close

Вход

Забыли?

вход по аккаунту

?

Complex Reducing Agents (CRA's)ЧVersatile Novel Ways of Using Sodium Hydride in Organic Synthesis.

код для вставкиСкачать
Complex Reducing Agents (CRA’s)Versatile, Novel Ways of Using Sodium Hydride in
Organic Synthesis
By Paul Caubere*
Why do we hardly use the simplest and, at the same time, inexpensive reducing agent sodium hydride in organic chemistry? To this question the answer is invariably: “It is too basic’’. In this progress report we describe work we have performed aimed at controlling the
basicity of NaH using sodium alcoholates and metal salts. The complex reducing agents
(CRA’s) developed (symbolized NaH-RONa-MX,) aHow organic halides, alkenes, alkynes
and ketones to be reduced selectively. Highly regioselective 1,4- and 1,2-reductions of a&
unsaturated ketones are easily performed using appropriate metal salts. Modified CRA’s
have proved to be excellent hydrosilylating reagents for carbonyl groups, non-pyrophoric
heterogeneous hydrogenation catalysts, coupling reagents for aryl and vinyl halides, and
reagents for the carbonylation of organic halides under very mild conditions. The study of
these reactions opened up the field to phase-transfer-catalyzed photostimulated carbonylations as well as to SRNlreactions of meta1ates.-Thus, starting from the simple sodium hydride a large number of useful reagents have become accessible.
base reactions occur (reaction (3))[6,71. Mechanistic studiesC6]led us to propose an ionic mechanism (Scheme 1).
1. Introduction
1.1. General Considerations
Sodium hydride is a long-known, inexpensive reagent
easily prepared on a large scale from sodium and hydrogen. It is generally used in organic chemistry only as a base
for proton abstraction[’]. Although a few substrates may be
reduced by NaH, the reactions are often slow and the
yields far from excellent[’,21;nevertheless, these reductions
indicate the potential reducing properties of NaH which,
unfortunately, are masked by the basic ones.
Thus to be able to use this reagent for reductions the following have to be achieved: 1) its ability to undergo oneelectron transfer must be increased and 2) its basic properties must be masked. In the present review we shall describe how these problems could be solved using a combination of observations, experiments, and intuition.
1.2. Reducing Properties of Sodium Hydride in
Aprotic Solvents
During investigations on base-promoted eliminations in
hexamethylphosphoric triamide (HMPA)13](Warning: suspected carcinogen[41)it was observed that NaH reacted
with benzyl bromide to give bibenzyl and a small amount
of toluene (reduction product), instead of the expected
trans-stilbene (elimination product). This unexpected reaction was thoroughly studied using substrates possessing, simultaneously, a reactive halogen and hydrogens of varying
acidity. With less acidic hydrogens the soft base[’] He attacks the halogen atom (soft acid) and the halogen-substituted carbon atom, leading to couplings (reaction (1)) and
reductions (reaction (2)), respectively. On the other hand,
if sufficiently acidic hydrogens are present normal acid[*] Prof. Dr. P. Caubtre
Laboratoire de Chimie Organique I, ERA CNRS no 476,
Universite de Nancy I,
B.P. 239, 54 506 Vandeuvre-les-Nancy Cedex (France)
Angew. Chem. Int. Ed. Engl. 22 (1983) 599-613
Coupling
products
Reduction
products
Elimination
products
Scheme 1. X = halogen
One-electron transfer from He to the substrate is presumably the preliminary step to halide loss. To confirm
this hypothesis we studied the reaction of NaH with gemdihalocyclopropanes in HMPA: a reaction which has been
well explored in connection with radical and “ionic” reductions[’]. We f ~ u n d [ ~
that
. ~ dibromides
]
could be reduced
to monobromides and that the reductions could take place
by way of radicals trapped in solvent cages. Support for
one-electron transfer from H Q was provided by the reduction of carbonyl derivatives in aprotic solvents[’01.Indeed,
ketyl radicals were detected by ESR spectroscopy in the reduction of benzophenone and fluorenone. Moreover, benzophenone incorporated an electron and a hydrogen atom
from NaH, but fluorenone accepted two electrons to form
a dianion. These experiments are concordant with the
mechanism depicted in Scheme 2.
An extremely unusual observation convinced us of the
propensity of NaH to undergo one-electron transfer reactions: careful addition of aqueous HMPA to a suspension
of NaH (free from Na) in HMPA led to blue solutions of
solvated electrons@].This indicates that in the presence of
water solvated by HMPA hydride ions could lose electrons, which in turn are solvated by HMPA (for interpretation of this reaction, see Section 2.10).
0 Verlag Chemie GmbH, 6940 Weinheim, 1983
0570-0833/83/0808-0599 $02.50/0
599
RONa. The following features emerged[7~9.’0.’21:
1) sodium
tert-amylalcoholate (tAmONa) was found to be an excellent activating agent; 2) in HMPA, NaH-tAmONa was a
stronger base than either NaH or tAmONa; 3) although
NaH was very weakly reactive in tetrahydrofuran (THF),
NaH-tAmONa showed marked reducing properties
(Scheme 3); 4) other things being equal, NaH-tAmONa in
THF had comparable or even better reducing properties
than NaH in HMPA (Scheme 3). Finally, since ketyl radicals were observed during the reduction of aromatic ketones by NaH-tAmONa in THF”’] we concluded that in
THF one-electron transfer from NaH could be amplified
by sodium alcoholates.
Unfortunately, all the results obtained up to this point,
interesting though they were, did not give us any hope of
finding general synthetic applications for the reagent; indeed, even in THF, NaH-RONa was still too basic. Fortunately, as we shall see later, a solution was found to this
problem in the course of investigations directed towards a
somewhat different target.
Scheme 2. Solv. =solvent = HMPA.
2. Complex Reducing Agents (CRA’s)
“NaH-RONa-MX,” and Related Reagents
1.3. Activated Sodium Hydride “NaH-RONa”
2.1. Introductory Remarks
In addition to investigating the reducing properties of
NaH, we studied the system NaH-RONa (we term the alcoholate RONa an activating agent). Indeed, at that time we
had already established the concept of the synergism of
bases and had shown that NaNHZ-RONa (“complex
had properties remarkably different from those
of the individual components. If such a synergism were to
occur between NaH and RONa, a parallel increase of the
basic or reducing properties would result.
Major differences appeared between the properties of
activated NaH (NaH-RONa (2:1)), which is as readily
prepared as complex bases‘”], and those of NaH and
Ph
Ph
PhHPh
Ph \(y
IAmONd
NaH
‘F\
55 C 4h
, / y60 ’C. P3 h h ’
\Rr
Ph
Ph
’ PhHPh
Ph
Ph
Ph
859’0
traces
In the hope of preparing transition-metal hydride derivatives, we studied the reactivity of NaH with Cu salts.
The results were disappointing and difficult to reproduce.
We therefore tested whether NaH-tAmONa could be used
in place of NaH. Moreover, it was felt that the presence of
alcoholates in the reagent might enhance its reducing
properties.
This time we were not disappointed. First experiment^^'^'
showed that addition of copper salts to NaH-tAmONa at
50-60°C in THF reproducibly led to formation of brownblack reagents which slowly reduced aromatic halides.
Since these first results the number of reagents and applications have grown considerably.
Nomenclature: complex reducing agents will be abbreviated as CRA’s or MCRA’s (metal atom specified) or as
RONa.MCRA’s (alcoholate and metal atom specified); if
necessary, we symbolize RONa. MCRA as NaH-RONaMX,, although, of course, MX, is not present as such in
the reagent. The abbreviation RONa. MCRA(x/y/z) indicates that NaH/RONa/MX, (in that order) is equal to the
molar ratio x/y/z. This ratio refers to the ratios of the
starting reagents and has nothing to do with the actual
constitution of the CRA (see Section 2.10).
2.2. Main Characteristics of CRA’s
70aa
l,e+g%hie
<N~H-~A~ON~/
THF
Rr H
55qb
Scheme 3.
600
66-C 6 h
\
N ~ H
THF
66% 6 h
~
no
reaction
CRA’s are readily prepared from commercial reagents.
Thus, it is sufficient to add, in the chosen solvent, anhydrous MX, to NaH-RONa, or to add the alcohol to the
suspension obtained by mixing NaH and MX,. After
warming, the reagents are ready for
CRA’s are heterogeneous reagents, whose reactive moiety seems, generally, to be located in the solid fraction. Their reactivity is a
function of the solvent; the ratio of the reagents; the naAngew. Chem. Inr. Ed. Engl. 22 (1983) 599-613
ture of the alcoholate; and, above all, the nature of the metal salt. Advantage may be reaped from this: a palette of
reducing agents with graduated properties can be obtained. Considering the large number of metal salts which
led to CRA's and their various areas of application, we are
presently far from having exhausted all the possibilities. The
properties of CRA'S~'~-*~]
are briefly summarized below.
CRA's can be prepared starting from the halides or acetates of Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Cd, Zr, Mo, Pd,
and W. Using the chosen test reaction, the reduction of 1bromonaphthalene, we observed that, besides naphthalene, noticeable amounts of 1,l '-binaphtyl were formed
with Fe-, Cu-, Cd-, and ZnCRA's.
The influence of the solvent is not easy to predict; however it may be stated that dimethoxyethane (DME) usually
leads to more reactive CRA's than THF. In general, it appears that the reducing properties decrease with the polarity of the solvent (from DME, T H F to benzene and cyclohexane).
When metal complexes are used as starting materials instead of salts, the ligands seem to have a strong influence
on the properties of CRA's. For example, the CRA prepared from NaH, tAmONa, and NiCI, reacted poorly with
1-bromonaphthalene, whereas the reagent obtained from
(Ph3P)2NiC12led to very good yields of naphthalene and
binaphtyl. The formation of large amounts of binaphtyl
was surprising, since all other reagents prepared from Ni
salts, such as Ni(OAc),, only afforded naphthalene. The
following interpretation is proposed: during preparation
of CRA from Ni salts, low oxidation-state Ni species are
formed which occur as reducing species later in the reaction. On the other hand, in the presence of Ph3P [from
(Ph,P),NiCI,] a fraction of the low oxidation-state metal
species must be trapped as complexes capable of coupling
with organic halides. Thus, in the presence of sufficient
amounts of appropriate ligands formation of CRA's could
lead to formation of new reagents possessing properties
other than reducing. This hypothesis has been verified experimentally (see Sections 2.4 and 2.9).
We must now underline a very important point concerning the complexes thus obtained. Even if they are wellknown complexes it must be borne in mind that they are
generated in a somewhat unusual environment! Thus,
completely unexpected properties may occur. For an example see Section 2.9.
For CRA's themselves, the influence of reactant ratios
has been determined experimentally for each type of reaction. Presently we are only at the initial stages of their interpretation (Section 2.10).
For alcoholates a short study showed that the variation
of activating properties follows a trend which resembles
that found for complex bases'"]. Thus tAmONa, tBuONa,
iPrONa, and Et(OCH2CH2)20Naexhibit excellent activating properties. Most of the reactions studied were performed using tertiary alcoholates to avoid possible redox
side reactions.
Interestingly, a number of CRA's reactions may be performed catalytically or semi-catalytically.
Finally, it should be noted that the basicity of CRA's appears to be weakened-a very important point for synthetic uses.
Angew. Chem. Inr. Ed. Engl. 22 (1983) 599-613
2.3. Reduction of Organic Halides by CRA's
All reagents used in this section were prepared from
t AmONa.
NiCRA(4/2/1) (NaH, tAmONa, and Ni(OAc), in molar
ratio 4 :2 :1) reduced aromatic iodides, bromides, chlorides, and even fluorides under mild conditions and in
good yields (Scheme 4)1'51.CoCRA (4/2/1) did not reduce
fluorides and reduced chlorides slowly, although in good
yields.
z
X
X
F
F
c1
CI
C1
Z
H
4-Me0
2-Me
3-Me
4-Me
[hl
X
Z
6
12
1
1
1.5
Br
Br
Br
Br
Br
Br
4-Me
2-Me0
4-Me0
2-COOH
3-COOH
4-COOH
f
f
Ihl
1
1
1
1-2
1-2
12-18
Scheme 4. Reaction times r are for comparable amounts of substrate
gem-Dibromocyclopropanes were easily reduced to the
corresponding monobromo derivatives by NiCRA and CoCRA. Even the reduction of 7,7-dichloronorcarane took
place under mild conditions with NiCRA (4/2/1)1'61. The
values of the cidtrans-ratios of 7-bromonorcaranes obtained from 7,7-dibromonorcarane vaned between 2.3 and
3, indicating the intervention of free radicals['I in the reductions. It was also shown that these reductions could be
performed with catalytic amounts of metal salts (yields
were ca. 400% with Co and ca. 500% with Ni).
Finally, the reduction and selective reduction of alkyl,
vinyl, benzyl, and ally1 halide^['^.''^ were examined in more
detail. The following salient features emerged:
Of the CRA's capable of reducing alkyl halides under
mild conditions (particularly CRA's containing V, Fe, Co,
Ni, Zn, and Cd), NiCRA(4/2/1) and ZnCRA(4/2/1) were
selected: the reactions were performed in THF or DME.
RX
R
X
T"W
I-Octyl
I-Octyl
2-Octyl
Cyclopentyl
Cyclopentyl
Cyclohexyl
I-Adamantyl
I-Adamantyl
1-Adamantyl
1-Adamantyl
1 ,I -Dimethylundecyl
1-Methylcyclohexyl
1-Methylcyclohexyl
Benzyl [b]
Benzyl [b]
2-C yclohexenyl
Br
CI
Br
CI
CI
Br
Br
Br
Cl
CI
CI
Br
CI
Br
CI
Br
20
65
20
20
65
20
20
65
20
65
65
65
20
20
65
-40
I
NICRA. DhlE
RTI
lhl
Yield [%]
0.5
0.5
0.33
1.5
0.25
0.5
90-95
I00
90-100
90
91
95
95
95
95
92
90
66 [a1
60 [a1
85- 100
85-100
90
I
0.25
4
0.75
48
0.0s
2
0.5
3
0.08
[a] In addition, 30-40% I-methylcyclohexene is formed. [b] Reduction with
ZnCRA in THF.
Scheme 5. Reaction times f are for comparable amounts of substrate.
60 1
With NiCRA, the general reactivity followed the trends
RI > RBr > RCI and primary, secondary > tertiary. It is noteworthy that even tertiary halides were reduced under
rather mild conditions. ZnCRA reduced only primary halides, following the trend RI > R B r g RCI.
Benzyl and ally1 halides were reduced by NiCRA and
ZnCRA. In Scheme 5 some significant examples are given.
Vinyl chlorides and bromides were reduced highly stereoselectively under extremely mild conditions in excellent
yields by NiCRA (Scheme 6).
Reductions could be performed with NiCRA but not
with ZnCRA in the presence of catalytic amounts of metal
salts.
H
H
H
4-Me
Br
Br
c1
CI
>--("'
Et
H
2
3
2
2
+
9
NKRA
H
Br
:
Et
2'
X
Z
Br
Br
Br
CI
H
H
Z
n
T ["Cl
Br
Br
CI
OH
COOH
COOH
10
9
3
65
20
20
10
COOH
H
H
9.3
l
E
[min]
5
2
20
NiCRA, DME
M e-( C Hz )"-C H(
H
5
5
5
20
65
65
65
X
,
DME
2O0C, 1h
65
OTetrahydropyranyl
OMe
-O-(CHJ-O-O-(CH&O-
X
H~~
Et
30
60
25
20
20
45
45
Finally, the unique reactivity of NiCRA and ZnCRA
with organic halides is such that carbon-halogen bonds
can be selectively reduced in the presence of numerous
functional groups (Scheme 8).
Me-(CHz) -CHzCOOH
90%
t
0.7
X
n
T ["el
f
Br
CI
8
I
20
20
2
30
[rnin]
90%
Scheme 6.
The differences in reactivity between the various halides
were sufficient to permit selective reductions (Scheme 7).
Particularly striking was the reduction of o-bromochlorobenzene to chlorobenzene, which took place readily without formation of benzyne.
RI e-(CH2) ,-C IIz R r + 51e-(CFiz) ,-C H,C1
-
+
4OoC, 30 rnin
Me-(CH2)6-Me
NICRA, DME
M e-(CH2),-CH2C1
1004ro
100%
0""
+
100%
6S0C.3mn
21 e-(CH,),-Me
ZnCRA, THF
M e-(CH2) &-CHz B r +
98%
0 oo 0 ooooH
Br
+
+
NlCRA,THF,
+
2 0 T , 2 0 nu"
1LIe-(CH2),-Ible + ~ile-(CI12)Z-CH-(CH2)3-;11e
I
Rr
l00gb
95%
OCO-CHaC1
NCRA, THF
6S°C, 15 min
ZnCRA, THF-C6H6
NZC O C H , B r
-
2 0 T , 2h
*
N G C O M e
56%
NfRA-WC16, DME
ZnCRA, THF
C1CII2-(CH2),-CH2Br
7%
93%
100%
C1CH2-(CH2)2-Me
93
75-8070
X = Br(20°C, 4h),
X = C1(20°C, 3h)
65T.55rnin
BrCH2-(CH2)2-COOEt
NiCRA-WC16, DME
*
Me-(CH2)2-COOEt
65OC. 2.5h
traces
Scheme 7.
602
82%
Scheme 8. Reaction times t are for comparable amounts of substrate.
Angew. Chem. Int. Ed. Engl. 22 (1983) 599-613
It is notable that no eliminations were observed in the
reduction of !3-haloethers and a-haloketals. A number of
a-haloketones and y-haloesters were too sensitive to be reduced by the usual CRA's; addition of wc16 to NiCRA
enabled "normal" reductions to be performed (cf. the last
two equations of Scheme 8), but the role played by wC16
remains obscure.
How should CRA's be classified among other known reducing agents?
Considerable effort has been devoted to the use of metal
hydrides as reducing agents for organic halides[19a1;alkali
metal aluminum hydrides have been extensively studied['91.
LiAIH, itself readily reduces reactive halides and, less
readily, secondary bromides, e. g. bromocyclohexane is
particularly resistant to reduction. Tertiary, aromatic, and
alkenyl halides are reduced with difficulty, unless they are
iodides. More efficient are a number of alkoxyaluminum
hydrides["".'."', but these are of limited value for the reduction of tertiary, vinyl, and aromatic halides.
Finally, the ability of aluminum hydrides to reduce numerous functional groups limits their use for the selective
removal of halogen atoms.
The less reactive borohydrides LiBH, and NaBH>'9a,b,g-rl
are more selective, but more drastic conditions are often
h. 2oa-4 . N ote that with these reagents "indirect
reduction" (dehydrohalogenation-hydroboration) may be
hl; these hydrides lead to low yields of reduction products with a number of bromides, chlorides and,
above all, with aromatic and alkenyl halides.
Modification of simple borohydrides leads to more efficient reducing agents such as M[BH3CN] (M=Na, K,
Bu4N),
9-BBNCN
(9-cyano-9-hydrido-9-borabicycIo[3.3. llnonane), or Li[BHEt,] (superhydride).
In HMPAf4]benzyl, allyl, as well as primary and secondary halides are reduced by an excess of cyanoborohydrides["'I (treatment for 24 h at 70-100°C is necessary with
the less reactive halides). These hydrides are much less
successful in reducing alkenyl, aromatic, and a number of
tertiary halides. On the other hand, they are selective, since
most other functional groups do not react under the conditions used.
Note that reducing agents such as the ate-complex prepared from 9-borabicyclo[3.3.1]nonane selectively reduce
benzyl, allyl, and tertiary halides. There is evidence to suggest formation of a carbocation intermediate''Ofl.
Li[BHEt3]['9a,g-'1
is a convenient reagent for the reduction
of reactive halides. Tertiary halides are reduced slowly,
since elimination can be a competing reaction and vinyl or
aryl halides are inert.
and organotin hydride~[''",~~]
are valuable
Cr"
reagents for the selective reduction of aliphatic and reactive halides, but are much less reactive with alkenyl and
aromatic halides. Moreover, a number of Cr" salts are unstable and use of tin hydrides often necessitates high reduction temperatures. Also, some problems can arise in disposing of the tin residues (see the comments given in
The results given in Schemes 4-8 show that CRA's
compare favorably with the above reducing agents. Generally, our reagents are more efficient with less reactive halides; in their selectivity they appear, with a few exceptions,
to be as selective as other reducing agents.
Angew. Chern I n t Ed. Engl. 22 (1983) 599-613
A large number of transition-metal species have also
been reported to reduce organic halide^^'^".^^]; among
these the reagents obtained by addition of metal salts to
aluminum hydrides or borohydrides are directly comparable to CRA's.
Lithium copper hydrides Li,CuH,+ I (from MeLi, CuI,
and LiA1H,)[24a1exhibit different reactivities. Li,CuHs (the
most reactive) reduces cyclohexyl, vinyl, and aryl chlorides
poorly at room temperature, even after 24 h. The reagents
obtained by addition of a series of transition-metal salts to
LiA1H,[24b,c1reduce aliphatic as well as aromatic halides,
including chlorides; however, unreactive halides necessitate long reaction times, and no selective reductions have
been studied. The reagent prepared from Li[AIH(OMe),]
and C U I ~ ' essentially
~~]
reduces organic bromides. Parallel
experiments have been performed with NaBH>19a,'S1.The
most reactive reagents were obtained with Ni0[2sb1and,
above all, Pd derivatives[25a1;
however, weakly reactive halides, such as arenes, require a very large excess of NaBH,
and reaction conditions conducive to the decomposition of
the borohydride[2sa1.More reactive is the reagent obtained
from K[BHsBu3] and CUI[~~'];
however, it must be used
in large excess with aromatic chlorides and also reduces
keto and ester groups.
Here too, it seems that CRA's compare favorably with
the last mentioned reagents. As a general comment it must
be underlined that CRA's can be prepared readily from
commercial starting materials and are, on the whole, less
expensive than most other reducing agents.
2.4. Use of Ligand-Modified MCRA's (MCRAL's) to
Couple Organic Halides
Stimulated by the reactions observed with (Ph3P),NiC12
(Section 2.2), we first prepared tAmONa. NiCRA(4/2/1)
in the presence of Ph3P. The addition of l-bromonaphthalene to the reagent thus obtained led to 70% of 1,l'-binaphthy1 and to only 25% of naphthalene["]. In order to avoid
side reactions, such as transfer of phenyl from Ph3P, the
latter was replaced by 2,2'-bipyridine, leading to satisfying
results (Scheme 9)[17.261.
NICRA (4/2/l)-bpy
THF, 63OC
X
R
1
Br
Br
Br
Br
H
2-Me0
3-Me0
4-Me0
0.5
Lhl
0.5
0.5
0.5
X
R
CI
CI
CI
CI
H
2-Me
3-Me
4-Me
Lhl
2.5
8
8
8
NiCRA (4/2/1)-bpy
>
i
THF, 63°C
xX = CI:
Br 2105mmnn
70-90%
10-30qo
Scheme 9. Reaction times f are for comparable amounts of substrate.
A comparison of Schemes 4 and 9 underlines the substantial difference between CRA's and the new reagents,
which we term MCRAL (L=ligand).
603
Some problems arose with NiCRAL and their reactions
with vinyl halides: NiCRA is too powerful a reducing
agent. Since NiCRA (4/2/1) contains NaH (see Section
2.10) but NiCRA (2/2/1) does not, we used the latter reagent and succeeded in performing coupling reactions
(Scheme
80-8770
5-100/0
pm
X = C, n = 2 : 1.5 h; X = Br, n = 4: 30 min
pH
THF,
WCRA3OoC,
(2/2/1)-bpy
2h
*
+
H
Ph
Ph
Br
H>=(,
N I C M (2/2/l)-bpy
THF,30°C, I h
*
H
7 5%
H
970
15%
Scheme 10.
It is noteworthy that the reactions took place without
isomerization.
Although our studies of coupling reactions are only just
beginning, comparison with the literature is very encouraging. Coupling of aryl or vinyl halides by Nio complexes or
Nio species generated in situ has been described (see e.g.
["I).
Not only are NiCRAL's much more thermally stable
and less air-sensitive than conventional Nio complexes, but
they are also more efficient in coupling bromides and,
above all, chlorides. Moreover, the coupling of simple alkenyl bromides appears to be stereospecific with NiCRAL
but not with Nio complexes[27d1.
2.5. Reduction of Alkynes and Alkenes by CRA's
Here we shall elaborate only upon those results which
throw light on the nature of CRA's. Mono- and disubstituted alkynes are smoothly reduced to the corresponding
alkanes by tAmONa. NiCRA (4/2/1) and tBuONa. FeCRA (4/2/1) (from FeCI,) in THF or anisole["'. Moreover,
the use of FeCRA avoids the inconvenience presented by
NaH-FeCl,, which necessitates a large excess of NaH and
of FeCI, and causes oligomerization of monosubstituted
alkyne~['~"].This result emphasizes the improvement
wrought by addition of alcoholates.
Interestingly, using appropriate conditions NiCRA allows the reduction of alkynes to alkenes. The alkenes obtained from disubstituted alkynes were predominantly cis,
corresponding to a syn-addition of the reagent. With
monosubstituted alkynes small amounts of 2-alkenes arising from isomerization of 1-alkenes were detected. These
observations strongly suggest the presence of nickel hydride species[301.
NiCRA and FeCRA also reduce alkenes[281.Depending
on the environment of the double bond they behave differently, thus permitting selective reductions (Scheme
11)128b.31al604
0
NiCRA, DME
%GE-+
0
100%
looq0
NCRA, DME,4S°C, I h
or
FeCRA, THF, 4 5 T , 2h
Ph-CH=CH,
+ Ph-CH=CH,
*H
ph
4 5 T , l5mn
,(C HZ)3-M e
H2C=C,
Me
Scheme 11
15%
2
NiCRA, DME
-CH=CH2
Careful study of the reaction with NiCRA Ied to an intriguing observation: reductions of alkenes are catalytic
with respect to Ni only if stoichiometric amounts of the alcoholate and alkene are used; they are catalytic with respect to alcoholate only if stoichiometric amounts of Ni
and alkene are ~ s e d [ ~ ' ~ ' .
Finally, even though NaH-FeC13 reduces ketones[29b1,
FeCRA does not: this permits selective reductions
(Scheme 12)[28b1.
+ R'-CO-R2
Me-(CH2)5-CH=CHa
Me-(CHa)6-Me
98%
R1, Ra = -(CHz),-:20
FeCRA
THF. 4592
+ R'-CO-Rz
+ R'-CHOH-R2
90 - 9570
1-570
h;R' = R2 = Me-(CH2)3: 16 h
Scheme 12.
CRA's are as efficient as reagents containing LiAIH4 or
NaBH4 and a metal salt (216, 321. Their selectivity in competitive reduction of alkenes is also comparable. On the
other hand, CRA's as well as LiAIH,-containing
are more selective than NaBH,-containing reagents[32b1
in the reduction of alkynes to alkenes. Finally CRA's
are much more efficient than LiH-metal salt systems in
which the very reactive and pyrophoric LiH is prepared
from tBuLi and hydrogenc2"'.
2.6. Reduction of Ketones and Aldehydes and
Regioselective Reduction of a,p-Unsaturated Ketones by
CRA's
Ketones and aldehydes, whose reductions are of synthetic interest, constitute ideal substrates to ascertain
whether the basic properties of CRA's really have been
controlled.
Exploratory experiments performed with tAmONa .NiCRA (4/1/1)128a1showed that enolizable ketones might be
reduced ; however, these reductions suffer from long reaction times, only moderate yields with some substrates, and
retro-reactions of varying rate (reoxidation of the alcoholate produced). We were able to
that reoxidation
was due to decomposition of transition-metal alcoholates[23.341
(formed in addition to sodium alcoholate), leading to ketones and metal hydrides. These results suggested
that the addition of alkali metal or alkaline earth cations
could increase the reduction rates by electrophilic assistAngew. Chem. Int. Ed. Engl. 22 (1983) 599-613
ance; moreover, these cations might efficiently compete
with Ni" for complexation at the oxygen atom, hence,
helping to limit reoxidation. Among others, Li and Mg halides were found to be particularly suited for this purpose.
They allowed fast NiCRA reduction of ketones in excellent yields, and dramatically retarded the reoxidation process (Scheme 13). Among other CRA's examined under the
same conditions, tAmONa- CdCRA (4/1/1) and, particularly, tAmONa.ZnCRA (4/1/1), were found to be very efficient reducing agents for ketones.
Reduction of aromatic or a-alkylated aliphatic aldehydes by NiCRA (with Li or Mg salts) presented no difficulties, but low yields of reduction products were obtained
with primary aliphatic aldehydes. In these cases, satisfying
results were obtained using ZnCRA in the absence of
added salts (Scheme 13)[331.
R:
NKRA-MgBrz
R1,
RZ
THF,40°C
R2
Rl
R2
nBu
iPr
nBu
/C=O
,CH- -OH
It was also shown that reductions performed with NiCRA could be catalytic with respect to Ni, tAmONa, and
MgBrz[331.On the other hand, with ZnCRA yields were
never higher than 180% relative to ZnL31b1.
Among the myriad of reagents able to reduce ketones
] CRA's occupy a conand aldehydes [see e. g. [19b,h,n,20d,Z1b1
solidated position among those species capable of operating at not too low a temperature. However, their stereoselectivity has to be improved. We already know that it is possible to modify the stereochemistry of the reduction by
varying the nature of the alcoholate, and this, therefore,
constitutes one of the ways we have at our disposal to improve the performance of these reagents.
It should be noted that CRA's are much more efficient
than NaH-FeCl,, NaH-FeC13[29b1,
or LiH-transition-metal
salt systems['"]; moreover, commercial NaH is much less
93-1001
"',,CH-CH,
NCRA-MgBrz
R3
FR4
0
f
IhI
0.75
2.5
iPr
R'
R2
f
[hl
tBu
Me
tBu
Ph
6.5
0.5
\
Ihl
n
Ihl
n
2
3
0.5
0.25
4
1
8
0.5
R'
Me
tBu
Me
Ph
Me
Ph
NCRA-MgBrz
R2
R3
H
H
H
Me
H
Me
H
H
H
H
Me
H
R4
Me
fBu
Me
Me
Me
Ph
ZnCRA-MgBr,,
R 1 ~ R 3
THF. 2 0 T
Ra
t [hl
M=Ni
Yield
0.5
2.5
0.25
0.25
0.75
3.5
30
93
98
85
99
50
[%I
I
CHXR4
[h]
M=Zn
Yield [%]
0.5
3
2
20 [a]
2
4
55-60'"'
93
98l"'
45-50["]
93 [b]
98 [a]
[a] 20°C. [bI 40 "C
R
t [h]
2-Me
4-tBu
3,3,5-Me3
3,3,5,5-Me4
1
1
23
3
Yield
[%I
Isomer ratio
OH
25c : 751
25c : 751
60ax : 40eq
98
95
83
85
R
t [hI
Yield [%]
ex0 : endo
H
0.5
5
100
98
32 : 68
50 : 50
1,7,7-Me3
R-CHO
____t
THF. 4 0 ° C
H
H
3,5,5-Me3
3,5,5-Me3
Reagent
NiCRA-MgBr2
ZnCRA-MgBr,
NiCRA-MgBr2, THF
ZnCRA-MgBr,, THF
R
B
A
R
6
0
R
T["Cl
20
20
40
40
C
I
[h]
Yield [%]
A
B
-
-
50
0.5
98
93
-
25
5
3.5
-
90
-
0.5
6
C
R-CHZOH
R
Reagent
f
[hI
Ph
I-Methylbutyl
Cyclohexyl
Hexyl
NiCRA-LiCI
NiCRi-MgBr2
NiCRA-MgBr,
ZnCRA
1
2.5
1
3
Yield [Yo]
98
98
98
60
ZnCRA-MgBrz
THF,40°C. 3h
'
+OH
930J0
A
Scheme 13. Reaction times t are for comparable amounts of substrate.
Angew. Chern. Inl. Ed. Engl. 22 (1983) 599-613
Scheme 14. Reaction times 1 are for comparable amounts of substrate
605
-“:
expensive and easier to handle than the very reactive,
ZnCRASi
,C=O
,CH-0-SiMe3
90-9570
highly pyrophoric LiH used by Ashby et al.
THF, 20-25’C
R2
R2
The results obtained in reducing alkenes and saturated
ketones encouraged us to attempt the regioselective reducR‘
Rl
I [min]
tion of a,b-unsaturated ketones. Under the conditions
Me
20
Me-(CH2)~
found to be satisfactory for saturated ketones it was disMe
Me2C=CH-(CH2)>
15
covered that highly regioselective reductions could be perMe
Ph
45
formed using NiCRA, leading to saturated ketones, and
H
Me-(CH2)s
15
H
15
Ph
with ZnCRA leading to allylic alcohols (Scheme 14)[351.
H
p-CIC6H,
15
It is noteworthy that, with two exceptions, saturated alcohols are formed, if at all, only in trace amounts.
Broadly speaking, reagents which reduce a,B-unsaturated carbonyl compounds may be divided into three
(H2@
:o
ZnCRAL,THF
92 - 9 5%
(H2C0°siMe3
groups:
2O-2S0C,15mm
The first contains reducing agents which mainly or exR = H,n = 2,3; R = 4-tBu. n = 2
clusively lead to saturated carbonyl compounds in moderate to excellent yields: among them are M@[HFe(CO),]@or
ZnCRASi, THF
M@[HFez(CO),lQ(see e . g . literature cited in [’9b1 and [231),
@
M
;
e
2O-2S0C, 7 5 m n
R3SiH-F3CCOOH[361,
NaBH,-~yridine[~’l,
NaHTe (from tellurium powder and NaBH4)l3*],LiBHEt3[lPh1,R C U H L ? ~ ~ ] ,
OSiMe3
or L~AIH,-CUI[~~].
Z~CRASLTHF,
The second group contains reagents which give excellent
91%
/
20-ZSOC. 1.5h
yields of ally1 alcohols. Here are found reagents such as 9BBNr4’]or NaBH,-lanthanoid
The third group comprises reagents which regioselectively reduce a,P-unsaturated carbonyl derivatives to varying extents and/or whose regioselectivity depends on the
structure of the carbonyl compound, on the experimental
R = Me: 30 min, 95%;R = H: 15 min, 91%
conditions, or on the nature of the reagent. In this group
Scheme 15. Reaction times t refer to comparable amounts of substrate.
are a number of aluminum hydrides and borohydrides,
either with or without transition-metal ~ a l t ~ [ ’ ~ ~ , ~ . ~ ~ . ~ ‘ . ~ ~ - ~ ~ ~ ,
pensive reagents, of frequently high reaction temperatures,
as well as alkylal~minurns[~~~,
Li,CuH, + ,I2,=],NaH-Fe” or
and the lack of selectivity towards CC double bonds in unNaH.Fe”’[29a].
saturated carbonyl derivatives. Recently[49b1
an interesting
Comparison of the results shown in Scheme 14 with the
highly regioselective hydrosilylation of a,B-unsaturated
data given in the above cited literature references indicates
carbonyl compounds has been described: with
that 1) the regioselectivity of the inexpensive NiCRA and
R3SiH-(Ph3P)3RhC1 1,4-addition products are formed
ZnCRA reagents is such that they may be classified among
and
with R2SiH2-(Ph3P)&hC1 1,2-addition products.
the best reducing agents of the first and the second groups,
ZnCRASi
compares favorably with the second system and
respectively, and 2) that, on average, the overall yields with
is
much
less
expensive.
CRA’s must be considered as very good.
~
&
2.7. Hydrosilylation of Carbonyl Groups in Saturated
and a,B-Unsaturated Ketones and Aldehydes by
Me,SiCl-Modified MCRA’s (MCRASi’s)
Condensation of three equivalents of Me3SiCI with
tAmONa.ZnCRA (4/1/1) led to a new reagent (ZnCRASi)
which hydrosilylated carbonyl groups[481.The active silicon component appears to be bound to the reactive aggregates. ZnCRASi, an inexpensive reagent, rapidly reduces
ketones and aldehydes to give excellent yields of trimethylsilyl ethers. This reagent specifically reduces carbonyl
groups and does not reduce or isomerize CC double bonds
(Scheme 15).
These results indicate that ZnCRASi is one of the best
reducing reagent for the selective reduction of the carbonyl
group of a,P-unsaturated carbonyl derivatives.
The most general method of hydrosilylating carbonyl
cagroups is the reaction of alkyl, aryl or alkoxy~ilanes[~~’
talyzed by group VIII metal derivatives. The common
drawbacks of these reactions lie in the use of generally ex606
a
2.8. Use of MCRA’s in the Preparation of Selective
Catalysts (Mc’s) for Heterogeneous Hydrogenation
Our results suggest that CRA’s could behave like metal
hydride species (see Section 2.5). Considering that M-H
bond formation is one of the postulated key steps in numerous catalytic hydrogenations[34b1,it might be anticipated that CRA’s could be sources of new hydrogenation
catalysts: exploratory experiments[s01 performed with
tAmONa. NiCRA (4/2/1) were successful. Catalysts obtained from CRA’s are abbreviated “Mc’s” (M = metal).
Nic’s are NiCRA’s (4/2/1) prepared at temperatures
lower than usual, in which residual NaH has been neutralized‘”]. In special cases, soluble alcoholates may be removed easily by washing (Nic,).
Even when dry, Nic’s are not pyrophoric. Nickel constitutes 60 to 70% (by weight) of the ethanol-washed catalyst.
Thus alcoholates, on which the efficiency of Nic’s depends, could be included as part of the catalyst framework.
Nic’s may be used in protic (MeOH, EtOH, etc.) or aprotic
Angew. Chem. Int. Ed. Engl 22 (1983) 599-613
(AcOEt, THF, toluene, etc.) solvents. Interestingly, they
are only very slightly sensitive to catalyst poisoning and
lead to good results with Ni/substrate ratios varying from
1 :20 to 1 :200. Finally, in the presence of NaH and under
an inert atmosphere they are stable for long periods.
2.8.1. Hydrogenation of C C Double Bonds by
tAmONa Containing Nic at Standard P r e ~ s u r e [ ~ ’ ~ ’ ~ ~
Nic is very sensitive to the environment of the alkene
double bond, and selective hydrogenations are easily performed. Moreover, hydrogenation of CC double bonds can
be performed without difficulty in the presence of various
functional groups (Scheme 16). It is noteworthy that although carbonyl groups may be reduced by Nic (see Section 2.8.3), its affinity for the CC double bond is strong
enough to allow selective hydrogenation of a,B-unsaturated ketones.
Hz-Nic, EtOH
Me-( C H, )5-CH=CH2 + Me-( CH2 ),-C=CH,
I
Me
Me-(CH,),-Me
+
2-Octene
770
93%
+
*
2SoC, 9 m n
Me-(CH2),-C=CH2
I
Me
100%
2S0C,7m!n
Complete hydrogenation of mono- and disubstituted alkynes is readily performed using Nic. The reaction proceeds in two steps ; the catalyst preferentially complexes
the a l k ~ n e ‘ ~Taking
~’.
advantage of this, partial hydrogenations of alkyne derivatives can be performed in excellent
yields in the presence of quinoline (Scheme 17)
R‘
R2
Pr
Et
Ph
Et2NCH2
HOCH2
Ph
Et N C H
OH
I
Et-C
Reagent
MeOH
MeOH
EtOH
MeOH
EtOH
EtOH
MeOH
Me-(CH2),-C=CH2
I
Me
98’70
Me
l-cyclohexenyl
Yield
Purity
“1
IW
82
80
80
93
80
86
90
96
91
90
91
95
91
90
Me
Et
Me
Et
CH20H
H
H
H2-Nic,,
H2-Nic,,
H2-Nic,,,
H2-Nic,,
H2-Nic,,
H2-Nic,,
H2-Nic,,
H
H2-Nic, EtOH
90
81
H
H2-Nic, EtOH
84
80
I
Hz-Nit, EtOH
H,C=CH-(CH,),T=CH,
Nle
2.8.2. Partial Hydrogenation of CC Triple Bonds by
tAmONa Containing Nic at Standard Pressure[’”521
Scheme 17. In most cases the impurity was the saturated product
Reagent
Educt
la]
[a]
[a]
[al
Ibl
Ibl
4-Vinylcyclohexene
1,3-Cyclohexadiene
1,5-Cyclooctadiene
1,3-Cyclooctadiene
Allylamine
Ethyl crotonate
[a] H2-Nic, EtOH. [b] H,-Nic,,
Product
f
Yiled
[04
[min]
7.5
12
14
4-Ethylcyclohexene
Cyclohexene
C yc Ioocten e
Cyclooctene
Propylamine
Ethyl butyrate
10.5
22
21
98
98
93
98
95
Taking into account that the yields in Scheme 17 refer to
isolated products, it should be noted that a number of results
obtained with Nic (particularly with 1,4-b~tynediol[’~~
and
l-ethynyl~yclohexene[~~1)
compare favorably with those
obtained with Lindlar palladium. Nic is superior to Raney
100
MeOH
2.8.3. Hydrogenation of C O Double Bonds by
tAmONa Containing Nic at Standard Pressure[’
Z
Reagent
CH20H
CHO
COMe
COOH
H2- Nic,,
H2-Nic,
H2-Nic,
H2-Nic,
Me\
EtOH
or Nic, EtOH
or Nic, EtOH
or Nic, EtOH
-
z [min]
Yield [Oh]
25
95
12
58
92
83
94
91
Hz-NIc, EtOH
C=CH-CO-Me
Me’
29~,45h
93-10oqo
Me
OH
521
It is well known that only a few nickel catalysts allow
the reduction of carbonyl compounds under mild conditions1’61.Among the more efficient are W7-, W6- and modified W4-Raney nickels; however, W7- and W6-Raney nickels have very poor storage stability, W,-Raney nickel appears to be hazardous[’61,and the modifications of W,-Raney nickel needed for hydrogenations of carbonyl groups
are not simple to prepare[”].
Nic does not suffer from these drawbacks and is able to
reduce carbonyl groups under mild conditions, frequently
regioselectively (Scheme 18).
HZ-NLC?THF-EtOH
0
2.8.4. Comparison of Me’s with
other Hydrogenation Catalysts
Y
4 - Cholesten-3 - one
0
0
5-Cholesten-3 - o n e
Scheme 16. Reaction times f refer to comparable amounts of substrate.
Angew. Chem. I n f . Ed. Engl. 22 (1983) 599-613
Nic’s are easily and reproducibly prepared, are non-pyrophoric, can be stored, and are nonhazardous; they allow
selective reduction of double bonds, partial hydrogenation
of triple bonds and hydrogenation of carbonyl groups under ambient conditions. The affinity of Nic’s for CC double bonds and carbonyl groups is sufficiently different, on
607
,c=o
R2
R
=
-”’\
cases, selective hydrogenation of the CC double bond of
unsaturated ketones; however, perhaps P,-nickel is more
selective in special cases.
Since Nic’s are prepared from NaH-RONa, their properties can be varied simply by changing the activating alcoholates. Moreover, many other catalysts may be prepared
in a similar manner to Nic, and some of them (COC,P ~ c [ ~ ’ ] )
appear to be very promising for selective hydrpgenations.
Hz-NIc. EtOH
88-9270
25’C
R2 = Bu: 9 h; R’ = P h , R2 = Me: 2.5 h
HrNlc, EtOH
R D o 25oc
‘DH
R
t [h]
Yield [%]
2-Me
3-Me
4-Me
4-tBu
2-cyclohexyl
2,6-Me2
3,3,5-Me,
3,3,5,5-Me4
4.5
4.2
3.8
4
20
16
6
16
83
82
82
92
94
85
88
90
2.9. Carbonylation of Halogen Derivatives with
CO-Modified MCRA’s (MCRACO’s) at Standard Pressure
2.9.1. Carbonylation of Aryl Halides with CoCRACO Consequencesfor Phase Transfer Catalyzed Carbonylations
H2-Nic, THF-EtOH
>
0
W C , 70rnin
A
HO
fi
Hz-Nrc, EtOH
Ph-CHO
Ph-CHZOH
91%
25OC, 1.6h
Hz-NK,,,, EtOH
Me-(C H2)5-C HO
25% 14h
* Me-(CH2)5-CH20H
88%
Scheme 18. Reaction times t refer to comparable amounts of substrate
Starting from the same ideas as in Section 2.4, we examined the preparation of CRA’s using CO at atmospheric
pressure. For the following reasons we chose to study the
possible preparation of carbonylcobalt compounds and
their ability to carbonylate aromatic halides: 1) the preparation of carbonylcobalt compounds from Co salts in
aprotic media under mild conditions was rare in the literaturefaaJ; 2) contrary to an erroneous
extreme
reaction conditions are required for carbonylation of aryl
halides in the presence of Co
After careful systematic
we succeeded in preparing a series of reagents (termed CoCRACO’s) from CoCRA’s (4/2/1), which reproducibly carbonylated aryl halides at atmospheric pressure of CO (Scheme 19). In the
presence of amines, amides were obtained in moderate to
good yields, and the carbonylations were found to be catalytic with respect to Co.
the whole, to allow selective hydrogenation of the CC double bonds of unsaturated ketones or aldehydes.
As far as efficiency, selectivity and handling are concerned, Nic’s are more interesting than Raney nickels. The
properties of Nic’s are more directly comparable with
Me3CCH20Na<oCRAC0, CO
>
those of Brown’s PI- and P , - n i ~ k e l [ ~ ~ ‘although
* ~ ~ ~ , their
THF, 63OC
compositions are completely different (P-nickels contain
large amounts of nickel b ~ r i d e ‘ ~ ~In] )fact,
.
to a certain degree, they are complementary; P-nickels and Nic’s are very
z~COO-CH2CMe,
+
’ D C O O H
sensitive to the environment of the double bond, show
i
i
rather low propensity toward isomerization or hydrogeno80 - 100%
lysis, and are not subject to poisoning by a m i n e ~ ‘ ~ Use
~.~~].
of P2-nickel, however, allows hydrogenation of unsatu2 = H :lO . h; 2-Me :25 h ; 3 - M e : 2 0 h; 4 - M e : 2 0 h;
rated nitriles to saturated ones[6o1,whereas use of Nic’s
2 - M e 0 : 15 h;3-MeO: 15 h; 4 - M e 0 : 1 5 h; 4 - F : 2 0 h;
does notf6’’. In contrast, carboxyl groups (as sodium salts)
4-Me CO : 2 5 h
do not prevent hydrogenations by Nic’s, whereas cinnamic
Scheme 19. Reaction times f refer to comparable amounts of substrates
acid cannot be reduced using P2-nicke11621.
P-nickels[56c~58~59~631
and Nic’s permit selective hydrogenation of alkenes, but their respective selectivity toward cyInvestigations of the constitution of CoCRACO’s led to
the following conclusions: 1) CQ. 90% of the cobalt was in
cloalkenes is different. On the other hand, unlike P,-nicksolution; 2) only a fraction (cu. 13%) of the soluble cobalt
Nic’s do not have a propensity to disproportionwas present as a carbonyl species; 3) the only carbonylcoate.
balt species present was NaCo(C0),‘681. However, this anPartial hydrogenation of acetylenes can be performed
ion alone does not carbonylate aryl halides under mild
highly selectively. However, in the presence of ethyleneconditions[691.We showed that this apparent contradiction
diamine P,-nickel[641could be slightly better for hydrogenwas due to an SRN1[”J-initiated process (one-electron
ating disubstituted alkynes, whereas Nic’s could be slightly
transfer from He). This result was confirmed when we
better with I-alkynes.
were able to photostimulate the carbonylation of aryl halP-nickels promote hydrogenation of carbonyl groups
ides in the presence of NaCo(C0)4-tAmONaf681.From
only very
whereas Nic’s do so fairly rapidly,
these results, reactions involving CoCRACO’s should prothereby allowing carbonyl hydrogenations to be performed
ceed as depicted in Scheme 20.
under mild conditions. Nic’s promote, even in difficult
Y
608
Angew. Chem. Int. Ed. Engl. 22 (1983) 599-613
mary and secondary halides are catalytic with respect to
the
FcCRACO, CO
Rx
D M E, th en H, O'
' RCOOH + RCOOtAm + RCHO
B
A
L
X
61
QCo(CO),
-Products
X
T ["C]
t
Me-(CH2h
Me-CH-(CH2),-Me
Br
Br
20
20
20
53
75 [a1
68 la1
Cl
65-70
13
60 [a1
Br
CI
Br
C1
30
65-70
40
65-70
40
36
84
80
70 [bl
75 [bl
88 [bl
80 Ibl
Me-CH-(CHz),-Me
etc
[a] Relative yields A
Scheme 20.
[h]
I
Cyclohexyl
Cyclohexyl
1-Adamantyl
I-Adamantyl
L
Yield [Yo]
R
I
+
C
+ B 58-90°/0, C 38-10%. [b] Only A + B are formed.
Scheme 22. Reaction times I refer to comparable amounts of substrate.
The ability of Co(CO)," to react via an SRNlprocess has
an interesting consequence. It is known that Co(CO),"
might be generated from C O ~ ( C Oin
) ~a FTC process[711;
however, under these conditions only organic halides
which are particularly reactive in SN2processes can be carbonylated. These results led us to think that simple irradiation of such systems might allow the carbonylation of aryl
and vinyl halides. This was in fact the case, and we can
now perform numerous reactions of this type (Scheme 21)
by irradiating a FTC system in a pyrex reaction flask with
a simple commercial "sun lamp"172'!
UX
4
RCOOH
X
R
~~
Reagent and
conditions
f [h]
Yield [%]
1-2
1-2
1
2
2
1
2.5
2.5
4.5
2
95-96
96
97
98
97
98
98
98
97
95
~
C~HS
2-MeC6H4
4-MeC6H,
4-MeOC6H4
4-FC6H4
4-C1C6H4
1-cyclohexenyi
1-cyclohexenyl
1-cyclooctenyl
Me3C-C=CH2
I
[a] 1. C O ~ ( C O(catalyst),
)~
CO, benzene/NaOH (aq) or NaOH (aq), Bu,NBr,
hv (sun lamp), 63°C; 2. H,O". P ] 1. C O ~ ( C O(catalyst),
)~
CO; benzene/
NaOH (aq), Bu,NBr, h v (sun lamp), 65 O C ; 2. H,O".
Scheme 21
2.9.2. Carbonylation of Organic Halides with FeCRA CO
FeCRACO is prepared from tAmONa.FeCRA (4/2/1)
using CO at 1 bar pressure. The soluble fraction contains
mainly Na2Fe(CO), and some NaZFe2(C0)8173!
Interestingly, this Na,Fe(C0)4 behaves very differently to pure
Na,Fe(CO),, which has been thoroughly studied by CONman et al.[23,741.
Thus, using FeCRACO primary, secondary, and even tertiary bromides and chlorides can be carbonylated, whereas under normal conditions Na,Fe(CO),
is moderately reactive with secondary halides; however, no
reaction is observed with I-bromoadamantane after 9 days.
Although these carbonylations have to be improved, some
excellent results have already been obtained (Scheme
22)["]. It should be noted that the carbonylations of priAngew. Chem. Inf. Ed. Engl. 22 (1983) 599-613
Of course, a radical mechanism must be suspected for
most of these carbonylations. This is supported by the carbonylation of C6HsBr by FeCRACO under co (40%
yield), the first example of carbonylation of bromobenzene
by Na,Fe(CO),.
2.10. Structure and Mode of Action of CRA's
The determination of the constitution of CRA's is a very
difficult task. However, interesting data have been obtained for tBuONa- NiCRA (4/2/1) and tArn0Na.ZnCR.A
(4,' 1/ 1)[751.
An analysis of the gas evolved during preparation of
C M s and their hydrolyses, and of the phase distribution
of their main constituents yielded important information,
which is briefly summarized in Scheme 23; the notation
defined in Section 2.1 is used. In both cases, the amount of
hydrogen evolved during the preparation was 10 mmol
in excess of that expected for the reaction of NaH and
ROH.
For NiCRA the excess of H, arises from the reaction
2 H Q + Ni"
-+
Nio + HZ
No reduction took place without alcoholate. The crucial
part played by the alcoholate may have two explanations:
1) activation of NaH (Section 1.3), and/or 2) formation of
nickel alcoholates with concomitant favorable shift of the
Ni" reduction potential1761,or electron relay between He
and Mnm.Both phenomena probably occur. Indeed, the
addition of NaH to RONa-Ni(OAc), in THF or DME
leads to a reagent similar to NiCRA, but the rate of formation is lower131b1.Likewise, tAmONa-Co(OAc), reacts
with NaH and CO to give NaCo(CO),; in contrast,
Co(OAc), does not[681.
The amount of hydrogen evolved in the hydrolysis of
NiCRA is consistent with that expected from the "free"
hydride. It is, further, noteworthy that tBuONa was found
only in the solid fraction despite its high solubility in
DME.
Comparison of X-ray powder patterns of the solid fractions with those of the possible constituents or precursors
leads to the following conclusions~751:1) at least some
NiCRA (if not all) is crystalline; 2) NiCRA consists of new
609
60 N a H
+
20 rBuOH
solvent: DME + rBuONa(traces)
DME, 63%
+ 10 Ni ( OAc ) ,
2h
DME
60 N a H + 10 Ni(OAc),
s o l i d : 10 Ni, 2 0 NaH,
19.8 rRuONa
-
20
Ir,
1 to 2 HZ +
6 3 O C , 2h
IZnCKA (4/1/1)
I
-
50 N a H + 10 ZnClz
then 10 tAmOH
50 N a H
+
‘11
-
s o l i d : 10 Zn, 2 0 NaH,
6-6.4 tAmONa
THF
10 ZnCl,
reflux
10 Zn + 10 H, + 30 N a H
20 TI,
J
+ 20 N a C l
very fast
Scheme 23. Above: the NiCRA (4/2/1) shown here consists of NaH, tBuONa, and Ni(OAc)2 in the molar ratio
4:2:1.-Below: the ZnCRA (4/1/1) shown here consists of NaH, IAmONa, and ZnCI2 in the molar ratio
4: I : 1 (numerical values in scheme: mmol).
species in which each constituent (or precursor) has lost its
own characteristics and participates in the formation of
new structures. From the data, NiCRA appears to be constituted of aggregates formed by association of matrices
(see Scheme 24) consisting of Ni atoms surrounded by
Na@,He, tBuOe, and perhaps AcO’.
Scheme 24.
The results for ZnCRAf7’] are interesting. The excess of
Hz observed in the preparation (Scheme 23) arises from the
reaction:
2He
+ Zn“
Zno + Hz
consistent with the results of Ashby et ~ 1 . ‘ Since
~ ~ ~ZnCRA
.
may be obtained by addition of tAmOH to the suspension
obtained after ZnC1, has been allowed to react with excess
of NaH, this suggests that ZnCRA is formed by the reaction of tAmOH with Zn-NaH! Moreover, tAmOH does
not react with Zn (prepared from NaH and ZnCIZ),and the
metal was found in the solid fraction. In contrast to
tBuONa with NiCRA about one third of tAmONa remained in solution with ZnCRA.
Analysis of the X-ray powder patterns study led to the
following conclusions: 1) most of the NaH and about 30%
of Zn are present as such; 2) the remaining Zn and
tAmONa form new species. In other words, ZnCRA has to
be considered as an association of Zn metal, NaH, and
species tentatively formulated as [(tAmO),ZnNa,],. Why
should this mixture be so efficient in reducing substrates
610
such as carbonyl compounds? In trying to answer this
question we noted the following
1)
“[(tAmO),ZnNa,],” obtained from NaH-tAmONa-ZnC1,
(3/1/1) does not reduce ketones; 2) Zn prepared from
ZnCI,-NaH (1/2) does not reduce ketones either; 3) curiously, Zn-NaH obtained from ZnCI,-NaH (1/5) slowly
reduces ketones such as cyclohexanone, but is much less
efficient than ZnCRA.
From these observations it appears that ZnCRA
could be tentatively considered as a mixture of NaH,
[(tAmO),Zn(Na),],,
a
small
amount
of
[(tAmO),(H),Zn(Na),+.1, (the active part), and some
Zn metal. Reductions would take place at the active
part, which could be regenerated from NaH and
[(tAmO),ZnNa,l,.
For chemists not intimately concerned with it, the work
described in this review may appear to be more akin to alchemy than chemistry. This is not completely true. In fact,
a unifying principle has directed the greater part of the research on complex bases and on CRA’s, and I shall attempt to rationalize this beginning with complex bases.
The activation of ionic bases or nucleophiles by alcohoIates must arise from formation of aggregates of the type
+M’-Nu”-MQ-6Rf,.
Of course, NueM@ also activates ROOM@,and once the aggregates are formed they
must be viewed as a single entity. I postulate that these
“supermolecules” presumably possess structures with
marked electron delocalization and strong one-electron
transfer tendencies. More generally, if Nu’M’ is a potential one-electron transfer reagent, an aggregation with appropriate ion pairs (or with itself) would lead to new species with much higher one-electron transfer abilities. It
should be noted that one-electron transfer from
NaNH,-tBuONa has been recently dem~nstrated[’*~.
The active part of CRA’s must also be viewed as consisting of NaH-RONa supermolecules, whose framework contains low oxidation-state metal cores. The metal species
must be stabilized by the surrounding ionic framework and
amplify the one-electron transfer ability of the whole.
Angew. Chem. Inl. Ed. Engl. 22 (1983) 599-613
3. Conclusions
Moreover, they must also participate in anchoring the substrates during the electron transfer process.
If this description is accepted, CRA’s appear to be complex polymeric hydrides with strong one-electron transfer
abilities. The reduction steps could then be: 1) fixation to the
“polymer”; 2) one-electron transfer; 3) hydrogen abstraction. The hydrogen source for the last step is of particular
interest. The low yields of coupled products in the reductions support the conclusion that free radicals are not very
abundant. However, some current experiments show that
some of the hydrogen fixed on the reduced substrates must
be supplied by the solvent. The absence of coupling could
indicate that the radicals formed on the aggregates are still
strongly bound to the reagent until the hydrogen abstraction step.
Thus, it appears plausible that in the reduction of substrates by CRA’s radical species are formed on the polymer. A small fraction of these radicals may be liberated,
resulting in coupling reactions and abstraction of hydrogen
from the solvent. Other radicals strongly bound to the reagent abstract hydrogen from the polymeric hydride. The
remaining fraction, consisting of more loosely bound radicals, abstracts hydrogen from the solvent. Presumably the
lower the amount of free NaH, the greater the degree of
solvent participation.
In view of what is known about reactions between organic substrates and low oxidation-state metal species[791,
together with the recent results obtained by Ashby et al. on
electron transfer from hydrides (and other reagent^)"^'.^^^,
the view of CRA‘s presented above appears to be a good
rationale.
Finally, in Section 1.2 we mentioned the generation
of solvated electrons in HMPA in the reaction between
NaH and water (or alcohols). This reaction, in my opinion,
may be related to the activation hypothesis. Indeed, in
HMPA the a$\e
part of NaH must take the form
(He--Nae--O-wn
(which resembles the aggregates in
complex bases or activated NaH), whose one-electron
transfer ability is increased. Electron transfer through
HMPA from active NaH towards a protic substrate (bound
to HMPA) must be the first step of the reaction (step 1,
Scheme 25). HMPA, a good electron stabilizing solvent‘”],
may trap some of the electrons (step (2)), while the others
follow the normal pathway (step (3)).
Before examining possible future developments of
CRA’s, some remarks must be made.
After we had published our first results on CRA’S[’~~,
some work appeared describing reducing agents involving
NaH-metal salt systems. Thus, Fujisawu et ul.[291showed
that NaH-FeCl, and NaH-FeCl, were able to reduce CC
double bonds and carboxy groups. However, long reaction
times and a considerable excess of reagents had to be used.
On the other hand, Ashby et al.[2e1,
who attempted to reduce I-octene with NaH in the presence of a series of metal salts, concluded their work with the following sentence:
“It was somewhat disappointing that NaH did not produce yields of octane larger than 5%, since NaH is much
less expensive than LiH and is easier to handle”. It must
be noted that even the highly reactive LiH did not lead to
very efficient reducing agents.
In fact, in the course of the hundred or so control experiments we performed with the NaH-metal salts systems
we also observed unsatisfactory reactivity in most cases;
however, with some substrates high yields of reduction
products were obtained. Nevertheless, the reactions were
not very reproducible, and above all, substitution of the
substrate by a similar one led to very low, or to no, yields
of reduction product.
It, therefore, appears that NaH alone can donate its
electrons, but its proton affinity impedes it with a few exceptions; activated by alcoholates or some metal salts its
tendency to electron transfer becomes better but erratic.
Moreover, the basicity of NaH-RONa still constitutes a
formidable drawback.
In CRA’s NaH, RONa and metal species occur together,
and a synergism between the constituents seems to arise
which concentrates their efforts toward a goal: to donate
electrons to organic substrates.
In the future some work which should be performed is
dictated by these observations. I am convinced that the
study of physical properties of solid CRA’s, such as conductivity and magnetic behavior, should lead to important
information and perhaps show that they constitute a new
class of electronic materials. Moreover, it is not unreasonable to assume that CRA’s could be the first members of a
new generation of electron or hydrogen reservoirs.
To generalize this view, it is proposed that in an appropriate medium radical formation as well as numerous
“ionic reactions” could be initiated by an electron transfer
through electron-transmitting, separated nucleophile-electrophile ion pairs.
Of course, the initial studies on the constitution of
CRA’s, which we have performed on NiCRA and ZnCRA,
must be extended.
The role played by the activating agents is so important
that it appears of paramount importance to continue the
Angew. Chem. Int. Ed. Engl. 22 (1983) 599-613
61 1
study of the influence of alcoholates on the properties of
CRA's. Above all, activating agents with heteroatoms other
than oxygen should be studied. Presumably elements such
as sulfur, for example, should be excellent electron regulators. The use of "chelating" and polymeric activating
agents would also lead to new CRA's.
A further point must be raised. We showed in Section
2.6 that CRA's containing two different metals have interesting properties. It is clear that the association of two (or
more?) complementary metals should open new possibilities for application of CRA's. Use of these multimetal-containing CRA's to obtain new catalysts should also be investigated.
For elucidating the mechanism of reduction by CRA's,
reactions with NaD should be very informative. Currently, the greatest problem is to find pure NaD. Sodium
deuteride is very expensive since it has few uses. If
deuteration were possible with CRA's, NaD would become less expensive than other deuterides. It should also
be noted that NaH, which is relatively cheap, would become still less expensive if proper packaging were available to facilitate its use in industry, and extend its use as a
reducing agent.
Among the problems which must still be considered is
the investigation of the possibilities offered by CRA's to
oligomerize or polymerize vinyl or dienyl
The use of asymmetric activating agents would also be of
interest to induce asymmetric reductions as well as asymmetric coupling reactions or carbonylations.
Carbonylations must also be developed further in
aprotic media and under phase transfer catalysis conditions. Specifically the concept of SRNl-condensation of
metalates should be extended.
To conclude it can said that the study of CRA's and related reagents is far from completion. If all the working
hypotheses expressed in the work covered in this review
are correct, then we have achieved the expected results in a
designed way. If, on the contrary they are wrong, then we
have been very lucky!
I express my sincere appreciation to my coworkers who are
individually identiJied in the references, particularly to J. J.
Brunet. I am indebted to them for their experimental as well
as intellectual contributions. Our work has been financially
aided by the Centre National de la Recherche Scientifique, as
E.R.A., DGRST (Proscom), Hoechst France, and SociPtk
Nationale Elf Aquitaine (Production). I am very grateful to
Professor K . G . Taylor, visiting Professor in Nancy University, for helpful discussions.
Received: February 18, 1983 [A 461 IE]
German version: Angew. Chem. 95 (1983) 597
[I] J. Plesek, S. Hermanek: Sodium Hydride, Iliffe, London 1968.
[2] See e. g. a ) . M. Z . Gladych, R. Hornby, Chem. Ind. (London) 1970, 652;
b) M. Natsume, S. Kumadaki, Y. Kanda, K. Kiuchi, Tetrahedron Lett.
1973, 2335; c) Y. J. Lee, W. D. Closson, ibid. 1974. 381; d) R. B. Nelson,
G. W. Gribble, J. Org. Chem. 39 (1974) 1425; e) E. C. Ashby, S. A. Noding, h i d . 45 (1980) 1041.
[3] P. Caubere, Bull. SOC.Chim. Fr. 1966. 1293.
141 HMPA is now thought to be a potent chemical carcinogen and should
be handled with extreme care; H. Spencer, Chem. Ind. (London) 1979,
728; On toxicity see also R. Kimbrough, T. B. Gaines, Nature (London)
211 (1966) 146.
612
[5] R. G. Pearson, J. Songstad, J. Am. Chem. SOC.89 (1967) 1827.
[61 P. Caubere, J. Moreau, Tetruhedron 25 (1969) 2469.
[7] P. Caubere, J. Moreau, Bull. SOC.Chim. Fr. 1970, 1986.
181 H. G. Kuivila, 0. F. Beumel, J. Am. Chem. SOC.80 (1958) 3798; H. G.
Kuivila, ibid. 84 (1962) 3584; R. Barlet, Y. Vo Quang, Bull. SOC.Chim.
Fr. 1969, 3729 Review; C . L. Osborn, T. C. Shields, B. A. Shoulders, C.
G. Gardenas, P. D. Gardner, Chem. Ind. (London) 1965, 766.
191 J. Moreau, P. Caubere, Tetrahedron 27 (1971) 5741.
[lo] P. Caubere, J. Moreau, Bull. SOC.Chim. Fr. 1971, 3270; 1971, 3276.
111) P. Caubere, Acc. Chem. Res. 7 (1974) 301; Top. Curr. Chem. 73 (1978)
S O ; For recent applications 1) Generation of carbanions: G. Ndebeka,
These d'Etat, Nancy (France) (1979) (registered at CDST Centre National de la Recherche Scientifique) 2) Anionic polymerizations: S. Raynal, S. Lecolier, G. Ndebeka, P. Caubtre, Polymer 23 (1982) 283, and references cited therein.
1121 P. Caubere, J. Moreau, Tetrahedron 26 (1970) 2637.
[13] P. Caubere, unpublished results.
[14] J. J. Brunet, R. Vanderesse, P. Caubere, J. Organomet. Chem. 157(1978)
12s.
[lS] G. Guillaumet, L. Mordenti, P. Caubere, J. Orgonomet. Chem. 92 (1975)
43.
1161 G. Guillaumet, L. Mordenti, P. Caubere, J. Organomet. Chem. 102
(1975) 353.
1171 B. Loubinoux, R. Vanderesse, P. Caubere. Tetrahedron Lett. 1977,
3951.
[IS] R. Vanderesse, J. J. Brunet, P. Caubere, J. Org. Chem. 46 (1981) 1270.
1191 a) A. R. Pinder, Synthesis 1980, 425; b) A. Hajos: Complex Hydrides and
Related Reducing Agents in Organic Synthesis, Elsevier, Amsterdam
1979; c) J. E. Johnson, R. H. Blizzard, H. W. Carhart, J. Am. Chem. SOC.
70 (1948) 3664; d) L. W. Trevoy, W. G. Brown, ibid. 71 (1949) 1675; e)
H. C. Brown, S. Krishnamurthy, J. Org. Chem. 34 (1969) 3918; f) C. W.
Jefford, D. Kirkpatrick, F. Delay, J. Am. Chem. SOC.94 (1972) 8905; g)
H. C. Brown, S. Krishnamurthy, ibid. 95 (1973) 1669; h) Tetrahedron 35
(1979) 567; i) S. Krishnamurthy, H. C. Brown, J. Org. Chem. 45 (1980)
849; j) S. K. Chung, ibid. 45 (1980) 3513; k) P. R. Singh, A. Nigam, J. M.
Khurana, Tetrahedron Lett. 21 (1980) 4753; 1) E. C. Ashby, R. N. de
Priest, A. B. Goel, ibid. 22 (1981) 1763; m) S. Krishnamurthy, H. C.
Brown, J. Org. Chem. 47 (1982) 276; n) J. Malek, M. Cerny, Synthesis
1972, 217.
[20] a) H. M. Bell, C. W. Vanduslice, A. Spechar, J. Org. Chem. 34 (1969)
3923; b) R. 0. Hutchins, D. Hoke, J. Keogh, D. Koharski, Tetrahedron
Lett. 1969, 3495; c) R. 0. Hutchins, R. J. Bertsch, D. Hoke, J. Org.
Chem. 36 (1971) 1568; d) C. F. Lane, Synthesis 1975, 135; e) R. 0. Hutchins, D. Kandasamy, C. A. Maryanoff, D. Masilamani, B. E. Maryanoff, J. Org. Chem. 42 (1977) 82; f) H. Toi, Y. Yamamoto, A. Sonoda, S.
I. Murahashi, Tetrahedron 37 (1981) 2261.
1211 a) J. R. Hanson, Synthesis 1974, 1; b) Tse-Lok Ho, ibid. 1979, 1.
1221 H. G. Kuivila, L. W. Menapace, J. Org. Chem. 28 (1963) 2161; H. G.
Kuivila, Ace. Chem. Res. I (1968) 299.
[23] J. P. Collman, L. S. Hegedus: Principles and Applicatzons of OrganoTransition Metal Chemistry, University Science Books, Mill Valley, CA
1980.
1241 a) E. C. Ashby, J. J. Lin, A. B. Goel, J. Org. Chem. 43 (1978) 183; b) E.
C. Ashby, J. J. Lin, Tetrahedron Lett. 1977, 4481; c) J. Org. Chem. 43
(1978) 1263; d) S. Masamune, P. A. Rossy, G. S. Bates, J. Am. Chem.
SOC.95 (1973) 6452.
[25] a) R. A. Egli, Helu. Chim. A d a Sl(1968) 2090; b) S. T. Lin. J. A. Roth, J.
Org. Chem. 44 (1979) 309; c) I. D. Entwistle, P. Bochm, R. A. W. Johnstone, R. P. Telford, J. Chem. SOC.Perkin Trans. 11980. 27; d) B. Meunier, J . Organomet. Chem. 204 (1981) 345; e) T. Yoshida, E. I. Negishi, J.
Chem. SOC.Chem. Commun. 1974, 762.
[26) a) R. Vanderesse, These d'Etat, Universite de Nancy 1981 (CDST Centre
National de la Recherche Scientifique); b) R. Vanderesse, J. J. Brunet, P.
Caubere, to be published.
1271 a) A. S. Kende, L. S. Liebeskind, D. M. Braitsch, Tetrahedron Lett. 1975,
3375; b) M. Zembayashi, K. Tamao, J. Yoshida, M. Kumada, ibid. 1977,
4089; c) T. T. Tsou, J. K. Kochi, J.Am. Chem. SOC.101 (1979) 7547, and
references cited therein. d) M. F. Semmelhack, P. Helquist, L. D. Jones,
L. Keller, L. Mendelson, L. Speltz Ryono, J. Gorzynski Smith, R. D.
Stauffer, ibrd. 103 (1981) 6460, and references cited therein.
[281 a) J. J. Brunet, L. Mordenti, B. Loubinoux, P. Caubere, Tetrahedron Lett.
1977. 1069; b) J. J. Brunet, P. Caubere, h i d . 1977, 3947.
1291 a) T. Fujisawa, K. Sugimoto, H. Ohta, Chem. Lett. 1976, 581; b) J. Org.
Chem. 41 (1976) 1667; T. Fujisawa, K. Sugimoto, H. Ota, Japan Kokai
7691,202, Chem. Abstr. 85 (1976) 1 4 2 2 9 9 ~ .
1301 P. W. Jolly, G. Wilke: The Orgunic Chemistry of Nickel, Vol. 2, Academic
Press, London 1975.
[31] a) P. Jacob, DEA, Universite de Nancy 1979; b) D. Besozzi, These 3e cycle, Universite de Nancy 1982 (CDST Centre National de la Recherche
Scientifique).
[32] a) E. C. Ashby, J. J. Lin, J. Org. Chem. 43 (1978) 2567; b) S . K. Chung,
ibid. 44 (1979) 1014.
I331 J. J. Brunet, L. Mordenti, P. Caubere, J. Org. Chem. 43 (1978) 4804.
Angew. Chem. Int. Ed. Engl. 22 (1983) 599-613
[34] a) H. D. Kaesz, R. B. Saillant, Chem. Rev. 72 (1972) 231; b) P. W/ Jolly,
G. Wilke: The Organic Chemistry of Nickel, Vol. 1, Academic Press, New
York 1974.
[35] L. Mordenti, J. J. Brunet, P. Caubere, J . Org. Chem. 44 (1979) 2203.
[36] D. N. Kursanov, 2. N. Parnes, N. M. Loim, Synthesis 1974, 633.
[37] W. R. Jackson, A. Zuraiyah, J . Chem. SOC.1965, 5280.
[38] M. Yamashita, Y. Kato, R. Suemitsu, Chem. Lett. 1980, 847.
1391 S. Masamune, G. S. Bates, P. E. Georghiou, J . Am. Chem. SOC.96 (1974)
3686.
1401 E. C. Ashby, J. J. Lin, R. Kovar, J . Org. Chem. 41 (1976) 1939.
1411 S. Krisnamurthy, H. C. Brown, J. Org. Chem. 40 (1975) 1864, and refer-
ences cited therein.
[42] J. L. Luche, L. Rodriguez-Hahn, P. Crabbe, J . Chem. SOC.Chem. Commun. 1978. 601; J. L. Luche, J . Am. Chem. SOC.100 (1978) 2226; A. L.
Gernal, J. L. Luche, ibid. 103 (1981) 5454.
[43] M. R. Johnson, B. Rickborn, J . Org. Chem. 35 (1970) 1041.
1441 M. F. Semmelhack, R. D. Stauffer, A. Yamashita, J . Org. Chem. 42
(1977) 3180.
[45] R. 0.Hutchins, D. Kandasamy, J . Org. Chem. 40 (1975) 2530.
[46] J. M. Fortunato, B. Ganen, J . Org. Chem. 41 (1976) 2194.
1471 A. M. Caporusso, G. Giaconelli, L. Lardicci, J. Org. Chem. 47 (1982)
4640.
[48] J. J. Brunet, D. Besozzi, P. Caubbre, Synthesis 1982, 721.
[49] See e . g . a) H. Matsumoto, Y. Hoshino, Y. Nagai, Bull. Chem. SOC.Jpn.
54 (1981) 1279, and references cited therein; b) I. Ojirna, T. Kogue, Organometallics 1 (1982) 1390, and references cited therein; c) C. Chuit, R.
J. P. Comu, R. Perz, C. Reye, Synthesis 1982, 981, and references cited
therein.
J. J. Brunet, P. Gallois, P. Caubtre, Tetrahedron Lett. 1977, 3955.
J. J. Brunet, P. Gallois, P. Caubere, J. Org. Chem. 45 (1980) 1937.
P. Gallois, J. J. Brunet, P. Caubere, J . Org. Chem. 45 (1980) 1946.
R. J. Card, C. E. Liesner, D. C. Neckers, J. Org. Chem. 44 (1979) 1095,
and references cited therein.
1541 See e.9. R. Romanet, C. R. Acad. Sci. 236 (1953) 1044, and references
cited therein; R. J. Tedeschi, G. Clark, J. Org. Chem. 27 (1962) 4323.
1551 a) E. N. Marvell, J. Tashiro, J. Org. Chem. 30 (1965) 3991; b) E. N. Marvell, T. Li, Synthesis 1973. 457.
[56] a) R. L. Augustine: Catalytic Hydrogenations, Marcel Dekker, New
York 1965; b) M. Friefelder: Practical Catalytic Hydrogenation, WileyInterscience, New York 1971; c) C. A. Brown, V. K. Ahuja, J . Org.
Chem. 38 (1973) 2226.
[57] A. A. Pavlic, H. Adkins, J. Am. Chem. SOC.68 (1946) 1471, and references cited therein
[SS] H . C. Brown, C. A. Brown, J. Am. Chem. SOC.85 (1963) 1003; C. A.
Brown, J . Org. Chem. 35 (1970) 1900.
1591 For an excellent review on NaBH,-reduced Co and Ni systems see R. C.
Wade, D. G. Olah, A. N. Hughes, B. C. Hui, Catal. Rev.-Sci. Eng. 14 (2)
[SO]
151)
[52]
[53]
(1976) 211.
[60] T. W. Russell, R. C. Hoy, I. E. Cornelius, J. Org. Chem. 37 (1972)
3552.
[61] P. Gallois, These Dr. Ing., Universite de Nancy 1979 (CDST Centre Na-
tional de la Recherche Scientifique).
[62] T. W. Russell, R. C. Hoy, J. Org. Chem. 36 (1971) 2018.
[63] W. Strohrneier, H. Steigenvald, Z. Naturforsch. B 30 (1975) 816.
[64] C. A. Brown, V. K. Ahuja, J. Chem. SOC.Chem. Commun. 1973, 553.
1651 T. W. Russell, D. M. Duncan, S. C. Hansen, J. Org. Chem. 42 (1977)
551.
[66] a) F. Calderazzo, R. Ercoli, G. Natta in I. Wender, P. Pino: Organic Syntheses via Metal Carbonyis, Vol. 1. Interscience, New York 1968, p. 1 ; L.
Cassar, G. P. Chiusoli, Ital. Pat. 868994 (1969); Chem. Abstr. 78 (1973)
147583; b) A. Mullen in J. Falbe: New Syntheses with Carbon Monoxide,
Springer, Berlin 1980, p. 295; references given deal with benzyl halids
Chem. Absrr. 85 (1976) 32638k; c) see [67] and references cited there-
in.
1671 J. J. Brunet, C. Sidot, B. Loubinoux, P. Caubere, J . Org. Chem. 44 (1979)
2199.
[68] J. J. Brunet, C. Sidot, P. Caubere, J. Organomet. Chem. 204 (1980)
229.
1691 I. Wender, P. Pino: Organic Syntheses via Metal Carbonyls, Vol. 1 and 2,
Wiley-Interscience, New York 1968 and 1977; R. B. King, Ace. Chem.
Res. 3 (1970) 417, and references cited therein (we verified this lack of
reactivity).
[70] For review on SRN1 reactions, see J. F. Bunnett, Ace. Chem. Res. 11
(1978) 413; R. A. Rossi, ibid. I S (1982) 164.
[71] Review: H. Alper, Adu. Organornet. Chem. 19 (1981) 183.
1721 J. J. Brunet, C. Sidot, P. Caubere, Tetrahedron Lett. 22 (1981) 1013; C.
Sidot, These dEtat, Universite de Nancy 1982 (CDST Centre National
de la Recherche Scientifique); J. J. Brunet, C. Sidot, P. Caubere, J . Org.
Chem. 48 (1983) 1166; ibid. 48 (1983) 1919.
I731 J. J. Brunet, C. Sidot, P. Caubere, J. Org. Chem. 46 (1981) 3147.
1741 See e.9. J. P. Collman, Acc. Chem. Res. 8 (1965) 342; J. P. Collman, R.
G. Finke, J. N. Cawse, J. I. Brauman, J. Am. Chem. SOC.99 (1977) 2515,
and references cited therein.
[75] J. J. Brunet, D. Besozzi, A. Courtois, P. Caubere, J . Am. Chem. SOC.104
(1982) 7130.
[76] For such ligands effects, see e. g . M. Mikuiya, M. Nakarnura, H. Okawa,
S. Kida, Chem. Lett. 1982, 839, and references cited therein
[77] J. J. Watkins, E. C. Ashby, Inorg. Chem. 13 (1974) 2350, and references
cited therein.
[78] M. Perdicakis, J. Bessiere, C.R. Acad. Sci. 295, I1 (1982) 879.
[79] See e . g . J. K. Kochi: Organometallic Mechanisms and Catalysis, Academic Press, New York 1978.
[SO] See e.g. E. C. Ashby, A. B. Goel, R. N. de Priest, J . Am. Chem. Soc. 102
(1980) 7779; Tetrahedron Lett. 22 (1981) 3729; E. C. Ashby, A. B. Goel,
ibid. 22 (1981) 1879,4783; E. C. Ashby, A. G. Goel, R. N. de Priest, H. S.
Prasad, J. Am. Chem. SOC.103 (1981) 973, and references cited therein.
1811 H. Normant, Bull. Soc. Chim. Fr. 1968, 791; J. K. Kochi: Free Radicals,
Wiley-Interscience, New York 1973.
COMMUNICATIONS
cleophile (Nu) can be oriented antiplanar (anti-reaction via
the transition state 1) or synplanar (syn-reaction via the
transition state 2). Of the SN2' reactions so far investigated
stereochemically the overwhelming majority proceeded
with high syn-preference"'.
On the Stereochemistry of the SN2'Reaction
By Wolf-Dieter Stohrer*
In bimolecular nucleophilic substitution with ally1 rearrangement (sN2' reaction) the entering and the leaving nu[*] Prof. Dr. W.-D. Stohrer
Fachbereich 2, Studiengang Chemie der Universitiit
Postfach, D-2800 Bremen 33 (Germany)
Angew. Chem. Int. Ed. Engl. 22 (1983) No. 8
V
1
Nub-
2
3
In most of the theoretical investigations''] on the stereochemistry of the sN2' reaction this syn-preference is rationalized, in that the quasi-cyclic transition state 2 (as a
Huckel arene) is sufficiently favored stereoelectronically
compared to the non-cyclic transition state 1, although
0 Verlag Chemie GmbH, 6940 Weinheim. 1983
0570-0833/83/0808-0613 % 02.50/0
613
Документ
Категория
Без категории
Просмотров
0
Размер файла
1 497 Кб
Теги
using, reducing, complex, ways, synthesis, sodium, organiz, agenti, hydride, чversatile, novem, cra
1/--страниц
Пожаловаться на содержимое документа