i1 I